The Centre de coopération internationale en recherche … · 2017-12-17 · The Centre de coopération internationale en recherche agronomique pour le développement (CIRAD) is a

The Centre de coopération internationale en recherche agronomique pour le développement(CIRAD) is a French research organization that specializes in agriculture in the tropicsand subtropics. It is a state-owned body and it was established in 1984 following theconsolidation of French agricultural, veterinary, forestry, and food technology researchorganizations for the tropics and subtropics.

CIRAD’s mission is to contribute to the economic development of these regionsthrough research, experiments, training, and dissemination of scientific and technicalinformation.

The Centre employs 1800 persons, including 900 senior staff, who work in about50 countries. Its budget amounts to approximately 1 billion French francs, more thanhalf of which is derived from public funds.

CIRAD is made up of seven departments: CIRAD-CA (annual crops), CIRAD-CP (treecrops), CIRAD-FLHOR (fruit and horticultural crops), CIRAD-EMVT (livestock productionand veterinary medicine), CIRAD-Fôret (forestry), CIRAD-SAR (food technology and ruralsystems), and CIRAD-GERDAT (management, common services and laboratories,documentation). CIRAD operates through its own research centres, national agriculturalresearch systems, or development projects.

The International Center for Tropical Agriculture (CIAT, its Spanish acronym) isdedicated to the alleviation of hunger and poverty in developing countries of the tropics.CIAT applies science to agriculture to increase food production while sustaining thenatural resource base.

CIAT is one of 16 international agricultural research centers sponsored by theConsultative Group on International Agricultural Research (CGIAR).

The Center’s core budget is financed by 27 donor countries, international andregional development organizations, and private foundations. In 1996, the donorcountries include Australia, Belgium, Brazil, Canada, China, Colombia, Denmark,France, Germany, Japan, Mexico, the Netherlands, Norway, Spain, Sweden, Switzerland,the United Kingdom, and the United States of America. Donor organizations include theEuropean Union (EU), the Ford Foundation, the Inter-American Development Bank(IDB), the International Development Research Centre (IDRC), the International Fund forAgricultural Development (IFAD), the Nippon Foundation, the Rockefeller Foundation,the United Nations Development Programme (UNDP), and the World Bank.

Information and conclusions reported in this document do not necessarily reflect theposition of any donor agency.

ii

Cassava Flour and Starch: Progress in Research and Development

Centre de coopération internationale en rechercheagronomique pour le développement,

Département des systèmes agroalimentaires et ruraux73, avenue Jean-François BretonBP 503534032 Montpellier Cedex 1, France

Centro Internacional de Agricultura TropicalInternational Center for Tropical AgricultureApartado Aéreo 6713Cali, Colombia

CIAT Publication No. 271ISBN 958-9439-88-8Press run: 1,000Printed in ColombiaDecember 1996

Cassava flour and starch : progress in research and development / D. Dufour,G.M. O’Brien, Rupert Best. -- Montpellier, France : Centre de CoopérationInternationale en Recherche Agronomique pour le Développement,Département des Systèmes Agroalimentaires et Ruraux ; Cali, Colombia :Centro Internacional de Agricultura Tropical, 1996.

409 p. -- (CIAT publication ; no. 271)ISBN 958-9439-88-8

1. Cassava -- Flour. 2. Cassava -- Starch. 3. Cassava -- Cassava as food. 4. Cassava --Research. 5. Cassava -- Action research. I. O’Brien, G.M. II. Best, Rupert. III. Centre deCoopération Internationale en Recherche Agronomique pour le Développement. IV. CentroInternacional de Agricultura Tropical.

Copyright CIAT 2002. All rights reserved

CIAT encourages wide dissemination of its printed and electronic publications for maximumpublic benefit. Thus, in most cases colleagues working in research and development should feelfree to use CIAT materials for noncommercial purposes. However, the Center prohibitsmodification of these materials, and we expect to receive due credit. Though CIAT prepares itspublications with considerable care, the Center does not guarantee their accuracy andcompleteness.

iii

Contents

CCCCCONTENTSONTENTSONTENTSONTENTSONTENTS

CONTENTS

Page

Foreword ix

PrefaceDany Griffon and Rupert Best xi

SESSION 1: INTRODUCTION

Chapter1 Adding Value to Products, Byproducts, and Waste Products

of Small and Medium-Scale Cassava-Processing IndustriesDany Griffon 3

2 CORAF NetworksG. Hainnaux 6

3 The Cassava Biotechnology Network and Biotechnologiesfor Improving the Processing Quality of Cassava

A. M. Thro, W. M. Roca, and G. Henry 10

SESSION 2: CURRENT USE AND FUTURE POTENTIAL

Chapter4 Starch Potential in Brazil

M. P. Cereda, I. C. Takitane, G. Chuzel, and O. Vilpoux 19

5 Producing Cassava Flour in Peru and Its Prospects forDevelopment

S. Salas Domínguez, Y. Guzmán, and S. Aquino 25

6 Cassava Starch in Northern Cauca, Colombia:Socioeconomic Evaluation of Its Production and Commerce

Liliana Mosquera P., Myriam Patricia Chacón P.,G. Henry, and G. Chuzel 30

iv


Page

Chapter7 Cassava Starch and Flour in Ecuador:

Its Commercialization and UseCarlos Egüez 42

8 Cassava Products for Food and Chemical Industries: ChinaJin Shu-Ren 48

9 Thai Cassava Starch Industry: Its Current Status andPotential Future

Boonjit Titapiwatanakun 55

10 Sweetpotato Flour and Starch: Its Uses and Future PotentialNelly Espínola 71

11 Prospects for Cassava Starch in VietnamDang Thanh Ha, Le Cong Tru, and G. Henry 78

12 Cassava Flour Processing and Marketing in IndonesiaD. S. Damardjati, S. Widowati, T. Bottema, andG. Henry 89

13 World Production and Marketing of StarchCarlos F. Ostertag 105

SESSION 3: PHYSICOCHEMICAL STUDIES OF FLOURS AND STARCHES

Chapter14 The Role of Common Salt in Maintaining Hot-Paste

Viscosity of Cassava StarchO. Safo-Kantanka and Rita Acquistucci 123

15 Amylographic Performance of Cassava Starch Subjected toExtrusion Cooking

Z. González and E. Pérez 128

16 Improving the Bread-Making Potential of Cassava Sour StarchD. Dufour, S. Larsonneur, F. Alarcón, C. Brabet, andG. Chuzel 133

17 Physicochemical Properties of Cassava Sour StarchC. Mestres, X. Rouau, N. Zakhia, and C. Brabet 143

18 Influence of Gelatinization Characteristics ofCassava Starch and Flour on the Textural Properties ofSome Food Products

S. N. Moorthy, J. Rickard, and J. M. V. Blanshard 150

v

Contents

Page

Chapter19 Two Rapid Assays for Cyanogens in Cassava:

Their Evaluation, Modification, and ComparisonG. M. O’Brien and C. C. Wheatley 156

20 Acute Poisoning in Tanzania: The Role of InsufficientlyProcessed Cassava Roots

N. L. V. Mlingi 166

21 Gari, A Traditional Cassava Semolina in West Africa:Its Stability and Shelf Life and the Role of Water

N. Zakhia, G. Chuzel, and Dany Griffon 176

SESSION 4: BIOCONVERSION AND BYPRODUCT USE

Chapter22 Fermentation in Cassava Bioconversion

M. Raimbault, C. Ramírez Toro, E. Giraud,C. Soccol, and G. Saucedo 187

23 Cassava Lactic Fermentation in Central Africa:Microbiological and Biochemical Aspects

A. Brauman, S. Kéléke, M. Malonga, O. Mavoungou,F. Ampe, and E. Miambi 197

24 A Lactic Acid Bacterium with Potential Application inCassava Fermentation

E. Giraud, A. Brauman, S. Kéléke, L. Gosselin, andM. Raimbault 210

25 Cassava Wastes: Their Characterization, and Uses andTreatment in Brazil

M. P. Cereda and M. Takahashi 221

26 Cassava Starch Extraction: A Typical Rural Agroindustrywith a High Contamination Potential

Olga Rojas Ch., Patricia Torres L., Didier Alazard,Jean-Luc Farinet, and María del Carmen Z. de Cardoso 233

SESSION 5: TECHNOLOGY DEVELOPMENT

Chapter27 Improving Cassava Sour Starch Quality in Colombia

C. Brabet, G. Chuzel, D. Dufour, M. Raimbault, andJ. Giraud 241

vi


Page

Chapter28 Investigating Sour Starch Production in Brazil

R. C. Marder, R. de Araujo Cruz, M. A. Moreno,A. Curran, and D. S. Trim 247

29 Implementing Technological Innovations in Cassava Flourand Starch Processing: A Case Study in Ecuador

Vicente Ruiz 259

30 The Influence of Variety and Processing on thePhysicochemical and Functional Properties ofCassava Starch and Flour

A. Fernández, J. Wenham, D. Dufour, andC. C. Wheatley 263

31 Establishing and Operating a Cassava Flour Plant onthe Atlantic Coast of Colombia

Francisco Figueroa 270

32 Improving Processing Technologies for High-QualityCassava Flour

D. M. Jones, D. S. Trim, and C. C. Wheatley 276

33 Cassava Flour in Malawi: Processing, Quality, and UsesJ. D. Kalenga Saka 289

SESSION 6: NEW PRODUCTS

Chapter34 The Potential for New Cassava Products in Brazil

G. Chuzel, N. Zakhia, and M. P. Cereda 299

35 Extrusion Processing of Cassava: Formulation of SnacksN. Badrie and W. A. Mellowes 304

36 Thai Cassava Flour and Starch Industries for Food Uses:Research and Development

Saipin Maneepun 312

37 Yuca Rava and Yuca Porridge: The Functional Propertiesand Quality of Two Novel Cassava Food Products

G. Padmaja, C. Balagopalan, S. N. Moorthy, andV. P. Potty 323

vii

Contents

SESSION 7: INTEGRATED PROJECTS

Chapter38 Integrated Cassava Research and Development Projects in

Colombia, Ecuador, and Brazil: An Overview of CIAT’sExperiences

B. Ospina, S. Poats, and G. Henry 333

39 The Cassava Flour Project in Colombia: From OpportunityIdentification to Market Development

Carlos F. Ostertag, L. Alonso, Rupert Best, andC. C. Wheatley 358

40 Women as Processors and Traders of Cassava Flour:The Philippine Experience

D. L. S. Tan, J. R. Roa, and E. A. Gundaya 364

41 Developing the Cassava Flour Industry in Rural Areasof Indonesia

A. Setyono, Sutrisno, and D. S. Damardjati 380

APPENDICES

I List of Participants 393

II List of Acronyms and Abbreviations Used in Text 402

Page

ix

Contents

FOREWORD

FOREWORD

The 1994 International Meeting onCassava Flour and Starch, held inCali, Colombia, focused on cassavaproducts and their use and potentialdevelopment. More than 130scientists, representing 29 countries,participated, presenting 45 papersand 80 posters, of which 41 papersare published in these proceedings.

The Meeting was co-chaired byCIRAD/SAR and CIAT, and was madepossible by the sponsorship of theEU, IDRC, MAE, NRI, ORSTOM, UBA,UNESP, and UNIVALLE.

Lodging and facilities wereprovided by CIAT, and we thankDr. W. Scowcroft, Director General ofResearch, and CIAT employees whoseefficient help was invaluable to ourpresentations.

We also thank Dominique Dufour,of CIRAD/SAR and stationed at CIAT,

who supervised the scientificpreparation of the sessions, andwhose dynamism was instrumentalfor the overall organization of theevent. Also much appreciated wasthe smoothly efficient logisticalsupport provided by Mrs. MaríaEugenia Cobo.

Not only the efficient organization,but also the number of participants,their diversity, the quality of theirpresentations, and their willingparticipation, contributed to thevalue of the discussions, makingthis Meeting a significant scientificevent.

CIRAD/SAR and CIAT, asco-publishers of these proceedings,were assisted by CIAT’sCommunications Unit. Despitecareful editing and production, errorsmay remain, for which we take fullresponsibility.

The organizers

xi

Contents

5,000 years ago—have studies infood demand begun to emphasizethe improvement of postharvesthandling, processing, and marketingof cassava and its derived products.Biotechnology research andopportunities are now also takeninto account in R&D programs onproducts, byproducts, and even thewastes produced by processingplants.

The 1994 International Meetingon Cassava Flour and Starch,organized in Cali, Colombia,demonstrated this burgeoningscientific interest in cassavaprocessing and its role in thesocioeconomic growth of developingcountries. Producers, researchers,processors, and consumers ofcassava products have never beforemet in such significant numbers toshare their experiences, presenttheir work and results, andexchange information. Thetechnological development ofcassava processing and conservationwill surely improve as participantsreturn to their work and apply theirnew knowledge.

The themes presented during theMeeting were:

• The existing and potential uses ofcassava in the world.

PREFACE

PREFACE

Until recently, most efforts in tropicalagriculture were focused onincreasing cereal productivity, thusneglecting root and tuber crops suchas cassava, long considered as a“primitive crop,” as “food for thepoor,” and as having “poor nutritionalvalue.” Cassava was rarely includedin R&D programs for tropicalagriculture.

But, with population increase andrapid urbanization in developingcountries, cassava has become moreimportant as a source of food securityand dietary calories for theinhabitants of these countries. Theunusual climatic variations witnessedin recent years, along with theprospect of global warming, highlightfurther advantages of this hardy,drought-resistant crop. Policymakers have therefore become moreaware of the crop’s significance andare encouraging its research.

The root’s remarkable capacity toadapt to various agroecologicalconditions and its potential for highstarch yields first oriented researchtoward increasing productivitythrough varietal improvement, newcultural practices, and cropprotection.

Only since 1985—which isremarkable, considering this edibleroot was domesticated more than

xii


• The physical and chemicalcomposition and functionalproperties of cassava flours andstarches.

• The possibilities of bioconversionof processed products andbyproducts.

• Technological improvement ofcottage and industrial processes.

• Development of new products.• Integrated development of

cassava products to supplymarket needs.

These themes set the scene formany stimulating discussions. Thenecessarily multidisciplinaryscientific approach, together withthe participatory researchapproach—both involving thevarious components of the cassavaproduction, processing, andmarketing system—emerged as a“recurrent pattern” for quality work.

Currently, these approaches form theonly way to contribute significantly tothe socioeconomic growth ofdeveloping countries.

The papers reported the mostrecent results of current researchprograms. They also pointed towardfuture research directions andsuggested ways of translating resultsinto socioeconomic benefits for allgroups involved in cassava.

With the publication of theseproceedings, both those who couldand those who could not attend theMeeting will be able to reap from thewealth of knowledge presented inthese papers, and so develop newmethodologies and new products andtechnology for their production, and,most importantly, better guide thedirection of thinking and planning fortheir communities’ development.

Dany GriffonDeputy Program DirectorCIRAD/SAR

Rupert BestLeader, Cassava ProgramCIAT

SESSION 1:

INTRODUCTION

3

Adding Value to Products, Byproducts, and...

CHAPTER 1

ADDING VALUE TO PRODUCTS,BYPRODUCTS, AND WASTE PRODUCTS OF

SMALL AND MEDIUM-SCALE

CASSAVA-PROCESSING INDUSTRIES1

Dany Griffon*

As urbanization increases in LatinAmerica, governments are becominginterested in markets forcassava-derived products. Nationaland international research projects oncassava and its products have beenset up, attracting new funding fortheir expansion.

The Need for TechnologicalResearch

Originally focused on improved yields,cultivation practices, and cropprotection, cassava research has,since 1985, also focused onprocessing, quality, and new productdevelopment. In 1988, a 3-yearEuropean Union (EU) project, ‘‘Qualityimprovement of cassava-basedfermented products,’’ involvingFrench, African, and Latin Americanresearch institutions, was set up.This project built up knowledge andstrengthened exchange between teamsinvestigating cassava conservationand processing technologies.Traditional fermented products suchas ‘‘gari’’ in Togo, ‘‘chickwangue’’ inthe Congo, and ‘‘sour starch’’ inColombia were chosen for the project.

A follow-up project was proposedto the EU in 1992 as a result ofinterest generated by the first project,especially in sour starch in LatinAmerica; the need to identify new uses

Introduction

The tropical root crop cassava(Manihot esculenta Crantz) isconsidered a ‘‘low risk’’ crop thatadapts readily to a wide variety ofagroecological conditions. It is highlyefficient in the conversion of solarenergy to starch.

Cassava serves as a subsistencecrop for marginal rural populations inthe tropics, because it efficiently usesthe mineral reserves of infertile soils; itcan withstand climatic variations; itcan stay in the ground unharvestedfor long periods; it resists drought;and it can function as a food-securitycrop in times of famine and otherdisasters.

Cassava’s importance in thesocioeconomic development of ruralareas has gained recognition duringthe last 20 years. Historically, its rolein Latin America, where the croporiginated, was that of a basicfoodstuff for rural inhabitants. Now itis also a source of income andemployment for rural populations.

* CIRAD/SAR, Montpellier, France.

1. No abstract was provided by the author.

4


for cassava, and improve marketknowledge; the need to involve smalland medium-scale processing plantsin minimizing their environmentalimpact by treating liquid and solidwastes; and the dynamism,motivation, and experience of theresearch groups assigned to the work.

A Multidisciplinary Project

Under the EU program ‘‘Science andtechnology of the living fordevelopment,’’ the EU commissionapproved a contribution of 760,000ECUs for a 3-year project entitled‘‘Value enhancement of products,byproducts, and waste products ofsmall and medium-scalecassava-processing industries in LatinAmerica.’’

Value enhancement involvesincreasing the value added duringprocessing; designing, developing, andmarketing quality products; andreducing environmental pollutioncaused by processing.

The project aims to help small-and medium-scale cassava producersand processors strengthen theirpositions in existing markets andpenetrate new markets. Researcherswould study markets for cassava andits derived products; match cassavavarieties with the specific technicalrequirements of users; improve thephysicochemical, functional, andnutritional properties of cassavaflours, starches, and other products;develop new second-generationproducts, and carry out feasibilityevaluations; and identify locallyfeasible technologies for treating wasteproducts.

The project has adopted amultidisciplinary—agronomy,economics, and biotechnology—andinterinstitutional approach to achieveoptimal impact. The project brings

together ORSTOM (France andColombia), NRI (United Kingdom),CIAT (Colombia), UNIVALLE(Colombia), the University of BuenosAires (Argentina), UNESP (Brazil), andCIRAD (France, Colombia, and Brazil),whose Rural and Food ProcessingSystems Department is in charge ofgeneral coordination.2 The 3-yearproject was approved in November1992, and funding began in March1993.

Scientific Organization

The project is structured around fivecomplementary research operations,each coordinated by a scientist:

Operation 1 characterizes rawmaterials and evaluates the quality ofcassava flours and starches forprocessing. (Managed by NRI andcoordinated by Dr. June Rickard.)

Operation 2 studies the treatmentof liquid and solid waste productsfrom processing. (Managed byORSTOM and coordinated byDr. Didier Alazard.)

Operation 3 studies thebioconversion of flours and starchesfor the development of new productsfor use in the food industry. (Managedby ORSTOM and coordinated byDr. Maurice Raimbault.)

Operation 4 focuses on improvingthe functional properties of cassavaflours and starches, and studies thephysicochemical and biochemicalproperties necessary for elaboratingnew products. Some of the newproducts being studied are modifiedstarches, cyclodextrins, glucose andmaltose syrups, extruded products,

2. For explanation of acronyms, see ‘‘List ofAcronyms and Abbreviations Used in Text,’’p. 402.

5

Adding Value to Products, Byproducts, and...

Dr. D. Dufour in Cali, Colombia, andDr. G. Chuzel in São Paulo, Brazil.)

Conclusions

The work plan, research teams, andfinancing became operative in 1993.The first results of the research arepresented in these proceedings,showing that added value isindispensable in the generation ofincome and employment. To obtain it,the following activities must be carriedout: varietal improvement to satisfytechnological applications;improvement in raw materialconservation and processing;innovation and diversification of finalproducts; attention to product quality;and marketing of the final products.

Cassava producers, processors,and traders can benefit from thescientific and technical knowledgegenerated by this project, thusobtaining a better market responsetoward this long-neglected tropicalstarchy food.

and fat analogs. (Managed byCIRAD/SAR and coordinated byDr. Gerard Chuzel.)

Operation 5 studies the traditionalmarkets for cassava and potentialmarkets for newly derived cassavaproducts. (Managed by CIAT andcoordinated by Dr. Guy Henry.)

The wide range of cassava clonesin the global germplasm collectionheld at CIAT is vital to the project.

The UNIVALLE team in Colombiaand the UNESP team in Brazil areinvolved in the research operationsmentioned above, and in forming linksbetween processors and productusers. The University of Buenos Airesin Argentina studies the bioconversionof flours and starches.

Accountable to the EU,CIRAD/SAR is responsible for theoverall scientific and financialcoordination. (Managed byDr. D. Griffon with Dr. Nadine Zakhia.In Latin America, coordinators are

6


CHAPTER 2

CORAF NETWORKS1

G. Hainnaux*

* Cassava Network, Institut français derecherche scientifique pour le développementen coopération (ORSTOM), Montpellier,France.


CORAF is run by a 10-manfollow-up committee, six who representAfrican national programs and fourwho are associate members fromEuropean countries. This committeeelects, from among its members, apresident and a vice president torepresent CORAF. They are assistedby the executive secretariat.

Associate Networks

An associate research network is agroup of researchers who worktogether on a research themerecognized as priority by CORAF. Thenetwork aims to:

(1) Strengthen existing agronomicresearch systems and give themregional and internationaldimension;

(2) Promote the acquisition ofscientific knowledge and optimaluse of results;

(3) Encourage joint action withInternational Agricultural ResearchCenters (IARCs) and with otherinternational and regionalorganizations;

(4) Prepare projects and submit themto external funding agencies;

(5) Encourage evaluation of researchin various agroecological andsocioeconomic conditions; and

(6) Facilitate the setting up ofinterdisciplinary teams, and thetraining of researchers.

What Is CORAF?

The Conférence des responsables derecherche agronomique en Afrique del’Ouest et du Centre (CORAF) is a toolfor cooperation in agronomic research.It provides a framework for collectiveaction and for the exchange ofinformation and experience. CORAFaims to:

(1) Promote cooperation, collectiveaction, and information exchangeamong member institutions;

(2) Define common researchobjectives;

(3) Prepare common researchprojects;

(4) Create, operate, and developassociate networks and regionalresearch workers’ teams; and

(5) Collaborate with internationalagronomic research centers,regional or internationalorganizations, and fundingagencies.

The institutional and operativeorgans of CORAF are the plenaryconference, follow-up committee,executive secretariat, and associatenetworks.

7

CORAF Networks

At present, six associatenetworks belong to CORAF, doingresearch on groundnuts, cotton,maize, cassava, rice, and resistanceto drought. The CORAF networkstake into account the bilateral andmultilateral relationships ofmember institutions.

Organization and operation ofassociate networks

An associate network has a generalassembly, and a steeringcommittee. The general assembly iscomposed of the coordinator,national correspondents, and one toseveral associate correspondents.The steering committee comprisesthe coordinator, correspondents,three members nominated by thegeneral assembly, two scientificauthorities outside the network andnominated by the general assembly,and donor representatives. Thesteering committee assists thecoordinator in managing thenetwork and in following up itsscientific activities.

The general assembly’s missionis to establish scientific prioritiesand research orientations. It liaiseswith scientific partners and withother networks, and convenes onceevery 3 years.

Research projects

The scientific activities of a givennetwork are divided into majorthemes that emerge according tonational program needs. Thesethemes are implemented asprojects, which take into account:

(1) The scientific priorities withineach theme, identified by thenetwork’s general assembly;

(2) The potential of each of thenetwork’s partners; and

(3) Acquired experience andexisting work.

The network appoints an authorityto lead each project and specifies thescientific objectives, duration,partners, and resources to beacquired. The network’s steeringcommittee determines the timing andmethodology for the internal scientificevaluation of the work.

Base Centers

A base center is an agronomicresearch center that belongs to anational network and is open toregional and international cooperationwithin the framework of a network. Itbrings together sufficient human,financial, and material resources toattain scientific objectives and achieveresults that are applicable oradaptable to other countries havingthe same development preoccupations.

Operation

A base center is placed under theaegis of an international network andof the national network that sheltersit. It:

(1) Provides the networks withsupplementary means forreinforcing a national program(scientific personnel, equipment,operations);

(2) Contributes to regionalcooperation by improving theworking relationships amongresearch workers of the sameregion (visits, workshops,seminars);

(3) Participates in the training andretraining of scientific andtechnical personnel of countries ofthe region;

(4) Provides expertise to third partiesin the form of support orconsultation; and

(5) Promotes the diffusion ofinformation and publication ofscientific and technicaldocuments.

8


Activities

Base center programs are planned withthe following factors taken intoaccount: national agricultural policies;development needs of each country;national research programs; prioritiesdefined by the respective network;scientific capabilities of members of therespective network; and other regionaland international arrangements inmember countries or outside. Theseprograms aim to:

(1) Improve crops and livestockaccording to socioeconomic,agronomic, biological, andedaphoclimatic conditions;

(2) Develop living collections to makepossible the sharing of availablegenetic resources among memberinstitutions; and

(3) Establish databases and encouragejoint studies of common interest.

The Cassava Network:An Example of an Associate

Network

Members

Network members number 156researchers from agricultural researchinstitutes of CORAF member (orassociate member) countries, that is,Benin, Burkina Faso, Cameroon,Central African Republic, Chad, theCongo, Côte d’Ivoire, France, Gabon,Guinea, Madagascar, Mali, Niger,Senegal, and Togo.

Associate network members areresearchers from agricultural researchinstitutes of countries who do notbelong to CORAF: Belgium, Colombia,Italy, Rwanda, Spain, United Kingdom,USA, Germany, and Zaire.

Other organizations connected withthe Network are the InternationalInstitute of Tropical Agriculture (IITA),based in Ibadan, Nigeria; the

International Board for Soil Researchand Management (IBSRAM), based inBangkok, Thailand; and theInternational Plant Genetic ResourcesInstitute (IPGRI), based in Rome, Italy.

Major research priorities

The Network has three main areasof priorities:

(1) Make an inventory of, characterize,and evaluate germplasm forselection;

(2) Develop technologies for promotinglonger shelf life, postharvesthandling, and improving nutritionalquality; and

(3) Study the management ofcassava-based systems to improvesystem productivity and conditionsfor propagation.

Major collaborative projects

CORAF has begun establishing thematicbase centers in the Congo and Togo.Four projects are under way:

(1) “Setting up and monitoring amultisite agronomic evaluation ofcassava in Africa.” Located in Togo,it has researchers from the Congo,Côte d’Ivoire, France, and Togo.

(2) “Improving African cassavacultivars.” Located in the Congo,the researchers come from theCongo, Côte d’Ivoire, France, Italy,and Spain.

(3) “Improving detoxification methods.”Also located in the Congo, theresearchers are from the Congo,France, and Togo.

(4) “Improving foodstuffs processedfrom fermented cassava.” Againlocated in the Congo, the researchersare from Belgium, Colombia, theCongo, France, Mexico, and Togo.2

2. For more information about the CassavaNetwork, contact the Coordinator, Dr. JosephMabanza, DGRST-ORSTOM, BP 181, Brazzaville,Congo; tel.: (242) 81 26 80 or 81 26 81; telex:5404 (Attn. ORSTOM); fax: (242) 83 22 05.

9

CORAF Networks

Summary of projects and activities carried out by the Cassava Network

Project Activity Country

(1) Improvement of production,processing, and nutritionaltransformation and quality ofcassava in Central and WestAfrica

(a) Create a base center toimprove cassava varieties andcropping systems

Cameroon, the Congo, Gabon,Zaire

(b) Search and evaluate localcultivars; set up a multisitetrial network to assess thegenotype-by-environmentinteraction

Central Africa, Cameroon, theCongo, Gabon, Guinea, Benin

(c) Improve cassava processingand conservation practices;improve nutritional quality ofproducts and byproducts

(2) Cassava agronomy in WestAfrica

(a) Create a thematic basecenter on the improvement ofcassava agronomy

Benin, Côte d’Ivoire, France,Germany, Ghana, Guinea,Senegal, Sierra Leone, Togo

(b) Improve management of soilfertility in cassava-basedfarming systems

Same countries as above

(c) Implement biological controlof cassava pests

Network member countries:France, Germany, Spain

Countries of the networks

10


CHAPTER 3

THE CASSAVA BIOTECHNOLOGY NETWORK

AND BIOTECHNOLOGIES FOR IMPROVING

THE PROCESSING QUALITY OF CASSAVA1

A. M. Thro*, W. M. Roca**, and G. Henry***

Introduction

Cassava plays two roles in tropicalagriculture: it provides food securityfor many countries; and is a source ofraw material for agroindustrialdevelopment. Because cassava is ahighly reliable crop, even on relativelypoor soils, it can play these roles inareas otherwise poor in resources.

The Cassava BiotechnologyNetwork (CBN) is one response byCIAT to cassava’s incognito outsidethe tropics. By 1984, powerful newbiotechnological tools for agriculturalresearch were developing rapidly butchiefly in countries where cassava wasnot grown. Thus, little was being doneto apply these new tools to cassavaeven though biotechnology couldsignificantly enhance cassava as atraditional staple and help developnew end uses for diverse markets.

The CBN was founded in 1988 toprovide a forum for cassavabiotechnology issues and to fostercassava biotechnology research onpriority subjects (CIAT, 1989). Since

then, many cassava biotechnologyresearch projects have been organizedand funded (Table 1).

CBN Objectives

(1) Identify priorities for cassavabiotechnology research.

(2) Stimulate complementary,collaborative biotechnologyresearch on topics of establishedpriority through (3).

(3) Foster free exchange ofinformation on cassavabiotechnology research, includingtechniques, results, and materials.

Defining Biotechnology

Among the many definitions ofbiotechnology is that formulated at theInternational Meeting on CassavaFlour and Starch (held 11-15 January,1994, at CIAT, Cali, Colombia):

“the deliberate use of anorganism, or part of anorganism, to make or modifyproducts or to improve plantsor animals.”

Biotechnologies in the context ofcassava processing include bothgenetic manipulation and the use ofmicroorganisms to effect desiredchanges.

* Cassava Biotechnology Network, c/o CIAT,Cali, Colombia.

** Biotechnology Research Unit, CIAT, Cali,Colombia.

*** Cassava Program, CIAT, Cali, Colombia.

1. No abstract was provided by the authors.

11

The Cassava Biotechnology Network and...

Table 1. Cassava biotechnology research projects, partial listing, 1994.

Research area Number of Countries andprojects international centersa

Tissue culture, Many Barbados, Brazil, Cameroon, Cuba, China, micropropagation Indonesia, Nigeria, Panama, Peru, Samoa,

Venezuela, Zaire, and othersCIAT, IITA

Regeneration 9 China, France, the Netherlands, UK, USA,ZimbabweCIAT, IITA

Transformation 7 Brazil, Canada, UK, USACIAT, IITA

Molecular mapping, 6 France, UK, USAmarkers, fingerprinting CIAT, IITA

Virus resistance 3 China, the Netherlands, USA, Zimbabwe

Cyanogenesis 7 Denmark, the Netherlands, Thailand, USACIAT, IITA

Photosynthesis 2 Australia, USA

Cryopreservation 2 FranceCIAT

Processing Many Argentina, Brazil, Colombia, the Congo,France, Ghana, India, Nigeria, South Africa,Tanzania, UK, and othersCIAT

a. IITA = International Institute of Tropical Agriculture, based in Ibadan, Nigeria.

CBN’s Interest in CassavaProcessing, Including

Flour and Starch

CBN’s interest in cassava processing,including flour and starch, tracesback to the cassava demand studiesCIAT conducted from 1984 to 1986(CIAT, 1987). These were extensivestudies of current and potentialconsumption of cassava, consumerpreferences for cassava versus otherstaples, income generation andemployment opportunities in cassavaprocessing, and use of cassava inanimal feeds. The possibilities forexpanding the cassava market werealso studied, including factors such asproduction costs, competing crops,and government policies.

The studies showed that cassavaproduction potential exceeded cassavaconsumption; that cassava

consumption, and so cassavaproduction, were falling in LatinAmerica; and that demand for cassavaseemed to be present but the marketwas generally unable to bring cassavato the consumer, either in fresh orprocessed form. This meant thatcassava’s advantages—high yields ofhigh-quality carbohydrate produced atlow cost and even on poor soils—werenot benefiting farmers nor urbanconsumers as they might. Thistranslated into a high priority forresearch on cassava processing andproducts to increase markets for thecrop and provide consumers withdesirable products at low cost.

The CBN conducted its ownpriority assessments in 1988 (CIAT,1989) and 1991. Its 1991 survey wasof experts on the value of differentpossible applications of biotechnologyto cassava (Henry, 1991). It revealed

12


that, among possible biotechnologicalinnovations for cassava, improvingstarch quality had a high, anticipatedimpact on small-scale farmers and onthe market value of cassava (Table 2).

In late 1994, the CBN began a5-year study to develop a versatileframework for using primary data forcassava research priority setting. Thiswill refer especially to assessing therelative advantages of biotechnologyas against other research approaches.

The CBN has also begun toestablish its own direct contacts withthe ultimate users of cassavaresearch. In 1993, a CBN case studywas conducted in northern Tanzaniawhere the staple food is a stiff porridgeof cassava flour. The flour is obtainedby pounding dried-then-fermentedcassava pieces. Villagers had specificquality preferences for the traditionalproduct; some women had alsoexperimented, but unsuccessfully,with mixing cassava and wheat flourto produce baked goods for sale in thesmall village restaurant (wheat flourwas expensive and often scarce).

CBN teams were often asked forsuggestions on improving the localmethods of cassava processing, or onmaking a greater variety of products.This perspective, gathered directly

from cassava users, suggests that astrong demand exists for research onthe quality of cassava flour, even inareas with near-subsistence farming,where such demand might not beexpected.

Current Research inBiotechnology with

Reference to CassavaQuality and Processing

Genetic transformation of cassava

Genetic transformation, or geneticengineering, refers to inserting DNA ofone genetic material into a cell ofanother genetic material; ensuring theDNA’s successful incorporation intothe cell’s genome; and, if the DNAencodes one or more genes,subsequently expressing those genesin the phenotype of the cell. The mostpromising methods used to geneticallytransform cassava include physicallybombarding cells with microprojectilescoated with DNA, and using thebacterial vector Agrobacteriumtumefaciens.

Although single cassava cells havebeen transformed, they have yet to beregenerated as uniformly transformedplantlets. Regenerating from singlecallus cells or from protoplasts—

Table 2. Relative importance of cassava constraints and opportunities for which biotechnology may have arelative research advantage, by region and by anticipated impact of biotechnological innovationson small-scale farmers and market value of cassava.a

Biotechnology Importance by region Impact of innovationsresearch topics

Africa Latin America Asia Yield Marketincrease advantage

Viral diseases +++ +++ + +++ +Insect pests +++ +++ + +++ +Cyanide toxicity +++ + ++ 0 ++Starch quality ++ ++ +++ 0 +++Postharvest root ++ +++ +++ 0 +++

deterioration

a. +++ = high; ++ = medium; + = low; 0 = no change.

SOURCE: Roca et al., 1992.

13


already successfully used for genetictransformation in other species—hasnot been reported for cassava. In vitroregeneration of cassava plantlets hasbeen achieved through somaticembryogenesis in a wide range ofgenotypes. These somatic embryosarise from multicellular buds and,when transformed, are chimeric.

Culture studies in embryogenicsuspension are so far promising,and the possibility of othersingle-cell-based regeneration systemsshould be investigated.

Mapping the cassava genome

A framework genetic map of cassava,based on molecular markers, is nowunder construction throughcollaborative interchange agreementsbetween CIAT and the U.S.Universities of Georgia andWashington—St. Louis. Several typesof molecular markers are being usedin the initial mapping work, includingRFLPs from both total genomic DNAand cDNA, and RAPD primers.2

A molecular map of markerslinked to traits of interest has theadvantages that molecular markersare found in all genotypes, they arenumerous (from hundreds tothousands in species so farinvestigated), and they arephenotypically neutral. This meansthat any normal plant will expressmany of them. A further advantage,and perhaps the most valuable toplant breeders, is that molecularmarkers are independent of externalenvironment or the organism’sdevelopmental stage. As a result,molecular markers, and any traitslinked with them, can be scored andselected in any environment and at

any developmental stage, even usingDNA from seedlings.

If, for example, molecular markerswere established for a certain desiredcooking quality of cassava, then abreeding population could be screenedfor that cooking quality even at theseedling stage, and even if thephysicochemical basis of the desiredcooking quality was unknown.

Cassava genomic and cDNAlibraries have been produced. Amapping progeny has been developedfrom the cross Nigeria 2 X ICACebucan, whose parents were selectedaccording to their variation for bothagriculturally interesting traits andmolecular markers. The first group ofuseful polymorphic markers has beenidentified. When completed, theframework map and the mappingpopulation will be made available tocassava breeders and otherresearchers.

Genes for starch quality in cassava

Several research groups, for example,in Brazil, the Netherlands, and CIAT,are interested in working ontransgenic approaches to cassavastarch quality and quantity. Toproduce transgenic cassava withappropriate characteristics,researchers need to control theproportions of amylose to amylopectinso to permit new or wider uses ofcassava starch. One form of control isthrough genes.

A private research group atWageningen University, theNetherlands, used their work withpotatoes to clone the starchbiosynthetic genes of cassava:granule-bound starch synthase(GBSS, responsible for amylosesynthesis), and branching enzyme(BE, responsible for the cross linkagesthat form amylopectin). This group isalso working on regenerating and

2. For explanation of acronyms, see ‘‘List ofAcronyms and Abbreviations Used in Text,’’p. 402.

14


genetically transforming cassava andis positioned to test the starch genesin cassava as soon as atransformation protocol is developed.Because the research is privatelysupported, the genes may not becomeavailable for public use, except in thelong term.

CIAT (whose research resultsbecome publicly available) may be veryclose to having a transformationprotocol for cassava, which, ifconfirmed, will then be optimized.CIAT is also investigating the priorityapplications that are the ultimateobjective of developing the technology.In accordance with cassava researchpriorities, CIAT is working withpublished sequences of the BE, usingpolymerase chain reaction (PCR)technology and a cassava genomiclibrary developed at CIAT. To date,CIAT has obtained several DNA clones,which may contain parts of the BEgene and is sequencing the clones toverify this. Confirmed clones will beused to “fish out” the complete genefrom cassava genomic DNA.

Cassava Cyanogenesis

Understanding the biochemistry ofcassava cyanogenesis has progressedsignificantly. Researchers at theUniversity of Newcastle (UK) havecloned for linamarase, a key enzyme inthe cyanogenesis pathway. When atransformation protocol is available,this cloned gene can be used toproduce acyanogenic cassavagenotypes for use in research on therole of cassava cyanogens and in plantbreeding.

Researchers must first understandthe implications of cyanogens forcassava production and use beforeapplying results of cassavabiotechnology research tocyanogenesis. For example, what isthe role of cyanogenic glucoside

compounds in the plant? Is there arelationship between root cyanogencontent and processing quality?Although research on these topics hasincreased, much more is needed.

Postharvest Deteriorationof Cassava

As cassava becomes more importantas an industrial crop, the logistics ofsupplying fresh cassava to processingplants becomes more critical. Cassavaroots that can be stored for more thana few days would let processors keep areserve of raw material and thusoperate more nearly at maximumefficiency.

A multidisciplinary approach hasbeen outlined for addressing rapidpostharvest deterioration of cassavaroots, a significant production andmarketing constraint. Four years agothis problem was insufficientlyunderstood to be consideredresearchable. Now, if funds wereavailable, research on cassavapostharvest deterioration wouldintegrate biotechnology, cropimprovement, and recent advances inmolecular genetics.

Microorganism-basedBiotechnologies for Cassava

This new area of interest for CBN iswell covered by other papers in theseproceedings (Session 4).

Outlook

Experiences with other crops suggestthat a genetic transformation protocolfor cassava is not far off. Starch geneconstructs, both publicly availableand private, will probably be ready fortesting in transgenic cassava plants assoon as a durable transformationprotocol is available, pending

15


observance of all applicable biosafetyregulations. Work on the frameworkmolecular map of cassava is inprogress.

References

CIAT. 1987. Global cassava research anddevelopment: the cassava demandstudies and implications for thestrategies for the CIAT CassavaProgram. CIAT Cassava ProgramStrategy Document prepared for theBoard of Trustees Meeting, June,1987. Cali, Colombia.

___________. 1989. Report on the foundingworkshop for the Advanced CassavaResearch Network, held at CIAT, Sept.6-9, 1988. Cali, Colombia.

Henry, G. 1991. Assessment of socioeconomicconstraints and benefits to small-scalefarmers from cassava biotechnologyresearch. In: CIAT. Proposal forDirectoraat Generaal voorInternationale Samenwerking (DGIS),Neths., funding of coordination andactivities of the Cassava BiotechnologyNetwork (CBN). Cali, Colombia.

Roca, W. M.; Henry, G.; Angel, F.; and Sarria,R. 1992. Biotechnology researchapplied to cassava improvement at theInternational Center for TropicalAgriculture (CIAT). Agric. Biotech.News Inf. 4:303N-308N.

SESSION 2:

CURRENT USE AND

FUTURE POTENTIAL

19

Starch Potential in Brazil

CHAPTER 4

STARCH POTENTIAL IN BRAZIL1

M. P. Cereda*, I. C. Takitane*,G. Chuzel**, and O. Vilpoux***

Cassava Starch Productionand Uses

Brazilian starch production is almost1 million tons per year: 76% frommaize (700,000 t/year), 23% fromcassava (220,000 t/year), and theremainder from other crops such aspotato and rice (500 t/year) (AdemirZanella, 1992-1993, personalcommunication). Being traditionalBrazilian foods, the last two crops areunlikely ever to play an importantrole in the starch market.

About 45% of maize starch isused raw (320,000 t/year), 40% asglucose and malto-dextrins(280,000 t/year), and 15% asmodified starches (100,000 t/year).In contrast, about 68% of cassavastarch is used raw (150,000 t/year),18% as modified starch(40,000 t/year), 10% as sour starch(22,000 t/year), and about 3% astapioca (8,000 t/year) (AdemirZanella, 1992-1993, personalcommunication).

Because of its high quality andhigh value (US$1.50/kg), arrowrootwill take a significant part of thefuture starch market. Cassavastarch, in contrast, is a low-valueproduct, with prices ranging fromUS$0.27 to US$0.40/kg (AdemirZanella, 1992-1993, personalcommunication).

Annual world productionof starch is currently about29 million tons, obtained from maize(12 million), wheat (10 million),potato (4 million), cassava(0.8 million), and others (2.2 million)(Chuzel, 1991). The main starchproducers are USA (maize), Canada(wheat), and the European Union(potato).

The USA imports 150,000 t ofcassava starch, the EU 50,000 t, andCanada 10,000 t, representing onlyabout 1% of world starch production,but 25% of the world’s cassava starchproduction. Japan imports another300,000 t of cassava starch(Lorenz Industry, 1990, personalcommunication). These countriesuse cassava starch to manufacturemodified starches (Table 1).

Knight (1974) lists differentstarches and their use in food(“waxy” starch has a high level ofamylopectin, a result of geneticmodification):

* Faculdade de Ciências Agronômicas (FCA),Universidade Estadual Paulista (UNESP), SãoPaulo, Brazil.

** CIRAD/SAR, stationed at UNESP/FCA.*** French Technical Cooperation, stationed at

UNESP/FCA.


20


Use Starches used Function

Spice for salads Maize + “waxy” Provide stability in acidity, cutting,starch mixtures and temperature

Filling for “Waxy” starch Provides texture, transparency,fresh-fruit pies and acid stability

Filling for “Waxy” starch Provides stability in acidity and frozenfrozen-fruit pies texture (does not coagulate), and

transparency

Maize-type cream “Waxy” starch Provides heat stability and high viscosity

Ready-made puddings Maize + “waxy” Provide stability in temperature, frozenstarch mixtures texture, and cutting

Baby foods “Waxy” starch Provides stability in frozen texture andhigh viscosity

Table 1. Applications (in percentage) of cassava starch in USA and the European Union.a

Crop Product

Glucose Fructose Alcohol Paper Modified Raw

Maize 30 20 10 10 20 10(40) (20) (-) (10) (20) (10)

Wheat 50 30 10 - 10 -(60) (20) (-) (10) (10) (-)

Potato - - - - 90 10(-) (-) (-) (10) (80) (10)

Cassava - - - - 100 -(-) (-) (-) (-) (100) (-)

a. Percentages in parentheses are values for the European Union.

SOURCE: Lorenz Industry, 1990, personal communication.

environmental conditions andcompetition with the tobacco industry,which has a quicker turnover of crops(cassava takes 1 year to mature).

Cassava Starch Industriesand Markets

Cassava starch industries are locatedin Santa Catarina, Paraná (78%), SãoPaulo, Minas Gerais, and Mato Grossodo Sul (Table 5) with 56 industriesregistered with the AssociaçãoBrasileira dos Produtores de Amido deMandioca (ABAM, 1992-1993). Butthe founding of many new industriesmay have increased this number to 70.Processing capacity is variable, for

Brazil, the world’s leadingproducer of cassava (Table 2), uses80% of its production in food.Although the national production ofcassava is spread over most Brazilianstates (Table 3), northern andnortheastern Brazil grow 67% of thenational crop. Most is used asfood—of the 1991 crop, only 4% wastransformed into starch.

Table 4 compares cassavaproduction in Paraná state with thatin Santa Catarina: planting area inthe first increased by 64%, as didproduction (65%), in the last 10 years.In contrast, in Santa Catarina,planting area dropped by 35%, as didproduction (-13%), because of

21


example, the average is 221 t/day inParaná state and 109 t/day in SantaCatarina. These industries haveequipment of international standard.

The Centro Raizes Tropicais(CERAT), Universidade EstadualPaulista (UNESP), researched 12cassava flour industries in SantaCatarina in 1993 through interviews,which showed an overall production of10,450 t. These results, however,differed from ABAM’s data of the sameyear (16,750 t).

Cassava starch production facesstrong competition from maize starch,the prices of which are stable, andquality is high and consistent. Suchcompetition inhibits the growth and

expansion of cassava starch use. Thestructure of the maize starch marketin Brazil is oligopolistic and is formedby three multinational enterprises:National Starch, Cargil, and CornProducts Corporation.

Maize and cassava starches arecommercialized in the same markets:foodstuffs (cheese breads, cookies,ice-creams, chocolates, processedmeat, and forcemeats), paper andcardboard, textiles, pharmaceuticalproducts, glues and adhesives, andmodified starches.

The biggest problem facing thecassava starch industry is a pricevariability that ranges between 60%and 70%. Prices for cassava roots

Table 2. World production of cassava roots (in millions of tons). Numbers are rounded.a

Producer 1961-1965b 1969-1971b 1991c

Major producers 50.0 (67) 63.5 (66) 99.4 (65)Brazil 21.9 (29) 29.9 (31) 24.6 (16)Thailand 1.7 (2) 3.2 (3) 20.3 (13)Nigeria 7.2 (10) 9.4 (10) 20.0 (13)Zaire 7.7 (10) 10.2 (11) 18.2 (12)Indonesia 11.8 (16) 10.6 (11) 16.3 (11)

Otherd 24.5 (33) 33.2 (34) - -

Total 75.0 (100) 96.7 (100) 153.7 (100)

a. Values in parentheses signify proportion of total by percentage.b. Compiled from FAO, 1990.c. CIAT, 1993.d. About 75 countries.

Table 3. Brazilian cassava production, 1991 crop, by region.

Region Area Output Proportion of Average(ha) (t) national crop yield

(%)a (t/ha)

North 328,792 4,461,354 18 13.5Northeast 1,132,889 12,005,948 49 10.5Middle west 68,819 1,082,950 5 15.7Southeast 134,775 2,118,052 9 15.7South 277,835 4,862,480 19 17.5

Total 1,943,110 24,530,784 100 -

a. Numbers are rounded.

SOURCE: IBGE and CEPAGRO, 1992.

22

Ca

ssava

Flou

r an

d S

tarch

: Progress in

Resea

rch a

nd

Develop

men

t

Table 4. Cassava production in the states of Paraná and Santa Catarina, Brazil, 1981-1993. Numbers are rounded.

Year of Area Growth rate Production Growth rate Yieldcrop (ha) (%) (millions of tons) (%) (t/ha)

Paraná Santa Paraná Santa Paraná Santa Paraná Santa Paraná SantaCatarina Catarina Catarina Catarina Catarina

1981/82 62,490 100 1.2 100 19.5

1982/83 69,870 12 1.3 13 19.7

1983/84 74,688 20 1.4 19 19.3

1984/85 85,800 88,443 37 100 1.7 1.1 41 100 20.0 13.3

1985/86 85,800 84,812 37 -4 1.7 1.2 39 4 20.0 14.4

1986/87 85,445 75,738 37 -14 1.8 1.2 52 3 21.7 16.1

1987/88 85,242 69,469 36 -21 1.8 1.1 52 -1 21.7 16.7

1988/89 77,839 74,756 25 -15 1.6 1.2 33 9 20.8 17.2

1989/90 101,854 67,596 63 -24 2.1 1.1 79 -2 21.4 17.1

1990/91 102,265 63,370 64 -28 2.2 1.0 86 -13 22.1 17.3

1991/92 100,000 56,873 60 -36 2.1 1.0 72 -13 21.0 18.0

1992/93 137,000 57,379 119 -35 2.0 1.0 65 -13 19.6 18.1

Average yield = 18.9 16.5

SOURCE: IBGE, various years.

23


of cassava starch (f.o.b. at factory) aremore stable than those of cassavaroots, which are vulnerable to theroots’ perishability and fluctuate withroot production. Products usingcassava and maize starches are elastic,that is, income positive, whereasproducts from cassava flour areinelastic.

Table 5. Brazilian starch production (in tons) for 1993, and estimated for 1994.

State Starch industries Production Estimated (no.) 1993 production 1994

Paraná 23 132,900 189,600Santa Catarina 21 31,550 56,600São Paulo 5 15,500 28,600Mato Grosso do Sul 4 23,000 29,300Mato Grosso 2 1,500 5,100Espírito Santo 1 3,000 5,000

Total 56 207,450 314,200

SOURCE: ABAM, 1993.

varied erratically between US$19.50(1983), $33.50 (1992), and $51.00(1989) per ton during 1980-1992(Ademir Zanella, 1992-1993, personalcommunication).

Other problems include the factthat the Brazilian cassava starchindustries must also stop workingfor 4½ months/year. Low rootproduction, a long vegetative cycle, andan inferior quality starch also makecassava starch production costly,compared with that of maize starch. Inthe last 3 years, maize prices havefallen against those of cassava roots,thus making the prices of maize starchmore competitive and maize starchmore available, and thus more used byindustries (Venturini Filho, 1993).

Large Brazilian agroindustrialcomplexes that use starch as a rawmaterial have invested in this area toguarantee an adequate supply of goodquality and suitably stored starch.Three examples can be cited: NationalStarch in Santa Catarina and Nestlé inParaná have just bought their owncassava starch industries. FleischmanRoyal in São Paulo has used its ownfactory, Júpiter, to manufacture itsown cassava starch for more than5 years.

Figures 1, 2, and 3 showdifferences between the real prices ofraw material (root), cassava flour, andraw cassava starch in Paraná. Prices

1980 82 84 86 88 90 92

Year

Figure 2. Real wholesale prices for cassava flour.Correct prices until August andSeptember 1993 (after readjustment forinflation). (After ABAM, 1993.)

CR

$/50 k

g

13,500

11,500

9,500

7,500

5,500

3,500

1,500

Figure 1. Cassava farmgate prices, Paraná state,Brazil. Correct prices until August1993 by general price index-internaldemand (deflator). (After FundaçãoGetulio Vargas, 1993, personalcommunication.)

CR

$/t

1980 82 84 86 88 90 92

Year

6,000

5,000

4,000

3,000

2,000

1,000

0

24


200

150

100

50

0

Figure 3. Real prices of raw cassava starch,industrial f.o.b., Paraná, Brazil. Correctprices until August 1993 by generalprice index-internal demand and inSeptember 1993 after readjustment forinflation. (After ABAM, 1993.)

Acknowledgments

We thank Ademir Zanella of HalotekFadel Industrial Ltd. for the data usedin the tables.

References

ABAM (Associação Brasileira dos Produtoresde Amido de Mandioca). 1992-1993.Produção; estoques, capacidadeindustrial das fecularias brasileiras.Paranavaí, Paraná, Brazil.

Chuzel, G. 1991. Cassava starch:current and potential use in LatinAmerica. Cassava Newsl. 15(1):9-11.

CIAT. 1993. Cassava: the latest facts about anancient crop. Cali, Colombia.(Pamphlet.)

FAO (Food and Agriculture Organization of theUnited Nations). 1990. Yearbook.Rome, Italy.

IBGE (Instituto Brasileiro de Geografia eEstatística). 1981-1993. Censoagropecuário. Fundação InstitutoBrasileiro de Geografia e Estatística(FIBGE), Rio de Janeiro, RJ, Brazil.

__________ and CEPAGRO (Centro Estadual dePesquisa Agronómica). 1992.Levantamento sistemático daprodução agrícola. Fundação InstitutoBrasileiro de Geografia e Estatística(FIBGE), Rio de Janeiro, RJ, Brazil.p. 46-47.

Knight, J. W. 1974. Specialty food starches.In: Cassava processing and storage:proceedings of an interdisciplinaryworkshop. Pattaya, Thailand. p. 77-87.

Venturini Filho, W. G. 1993. Fécula demandioca como adjunto de malte nafabricação de cerveja. Ph.Ddissertation. Faculdade de CiênciasAgronômicas, Universidade EstadualPaulista (UNESP), Botucatu, SP,Brazil. 234 p.

Although all Brazilian statesproduce cassava, only the states of thesouth (Santa Catarina and Paraná),southeast (São Paulo), and middlewest (Mato Grosso do Sul) aretechnologically prepared to producecassava starch.

Other constraints to expanding thecassava starch market includefarmers’ ignorance of the market, andlack of promotion of the virtues ofcassava starch. Promotionalpamphlets could be created by theCIRAD/SAR-UNESP project to targetspecific markets, potential markets, orgrowing existing markets.

An example of a growing marketfor cassava starch is beer manufacture(Venturini Filho, 1993). To make7,400 g of beer, 474 g of cassavastarch are needed. Brazilian beerproduction is 5.8 billion liters.Current mixes use malt with maizeand rice grits. If the grits marketcould be divided into three to includecassava starch, a potential120,000 t of cassava starch would beneeded for this sector alone.

1986 87 88 89 90 91 92 93

Year

CR

$/kg

25

Producing Cassava Flour in Peru and...

Peruvian Amazon and the humidtropics. The Instituto deInvestigaciones de la AmazoníaPeruana (IIAP) established a pilotplant for producing cassava flour inPucallpa, capital of the Department ofUcayali, in the center of the PeruvianAmazon. This flour is used for humanconsumption and as a substitute forinputs used in plywood andbread-making industries.

Cassava Production inUcayali, Peru

In 1991, national cassava productionwas 405,725 t, twice that of the1950s. In contrast, other staples suchas potatoes, wheat, and quinoa(Chenopodium quinoa) have decreasedby one-third. Ucayali produces20,000 t of cassava annually, fourthin national production. Consumptioncenters are located on differenttributaries of the Ucayali River and,although tributaries are navigable,most cassava is wasted becausedistances are long, and boats slow andsmall. The highly perishable andbulky roots therefore do not reachmarkets in time.

Yields in the Departments ofLoreto and Ucayali vary from 10 to35 t/ha. The little produce that doesreach urban markets has increased itsprice by 200% in relation to farmgate

Introduction

Concern is increasing worldwide aboutthe social problems of poverty,unemployment, hunger, and mountingchild mortality. In Peru, preliminarydata from the most recent censusshows that a population explosion hastaken place in the last few years. Thisfactor, together with Peru’ssociopolitical and economic problems,has depressed living standards,especially in rural areas, which has toproduce enough food to feed nine citydwellers for every rural inhabitant.But subsistence agriculture isprevalent because of agroecologicalconstraints, lack of infrastructure,and lack of technical and economicresources.

More than two-thirds of Peru hasagroclimatic conditions suitable fortropical crops that can grow in poorsoils, with little fertilization, and areresistant to disease. Such crops havebeen rapidly distributed, and are themost valuable resource in fightinghunger and the greatest hope for ruraldevelopment through agroindustry. Ofthese crops, cassava and plantain arethe most important, both in the

CHAPTER 5

PRODUCING CASSAVA FLOUR IN PERU

AND ITS PROSPECTS FOR DEVELOPMENT1

S. Salas Domínguez, Y. Guzmán, and S. Aquino*

* Instituto de Investigaciones de la AmazoníaPeruana (IIAP), Pucallpa, Peru.


26


prices. In rural areas, cassava istraditionally processed into productssuch as fariña and tapioca, butbecause of inferior quality, theseproducts are not sufficientlycompetitive for urban markets.

In the past, plants for producingcassava flour were installed inPucallpa and Iquitos (Department ofLoreto, north of Ucayali). These failedmainly because the technology did notaccord with the geographical andsocioeconomic conditions of therespective areas.

The inhabitants of Ucayali eatsufficient carbohydrates to complywith the minimum nutritionalrequirements set by the NationalNutritional Institute. That is,142,350 t of roots and tubers and98,550 t of cereals (mostly importedwheat flour) are consumed yearly.

The IIAP Cassava Flour Plant

Background

The farmers of Ucayali, especiallycassava producers, confront severesocioeconomic and political pressuresthat often force them to emigrate enmasse to cities or cocaine areas. In anattempt to keep people on the land,the IIAP looked for ways to proposeand generate appropriate technologies,employment, and organization. TheIIAP suggested integrating productionsystems to permit a more efficient andeffective use of small-farm resourcesand thus improve production.

In 1989, the IIAP, withcollaboration from CIAT (based inColombia), began developingtechnology and machine prototypes forcassava processing. A plant wasestablished at “Fundo Villarica,” IIAP’sexperiment station at Pucallpa,despite a recession in almost all

production areas, which occurred as aresult of the 1991 political andeconomic emergencies.

The pilot plant was conceived aspart of an integrated system.Activities were to complement eachother so to increase potential and thususe more effectively availableresources. The plant was to serve anarea that suffers multiple problems,and the Peruvian Amazon wastargeted.

The plant was complemented byvermiculture (farming of worms),agroforestry, fish farms. Theseactivities not only provide a market forcassava products, but also help slowdown the degradation of naturalresources, for example, worm humushelps improve poor soils. The raisingof small animals, based on productsand byproducts of rural agroindustry(such as cassava flour), helps resolvethe lack of protein in the regional diet.Efficient farm management (and thushigher productivity) reducesemigration.

Objectives

Through research, the plant was togenerate and adapt technologies forprocessing flour, and evaluate andestablish production and qualityparameters. The plant, however, hadto be a successful enterprise tointerest farmers in the potentialsocioeconomic benefits of cassavaflour production. Once farmers beganparticipating, the plant was to offertraining to cassava farmers andprocessors interested in integratingproduction, processing, andmarketing.

To fulfill its functions oftechnological research, flourproduction, product promotion, andtraining, the plant had the followingobjectives:

27


(1) To validate, adapt, and generatetechnologies for processingcassava and its products.

(2) To open markets for cassava-based products such as flour,flakes, and grains.

(3) To integrate the use of the entirecassava plant in animal feed.

(4) To increase the value of cassavaroots, which are underusedbecause of their perishability.

(5) To gradually substitute importedwheat flour.

(6) To provide technical andorganizational training for farmersand mid-level technicians.

(7) To encourage farmers to not onlyproduce cassava, but also toprocess and market it.

Plant facilities

The plant had four sections:(1) reception, storage, andpreparation; (2) washing;(3) chipping; and (4) preliminarysun-drying, artificial drying, milling,and storing the final product. Thearea for storing and preparing rawmaterial was built on higher terrainthan was the chipping area to makeuse of gravity in transferring rawmaterial. The dryer was a traysystem used by CIAT, with a burnerthat, for fuel, used wood discardedfrom sawmills.

To reduce drying time, flakesdestined for animal feed were firstdried in trays, and then sun-dried.The basic machinery was broughtfrom Colombia, but accessories andother equipment were built locally andelsewhere in Peru. The totalinvestment was US$27,000, includingbuildings, machinery, and otherequipment.

From the start, the IIAPencouraged the organizedparticipation of cassava growers sothey could evaluate the possibilities ofother plants under similar direct

management, and so sign agreementsthat permit mutual collaboration.Supplies of raw material came fromsome sectors of the Campo Verdedistrict, near Pucallpa.

Plant operation

The plant operated at 60% capacity, inaccordance with the goal set. Thefollowing five cultivars were used:Señorita, Huangana, Huanuqueña,Arponcillo, and Nusharuna. Bestresults have been obtained with cv.Señorita with a yield of 3.2:1 (root toflour), but is more perishable thanother roots (lasts 2 days). CultivarNusharuna has the most durableroots but its yields are low, 3.9:1, andthe flour is darker because the peel isdifficult to remove.

The percentage of loss from rootdefects after selection and preparationwas high (15%). Although thisproblem could be overcome bydifferentiating root prices, farmers hadto be taught the need for selection.

Overall, the equipment performedwell, except for the screen and dryer.The minimum drying time achievedwas 12 hours, including preliminarysun-drying. Raw material accountedfor 85% of production costs, fuel 7%,and labor 3%. Packaging,depreciation, and maintenanceaccounted for the remaining 5%.

Marketing

The plant targeted the local market,with some initial promotion in Iquitosand Lima. Currently, demand is 70 tof flour per month, of which only 16 tcould be supplied. About 60% ofproduction is sold to bakeries (whichsubstitute as much as 20% of wheatflour) through the Programa Nacionalde Alimentación (PRONAA) and to theprivate company, Cotrip, that makeswater biscuits. About 20% goes toplywood industries, 5% to Lima, and

28


another 5% to Iquitos. The bran,together with discarded roots, is usedfor animal feed.

Key market segments at a nationallevel are still to be identified, andcompetition from imported wheat flourhas to be resolved. Ucayali, forexample, uses 700 t/month, of which80% is for bread making and 20% forplywood industries.

Research

The plant lacked laboratory equipmentfor quality control, which was done byseveral universities andnongovernmental organizations(NGOs). Proximal and microbiologicalanalyses were carried out.

At first, because water quality wasinferior and vermiculture was locatednear the plant, microbiological qualitywas poor. Scientists found fungi,yeasts, fecal coliforms, andclostridium sulfite reducers inquantities above permissible levels,but no Escherichia coli nor salmonella.These problems have been identifiedand solved, and the flour is nowacceptable for human consumption.

Dry matter content of cv. Señoritais 34%. On the average, wholecassava flour contains 84.2% starch,1.4% protein, and 3.1% fiber.

Based on experiences in breadmaking, trials were conducted withbakeries to establish the followingformula for bread preparation:

Wheat flour 80 parts Yeast 3 parts

Cassava flour 20 parts Additive 1 part

Sugar 6 parts Salt 1 part

Fat 6 parts Water 30 parts

Color still has to be improved butflavor and consistency are good.Currently, artisanal modules for

making bread and pasta are beinginstalled to promote the establishmentof similar projects in different ruralsectors.

Training

Training focuses on three levels:(1) university theses; (2) training ruraldwellers to become qualified workers,or, through modular training courses,knowledgeable on any phase of theprocess; and (3) courses for the public,such as bread making for commercialbread makers and housewives.

Achievements

After 2 years of operation, the plantsuccessfully:

(1) Identified, analyzed, andimproved native technology.

(2) Built the productioninfrastructure, using locallyavailable resources. Machinesand equipment were simple,versatile, and adaptable toprocessing other products, suchas plantains, “sachapapa” ortaro, and cassava.

(3) Produced flour that was US$0.25cheaper than wheat flour.

(4) Found favorable local andregional markets. These werevermiculture, agroforestry,raising of small animals,horticulture, and pisciculture.

(5) Made the new technologyeconomic for small-scale farmersto invest and rapidly recuperatetheir investments, thusdiminishing risks whenconditions become unfavorable.

(6) Established a modular system forinstallation and operation, thusenabling each phase of theprocess to be totally independentand thus more efficient.

(7) “Passed the test” of adversepolitical and economic conditions,including violence, recession, andgeneralized poverty.

29


Prospectives

An agreement has been signed withthe Alto Huallaga Special Project tointroduce integrated productionsystems as an alternative tocultivating coca. Currently, the plant

at Tocache is being installed, withCIAT’s assistance. With collaborationfrom Caritas Peru, four plants will beestablished in Puerto Maldonado(southeast Peru), Iquitos andYurimaguas (Department of Loreto),and Tumbes (north coast).

30


CHAPTER 6

CASSAVA STARCH IN NORTHERN CAUCA,COLOMBIA:SOCIOECONOMIC EVALUATION OF ITS

PRODUCTION AND COMMERCE1

Liliana Mosquera P.*, Myriam Patricia Chacón P.**,G. Henry**, and G. Chuzel***

* Cassava Economics Section, CIAT, Cali,Colombia.

** Cassava Program, CIAT, Cali, Colombia.*** CIRAD/SAR, stationed at the Faculdade de

Ciências Agronômicas (FCA), UniversidadeEstadual Paulista (UNESP), São Paulo,Brazil.


Introduction

Cassava plays a major role insubsistence farming in northernCauca, Colombia. About 90% of rootproduction is used for extracting sourstarch, and as much as 80% of sourstarch production is used for makingthe breads “pandebono” and“pandeyuca.” Sour starch has its owncharacteristic functional properties,flavor, and aroma.

In northern Cauca, cassava starchextraction is mainly an artisanalactivity, although processing plantsare mechanized to some extent. Inimportance, this agroindustry ranksthird after the sugar, and editorial andpublishing industries.

Our study aimed to betterunderstand the different problemsaffecting cassava starch production inthe region, and help researchersidentify priority needs for possibletechnology intervention.

The study is part of a research anddevelopment (R&D) program on cassavastarch production being conducted byCIAT’s Cassava Utilization Section.The program’s objective is to offertechnological alternatives to small-scale,cassava starch producers. The programfirst began in 1989, and is based on aninformal network comprising variousLatin American laboratories andinstitutions involved with cassavastarch production. The program alsocomprises regional working groupsthat evaluate technology for starchproduction, study the technical andeconomic system, characterize andevaluate products, treat waste waters,and conduct basic research onfermentation and raw material (Chuzel,1991).

Objectives of the Study

The general objective was tocharacterize starch production andcommerce in northern Cauca, Colombia,and so assess the technical andeconomic performance ofsmall-scale, cassava starch factories.

Specific objectives were:

(1) To statistically analyze the surveyscarried out by the CassavaUtilization Section in 1990 on thetechnical performance of

31

Cassava Starch in Northern Cauca, Colombia:...

small-scale, starch factories, and tostudy the economic performance ofthese and their social implications.The economic survey of starchfactories carried out by universitystudents was used as a base.

(2) To characterize the productionprocess of cassava processingplants and determine the capacityof installed plants.

(3) To identify the seasons whensupplies of cassava sour starch areabundant or limited and determinedistribution channels.

(4) To analyze social characteristicsrelated to starch production andcommerce.

(5) To identify factors limiting starchproduction and commerce.

Methodology

The study began by surveyingsmall-scale, sour-starch producers andmiddlemen. The target populationconsisted of 99 small-scale starchfactories, first surveyed in 1990 whenthe R&D project began, in the towns ofSantander de Quilichao (86 factories)and Caldono (13). Of this group,35 processing plants were selectedand, for reasons of efficiency,stratified according to plant size(small = fewer than 3 workers;large = 4 or more), age of equipment(new = purchased in the last 15 years;old = more than 15 years old), andgeographic area (municipality ofSantander de Quilichao or Caldono).

Selection was randomized, butproportional to stratum size. TheStratified Sampling of Elementstechnique (Pardo Camacho, 1991; deServín and Servín Andrade, 1978),which gives proportional allotment,was used for sampling.

Cassava starch middlemen weresurveyed in Caldono, Santander deQuilichao, and Cali. Because nocensus of middlemen existed, no

specific sample was established. Thesurvey was therefore based on a listof 35 middlemen identified in theprevious survey on small-scale,cassava starch producers; of these35 middlemen, 20 were surveyed.

Numerous problems arose inobtaining comprehensive information,especially that on the volume of starchpurchases and sales. Becausemiddlemen were so reluctant to shareinformation on how they managedtheir businesses, a case study wasconducted, based on informationsupplied by the COAPRACAUCACooperative, Santander de Quilichao.

Starch Processing andCommerce

Root production is a key aspect in theprocessing and commerce of cassavastarch in northern Cauca. Althoughroots are used more to obtain starchthan for human consumption, whenmarket prices drop below the averageCol$32/kg (US$0.05/kg), then freshcassava is sold to cassava dryingplants for use in animal feed.

In the Cauca Department,6,290 ha were planted to cassava in1991, producing 71,624 t at a yieldof 11,387 kg/ha (DepartamentoAdministrativo Nacional de Estadística,Colombia, 1992, unpublished data).

The municipalities of Buenos Aires,Santander de Quilichao, and Caldonoplanted 4,080 ha of cassava,accounting for 64% of the total areaunder cassava cultivation in Cauca.Production reached 39,000 t, 54% ofthe department’s total production.Yields were 9.5 t/ha, almost 16%below the departmental average of11.3 t/ha. But only 2.3% of thenational crop (173,999 ha) is plantedin Cauca because the root’sperishability, and its low andfluctuating prices, among other factors.

32


The amount of cassava on offer tosmall-scale, starch factories averages556 t/year, for a consumption of456 t/year at Col$32,000/t(US$47.00). Although shortages of rawmaterial occur in Cauca at certaintimes of the year, the annual supply ofcassava often exceeds demand,especially for 18 of the factories inSantander de Quilichao. Located closeto the Pan-American Highway, theytend to be oversupplied.

Plant production during1990-1991 was irregular: some plantsoperated sporadically, according to theavailability of raw material andworking capital. For 1990, the averageminimum production was 4.3 t ofstarch per week and the maximum was175.0 t.

For 1991, the plants had anaverage production of 420 t of starchper year (8.7 t/week) and a maximumof 775 t/year (16.1 t/week). Suchfigures indicate that the plants do notwork at full capacity because of thelack of raw material in the area.Production recession caused by lack ofraw material can last from 2 to12 weeks. Of the processors, 51%stated that they required a constantamount of raw material.

Yield

In 1991, sour-starch production perfactory decreased considerably,averaging 97 t/year. Byproducts werebran (fiber and peel left over fromsieving starch) at 42 t/year and“mancha” (scum skimmed off surfaceof sedimented starch) at 8 t/year.Most small-scale, starch factories carryout sweet-starch extraction on request,but production is sporadic becausestarch quality does not always reachindustrial technical specifications.But, at the time of the survey, only onefactory was producing sweet starch(1.2 t/week).

The decrease in sour-starchproduction in 1991 was caused partlyby a lack of both raw material andworking capital. At the same time, theColombian Government beganimplementing a policy of “openeconomy.” Bank credits were closed tostabilize inflation at 22%. FromAugust 1990 to September 1991, the33 starch factories under studyprocessed 16,878 t of starch,producing 3,207 t of sour starch,1,333 t of bran, and 270 t of “mancha.”That is, every 100 kg of roots yielded19% starch, 8% bran, 1.7% “mancha,”and 71.3% of both water (whichcomprises 65% of roots) and waste,that is, peel and starch lost toinefficient processing techniques. Yielddifferences among factories are causedby, for example, cassava variety,harvest age, and postharvesthandling.

Producers can obtain as much as27% starch (wet basis) with 60%technological efficiency, according toexperiments by the Corporación paraEstudios Interdisciplinarios y AsesoríasTécnicas (CETEC), a Colombianorganization that provides technicalassistance to starch-producing farmers.Once the product is processed to 12%moisture content, these values can beobtained per 100 t of cassava. Forsmall-scale, starch factories innorthern Cauca, the cassava-to-starchconversion ratio is 5:1. The200 small-scale, starch factories of thisregion therefore produced a total of8,500 t of sour starch in 1994.

In Ecuador, the cassava-to-starchconversion ratio is 5-10:1. This ratiovaries greatly according to the time ofyear and cassava varieties used.Byproducts (bran and “mancha”) aresold for animal feed (Chuzel, 1991).

In Minas Gerais, Brazil, the“polvilho azedo,” or sour starch,enterprises can process from 1 to 40 tof cassava roots per day. Annual

33


production ranges from 20 to1,000 t/year, and yields from 200 to300 kg of starch per t of roots(Oliver Vilpoux, 1992, personalcommunication).

Procedures, Equipment, andMaintenance

In Colombia, small, semicommercial,cassava starch factories are called“rallanderías.” These factories typicallyhave a grater-sieve and washer-peeler,both motor-driven. Their processingcapacity ranges between 4.4 and 44 tof roots per week, with an overallaverage of 16.2 t/week. Figure 1demonstrates processing in amedium-sized, starch extractionfactory, beginning with the acquisitionof roots.

Root supplies. Small-scale starchprocessors do plant cassava, according

to surveys carried out in 1990 byCIAT’s Cassava Utilization Section.From August 1989 to August 1990,the total area planted to cassava byprocessors averaged 106 ha, of which43% corresponded to the processors’own plots and 57% to rented plots.For 1991, the percentage of processorsrenting land for cassava cultivationdecreased to 54%. That same year,the total area planted to cassava bythe 99 starch processors averaged80 ha. Thus, in the twomunicipalities, the 26 ha planted tocassava the previous year weredestined for other purposes.Furthermore, of the 51% growingcassava, only 33% owned the land and18% rented it; 48% lack land titledeeds, which reduced access to credit.The cost of leasing 1 ha rangesbetween Col$3,000 and Col$40,000(US$4.43-$59.00) per month,averaging Col$10,333.00 (US$15.34)per small-scale, starch factory.

Washing

Roots

Grating

Sieving

Sedimentation

Starch

Moiststarch

“Mancha” + Wastewaters

Fermentation

Dehydration

Drying

Sourstarch

Dehydration

Sweet

Drying

Drying

Bran

Animal feed

Water, external cortex, dirt

Figure 1. Processes performed in a medium-sized, cassava-starch extraction factory, northern Cauca,Colombia. (Modified after Chuzel, 1991.)

Water

34


Washing and peeling. These aredone either manually or in a rotatingdrum. Of the survey respondents, 93%have mechanized these operations,thus reducing women’s participation inprocessing. Before, women wereemployed to peel the roots.

Grating. Grating is carried out byrotors with perforated laminae that arechanged periodically. About 36% offactories change the laminae every90 days, 23% every 60 days, 21% every30 days, and the rest more than90 days.

Sieving. The starch dissolved inwater is separated from the pulp orbran, which is later used in animalfeed. Different types of fabric, placedon rotating screens, are used forsieving, the most common being nylon(58%), canvas (28%), and silk (3%).The fabrics are changed frequently:77% of processors change them every30 days, 11% every 60 days, and therest after 90 days. For 89% of theplants surveyed, sieves are less than10 years old.

Sedimentation. The slurry fromsieving is left to settle. Particles offiber and other fine materials that hadnot been removed during sieving areseparated to form “mancha,” anotherbyproduct used in animal feed.Sedimentation is carried out inconcrete tanks veneered with wood orglazed tile. On the average, processingplants have five sedimentation tanks,each with an average capacity to hold551 kg/day.

Fermentation. To obtain sourstarch, the moist starch is passedthrough a series of tanks, where itremains 15 to 20 days until thedesired acidity is reached. The averagefactory has five fermentation tanks,each with a capacity of 1,030 kg.Sweet starch is obtained bydehydrating and sun-drying the moiststarch after sedimentation.

Drying. Starch is usuallysun-dried on trays or terraces, or onconcrete floors previously covered withplastic to prevent farmyardcontamination. The dried starch isthen packed for market distribution.

Commerce

The typical distribution of cassavasour starch begins with the cassavafarmer who sells the roots either tomiddlemen or directly to the starchprocessor. Only 7% of processorssurveyed purchase roots only throughmiddlemen; 65% buy directly from thefarmer, and 28% from both. Starch isalso distributed through middlemen ordirectly to users. The middleman sellsto the wholesaler or retailer who, inturn, distributes to intermediateconsumers such as bakeries andindustries that, in their turn,distribute their products directly toconsumers or to distributors ofprocessed food products (Figure 2).

Ultimate consumers

Intermediate processors(bakeries and industries)

Middlemen

Starch processors

Cassava farmers

RetailersWholesalers

Figure 2. Typical distribution chain for cassavastarch in northern Cauca, Colombia.

Middlemen

Distributorsof finishedproducts

35


Starch processors sell most oftheir products in cash but may sell oncredit to regular clients. These clientsalso pay the processor in advancewhen they urgently need starch andthe processor does not have enoughworking capital to fulfill the demand.A credit period is usually 19 days.Although commercializing products isdifficult, stocks rotate quickly. Sourstarch is stored for an average of10 days, sweet starch 8 days, bran11 days, and “mancha” 14 days.

Processors also distribute sourstarch through the COAPRACAUCACooperative, which groups about30 small-scale, starch processors ofthe region, and through middlemen.The cooperative and intermediatemiddlemen, in turn, sell the starch towholesalers and retailers who thendistribute the product to majormarkets in the cities of Santander deQuilichao, Cali, Buga, Cartago, Tuluá,Pereira, Ibagué, Medellín, Bogotá,Cartagena, and Montería.

The municipality of Santander deQuilichao has the highest number ofmiddlemen—which explains why 42%of starch processors sell their productthere—followed by Caldono and Cali,each with 15%, and the other citieswith 28%. For 1991, the average priceper kg of sour starch was Col$230(US$0.39). The byproducts (bran and“mancha”) are usually sold on theretail market in Santander deQuilichao, being mainly used foranimal feed (Figure 3).

Economic Evaluation ofSmall-scale, Starch Factories

Table 1 shows the costs involvedin producing sour starch. Allsmall-scale, starch factories aremechanized to a certain extent soelectricity is necessary. An averagefactory pays US$220/year forelectricity, accounting for 1% of totalcosts. Because the factories mustperiodically change some of

Figure 3. Market channels of cassava starch in northern Cauca, Colombia. (From interviews withCOAPRACAUCA Cooperative members.)

Starch processors

S/der de Quil.1%

Pereira17%

Wholesalers and retailers

Intermediatemiddlemen

Middlemen

Wholesalers and retailers

Cooperative

Ibagué 8%

Cartago20%

Medellín37%

Cali20%

Cali25%

Bogotá24%

Armenia6%

Ibagué6%

S/der de Quil.7%

Medellín 7%

Other cities 8%

Palmira6%

Pereira8%

36


Table 1. Annual average costs involved in producing cassava starch (1991), northern Cauca, Colombia.

Item Col$ US$ Percentage of(1991 value) total costs

Fixed costs 2,882,000 4,836 12.1Energy 131,000 220 0.5Maintenance 487,000 817 2.0Rent 1,082,000 1,815 4.6Administration 855,000 1,435 3.6Others 327,000 549 1.4

Variable costs 20,632,000 34,618 87.9Raw material 16,708,000 28,034 71.5Labor 1,751,000 2,938 7.4Transport 1,958,000 3,285 8.0Packing 215,000 361 1.0

Total costs 23,514,000 39,454 100.0

the equipment they incurmaintenance costs equivalent to 2%of total costs.

Small-scale, starch producerssave money when they own thefactory rather than rent its premises(CIMMYT, 1993). Rental costsdepend on the factory’s location inthe region, but these are normallylow, with little tendency to increase.Only two of the starch processorssurveyed had explicit rental costs,accounting for 5% of total costs.

Administrative costs account for4% of total costs. Usually, the ownerhimself manages the factory, thisbeing his means of support. Anundetermined amount of the earningsfrom the sale of starch is used forhousehold expenses, that is, as hissalary. This fact, perhaps, mostinfluences the efficient operation ofthe factory.

Administrative costs are closelyrelated to the efficiency of anyenterprise. A small-scale, cassavastarch factory should have adedicated manager with sufficientexpertise to administer the factoryduring production time. Themanager should be assigned a salary

and, even though some starchprocessors feel that this representsan extra cost, it would guaranteemore efficient operation.

Of the 99 small-scale, starchfactories operating during the study,three had administrative costs thatexceeded labor costs. In most cases,labor costs were notably greater thanadministrative costs, indicating thata balance does not exist betweenthese two that would ensure adistribution of the economic benefitsof small-scale, starch factories.

The cost of buying andtransporting raw material accountfor 71% of operating costs. Severalsmall-scale, starch factories arelocated where it is easy to purchaselarge volumes of low-quality cassava,especially in flat areas, thus reducingthe average purchase price per kg ofraw material. In 1991, averagefreight charges were Col$1,958,000(US$3,285), accounting for 8% oftotal costs and exceedingadministrative and labor costs.

Labor accounts for 7% of totalcosts. Labor is inexpensive, andsmall-scale, cassava starch factoriesprovide a significant source of

37


employment. In several ruralcommunities of the region, thesefactories are the rural population’sonly source of income. Labor is alsoseasonal, being scarce during coffeeharvest. Most factories’ busiest timeis during cassava harvest.

Income Generated by SourStarch Sales

Processors sell everything theyproduce, with sour starch bringingthe highest yearly income ofCol$22,942,000 (US$38,493) or 89%of the total income. Byproductsbrought Col$2,909,000 (US$4,881)or 11% of the total income. Netprofit per factory was estimated atCol$2,337,000 (US$3,921), and thenet profit per ton of sour starch wasCol$24,300 (US$41).

Cost-to-Benefit Ratio

Profitability of cassava starchprocessing was compared with theinterest that the local agriculturalbank (Caja Agraria) pays to savingsaccounts (21% per year in 1991) as ameasure of opportunity cost. Thereturn on starch processing was only12%, although the opportunityinterest was 21%. That is, theprocessor lost 9%.

Profits generated by this type ofsmall enterprise are thereforeoperative in nature, not financial.Most small-scale, starch processorsearn only enough to satisfy theirbasic needs. Without an economicsurplus to reinvest in their business,processors cannot readily modernizethe infrastructure. Processorscontinue to participate in the marketbecause their basic necessities andfixed costs are covered and they cancontinue to sustain themselves inthe market despite the lack of profitsfor reinvestment.

Processing Constraints

Major constraints found in cassavastarch processing are:

Irregular cassava supply. Amajor constraint is irregular cassavasupply (Table 2), which is caused byinconstant cassava production, whichitself is related to unstable cassavaprices. As cassava prices rise, farmersintensify cultivation, thus increasingsupply and lowering prices.Processors do not control the flow ofraw material required to initiate theprocess; if they did, they could planproduction according to the marketand the output of each plant.

Working capital. The lack oftimely credit limits sour starchproduction and its subsequentcommercialization. Of the processors,61% had plans to obtain credit with abank. This credit was to pay suppliersfor raw material and to improve theinfrastructure, not only for plants thatprocess both coffee and cassava, butalso housing for the processors.

Often this credit is used forpurposes other than those indicated inthe initial request. The factory is soonleft without working capital and has toresort to informal lines of credit suchas suppliers giving extra days to pay.Middlemen may also lend money tothe processors, with the compromisethat, once the starch is processed, it

Table 2. Constraints to cassava starchprocessing, northern Cauca, Colombia.

Constraints Responses byprocessors

(no.) (%)

Irregular supply 27 57Supply vs. demand for starch 6 12Working capital 6 12Lack of water (climate) 6 12Tank capacity 3 6

Total 48 100

38


will be sold to them at favorableprices. Many loans granted bymiddlemen are used to make downpayments to cassava farmers toensure root supplies.

Tank capacity. Tank capacitiesare often very limited: on the average,a factory will have five fermentationtanks with an average capacity of 1 teach, and five sedimentation tankswith an average capacity of 551 kgeach.

Stock of spare parts.Small-scale processors do not keep astock of spare parts needed tomaintain their equipment, oftencausing holdups in starch production.

Commercial Constraints

Factories are affected by differentcombinations of several majorcommercial constraints (Table 3);these are:

Transport. Remote rural areascharacteristically have deficienttransport facilities, which delaydeliveries. Starch processors are thusoften obliged to rely on middlemen,which may go against their owninterests.

Plant site. Starch processorslocate their processing plantsaccording to where land is available,rather than where consumers aresituated. Control over the product istherefore lost and the distancebetween the two ends of the system(supply and demand) grows and sodoes the chain of middlemenparticipating in the commerce.

Starch quality. Processors havefew standard ideas on starch quality,making it difficult to determine criteriafor product quality. For 97% of thesurveyed processors, fermentation isimportant; this process should takefrom 15 to 20 days. For 70% ofprocessors, cassava variety is also amajor criterion. But processors tendto select varieties with high starchyields rather than for quality, partlybecause working capital is insufficientfor purchasing the more expensive,high-quality starch varieties (Table 4).

Water. During summer months,water is scarce and, in winter,processors have difficulty in dryingand transporting the starch. For 78%of surveyed processors, water qualityis an important criterion: it should becold. The water used by 60% of thesurveyed processors comes fromstreams and is untreated before use,

Table 4. Processors’ criteria for quality incassava starch, northern Cauca,Colombia.

Criterion Factories usingcriteriona

(no.) (%)

Color 33 33.3Fermentation time (acidity) 96 98.0Starch grain 54 54.5Cassava variety 69 69.7Age of cassava 30 30.3Water quality 78 78.8Climate 9 9.1Others 30 30.3

a. Total number of starch factories surveyed(weighted data) = 99.

Table 3. Constraints to cassava starchcommerce in northern Cauca,Colombia.a

Constraints Factoriesaffected

(no.) (%)

Transport 15 15.2Location 9 9.1Availability of raw material 36 36.4Availability of credit 24 24.2Starch quality 33 33.3Climate 33 33.3School vacations 21 21.2Others 39 39.4


39


thus contributing to low productquality.

Processors’ knowledge. Thelimited technical knowledge thatcassava farmers have of starchquality, and its processing andcommerce, also negatively affects thisagroindustry.

Social Characteristics Relatedto Cassava Starch Production

and Commerce

The following social issues areinvolved in the cassava starchagroindustry:

Improved living standards forrural, small-scale, starch producersand of the region as a whole.Table 5 shows that the starchagroindustry benefits both the peopledirectly involved in the industry andthe entire northern Cauca region.This small-scale enterprise increasesthe number of jobs (according to 76%of the processors surveyed) and betteruses available resources in the region,thus considerably energizing theeconomy of the Valle del CaucaDepartment. The region is becoming acenter of development for the entireDepartment, favored by its proximity

Table 5. How the cassava starch agroindustry contributes to the economic well-being of the individualfamily and of the region, northern Cauca, Colombia. Responses from a survey of 99 households.

Socioeconomic criterion Family Region

(no.) (%)a (no.) (%)a

Overall improvement 99 100 99 100Increased education 51 52Improved housing 69 70Improved living standards 48 49Vehicle ownership 27 27Improved roads 3 3Increased income 66 67Jobs 75 76Others 21 21 24 24

a. Percentages are rounded off.

to Cali, capital of the Department.Cali provides resources needed bysmall-scale starch factories,particularly spare parts for equipmentand financial resources.

Industrial security. Adequateindustrial security, to reduce risks foremployees during processing, does notyet exist within the organizationalstructure of small-scale starchfactories. Processors usually do notappreciate the risks and diseases thatcan occur during starch processingand rarely take minimum protectivemeasures.

Colds comprise the commonestailment (according to 39% of surveyedprocessors), a result of personnel notwearing dust masks during drying andpacking (Table 6). The personnel incharge of sieving should be fitted withgloves and goggles; 27% have sufferedeither cuts or eye ailments. Dryingsites located in high places, such asthe “eldas” (sliding overhead screens),should be constructed with protectivebanisters to prevent fractures andblows.

A related problem that affectsproduction continuity is frequent“Monday absenteeism” as a result ofhangovers after heavy drinking.

40


For example, CETEC is conductingstudies on treating waste waters.

Conclusions

Some conclusions from the study are:

(1) Cassava starch production is ofmajor importance in northernCauca, with 90% of cassava rootsproduced destined for starchproduction.

(2) Major constraints are, for starchproduction, irregular supply ofcassava, lack of timely credit, andmaintenance of equipment; forstarch quality, quality of waterused, fermentation time, andvariety and age of cassava; forcommerce, starch quality,climate, and transport.

(3) Small-scale processors cannot fixstarch prices, which thereforeobey the laws of supply anddemand. Cassava farmers needassistance in ensuring a constantsupply of roots for processors,which would help control pricefluctuations.

(4) Cassava starch production offerssocioeconomic benefits such asemployment. In 1990, 422 peopleand, in 1991, 345 people wereemployed.

(5) Over the long term, this study isexpected to benefit about3,000 households that subsist onthis agroindustry. Once theyunderstand and efficientlymanage the production andcommerce of cassava starch,these families will have betteropportunities of participating inthe market and improving theirsocial well-being.

Recommendations

The following list of recommendationsaim to help guide experts interveningin technical, economic, and scientific

Table 6. Incidence of diseases and accidents insmall-scale, starch factories, northernCauca, Colombia.a

Complaint Factory reporting

(no.) (%)

Ailmentb

Cold 39 39.4Backache 3 3.0Eye problems 3 3.0Sinusitis 3 3.0Nonec 27 27.3

AccidentsFractures 9 9.0Cuts 27 27.3Blows 3 3.0Nonec 60 60.6


b. A problem that causes absenteeism andindustrial accidents is the hangover. Twenty-four(i.e., 24%) factories reported on this problem.

c. That is, the factory either did not know, or did notanswer.

Environmental contamination.About 85% of residues producedduring starch extraction aredeposited in the streams (40%),rivers (27%), and ravines (18%) nearthe factories. Another 12% is used asmanure, and 3% enters the seweragesystem. As a result, the agroindustrynoticeably contaminates the region’srivers and affects its inhabitants’health. Even the processorsthemselves use this same water forwashing, drinking, and cooking, aswell as root processing. Thecontaminated water also affectsstarch quality and thus theprocessors’ income.

Given their usually loweducational level, processors do notappreciate the importance of caringfor rivers or for the adequate disposalof residues. To reduce environmentalcontamination, the departmentalgovernment and different institutionsinterested in regional economic andsocial development need to intervene.

41


decisions on behalf of cassava starchprocessors.

(1) Small-scale cassava starchprocessors working in rural areasshould be encouraged to planstaggered crops by taking intoaccount the vegetative period ofthe varieties they select. The cropshould satisfy, at least partly, thefactory’s requirements for rawmaterial so that it may reachequilibrium point or higher. Theremaining amount can beobtained from third parties withinthe factory’s area of influence byproviding incentives to cassavafarmers.

(2) Differential prices for cassavaroots should be fixed, dependingon quality and yield. This policywill allow processing plants tooperate more economically.

(3) Additional technical, financial,and administrative support,adapted to the processors’socioeconomic level, is needed.The processors can then benefitfrom real improvements in theirenterprise’s infrastructure andorganization.

(4) Operational schemes thatmaintain labor and administrativecosts at acceptable levels shouldbe incorporated. The small-scale,starch factory can then achieveequilibrium and will operateacceptably and economically.

(5) Measures should be taken toimprove factory infrastructure,thus improving cassava starchproduction while better conservingthe waterways. Examples of such

measures are draining definedareas and conserving riversides toprevent erosion.

(6) Activities aimed at improving thepopulation’s standards of living arealso needed in such areas ashealth, education, housing, andpublic services.

(7) The local government andcommunities should beencouraged to provide potablewater for human consumption andfor use in small-scale, starchfactories.

(8) Farmer associations should beencouraged to stimulate theirmembers to negotiate more andparticipate in setting cassavastarch prices. Farmers would thenhave increased financial, operative,and administrative capacity; beable to handle their own tradingneeds; and better understandmarket behavior.

References

Chuzel, G. 1991. Cassava starch: currentand potential use in Latin America.Cassava Newsl. 15(1):9-11.

CIMMYT (Centro Internacional deMejoramiento de Maíz y Trigo),Programa de Economía. 1993. Laformulación de recomendaciones apartir de datos agronómicos. Lisboa,Mexico.

Pardo Camacho, F. 1991. Diseño estadísticode muestreos. Universidad de losAndes, Santafé de Bogotá, Colombia.

de Servín, A. and Servín Andrade, L. A. 1978.Introducción al muestreo. EditorialLumusa, Mexico City, Mexico.

42


CHAPTER 7

CASSAVA STARCH AND FLOUR INECUADOR:ITS COMMERCIALIZATION AND USE

Carlos Egüez*

because either factories were poorlylocated in relation to productionzones, or national or imported rawmaterials were cheaper. In contrast,small-scale cassava starch extractiondates back to early this century, whileflour processing began 8 years ago.

About 200 family-run processingunits or “rallanderías” currentlyproduce between 2,500 and 4,000 t ofcassava starch per year. Thetechnology of drying cassava chips toproduce flour was introduced fromColombia to Manabí Province,Ecuador, in 1985, and has beenadopted mainly by the Unión deAsociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY), which produces1,000 to 1,500 t of flour per year.

The Ecuadorean IntegratedCassava Program, consisting ofUATAPPY and several national andinternational institutions, hasproduced 10 different cassavaproducts with a wide marketing range,including exports to Colombia over2 consecutive years.

The commercialization ofUATAPPY’s products has allowed it tocontinue its activities. But marketexpansion and consolidation remainsdifficult as Ecuadorean industriescontinue to use other starchy rawmaterials that are sometimes

Abstract

In Portoviejo, Ecuador, the Unión deAsociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY) produces cassavastarch and flour for a wide variety ofproducts, including animal feed,corrugated cardboard, plywood,cassava bread (pandeyuca), bakedproducts, and ice-cream cones. Theamounts of cassava starch or flourincorporated vary according tointended use. The most common usesare filling in plywood, carbohydratesource in balanced animal feeds, andas binder in cardboard boxes andshrimp feeds. Ecuadorean industriesare beginning to appreciate thepotential advantages of theseproducts. Recent studies estimatethat the potential demand greatlyexceeds the current supply, whichaugurs well for cassava rootprocessors.

Introduction

Attempts to produce cassava flour andstarch at the industrial level inEcuador have been unsuccessful,

* Cassava Program, Fundación para elDesarrollo Agropecuario (FUNDAGRO),Portoviejo, Ecuador.

43

Cassava Starch and Flour in Ecuador:...

Table 1. Comparison of current prices of cassava byproducts with wheat flour and maize starch inEcuador (factory prices).

Cassava product Current price Other product Current price(US$/t) (US$/t)

Cassava meal 175

Sieved whole flour 231

White cassava flour 236 Wheat flour 352

Sieved white flour(for human consumption) 275

Starch (for human consumption) 660

Industrial starch (first grade) 440 Domestic maize starch 400

Standard industrial starch 363 Colombian maize starch 305(second grade) (placed in Ecuadorean

factories)

Starch bagasse 113

Bran of sieved flours 88

SOURCE: UATAPPY, 1993, personal communication.

1. The cassava products and their uses asdescribed here are not registered; they reflectthe author’s research at processing andindustrial levels in Ecuador.

Table 2. Demand for wheat flour and maize starch by several markets, and current sales of cassavaproducts, Ecuador.

Product Market Annual demand (t) Current sales

Wheat flour Balanced shrimp feed 25,000 0Lumber industries 2,400 256

Maize starch Cardboard factories 6,000 0Colombia ? 200

SOURCE: Susan Poats, 1993, personal communication.

subsidized, such as wheat andmaize starch (Table 1).

Persuading industrial managersto use cassava products—sometimes the same product—instead of traditional materials for awide variety of individual uses hasalso been difficult (Table 2).

The situation is improving,however, with the current “freemarket” conditions, which allowcassava products to be morecompetitive in terms of quality andprice.

Descriptions and Uses ofCassava Products and

Byproducts1

Cassava meal is a coarse, brownpowder obtained from unpeeled chips,sun-dried on concrete, and ground ina hammer mill. It is used as acarbohydrate source in balancedfeeds, and as a pellet binder in shrimpfeed, replacing wheat flour andsynthetic binders and forming2%-12% of the formula, depending onthe manufacturer.

44


Sieved whole flour is a very fineoff-white powder obtained by sievingmeal through a no. 60 mesh. It isused in plywood, replacing 35%-40%of wheat flour (17% of the formula).

White industrial-grade flour is acoarse powder made from peeledchips, sun-dried on concrete, andhammer-milled. It is used as pelletbinder in shrimp feed.

White table flour is a very finewhite powder obtained from peeledroots that have been treated, chipped,tray-dried, and sieved through ano. 60 mesh screen under hygienicconditions. It is suitable for humanconsumption, used to partially replacewheat flour in cones for ice cream(25%-30%) and in noodles (10%).

First-grade industrial starch is avery fine, very white powder obtainedfrom peeled cassava roots that havebeen rasped, washed, sedimented, andsun-dried on concrete. It is used aspellet binder in balanced shrimpfeeds, and in cardboard boxes,replacing maize starch by as much as100%.

Second-grade industrial starch is oflower quality than first-grade starch,because the protein fraction remains.It is used in balanced shrimp feedsand cardboard boxes, replacing maizestarch by as much as 100%.

Starch for human consumption is avery fine white powder obtained frompeeled roots that have been rasped,washed, sedimented, and dried onpaper under hygienic conditions. It isused in bread, milk products, bakeryproducts, and sausages.

Ground bagasse is a coarse,white-yellow powder that is abyproduct of starch extraction. It isused as a carbohydrate source inbalanced feeds, and in shrimp feed,combined with meal and starches.

White bran is a coarse whitepowder that is a byproduct ofprocessing for white table flour. It isused as a fiber source in feeds forlivestock and pigs.

Whole bran is a coarse, brownpowder that is a byproduct ofprocessing for sieved whole flour. It isused as a fiber source in feeds forlivestock and pigs.

UATAPPY’s Production andMarkets

Since the program was established,UATAPPY has marketed more than8,000 t of cassava products fordifferent uses (Tables 3 and 4).The 50 t of cassava meal producedduring the first year were sold topoultry-feed plants, replacing maizegrain. Since then, both markets andproducts have become morediversified. Ten products are nowmarketed, for three to five differentpurposes, depending on annualnegotiations (Tables 3 and 4).

Between 1986 and 1989,cassava meal was almost the onlyproduct, finding a ready use as ashrimp feed binder. Between 1989and 1990, cassava meal for shrimpfeed was still being produced, butimportant industries began todemand cassava flour without peel.Since then, this market has beenthe most important, accounting for87% of the total volume produced.

In 1990-1991, UATAPPY’s totalproduction volume increased by 70%over that of the previous year. Butthe percentage of UATAPPY’s totalproduce destined for the shrimp feedmarket fell from 87% to 71% as twonew markets opened up: sievedwhole flour for the plywood industry,and starch for the cardboardindustry.

45

Ca

ssava

Sta

rch a

nd

Flou

r in E

cua

dor:...

Table 3. Total amount of cassava processed (ta) by UATAPPY, Manabí, Ecuador, 1985-1993.

Yearb Assns. Fresh Flours Starch Total annual(no.) roots production of

Indust. Indust. Indust. Sieved white Bran Indust. Human Bagasse Indust. Purchase Purchase cassavameal white sieved human 1st and consumpt. purchase human bagasse products and

meal consumpt. 2nd consumpt. byproducts

1985-86 2 - 50 - - - - - - - - - - 50

1986-87 4 19 96 - - - - - - - - - - 115

1987-88 10 28 500 - - - - 11 4 - - - - 543

1988-89 16 - 1,100 - - - - - 5 - - - - 1,106

1989-90 16 - 304 574 - 33 - 70 10 24 - - - 1,015

1990-91 17 - 258 982 200 6 52 119 2 51 69 4 - 1,743

1991-92 17 - 464 304 170 - 17 20 4 12 37 5 - 1,033

1992-93 17 - 127 631 292 33 80 101 17 60 155 - 26 1,522

1993-94c 17 - 300 512 - 21 - 89 25 56 - - - 1,003

Total 17 47 3,199 3,003 662 93 149 410 67 203 261 9 26 8,129

a. Values rounded to metric tons.b. Production estimated for the crop year 1 July-30 June.c. Preliminary data, subject to confirmation.

SOURCE: Susan Poats, 1993, UATAPPY Socioeconomic Monitoring Survey.

46


Table 4. UATAPPY markets and their share of cassava products, Ecuador.

Year Product Volume Market Share(t) (%)

1985-86 Meal 50 Poultry feed 100Total 50

1986-87 Meal 96 Shrimp feed 83Treated roots 19 Export human consumption 17

Total 115 Shrimp feed 2

1987-88 Meal 500 Shrimp feed 92Treated roots 28 Export human consumption 5Industrial starch 11 2Starch human consumption 4 Bread making 1

Total 543

1988-89 Meal 1,100 Shrimp feed 99.5Starch human consumption 5 Bread making 0.5

Total 1,105

1989-90 White flour 574 Shrimp feed 57Meal 304 Shrimp feed 30Industrial starch 70 Shrimp feed 7White flourSieved flour human 33 3consumption 10 Bread making 1Starch human consumption 24 Bovine feed 2Bagasse 1,015

Total

1990-91 White flour 982 Shrimp feed 56Meal 258 Shrimp feed 15Sieved whole flour 200 Plywood 11Industrial starch 188 Cardboard industry 11Bran 52 Bovine feed 3Bagasse 51 Shrimp feed 3Starch human consumption 6 Bread making 0.5Sieved white flour human 6 0.5consumption 1,743

Total

1991-92 Meal 464 Shrimp feed 45White flour 304 Shrimp feed 29Industrial starch 57 Shrimp feed 5Sieved whole flour 170 Plywood 17Bran 17 Bovine feed 2Bagasse 12 Bovine feed 1Starch human consumption 9 Bread making 1

Total 1,033

1992-93 White flour 631 Export to Colombia 42Industrial starch 292 Export to Colombia 19Sieved whole flour 256 Plywood 17Meal 127 8Bagasse 86 6Bran 80 Bovine feed 5Sieved white flour human 33 Ice-cream cones 2consumption 17 Sausages 1Starch human consumption 1,522

Total

SOURCE: UATAPPY, 1993, personal communication.

47

Cassava Starch and Flour in Ecuador:...

In 1992-1993, the Colombianmarket was the main client: 600 t ofwhite flour and 200 t of second-gradestarch were exported. Although theiruse has not yet been confirmed, theyappear to have been used to makeadhesives.

Constraints toCommercializing Cassava

Products

The major constraints are:

(1) Poor product quality, resultingfrom contamination at one or moreof the processing stages (mostimportant in relation to the moreprofitable, but more demanding,markets).

(2) Seasonality of supply (UATAPPYcan only produce during the8 “summer” months as thecassava is sun-dried).

(3) Competition from other rawmaterials, especially maize starchthat enters Ecuador fromColombia at low prices (Table 1).

(4) Lack of knowledge: industries donot yet know how to substitutewheat flour or maize starch withcassava products.

Conclusions

(1) The current supply of cassavaproducts is small in relation tothe potential demand.

(2) Cassava products compete well interms of quality and price withother raw materials, except formaize starch, which is cheaperimported from Colombia.

(3) “Free market” and “open border”conditions favor thecommercialization of cassavaproducts.

Bibliography

Brouwer, R. 1992. The cassava flour demandin the plywood industry in Ecuador.Wageningen, the Netherlands.

CENDES (Centro de Desarrollo). 1993.Estudio de mercado para conocer lademanda potencial de productoselaborados de yuca. Unión deAsociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY) and CENDES,Quito, Ecuador.

Egüez, C. 1992. Informe anual del Programade Yuca, 1992. Fundación para elDesarrollo Agropecuario (FUNDAGRO),Portoviejo, Ecuador.

__________. 1993. Revisiones de los archivosdel Departamento de Contabilidad dela UATAPPY, 1985-1993. Unión deAsociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY), Portoviejo,Ecuador.

48


CHAPTER 8

CASSAVA PRODUCTS FOR FOOD AND

CHEMICAL INDUSTRIES:CHINA

Jin Shu-Ren*

cassava must also increase—mosteffectively by developing the range ofits products through adopting andexpanding secondary processingtechniques.

Changes in CassavaProcessing

During the 1960s

Cassava processing in China wasmainly small-scale: production groupsof 20-30 families in rural areas wouldplant cassava in unused areas or onsloping land as insurance for foodscarcity. Because such land wasusually of low fertility and received nofertilizer, cassava yields were low: inthe 1960s, in Guangxi, China’s largestcassava-producing area, the averageyield per mu (1 mu = 665 m2 or15 mu = 1 ha), expressed as driedchips, was only 46 kg, that is, about0.7 t/ha. The area planted to cassavavaried from year to year: in 1967, inGuangxi, cassava was grown on1,054,000 mu. Because of climaticconstraints, cassava is a seasonal crop,and the small-scale processing plantsoperated only 3 to 4 months a year.The major product was poor qualitystarch.

Cassava was processed by firstcrushing fresh roots in a grinder andallowing the resulting mash to settle in

Abstract

Cassava is grown in China basically asa food security crop. But, in the last20 years, yields have increased sharplyin answer to demand from small-scaleand, more recently, large-scaleindustries. Since the 1980s, Chinahas seen rapid development in thecommercial prospects of a wide rangeof cassava derivatives, includingfructose-series products, sorbitol,maltol, fermentation products (suchas alcohol, MSG, and citric acid),denatured starch, glucose, andglucose syrup. A hillside crop, cassavaplays a key role in the economy andagroindustry of southern China.However, local economies andproduction in poorer rural areasurgently need modernizing if they areto fully benefit from these newdevelopments. Recommendations aremade regarding appropriate scale andtechnology, given the variousconstraints (e.g., transportationthrough hilly terrain and seasonalavailability of fresh roots). Relevanteconomic factors are also reviewed.Through improved cultivars andfarming practices, cassava yields canincrease significantly. But, toencourage production, the value of

* Guangxi Nanning Cassava TechnicalDevelopment Center (GNCTDC), Nanning,Guangxi, China.

49

Cassava Products for Food and Chemical Industries: China

Production capacity of some factoriesreached 10 t/day. Guangxi had morethan 270 starch factories, althoughthe total output was low—probablyless than 80,000 t, or less than300 t/factory.

During the 1980s

Cassava production improvedmarkedly, because:

(1) More land was made available inrural areas;

(2) Farmers were allowed to planthigh-value crops, leading to majorincreases in the area planted tocassava;

(3) As the production of other cropsimproved, cassava’s role shiftedaway from being a food securitycrop to providing raw material foranimal feed and industry;

(4) As the national economy developed,the demand for starch increased;and

(5) Capital and imported equipmentwere made more readily available.

The combination of these factorscreated an unprecedented expansion inthe scale and technology of cassavaproduction.

The last five years

The two main cassava-producing areasin southern China now have severallarge-scale starch factories. By 1992,at least 10 factories had an outputcapacity of 40 t/day, the largestcapable of producing 60 to 80 t.Overall, the factories produced morethan 30,000 t. Three types of factoriesco-exist:

(1) Plants newly constructed oradapted, and using domestictechnology. Features include aroller cleaner, two-stage crusher,countercurrent-washer, rapidblancher, whirlpool sand remover,dish-separator, dewatering

water. The starchy mass was thenstrained through a fine-mesh clothbag to separate the starch, which wasthen sun-dried and pulverized.

Some high-quality starch was alsoproduced, although outputs were low.Fresh roots were crushed in a grinder;passed through a second, finer,grinder, and then through a vibrating,or octagonal, sieve that removedcoarse residues; and, finally, passed toan open-ended, horizontal-flow,sedimentation trough that was30-50 m long, 35 cm deep, and 40 cmwide. High-density impurities, suchas sand and gravel, were depositedfirst, starch farther toward the middleof the trough, and low-densityimpurities, such as fiber and protein,at the far end, or flushed out. Thestarch was then removed from thetrough, dewatered by centrifugation,dried, and pulverized. If desired, ableaching agent, such as potassiumchlorate or hypochlorite, was addedbefore the starch entered the trough.

Starch was produced from freshcassava for only 3 to 4 months a year.During the rest of the year, driedcassava chips were used, requiring anadditional 1 to 2 days of immersion inwater (or longer in cold weather) beforefiltering. During fine grinding, anadditive would be introduced toimprove the starch extraction rate.

Most of the cassava not used forstarch extraction was used as pig feed.The roots were first peeled, soaked inwater to remove hydrocyanic acid, andthen boiled.

During the 1970s

The technology of producingquality cassava starch improved. Thewooden lining of the trough wasreplaced by marble or glass, thecentrifuge assumed a horizontal ratherthan vertical structure, and a cranksieve replaced the octagonal sieve.

50


centrifuge (sometimes imported),and two-stage forced-air drier.Product quality and cost efficiencyare adequate. Such factoriesaccount for 25% of cassava starchproduced in China.

(2) Plants where equipment andtechnology are entirely imported,for example, from Japan, Germany,and Thailand. Features include aneedle grinder, high-pressurecrank sieve, whirlflow separator,centrifugal layer-separator, waterremover, and airflow drier.Although they producehigh-quality starch, such plantsare economically less viablebecause of high equipment costsand associated steep depreciation.In addition, they compareunfavorably in performance whendried cassava chips are used inseasons when fresh roots areunavailable.

(3) Small-scale, low-technologyfactories that, technologically andeconomically, compare poorlywith (1) and (2). The averagestarch-extraction rate is estimatedas being 20% lower. Becausethese factories currently accountfor about half of southern China’sannual starch production, theirtechnology urgently needsmodernizing.

With fresh cassava available foronly 3 to 4 months a year, the use ofdried chips has been inevitable, eventhough costs are higher, the starch ofpoorer quality, and the recovery ratelower. To counter these problems, atechnology has been recently adoptedthat would produce “high starch, highextraction, and high storage.” Itinvolves bulk-buying fresh cassavawhen starch content is at its highest,and crudely processing the roots into a“paste pool.” The starch can thereforebe extracted in due course, extendingthe annual period of cost-effective,optimal quality starch production from3 to 5 months.

By 1992, cassava starchproduction in southern Chinaaccounted for 23% of the nationalproduction. Cassava yields hadincreased notably, the regional totalexceeding 1,200,000 t of dried chips.Yield per mu increased to 500 kg forfresh cassava and 200 kg for driedchips.

Secondary processing

Since the 1980s, the Government hasshown more interest in developingcassava products derived fromsecondary processing. Cassavadevelopment and utilization are listedamong the key projects of the sixth5-year plan drawn up by the StateScience and Technology Commission.Several national centers are alsoinvolved in the development andutilization of cassava, including theGuangxi Nanning Cassava TechnicalDevelopment Center (GNCTDC).

Developing Cassava Productsfor Food and Chemical

Industries

Industries began using cassava-basedproducts, developed from secondaryprocessing, during the 1980s. Theseinclude:

Fructose-series products

Fructose emerged in the 1970s as ahealthier alternative to sucrose.Technology using starch as a rawmaterial was developed soon after, and1980 saw the first factory, with anannual fructose output capacity of10,000 t, set up in central China. Thetechnology of “third-generation”fructose (i.e., fructose containing notmore than 10% glucose) has sincebeen mastered in China. Through acollaborative project, the GNCTDCfinished testing a pilot plant in 1986,and, in 1992, set up the firstindustrial plant to produce crystalline

51


fructose. Third-generation fructose isused medically, in cases of glucosecontraindication. Clinical tests on100 diabetic patients given 25-50 g ofhigh or crystalline fructose showed nosignificant changes in blood sugarlevel. Thus, its safety, sweetness,pleasant taste, and few calories makefructose particularly suitable fordiabetics.

Chemico-industrial products

Sorbitol. A hexan-hexol, sorbitolis made from glucose byhydrogenization in a high-pressurereactor. Because it readily absorbsmoisture, it can replace glycerine inthe manufacture of toothpaste,cosmetics, and oil-based paints. Itserves as raw material in themanufacture of vitamin C byfermentation, first into hygric acid andthen into ascorbic acid. Every ton ofascorbic acid produced requires2.7 t of sorbitol. More than 10 sorbitolfactories operate in China, the largestof which has an annual productioncapacity of 13,000 t. A unit capable of30,000 t is being planned, while somehave recently begun using continuous-hydrogenization technology.

Production of solid sorbitol(3,000 t/year) has been successfullyestablished in Nanning, Guangxi.Glucose produced from cassava chipsis hydrogenized under high pressure(continuous process) to produce liquidsorbitol. The liquid is concentrated to98 Brix, seed crystals are introduced,and the sorbitol spray-dried andcrystallized. Solid sorbitol is easier totransport and store.

Mannitol. Another hexan-hexol,but with little moisture-absorptioncapacity, mannitol is usually abyproduct of iodine extraction fromkelp. But it can also be producedcommercially by hydrogenizingfructose, of which 50% converts into

mannitol, which is then purified bycrystallization. Mannitol is usedmedically in blood-vessel diastolicpreparations, as a dehydrating agent,and in the treatment of cerebralthrombosis and other circulatorydisorders. In industry, it can be usedas raw material for the production ofpolyester, polyethylene, and solid-foamplastics.

Maltol. A sugar alcohol, maltol isproduced by incomplete hydrolysis ofstarch, using the enzyme maltase, andsubsequent hydrogenization. It is asyrup that is as sweet as sucrose, andis used in confectionery.

Fermented products

Fermented cassava products form asizeable industry in China, and includealcohol, monosodium glutamate (MSG),and citric acid. Cassava wine wasproduced in the 1960s when grain wasscarce, but has now become obsoletebecause of poor quality.

Alcohol. After 2 days offermenting, the alcohol content incassava can reach 10%-11%. Mostfactories were established in the 1970sand have an annual output capacity of10,000 t. New factories with a30 to 50-thousand-ton capacity arenow being planned. Sugarcane andcassava growing areas usually coincideand cassava alcohol is almost alwaysproduced by sugar mills, which usemolasses during the sugarcane season(November-April) and, using the sameequipment, cassava roots for the restof the year. Because cassava is low inprotein and nutrients needed forgrowing yeast (the fermentative agent),it must be supplemented. A mixture ofcassava and molasses is often used togood effect.

MSG. Also known as gourmetpowder, MSG is a popular flavorenhancer in Chinese cuisine. Nationalproduction exceeds 200,000 t/year. Of

52


these, about 25,000 t are obtainedfrom cassava starch, which firstundergoes acid hydrolysis, and is thensupplemented with growth factors andleft to ferment for 4 or 5 days. Duringthis time, ammonia salt is addedcontinuously. When the glutamic acidcontent reaches 7%-8%, the mixture isfiltered and the acid precipitated byiso-electric points. The acid is thenpurified by ion exchange, neutralizedto produce the sodium salt, andcrystallized.

Citric acid. In China, citric acidis mainly produced by fermentingsweetpotato. In 1990, more than80,000 t were produced. Recently,however, citric acid is increasinglybeing produced from molasses andcassava, using an Aspergillus strain,known as Citrobacter, which wasdeveloped by the Shanghai IndustrialMicrobiology Research Institute.Cassava starch liquefies easily to alow-density liquid and, after a 4-dayfermentation, the citric acid contentexceeds 15%. An extraction rate ofmore than 92% is possible. The shortfermentation period, and ease ofliquefying the starch and extractingthe acid keep production costs low.

Denatured starch

Since the 1980s, research ondenatured starch has developedrapidly, allowing some processes tobecome industrialized. The currentannual yield of denatured cassavastarch is about 7,000 t, and includesacid-denatured starch, α-starch,ethylic starch, phosphate ester starch,and co-polymerized starch. Althoughcurrent outputs are low, the futureprospects of this industry arepromising.

Glucose and glucose syrup

Crystalline glucose in southern Chinais produced primarily from cassavastarch, as are injection glucose (used

in medicine) and glucose syrup(DE42) (confectionery). More than100,000 t are produced annually.

Market for CassavaProducts

While the Government does notrestrict sales within China, it controlsexports. Fresh cassava or driedcassava chips are sold to domesticmarkets by farmers or by localsupply-and-marketing cooperatives.The higher value chips are cut 0.5 to1.0 cm thick, peeled, and sun-dried.The price of fresh cassava sold tofactories varies according to season,starch content, and transportationdistances. More recently, prices havebeen affected significantly by grainprices. Cassava starch costs10%-15% less in winter, theproduction season, than at othertimes, reflecting the fact that mostfactories are small-scale and lackcapital.

In total, about 500,000 t ofcassava (based on dried chips) areused in starch production, 80% fromfresh roots and 20% from dried chips.The glucose industry uses the largestamount of starch (55%), followed byMSG production (20%), familyconsumption (4%), and sales tonorthern China or abroad (11%).

Only about 15% of the cassavagrown is used for alcohol and otherproducts. Alcohol producers incassava-producing areas have accessto, and prefer, molasses fromsugarcane. In northern Chinatransportation difficulties constrainalcohol producers from buyingcassava.

Citric acid production accounts forabout 3% of cassava grown.

About 600,000 t of dried cassavachips are exported annually, but much

53


is used locally as animal feed, both intraditional form and, more recently, incompound feeds.

Although yields increased sharplyin the 1980s, processing remainedbackward and markets were few. Morerecently, however, the processingindustry has been modernizing andsupply and demand have increased intandem.

Opportunity andCompetition

The development of cassava productionand processing in China, alreadyinhibited by strong domestic grainproduction, is further restricted by thenatural coupling of major cassava andsugarcane producing areas. Thus,market prices of sugarcane largelydictate the extent of cassava farming.Furthermore, where sugarcane yieldsare high, markets stable, and farmersexperienced producers of sugarcane,cassava is unlikely to be planted inpreference. However, where land isless fertile and cane yields low,cassava’s potentially higher productionis more attractive.

Although its starch is used as anadditive in cooking, cassava is rarelyused as a food in China. Most cassavais destined for the textile, papermaking, and chemical industries,where it must face competition frommaize products.

Cassava’s future prospects aregood, even though production yieldsare still low: about 500 kg of freshroots per mu (or 200 kg per mu of driedchips). Although most farmers still usean old variety, Nanyang Red, freshcassava yields may eventually reach 2or 3 t per mu with the adaptation ofimported improved varieties. That is,improved technology would increasethe current average yield per unit areaby an estimated 500%.

The rapid development of industryin China provides an ideal opportunityfor cassava. Yields of starch increasedfive-fold between 1981 and 1989 as theannual growth rate exceeded 15%.Those industries using starch as a rawmaterial, such as MSG, maltol,glucose, and fructose, anticipate rapidexpansion while the domestic marketremains unsaturated. Becausemaize-growing areas are far fromcassava-producing areas, cassava hasthe advantage of lower transportationcosts. While the price differential ismaintained, cassava has theadvantage in southern China.

The market for cassava products ispotentially rich, particularly fordenatured starch, which is still new.Although the use of cassava starch infoodstuffs is currently limited, itsfuture potential is high.

Conclusions andSuggestions

Cassava, a hillside crop, plays a keyrole in the economy and agroindustryof southern China. Cassava productsare important both in their own rightand as industrial raw material. Theirfuture development is integral to thatof the economy of rural and poorerareas. Two objectives should bepursued in parallel: to increase yields,and to raise the value of cassava. Forthe first, emphasis must be given toimporting and adapting improvedvarieties, developing farmingtechnology, and improving fertilizers.The second objective requirescommitment to developing cassavaproducts such as those discussedabove. If cassava is processed onlyat the primary level and commandsonly basic prices, then its prospectsare severely limited. Governmentalpolicy in this area must thereforepromote research into secondaryprocessing techniques to accompanythe improvement of cassava cultivars.

54


The modern cassava starchfactory, in scale and in the technologyemployed, should correspond closelyto local needs and to take account ofsuch factors as transportationdifficulties, and seasonal variations inproduction and availability of freshroots. The small, largely undevelopedfactories of southern China urgentlyneed modernization. At present,Chinese-manufactured machinery isadequate for such factories, whereasinvestment in expensive machineryfrom developed countries can increaseproduction costs out of proportion tobenefits in output and quality.Secondary processing of cassava does

not need a high degree of automationnor large factories (which arepenalized by higher transport costs),and is readily adaptable to local ruralconditions.

Most cassava industrial activity inChina has so far been related tohigh-value products, such as MSGand sorbitol, and, although progress isevident in some areas, effort isrequired in others, especially in thedevelopment of denatured starch.Given realistic processing andtechnological transformations, cassavahas much to offer the farmers andlocal economies of southern China.

55

Thai Cassava Starch Industry:...

CHAPTER 9

THAI CASSAVA STARCH INDUSTRY:ITS CURRENT STATUS AND POTENTIAL

FUTURE1

Boonjit Titapiwatanakun*

A Brief History

Early development

Cassava first came as a subsistencecrop, probably to southern Thailand,through Malaysia, from West Java inIndonesia. Industrial cassavaprocessing began in the 1920s inChonburi Province, on the EasternSeaboard. The first plants used asimple sedimentation process toextract starch, which was destinedmostly for household consumption. Atthe end of World War II, starch millingwas introduced, thus catalyzing thedevelopment of Thailand’s nowmodern cassava starch industry.Most of the starch was exported,together with certain byproducts.

In the early 1950s, starch wascassava’s most important exportproduct. In 1955, for example,54,122 t of cassava products wereexported, with a total value of69.1 million baht. Of this value,cassava starch accounted for about76% (i.e., 54% of the tonnage) andbyproducts about 22% (i.e., 44% of

tonnage). Most of the starch wasexported to USA and Japan, and thebyproducts to Malaysia andSingapore. Cassava starch exportsincreased every year from about29,000 t in 1955, peaking at227,000 t in 1961, and droppingslightly in 1973.

No published data are available ondomestic cassava starch consumption,but it was probably less than theamount exported, indicating a highlyexport-oriented industry thatdeveloped in response to exportmarkets.

Entering the animal feed market

The value of cassava starch exportsincreased to 220 and 223 million bahtin 1960 and 1965, respectively, but,as exports of cassava products foranimal feed expanded rapidly, theirpercentage share of the total value ofcassava exports decreased from 76%in 1960 to 33% in 1965.

Thailand began processingcassava for animal feed in the late1950s in response to heavy demand(triggered off by the CommonAgriculture Policy) for nongrain feedingredients (NGFI)—including cassavaproducts—from the then EuropeanEconomic Community (EEC, now theEuropean Union). In 1957, cassavabyproduct exports to the Netherlands

* Department of Agricultural and ResourceEconomics, Faculty of Economics, KasetsartUniversity, Bangkok, Thailand.


56


and the then West Germany were1,400 and 7,000 t, respectively.

At first (1955-1968), Thailandexported, as animal feed ingredients,cassava byproducts, meal, and chips;during 1969-1982, native or softpellets were exported; then from 1983to the present, hard pellets. Theseproducts evolved as efforts weremade to minimize transport costsand contamination during loadingand unloading. Such evolutionreflects the increasing efficiency ofthe industry’s marketing system,stimulated by coordination with, andtechnological transfer from,importing countries.

Exporting cassava products foranimal feed had a significantsocioeconomic impact on Thailand,for example, farm income,employment, and foreign exchangeearnings all increased. The areaplanted to cassava increaseddramatically from 38,400 ha in 1957to 171,000 ha, producing 2.6 milliontons, in 1968. Although about 75%of the expanded area first occurred inThailand’s eastern region, by 1977,the northeast was producing morethan 50% of the national production.The starch industry also benefited,setting up plants in the new areas.

The increased export trade alsoencouraged the development ofequipment that enabled high-speedloading of pellets, and permittedspecialization within the exportbusiness. Thai exporters could setup trading companies in the EEC,thus opening up investmentopportunities in the industry.

In July 1982, however, the EEC,through a Cooperation Agreementwith the Kingdom of Thailand oncassava production, marketing, andtrade, set an annual maximum of5 million tons of imports of Thaicassava products as animal feed.

The Agreement slowed down theindustry’s development, and createdthe need to explore new ways of usingthe cassava root and its products,particularly as starch.

Modified starch

Even with Bangkok as a significantmarket, Thailand was alreadyexporting native starch to USA andJapan, who processed some of it intothe more valuable modified starch. InThailand, the earliest processing ofnative starch into modified starch(glucose syrup) was in 1950, followedby monosodium glutamate (MSG) in1960. In the late 1970s, USA andThailand collaborated to producemodified starch for export, followed byjoint ventures with European andJapanese firms. At the same time,exporters of cassava for animal feedintegrated with native and modifiedstarch processing enterprises.

When modified-starch processingbegan in Thailand, it was typically aclosed industry, but with the need fornew export products, the industryopened up to the extent that evenplants in cassava-producing areas areproducing modified starch, usingsimple processing techniques. Plantsusing more complex techniques toproduce chemically modified starchare located mostly in the provincesaround Bangkok, where mostindustries that use modifiedstarch are located. Seventeenmodified-starch plants, with anestimated total capacity of300,000 t/year, were operating in1990, producing about 250,000 t.

The Thai modified-starch industrydeveloped rapidly during the lastdecade, because, first, theinternational trade in native starchwas hampered by import barriers,imposed to protect domestic nativestarch industries. In contrast, fewimport barriers operated against

57


modified starch. Second, Thailand’simpressive economic growth during1980-1990 made possible theinvestment in high-level technology forprocessing. Third, modified starchprovided an outlet for the foreseeableoverproduction of cassava, caused bythe EEC’s restricted imports ofcassava pellets.

Production, Marketing, andPrice Formation

Root and starch production

During the past decade, data from theMinistry of Agriculture andCooperatives (MOAC) showed thattotal cassava root productionincreased from 19 million tons in 1983to about 20 million tons in 1992, thatis, at an annual growth rate of only0.7%. Yield per ha decreased fromabout 18 t in 1983 to 14 t in 1992,mainly because fertilizers were notapplied in most cassava-growingareas, especially in the northeast.

The national average productioncosts per ton of cassava increasedfrom 450 baht in 1989/90 to 470 in1990/91 and 540 in 1991/92. Thenational average farmgate price perton (i.e., the price received bycassava growers) was 620 baht in1990, 830 in 1991, and 770 in1992. Between 1990-1992, theaverage farmgate price increased by24%, which compares with aproduction cost increase of 20%during the same period. Cassavafarmers made, on the average, aprofit of 253 baht/t. But ifproduction costs continue to increaseat their current annual growth rateof 7%, the competitiveness of Thaicassava products in the future worldmarket will be jeopardized.

During the last 5 years, about14-15 million tons of root wereprocessed into animal feed products

(chips and pellets), which were mostlyexported, and 5-6 million tons wereprocessed into cassava starch.

Statistics on starch production arenot available, although the ThaiTapioca Flour Industries TradeAssociation (TTFITA) estimates thatthe total cassava or native starchproduction was about 1.2 million tonsin 1989, 1.3 million in 1990, and1.4 million in 1992.

As for many agroindustries, thetotal number and capacity of cassavastarch plants are not updated byofficial sources. Official records,especially those of the Ministry ofIndustry (MOI), register data as plantsare established, but conduct fewsurveys for updating. For example,MOAC reported the total number ofstarch plants in 1970 as 50 and in1973 as 128. The MOI then reported146 plants in 1978, but, in 1987, only82, with an estimated capacity of1.5 million tons. Although, in 1990,the number dropped further to 45plants—a decrease of 45% (Table 1)—the estimated capacity dropped only13% to 1.3 million tons.

The starch industry may havesuffered from overcapacity since1978. Even in 1990, the industryoperated for only 8 months. If theindustry were to operate 10 or11 months/year, then its potentialproduction would be 1.7-1.9 milliontons of starch. The capacity of someplants could be expanded, especiallyin the eastern and northeasternregions.

During 1978-1990, the number ofstarch plants in the eastern regiondecreased dramatically from 121 to17, whereas in the northeast itincreased from 12 to 22, suggesting ashift of cassava-producing regionsfrom eastern to northeasternThailand, as the animal feed marketexpanded.

58


But plant numbers may havedropped as the many small,family-operated businesses, using oldand mixed technologies, lost out in thecompetition with modern plants,which had comparatively higherproduction efficiency. Most modernstarch plants were constructed bylocal firms, and more than 80% of thematerials and machinery used forconstruction were locally producedand assembled. Thai firms even builtstarch plants in Indonesia.

Marketing and price formation

The marketing of cassava roots issimple: through local truckers toboth chip-and-pellet and starchplants. Because the Governmentallocated quotas of exports to theEEC, based on accumulated stock in1988, the price of roots wasdetermined by that of chips andpellets during the time exporterswere accumulating their stock andbefore stock checking. Once stockchecking was over, root prices weremore or less determined by nativestarch prices. When domestic pricesof starch were high, following highworld prices, then only those rootsthat could not meet starch contentrequirements would be sold to thechip-and-pellet plants. In otherwords, roots are sold first to thestarch plants, until their demand is

satiated, and then to chip-and-pelletplants.

Two major outlets are available fornative starch: (1) domesticconsumption, and (2) export.Domestic consumption includesfoodstuffs and industrial use(discussed below), but also rawmaterial for domestic modified-starchplants located in provinces aroundBangkok. Their products, however,are mostly exported. Native starch issometimes sold as wet starch tomodified-starch plants located incassava-producing regions.

In terms of market share, starchhas been more or less equally dividedbetween export and domesticconsumption for the last 5 years. Butprices of domestic native starch arestrongly influenced by export marketsfor both native and modified starch.Prices in importing countries competewith prices of other starches,especially maize, and the eventualprices reached, in turn, influence Thaidomestic prices.

In the domestic market, nativestarch is, comparatively, thecheapest starch available and usedat costs that comprise a relativelysmall percentage of the value of thefinal products. This enablesdomestic industries that use native

Table 1. Number of cassava starch plants and productiona by region, Thailand, 1989 and 1990.

Region 1989 1990

Plants Production Plants Production(no.) (1,000 t) (no.) (1,000 t)

Northern 4 39 4 39Western 2 27 2 27Eastern 18 263 17 263Northeastern 23 936 22 1,024

Total 47 1,265 45 1,353

a. Annual production figures were estimated by multiplying the daily capacity of plants by 240 days.

SOURCE: TTFITA, various years.

59


starch to absorb price fluctuationswithout seriously affecting the priceof the final product.

The links among prices formodified starch, native starch, androots are shown in the following list of1990 average prices, and marketingand processing costs, obtained from a1991 industry survey:

Items US$/ta

Modified starch price c.i.f., Japan 405.0

Freight and insurance costs,Thailand-Japan 45.0

Modified starch price f.o.b., Bangkok 360.0

Exporting costs 20.0

Modified starch prices at plant,Bangkok 340.0

Processing costs of modified starch,including 5% weight loss 117.8

Native starch prices at Bangkok plant 222.2

Transport costs, Nakhon-Ratchasimaand Bangkok 9.0

Native starch prices at plant inNakhon-Ratchasima 213.2

Processing costs of native starch(conversion rate of starch toroot = 1:5) 52.0

Value of roots per ton of starch at plant 161.2

Value of wastes (10% of the valueof roots) 16.1

Total value of roots per ton of starchat plant 177.3

Price of roots per ton (root prices atthe plant in Nakhon-Ratchasima) 35.5

Production costs of roots in 1989/90(published by MOAC) 17.6

______________________________________________________________a. Exchange rate is 25.50 baht = US$1.00 (1994).

The above list also shows therelationships between Bangkok pricesand Nakhon-Ratchasima Provinceprices for native starch and roots.Under normal conditions, cassavastarch plants derive their daily buyingprice of cassava roots from this type ofinformation.

Current and Future DomesticUse and Export

Domestic use

Overall, domestic use of cassavastarch can be classified into eitherfood or nonfood industries. No officialrecords are available of the totalcassava starch consumption by bothgroups. In 1991, an industrial surveywas conducted by the ThailandDevelopment Research InstituteFoundation (TDRI) to compile andestimate starch consumption for thatyear, and to project future use.

Starch consumption wasestimated by calculating thepercentage of starch consumption perunit of final product, whether food ornonfood. The total starchconsumption of each final product wasthen computed by multiplying thepercentage use by the total annualproduction of each final product. Theannual total cassava-starchconsumption for producing the finalproducts was obtained during thesurvey as a discrete series. Thecomplete series was then constructedby using growth rates betweenperiods.

Once the series of starchconsumption data was constructed,the future consumption of starch ofeach product was forecast by a simpledemand projection equation:

D = R + NY (1)

where:

D = growth rate of quantitydemanded for the finalproduct

R = growth rate of population (theannual population growth rate,estimated by the TDRI as 1.3%,for 1991 to 2001 was used)

60


N = income elasticity of demand forstarch for the final product(using the 1972 per capitaincome as the base year)

Y = growth rate of income percapita, using 1972 as the baseyear (the TDRI projection of6.4% for 1991 to 2001 wasused)

The estimate of starchconsumption for each final product isdiscussed below:

Food industries

Monosodium glutamate andlysine. In 1991, three MSG plantsoperated: Ajinomoto, Raja, and ThaiChuros. In 1960, Ajinomoto set up thefirst MSG plant in Thailand and wasthe first to use modern technology tomodify cassava starch. Ajinomoto isalso the only MSG plant to use cassavastarch as the major raw material,which it does at a rate of 2.4 t of starchper ton of MSG. The other two plantsuse molasses.

In 1986, the Ajinomoto group setup the first and only lysine plant, notjust in Thailand, but also in SoutheastAsia. To produce lysine, the plantconsumes cassava starch at the samerate as for MSG.

To produce both MSG and lysine,the Ajinomoto group consumed28,000 t of cassava starch in 1980;33,000 t in 1985; and 87,000 t in1990. Growth rates were 3.3% during1980-1985, and 21.4% during1985-1990. Based on these data, astatistical series of starch consumptiondata for MSG and lysine productionduring 1980-1990 was constructed. These data were used to estimate theincome elasticity of demand for starchfor MSG production, that is, 1.75. Theestimation equation, of which theautocorrelation was corrected, was asfollows:

1n(STl) = -16.043 + 1.749 1n(GDP) (2)(-9.125) (9.034)

R2 = 0.950D.W. = 1.244

where:

STl = per capita demand for starchin producing MSG and lysine

GDP = per capita income in the 1972base year

The growth of demand for starchto produce MSG and lysine can beapproximated by equation (1). That is,if D = 12.617% (or 1.33 + 1.75 x 6.45),then starch consumption forproducing MSG and lysine in 1991 =97,977 t (87,000 x 1.12617).

Sweeteners (excluding fructose).Domestic production of glucose syrupbegan in 1950, glucose powder in1976, and sorbitol in 1980. Theconversion ratio of each product andthe estimated annual productionobtained from the survey are asfollows:

Product Ratio of Estimated annualcassava production of

starch to final productproduct (tons)

Glucose syrup 1:0.92 30,000Sorbitol 1:1.20 28,000

The glucose syrup producersestimated that sweetener productionconsumed 28,040 t of cassava starchin 1950; 42,060 t in 1980; and70,100 t in 1990. Based on thesedata, a statistical series of cassavastarch consumption data wasconstructed, and the income elasticityof demand for starch was estimated at1.16. The annual growth rate ofdemand for sweeteners was calculatedat 8.812%, which was used to projectdemand for cassava starch forproducing sweeteners (but notfructose), during 1991-2001.

61


Pearl sago. Pearl sago or tapiocawas produced by many small, and afew large, cassava starch plants. In1990, the TTFITA listed 12 pearl-sagoplants, five of which were large. Butmany small, family-operated plantsmay not have been counted.Processing involves mixing cassavastarch with water, pearling themixture, and sun-drying it. Theconversion rate of cassava starch topearl sago is 1:0.9. Starchconsumption was about 23,000 t in1986 and 30,000 t in 1990, an annualgrowth rate of 6.7%. Pearl-sagoproducers expect that the rate will bemaintained for the future, becauseboth domestic and export markets areexpanding. The 6.7% growth rate wastherefore used to project cassavastarch consumption in pearl-sagoprocessing.

Household consumption. Threekinds of starch are consumed by thehousehold: rice starch, sticky ricestarch, and cassava starch. Totalstarch consumption was reported at7.12 kg/person per year. Assumingequal proportions of starchconsumption, then per capita cassavastarch consumption would be 2.37 kg.A statistical series of household starchconsumption data was constructed for1991-2001 by assuming a constantper capita consumption at 2.37 kgand using the TDRI’s populationprojection. The constant wasapproximated from a householdsurvey conducted by the Office ofAgricultural Economics, MOAC,during 1970-1971.

Other food industries. Cassavastarch is used as a raw material oringredient by canning and other foodindustries that make, for example,instant noodles, vermicelli, sauces,soups, sausages, and candies. Theannual cassava starch consumptionwas estimated to be 17,960 t in 1980and 31,986 t in 1990. Based on thesedata, the income elasticity of demand

for starch was calculated at 0.64. Equation (1) was used to projectfuture starch consumption.

Nonfood industries

Paper industry. In 1989, theThai Pulp and Paper IndustriesAssociation (TPPIA) reported that38 paper mills were operating, 12 ofwhich received the Board ofInvestment (BOI) privilege. The totalannual capacity was 870,000 t ofpaper, proportioned as follows:521,000 t in kraft paper; 193,000 inprinting and writing paper; 110,000 inpaperboard; and 46,000 in sanitarypaper. Although Thailand importsnewsprint, by the end of 1993, threeplants with total annual capacity of300,000 t were operating.

Of these five types of paper, onlythe plants producing kraft paper,printing and writing paper, andpaperboard used cassava starch as araw material in production. Theaverage consumption rate of starchwas about 5% of the total paperweight, with paper productionexpanding at a rate of 13% per year.From these data, cassava starchconsumption in the paper industrywas estimated at about 42,000 t in1990, and projected by using a 13%annual growth rate.

Plywood industry. In 1990,35 plywood manufacturers wereoperating. One piece of plywood usesabout 370 g of cassava starch. As faras can be ascertained, the averagemetric ton of plywood contains80 pieces. Total plywood productiontends to be underreported becauselogs are imported illegally fromneighboring countries. More accurateestimates may be obtained byexamining the relatively constantplywood market share of the ThaiPlywood Company Limited, a stateenterprise, which share held at 10%during the last few years. Estimates

62


for cassava starch consumption were4,775 t in 1989, 6,924 in 1990, and6,700 in 1991.

For the next 3 years, the annualcassava starch consumption in theplywood industry may stay at about6,700 t, because, first, importing logsfrom neighboring countries willbecome difficult as these countriesestablish their own plywood industriesand the prices of logs rise. Second,other boards are substituting plywood,such as hardboard, medium board,medium density fiber board (MDF),and soft board. Some of theseproducts are made from sugar fiber.Third, the comparative advantage inplywood production of Thailand willdecrease over the years as that ofIndonesia and Malaysia increase.Fourth, some plants are replacingcassava starch with phenolic resin,which provides a better adhesivequality. Total cassava starchconsumption in the plywood industrywill therefore decrease by 30%-40%from the 1993 level and then remainstable until year 2000.

Textile industry. Cassava starchis applied to the yarn in the warp

before weaving, at about 1% of thewarp’s total weight. Modified starch isalso used in dyeing, an industry thatis not yet well developed in Thailand. The estimated current consumption ofcassava starch in the textile industryis therefore minimal. A statisticalseries on cassava starch consumptiondata was constructed for 1985-1990,and used to estimate a simple trendregression. The simple trend equationis as follows:

STH = 9657.5 + 816.5 Y (3)t-vale = (26.699) (6.182)

R2 = 0.9508

D.W. = 2.0012

where:

STH = total annual cassava starchconsumption

Y = year 1985 = 1

Other industries. Otherindustries that use cassava starch asa raw material are those thatmanufacture glues, paper products,and chemicals. The estimated cassavastarch consumption are about15,000 t in 1980, and 60,000 t in

Table 2. Projected consumption of cassava starch by Thai food and nonfood industries.a

Industry 1991 1996 2001

Food-processing industries 375,071 (73) 516,463 (70) 772,819 (65)MSG and lysine 97,977 (19) 170,456 (23) 322,194 (27)Glucose syrup 76,375 (15) 113,368 (15) 177,490 (15)Pearl sago 32,060 (6) 44,690 (6) 62,295 (5)Household consumption 134,908 (26) 144,582 (19) 153,645 (13)Other food industries 33,751 (7) 43,367 (6) 57,195 (5)

Nonfood industry 136,151 (26) 226,357 (30) 411,634 (35)Paper 47,098 (9) 86,776 (12) 159,879 (13)Plywood 6,700 (1) 2,010 (<1) 2,010 (<1)Textiles 14,557 (3) 18,640 (3) 22,722 (2)Other industries 67,796 (13) 118,931 (16) 227,023 (19)

Total 511,221 (100) 742,818 (100) 1,184,453 (100)

a. Figures in parentheses are rounded percentages of the total.

SOURCE: TDRI, 1992.

63


1990, indicating an annual growthrate of about 15%. This growth ratewas used to project futureconsumption of cassava starch.

Estimates of cassava starchconsumption in Thailand arepresented in Table 2. In 1991, about511,221 t of cassava starch wasconsumed, 73% of which wasconsumed by food industries andhouseholds. When the data arebroken down, household consumptionis highest with 26%, followed by MSGand lysine (19%), sweeteners(excluding fructose) (15%), “other foodindustries” (7%), and pearl sago (6%).

Nonfood industries consumed136,151 t of cassava starch or 26% ofthe total. “Other nonfood industries”consumed the highest percentage(13%), followed by paper industries(9%), textile industries (3%), andplywood industries (1%).

The estimated total domesticstarch consumption in 2001 is1.18 million tons. Although starchconsumption by nonfood industrieswill have increased to more than400,000 t (35% of the total), mostdomestic starch consumption will stillbe in the food industries (65%).Among the industries, theMSG-and-lysine industries willconsume the most (27%), followed by“other nonfood industries” (19%),sweeteners (15%), and paperindustries (13%).

The fructose industry used9-15 thousand tons of cassava starchduring 1988-1990. Once existing foodregulations permit the use of fructosein the domestic soft drink industry,then demand for fructose will increaseby about 20% per year. This willmean an extra 17,600 t of cassavastarch in 1991; 38,000 in 1996; and92,200 in 2001.

Exports and major markets

As mentioned above, export marketsfor Thai cassava starch stronglyinfluence domestic price formation.The future prospects of export marketsare therefore highly significant for thedevelopment of, not only the starchindustry, but the entire cassavaindustry in Thailand.

On the whole, the starch industryhas been export oriented since the1940s. Although the quantities ofexported cassava starch havefluctuated, an upward trend isobvious. Data from the TTFITA showthat exports of cassava native andmodified starches increased from459,048 t in 1985 to 656,291 in 1990.Exports to Japan increased from143,619 to 204,572 t, and to Taiwanfrom 124,926 to 248,434 t. That is,the export share of Japan and Taiwanincreased from 58% to 69% of allexports. These countries are expectedto remain major export markets in thefuture.

Japan. Although data from theJapanese Ministry of Agriculture showthat Japan’s total annual starchconsumption increased from2.4 million tons in 1986 to 2.7 millionin 1990, other sources suggest thatJapan consumes at least 3.5 milliontons of starch annually. The JapaneseGovernment has set 0.2 million tonsas the maximum annual import quotafor starch to protect the domesticstarch industry, which is based onsweetpotato and white potato.

Starch in Japan is consumedmostly by manufacturing andprocessing industries, especially forsyrup dextrose (60%). Otherindustries, in descending order, arechemicals (including medicines) ormodified starches, fibers, foodstuffs,paper and adhesives, beverages,

64


fish-paste products, and MSG.Sources of starches are maize (79%),white potato (10%), sweetpotato (5%),imported starch (4%), and wheatstarch and/or flour (2%). Some ofthe products of processing arere-exported.

The manufacturing andprocessing industries are not only themajor consumers of starch, but theyare also the major importers,especially the syrup dextroseproducers, modified-starchprocessors, re-export processingindustries, and manufacturers ofMSG, medicines, and adhesives.

In 1990, the average wholesaleprices of starch in Japan showed thatnative cassava starch was thecheapest (Table 3). If there were noimport barriers, cassava starchimports would increasetremendously.

Imported modified starch issubject to an 8% import duty if fromdeveloped countries. Althoughdeveloping countries pay 0% tariff,they face an import ceiling, imposedby the Japanese Government since1989. The ceiling is based on a totalvalue per year. During the earlystages of implementation, theJapanese Government was flexible,and some groups of modified

starches were imported at muchhigher rates than the set ceiling.Imported modified cassava starchmust compete with domesticmodified starch made from maize.

Thailand’s competitive position inthe Japanese market is determinedby its status in two categories: first,the native starch market, in whichThailand still has the strongadvantages of low prices andcontinuous supplies; and, second,the modified-starch market, in whichThailand faces not only competitionfrom domestic modified starch, butalso from modified starch importedfrom the EEC (which makes it fromlow-priced starch, itself importedfrom Eastern Europe). Futureprospects in the Japanese market,however, depend heavily on Japan’strade protectionist policies.

Taiwan. Being a newlydeveloped industrializedcountry, with a rapidly growingeconomy, Taiwan has had torestructure its agricultural sector.From producing basic raw materials,it now produces high-value productssuch as fruit and those fromlivestock and fishery. Consequently,Taiwan expects to import more ofboth raw and finished agriculturalproducts. Although Taiwan does notimpose a tariff import barrier oncassava starch, it does on importedcassava products (Table 4).

Thai cassava starch productshave good prospects in Taiwan,where Thai exporters and concernedgovernmental agencies have activelypromoted cassava products in theTaiwan market.

Projections of cassava starchexports to Japan and Taiwan, andof total exports. Simple lineartrends (Table 5) were used to projectcassava starch exports to Japanand Taiwan, and in total. The

Table 3. Average wholesale prices of starches inJapan, 1990.

Starch Price(yen/kg)

Domestic starches produced from:Sweetpotato 65.00White potato 140.00Maize 62.00

Imported starchesNative cassava starch 33.00White potato starch 63.00

SOURCE: TDRI, 1992.

65


will reach about 1.82 million tons(9.1 million tons of roots), of whichdomestic consumption would accountfor 41% (data not shown). In 2001,total demand would increase to2.6 million tons (13 million tons ofroots), of which domestic consumptionwould account for 46%. The future ofthe cassava starch industry willtherefore still be export oriented.

Scenario of FutureIndustrial Adjustment

As mentioned earlier, the EEC’sCommon Agricultural Policy (CAP)triggered off the development of Thaicassava products for the animal feedindustry. The EEC has been the onlymajor market for these products, aresult of the EEC’s high cereal prices.Hence, any changes in the CAP willhave a strong impact on the Thaicassava industry. Analyses of CAPreforms will therefore be imperative forpredicting the industry’s futureprospects and development.

The CAP reforms

Overall, the CAP has fulfilled theEEC’s objective of reachingself-sufficiency in food, but at the highprice of subsidizing the agriculturalsector. The CAP also created severalproblems, especially theoverproduction of cereals, andlivestock and dairy products, whichcost more than ECU 79,000 million.

One reason for the overproductionof cereals was their reduced use in theanimal feed industry, whichsubstituted the highly priced cerealswith cheap NGFI imports. The EEChas tried to limit and reduce NGFIimports by setting quotas for cassavaimports from Thailand, Indonesia,Brazil, and China. Many other NGFIproducts, however, were importedwithout restrictions or tariffs.

Table 4. Import duties imposed on cassavaproducts by Taiwan.

HS code Tariff rate

0714.10 Manioc (cassava) 20%

1108.14 Manioc (cassava) starch 17% or NT$1,200/t

1903.00 Tapioca and substitutes 17% orprepared from starch NT$1,306/t

3505.00 Dextrins and other 7.5%-20%a

modified starches 7.5%-17%b

a. Imposed for all countries.b. Applied to countries with reciprocal benefits, such

as Thailand.

SOURCE: TDRI, 1992.

Table 5. Projected cassava starch exports toTaiwan and Japan and total exports(in tons), Thailand, 1993-2001.

Year Taiwan Japan Total exports

1993 355,673 259,837 872,614

1994 390,922 278,065 939,709

1995 426,171 296,293 1,006,805

1996 461,420 314,520 1,073,901

1997 496,668 332,748 1,208,093

1998 531,917 350,976 1,275,189

1999 567,166 369,204 1,342,285

2000 602,415 387,431 1,342,285

2001 637,664 405,659 1,409,381

SOURCE: TDRI, 1992.

projection of total starch exports,however, did not include thepossibility of new markets, such asSouth Korea’s paper industry. Atpresent, cassava starch imports underthe international HS code 1108.14 arenot restricted by South Korea.Another potential market is Russia, ifspecial export credits can be madeavailable to Thai exports through theestablishment of an export-importbank. At least 10,000 t of cassavastarch would then be exported.

Based on the above estimates,demand for cassava starch in bothdomestic and export markets in 1996,

66


Impact of CAP reforms on prices ofcassava products and roots

The CAP reforms will probablystrongly effect NGFI imports,especially those providing sources ofenergy in animal feed, such ascassava products. As cereals becomecheaper, substitutes will be used less.The EEC commission reported that thesubstitution effect would be6-7 million tons (EEC, 1993b).

Given the price relationship ofECU 24 per ton between wheat andcassava products used in compoundfeed in previous years, Thai cassavaproducts are expected to becompetitive in the EEC market and tobe consumed by the animal feedindustry at the current rate of about5 million tons. Prices, however, woulddecline to the following levels:

Season Wholesale prices of cassavaproducts in the EEC

(ECU per ton)

1993/94 93

1994/95 84

1995/96 76

However, the above price levelsshow the worst scenario. Given theexchange rate of ECU 1 = US$1.19and US$1.00 = 25.30 baht, farmgateprices of cassava roots in NakhonRatchasima Province, Thailand, wouldbe as follows:

Season Farmgate prices inNakhon RatchasimaProvince, Thailand

US$/ton baht per ton

1993/94 22.81 577.14

1994/95 18.78 475.26

1995/96 15.93 403.00

As well as the problems created bythe CAP, the EEC also has had to facepressure from the GATT UruguayRound of trade negotiations toliberalize trade. CAP reforms weretherefore inevitable—concentrating ondecreasing agricultural subsidies toreduce grain and meat surpluses. Thestrongest impact on NGFI importscame from the drastic decrease ofintervention prices for cereals, whichseverely reduced domestic wholesaleprices of cereals. Three major changesfrom the existing system occurred:

(1) Agricultural support shifted frombeing solely price subsidies tobeing compensatory payments toproducers;

(2) Measures for increasingproduction for self-sufficiency wereno longer emphasized; and

(3) Free trade was encouraged whilemaintaining the basic principlesand instruments of the CAP.

Under the CAP reforms, cerealprices will change as follows:

(1) Buying-in prices and interventionprices will be the same for everycereal; and

(2) From the 1993/94 season (July)onward, all cereal prices (ECU perton) will be:

Season Intervention Target Thresholdpricea priceb pricec

1993/94 117 130 175

1994/95 108 120 165

1995/96 100 110 155

a. The price at which the EEC is prepared to buycereals if the market price is below it.

b. The price the EEC wants producers to receive(and consumers to pay). The EEC will intervenethrough import levies (taxes) or by buyingsurpluses to ensure that prices do not fall belowthe target level.

c. The price at which cereal imports enter the EEC,i.e., the world price plus the variable import levy.

67


Based on MOAC statistics, thenational average cost of producingcassava roots in 1991/92 was540 baht/t (US$21.34/t). Thisimplies that farmers received only37.14 baht/t (US$1.47/t) in1993/94, and that, in 1994/95 and1995/96, farmgate prices will be lessthan production costs. Obviously, ifthese price levels become reality,cassava farmers will switch to othercrops.

Also obviously, hard pellet pricesin Rotterdam (ECU 76-93/t) and rootprices in Nakhon Ratchasima(US$15.93-22.81/t) will discourageThai exports to the EEC. That meansthe quota rent of export quota in theEEC will vanish, making it difficult forThai exporters to export pellets tonon-EEC markets at such low pricesthat they would obtain export quota tothe EEC and thus sell pellets at highprices. In fact, current non-EECmarkets for Thai pellets are subsidizedby the quota rent to the EEC. Thesemarkets are not potential markets forcassava pellets, unless cereal suppliesbecome drastically short and worldprices of high-protein ingredients foranimal feed (soybean meal) becomevery low.

Impact of CAP reforms on rootsupplies to the Thai cassava starchindustry

This section tries, perhapsprematurely, to project what wouldhappen if exports of Thai cassavaproducts as animal feed to the EECwere decreased drastically. Theprojection is based on observations ofevents in Nakhon RatchasimaProvince.

After the new CAP wasimplemented in July 1993, buyingprices for roots offered by plants in theprovince decreased from 740 baht/t inJuly to 700 in October. This was aresult of adjustments in the EEC’s

compound feed industry and of theThai export industry to the CAPreforms, which dropped export pricesfor cassava products for animal feed inthe EEC. Consequently, producers ofcassava chips and pellets had to lowertheir buying prices for roots. As rootprices decreased, root supplies alsodecreased.

The immediate impact of CAPreforms on the Thai cassava starchindustry was to create competitionamong cassava starch plants to obtainthe cheapest root supplies. Thismeant that, if cassava farmers delayedtheir harvests, root prices wouldincrease in the short run. But, ifprices of cassava products in the EECdropped to ECU 93/t, then exports tothe EEC in 1993/94 would be reduceddrastically. Eventually, surpluses ofcassava roots would develop andprices will drop below 700 baht/t in1993/94.

If prices remain at 700 baht/t (orUS$27.67/t), which would give a netfarmgate price of 580 baht/t(US$22.92/t), farmers would find itunprofitable to grow cassava. Forexample, the 1993/94 root productioncosts in Nakhon Ratchasimawere 578.50-664.80 baht/t(US$ 22.87-26.28/t). This impliesthat cassava root production will begindecreasing.

The Thai Department ofAgriculture reported that productioncosts for cassava would decrease to509.00 baht/t (US$20.12/t) if farmersfollowed appropriate agriculturalpractices and used the new Rayong60 variety. Trade associations ofNakhon Ratchasima Province are nowworking closely with concernedgovernmental agencies to provideextension services for cassava farmersto encourage them to adopt newagricultural practices and varieties.Extension services, however, are notsufficient if farmgate prices are not

68


high enough. To ensure their rawmaterial supplies, therefore, cassavastarch plants have begun contractfarming.

To avoid cassava root surpluses,the Government had already, in early1993, launched a program toencourage farmers to reduce plantingareas (now totalling 400,000 rai, or64,000 ha). It is still too early toassess the program’s success, butcassava production will decrease inany case, if the above price level of700 baht/t is realized for 1993/94.

Cassava starch plants willtherefore face problems of rootsupplies, and their period of operationmay become smaller than 8 months ifcontract farming and extensionservices for improved varieties andagricultural practices are not realized.

As the production of cassavaproducts for animal feed decreases,the cassava market will becomedominated by starch plants operatingin cassava-producing areas. Duringpeak seasons, local starch plants willnot be able to buy all available roots.Root prices will therefore drop to levelsat which chip-and-pellet plants findprofitable to start their operations.Thus, a new market equilibrium ofroot prices will be established atlevels profitable for farmers andchip-and-pellet producers. The levelwill depend heavily on the exportprices of chips and pellets and ondomestic demand for these products.Even so, both farmers and starchplants would mutually benefit fromsetting up a system that regulates rootsupplies.

Starch processing

As low prices and decreasing demandfor roots force reductions in cassavaproduction, the root marketing periodwill shorten and adjust to the seasonaldemand for cassava products in the

EEC market. Cassava starch plantswould have fewer operational days andhigher average production costs. Toovercome such problems, the plantsmay either increase capacity per day,or minimize production costs whereverpossible.

The first alternative may bepossible by merging plants. Thus,only large and efficient starch plantswill survive, and their operationswould also be further integrated withhigh-value processing activities suchas modified starch. The plants mayalso be forced to diversify intocommodity trade.

Production costs may beminimized through joint efforts inobtaining special rates fromgovernmental authorities for utilitiessuch as electricity, which accounts formore than 35% of total processingcosts.

Governmental policy

Although concerned governmentalagencies realize that the CAP reformswill generate negative impact on theThai cassava industry, especially foranimal feed, the only policy so farimplemented is that of reducingcassava planting areas. Short- andlong-term policies for the cassavaindustry are yet to be formulated. Inaddition, the Government has still todecide whether to renew or renouncethe Agreement, which will expire in1994, between Thailand and the EECon cassava exports.

Summary, Conclusions,and Recommendations

The cassava starch industry hasdeveloped largely under a free marketsystem, with limited governmentalintervention. The EEC’s CAP triggeredoff the rapid development of cassavaexports for animal feed in the 1970s,

69


causing the whole industry to shiftfrom starch processing to theprocessing of cassava exports foranimal feed.

Although, by percentage, theproportion of cassava starch exports tototal cassava exports decreased from25% in 1966 to 11% in 1991, starchexports themselves increased at anannual growth rate of 5.5%. USA andJapan have formed the major marketfor Thai cassava starch since 1966,despite competition with domesticmaize starch. During the 1980s,Taiwan became the third mostimportant market for Thailand, usingThai starch in modified-starchprocessing and other industries.

In 1982, the EEC-Thai CooperativeAgreement was signed; it set amaximum import quantity of21 million tons over 4 years. TheAgreement also obliged Thailand toactively search for other uses ofcassava, finally settling onvalue-added cassava starch, that is,modified starch, for Japan.

Cassava starch was alreadyproduced for domestic consumption,both as food and industrial rawmaterial, and, in relatively largerquantities, for export. In 1965, theestimated total domestic consumptionwas 44,557 t, and exports were148,206 t. During 1965 to 1980,starch was used mostly in foodindustries (27%), the manufacture ofMSG (22%), paper industry (16%), andhousehold consumption (16%).

Thailand’s outstanding economicperformance during 1980-1990 in bothindustrial and agroindustrial sectorsdrew the attention of cassava starchentrepreneurs to the domestic use ofstarch and its potential. During1990-1991, a survey was carried out toestimate domestic starch consumptionin various Thai industries, and toproject starch use in the next decade.

In the early 1990s, fewer than50 cassava starch plants wereactively operating, with a totalcapacity of about 1.4-1.6 milliontons of starch per year. Thiscompares with the 2 million tonsthat 84 plants produced in the late1980s. Of the plants remaining, 17were modified-starch plants, with anestimated capacity of 300,000 t/yearand an actual production of250,000 t.

Domestic cassava starchconsumption was projected (asdescribed in “Current and FutureDomestic Use and Export,” p. 59-65)to the year 2001 as almost1.2 million tons. Domesticconsumption and use in foodprocessing will decrease to 18%.Use in textiles will decrease to 2%,and in plywood to 0.2%. In contrast,starch consumption in themanufacture of MSG and lysine willincrease to 27%, and in the paperindustry to 13% (Table 2).

Total cassava starch use in1991, that is, the sum of totaldomestic consumption plus totalexports, was more than 1.2 milliontons. It may increase to more than2.5 million tons by the year 2001,assuming Japan and Taiwan as theonly two major export markets.

Despite the fact that domesticconsumption of cassava starch hasincreased over time, domestic pricesdepend heavily on export prices,especially those of modified starch inrecent years. For the future, thecassava starch industry, and thecassava industry as a whole, willstill be export oriented. The EEC’sCAP reforms, which reduceddomestic cereal prices by 29% forJuly 1993 to June 1996, willtherefore strongly influence the Thaicassava industry.

70


The impact of reduced cerealprices in the EEC (to ECU 117/t in1993-1995) on Thai pellet prices inRotterdam was to reduce them toECU 93/t. This reduced price, inturn, reduced the farmgate price ofcassava roots in Nakhon RatchasimaProvince, Thailand, to US$23/t in1993/94, only slightly aboveproduction costs. These reducedprices may make the cassava starchindustry the major buyer of roots inthe domestic market.

But if the CAP reforms drasticallydecrease the exports of Thai cassavaproducts to the EEC, then cassavaproduction would decrease in thefuture, creating problems of suppliesfor cassava starch plants. Toovercome these problems, starchplants and cassava farmers may findthat contract farming would bemutually beneficial.

Despite the uncertainty of the kindof impact the CAP reforms will have onthe Thai cassava industry, bothdomestic cassava starch consumptionand starch exports are likely toincrease. As a whole, the Thaicassava industry is anexport-dominated industry that hasfaced many trade restrictions. Theoutcome of the GATT Uruguay Roundof trade negotiations will stronglyinfluence the cassava industry,especially the starch sector.

Each cassava-producing countryshould take this opportunity to reviewthe potential of its cassava starchindustry in terms of its economiccomparative advantage over otherstarches produced domestically and ofits international economic comparativeadvantage.

As far as the future development ofthe Thai cassava industry as a whole,and its starch industry in particular,is concerned, the followingrecommendations are suggested:

(1) Research on new uses for bothcassava roots and starch should becarried out as a joint effort betweenprivate and public sectors;

(2) Research on cost-reductiontechnologies in cassava productionshould be enhanced anddisseminated to farmers as soon aspossible;

(3) Coordination and cooperationbetween public and private sectorsshould be strengthened throughfrequent dialog and consultation;and

(4) Short- and long-term governmentalpolicies on the cassava industry asa whole should be formulated.

Bibliography

EEC (European Economic Community),Commission of the EuropeanCommunities. 1993a. Agriculture inthe GATT negotiations and reforms ofthe CAP. Brussels, Belgium.

__________. 1993b. CAP reforms and the GATTcompatibility. DG VI. Brussels, Belgium.

Jones, S. F. 1983. The world market for starchand starch products with particularreference to cassava (tapioca) starch.Tropical Development and ResearchInstitute (TDRI), London, UK.

TDRI (Thailand Development Research InstituteFoundation). 1992. Cassava: a scenarioof the next decade. Bangkok, Thailand.(In Thai.)

Titapiwatanakun, Boonjit. 1983. Domestictapioca starch consumption in Thailand.In: TTTA year book 1982. The ThaiTapioca Trade Association (TTTA),Bangkok, Thailand.

__________. 1985. Analysis of the short- andlong-run demand and supply prospectsof tapioca products: report submittedto UN/ESCAPE. Bangkok, Thailand.

TTFITA (Thai Tapioca Flour Industries TradeAssociation). 1989. Thai tapiocaindustries. Bangkok, Thailand.

__________. Various years. Thai TapiocaAssociation yearbook. Bangkok,Thailand.

71

Sweetpotato Flour and Starch:...

CHAPTER 10

SWEETPOTATO FLOUR AND STARCH:ITS USES AND FUTURE POTENTIAL1

Nelly Espínola*

Introduction

In terms of production, sweetpotato(Ipomoea batatas) is the fifth mostimportant crop in developing countries(Table 1). Latin America, its place oforigin, paradoxically accounts for only1.8% of world production (Table 2)(Scott, 1992).

In Peru, this root crop is usedmainly for direct human consumption(96%) and its foliage or lianas are fedto animals (90%). In other countries,like the Philippines, the tender parts ofthe foliage are also eaten as avegetable (Woolfe, 1992).

In China, the world’s leadingproducer of sweetpotato, its use hasvaried over the last 20 years,decreasing for human consumption(from 60% to 40%), and increasing foranimal feed (30% to 45%) andindustrial use (5% to 10%).

About 90% of sweetpotato foliageis used for animal feed. Although ithas a low carbohydrate content, itslevels of fiber, protein, and vitaminsare higher, thus stimulating milkproduction in cattle. Sweetpotatoproduction may vary from 4.3 to

6.0 t/ha of dry matter, depending onthe variety (Ruiz et al., 1980). It canalso be grown as a perennial crop,tolerating foliage cutting every 3 to4 months, depending on where it is

* Physiology Department, Centro Internacionalde la Papa (CIP), Lima, Peru.


Table 1. Food crop production in developingcountries, 1961-1988.

Crop Production(thousands of tons)

Paddy rice 449,968Wheat 214,119Maize 184,927Cassava 137,412Sweetpotato 125,359

SOURCE: Food and Agriculture Organization of theUnited Nations (FAO), Basic Data Unit,unpublished data.

Table 2. Sweetpotato production in developingcountries by region, 1961-1988.

Region Production

(thousands of tons) (%)

Africa 6,263 5.0 (Sub-Saharan) 6,192

Asia (China) 116,811 93.1

108,062

Latin America 228 1.8

Total 125,359 100.0

SOURCE: Food and Agriculture Organization of theUnited Nations (FAO), Basic Data Unit,unpublished data.

72


grown (either coastal regions orhighlands) (Beaufort-Murphy, 1993,personal communication).

A new variety is to be released tofarmers; it is high yielding for forage,with high leaf protein content(up to 19% dry base) and very lowtrypsin-inhibitor content. This varietycan also be grown in the highlandsduring the dry season when naturalpastures become scarce.

In a study carried out in Peru, thecosts of cultivating sweetpotato werefound to be lower than those of othercrops like potato and maize. It can beharvested two or three times a yearand so is considered a staple crop forboth human and animal consumption(Achata et al., 1990).

As a food product, sweetpotato is asource of energy, proteins, provitaminA (β-carotene), vitamin C, and iron. Itranks among the crops generating thehighest carbohydrate content perhectare (156 MJ/day) over a relativelyshort period (120 days), even when lowlevels of fertilizer and pesticide areapplied.

The germplasm bank at the CentroInternacional de la Papa (CIP), Peru, isa source of important genetic diversitywith great potential for improving thecrop for human and animalconsumption and industrial purposes.

Peru has some experience insweetpotato cultivation, but this crop isgrown mainly in countries like Japan,China, Vietnam, and the Philippines,where it is used in foodstuffs, animalfeed, and industry. Sweetpotatoprocessing is increasing in importancein these countries and considerableinformation is available on the topic. Ifthis information were adapted toPeruvian conditions, it could yieldrapid results, making investment insweetpotato products comparativelycheaper than that in other crops.

This paper reviews sweetpotatoprocessing at rural and industriallevels for both human and animalconsumption; and the effects itspromotion would have on production,rural and urban employment, savingsin foreign exchange, and stimulatingagroindustrial activity.

Unprocessed Sweetpotato

In Peru, sweetpotato is boiled, fried,baked, or mashed. Its raw roots can begrated and used to make bread orsweets like camotillo. Sweetpotato rootscan also be used in cattle feed as rawchips mixed with fibrous feedstuffs.Animals seem to find sweetpotatopalatable, searching in their mixed feedto consume it first.

A dairy in the Cañete River area,the largest sweetpotato producingregion of Peru, used sweetpotato in itscattle feed, thus considerablyincreasing the daily milk productionfrom 25-26 L/head (1992) to 30 L/head(1993) (Espínola, 1992).

Bread made with sweetpotato purée

Bread is an important item in thePeruvian diet, but imports of wheatflour for bread making are costly. Sincethe 1960s, researchers have soughtways of substituting wheat flour withsweetpotato flour. The UniversidadNacional Agraria “La Molina” and abakery in Chincha (near Lima), forexample, have started manufacturingcomposite-flour specialty breads, bothloaf and cake-type, which includeboiled, peeled sweetpotato, substitutingas much as 40% of the wheat flour(Peralta et al., 1992).

Bread made with raw, gratedsweetpotato

Peru. The use of raw, unpeeled,grated sweetpotato to make an

73


economic bread is being introduced;sweetpotato substitutes for 30% ofthe wheat flour. A “grainy” bread isobtained, with a slightly sweet taste,resulting from the very sweetcommercial varieties used. Itsnutritional quality is similar to that oftraditional bread made with 100%wheat flour (Denen et al., 1993). CIPis collaborating with the UniversidadNacional Agraria “La Molina” toimprove this sweet bread’s recipe,appearance, shelf life, and crumb (or“grain”) size. Initial efforts, in whichthe peel was discarded and the rootsfinely grated, succeeded in substitutingas much as 50% of wheat flour(20% dry base) with sweetpotatoes.

Burundi. Bread preparation withraw, grated sweetpotato wasintroduced to Burundi, eastern Africa,where the technology was adapted.White varieties, with dry mattercontents ranging from 25% to 30%,were used to replace 30% of the wheatflour. Modifications included finegrating of peeled roots, oil instead oflard, and the elimination of enhancers.Egg whites were added as a finaltouch, to give the bread a look similarto the Peruvian pan de yema.

Costs were lower than for 100%wheat bread but the selling price wasthe same. This meant a higher profitfor the baker and a reduced use ofwheat flour, with subsequent savingsin foreign exchange (Berrios andBeavogui, 1992).

Processed Sweetpotato

Processed sweetpotato productsinclude snack foods, such as friedchips and caramel-coated chips, andindustrial products such assweetpotato flour, purée, and starch.A large variety of starch-basedproducts exists, and Japan has givenhigh commercial value to rawsweetpotato starch.

Flour: Peru

If sweetpotato varieties with a highdry-matter and starch content, and alow oxidation rate are processed, ahigher percentage of flour is recovered(Table 3).

CIP is collaborating with theUniversidad Nacional Agraria “LaMolina” to test both sun- andoven-dried flour in poultry rations. Theoptimal level of maize substitution andthe better type of flour (raw or cooked)will be determined. In another study,with pigs, conducted by the sameuniversity, oven-dried sweetpotato wasused to determine the optimalsubstitution level, according to animalage.

Commercial production ofsweetpotato flour, supported by theprivate sector, is just beginning, andnew products for subsequentcommercialization are being developed.Commercially, sweetpotato flour can beused to substitute wheat flour in breadmaking or maize flour in balancedfeeds (in 1992, maize imports weremore than US$82 million).Sweetpotato flour will be analyzed toidentify its chemical components andcontaminants, and to determine itsdigestibility and potential uses inhuman and animal consumption.

Companies are producingsweetpotato flour through a novel,low-cost industrial process thatprocesses the roots either continuously

Table 3. Percentage of flour recovered fromdifferent varieties of sweetpotato.

Country Variety Flour yield(%)

Philippines Georgia Red 12.0Ilocos Sur 37.0

Peru Jonathan 28.6(commercial variety)

74


(500 kg/h), taking 8 to 10 min; or inbatches (less than 500 kg/h), taking45 to 60 min. Flour yield forcommercial sweetpotato varieties is28.6%, that is, 3.5 kg of freshsweetpotatoes produces 1 kg of flour(Vásquez, 1994).

First, sweetpotatoes are selected,then washed and peeledmechanically, only partially removingpeel and foreign material by abrasion.Milling, that is, rasping the root athigh velocity, follows. A very fineproduct is obtained, which is thenpressed to eliminate water, resultingin a pressed cake with 38% moisture.Some solids containing starch,β-carotene, sugars, and proteins areeliminated with the water but arelater recovered by decanting andincorporated into the flour duringdrying. The flour is dried at 40 °C for8 seconds in a current of hot airproduced by propane gas (for tableflour) or by carbon briquettes (foranimal feed). The flour is packed in50-kg plastic bags through a feederhopper.

Production costs for a volume of700 t are US$150/t, and the sellingprice is US$190/t. To maintainoptimal sanitary conditions, stainlesssteel materials are used in theequipment.

Purée

Philippines. Because thisprocess involves advanced technology,high-quality varieties should be usedto obtain a maximum yield.Developed countries, such as the USAor Japan, produce sweetpotato puréeon a large scale. In the Philippines, apowdered product for preparingsweetpotato mash and “Cantonese”noodles, has developed through acollaborative project between thenational agricultural program andVisayas State College of Agriculture(ViSCA). This powder is also the main

ingredient for instant soups andtraditional breakfasts.

Peru. The sweetpotato processingindustry is just beginning and onlyone company is producing flakes fromsweetpotato purée under thecommercial trade name “Menú.” Thisproduct can be used to prepare babyfoods or school breakfasts, but highproduction costs are still a constraint(Denen et al., 1993).

Sweetpotato starch

The granule size of sweetpotato starchis similar to that of cassava, with only5% of the total being very small andcolloidal, and easily lost during waterextraction. Both sweetpotato andcassava starches, when submitted toX-ray diffraction, have type-Astructures commonly found in cereals.The ratio of amylopectin to amylose is3:1 and, in some cases, 4:1 (Woolfe,1992).

Gelatinization temperature andtype are important in feed formulas,varying with variety. Gelatinizationtemperatures from 58 to 69 °C, 58 to75 °C, and 65 to 80 °C have beenreported. The degree of associationbetween molecules is much greaterthan in potato and similar to thatfound in cassava.

Raw sweetpotato starch is muchmore resistant to the action ofdigestive enzymes (2.4%) than aremaize (9.2%) and wheat (17.6%)starches. The degradation time ofraw sweetpotato starch is 15% in 6 hcompared with 20% for cassava and10% for yam. Amylose degradationcapacity increases when starchgranules are ruptured, improving inpelletization (animal feed) andreaching a peak in cooking.

Asia. In countries like China,Taiwan, and the Philippines,

75


sweetpotato is grown on small farms,where the raw roots are cut into longchips or flakes and sun-dried. Thedried product is then sent to distilleriesand starch factories for furtherprocessing.

In Korea, Taiwan, and Japan,about 8%, 16%, and 28%, respectively,of the sweetpotato production is usedas raw starch in the food industry, formaking bread, biscuits, cakes, juices,ice cream, and noodles. The starch isalso converted into glucose syrup orisomerized glucose syrup (where someof the glucose has been converted intofructose to sweeten it further).

In Japan and Korea, starches andother fermentable carbohydrates areused to distil a typical liquor calledsocchu. Lactic acid, acetone, butanol,vinegar, and leavenings are alsoproduced by fermentative processes.

Japan has developed a cyclodextrinwith diverse, high-value uses in thefood and pharmaceutical industriesand in blood tests.

China. Sichuan Province is theworld’s largest producer ofsweetpotato, most of which is used inprocessed products and animal feeds.The crop forms the major source ofincome for most of the inhabitants.Simple, small-scale technology is usedto produce starch and noodles. Thenoodles are similar to the rice noodles

(or Chinese noodles) commonly used inthe dish chifa.

Three different methods of starchextraction are employed, with differingpercentages of recovery: the waterprecipitation method (12%-14%); thenatural precipitation method(16%-18%); and the liquid acid method(17%-20%) (Timmins et al., 1992).

Residues are peelings and fiber,which are fed to pigs either directly orafter fermentation, or after being driedand mixed with other types of foragessuch as maize stalks or rice hulls (CIP,1991).

Peru. The prospects for starchextraction in Peru are good, especiallyin rural agroindustry, as are potato andmaize starches, both widely used in thecountry. Sweetpotato starch could beused in the food, textile, glue, paint, andcardboard industries. A company hasbegun manufacturing starch extractionequipment adapted to small- andmedium-scale farming conditions.

CIP’s Plant Breeding Program hasdeveloped advanced clones, withadequate processing characteristics,that are undergoing final testing andselection for release by the InstitutoNacional de Investigación Agraria (INIA).The clones should have high dry-mattercontent, high yields, low oxidation rate,low fiber content, and very low latexcontent (Table 4).

Table 4. Preliminary data of clones selected for starch contenta and suitable for industrial processing,Peru.

Clone Color Dry matter Starch(%) (%)

SR 87.070 White 37.6 25.83SR 90.012 Cream 38.6 23.66SR 90.021 Cream 30.0 20.30SR 90.323 Orange 42.8 22.80YM 89.240 White 35.2 23.41YM 89.052 White 29.8 20.47

a. Analyses carried out by Derivados del Maíz, S. A. (DEMSA).

76


INIA is also assessing about1,000 clones for humanconsumption, processing, andforage purposes in the Cañeteregion. A private company hasevaluated 42 of these clones fordry-matter and starch contents. Sixclones had high yields, and highdry-matter and starch contents,indicating broad potential forfurther research (Table 4).

Agroindustrial Prospects

The following list summarizes theagroindustrial prospects ofsweetpotato:

(1) Grated, raw sweetpotato forpreparing an economic bread.

(2) Sun-dried flakes for producingstarch, alcohol, or flour forhuman consumption, animalfeed, and industry.

(3) Sweetpotato flour forpreparing porridge, breads,biscuits, and balanced feedsfor poultry and swine(replacing maize).

(4) Sweetpotato starch forpreparing noodles and glucosesyrup (dextrins). Peru importsglucose to producepharmaceutical syrups,caramels, and gum drops. Itis also used by the textile andglue industries.

(5) Production of alcohol (Japan,Korea).

(6) Sweets (e.g., flakes andcaramels).

(7) Ketchup (Philippines).(8) Fruit-flavored juices

(Philippines).(9) Liquor (Japan, Korea).

(10) Balanced feeds for poultry andswine (replacing maize).

(11) Extracting anthocyanin, anatural purple coloring usedin preparing ice cream, yogurt,and pastries.

(12) Sweetpotato pastes with highβ-carotene content for babyfoods.

Conclusions

The potential of sweetpotatoprocessing will be realized if:

(1) Sweetpotato breeding is directedtoward processing. This meansusing the world collection ofgermplasm and installingfacilities for conductingproduction trials at differentsites.

(2) Collaborative research projectsare conducted to develop newproducts.

(3) Information on experiences inother countries is collected andorganized to help designstrategies and policies thatwould support sweetpotatoprocessing.

(4) The private sector participatesactively. Links must thereforebe established with the privatesector.

References

Achata, A.; Fano, H.; Goyas, H.; Chiang, O.;and Andrade, M. 1990. El camote(batata) en el sistema alimentario delPerú: El caso del Valle de Cañete.Centro Internacional de la Papa (CIP),Lima, Peru. 63 p.

Berrios, D. and Beavogui, M. 1992. Trials forthe introduction of sweetpotato inbreadmaking in Burundi. In: Scott,G. J.; Ferguson, P. I.; and Herrera,J. E. (eds.). Product development forroot and tuber crops, vol. III—Africa:proceedings of an internationalmeeting held 26 October-2 November,1991, at IITA, Ibadan, Nigeria. CentroInternacional de la Papa (CIP), Lima,Peru. 506 p.

CIP (Centro Internacional de la Papa).1991. Informe anual. Lima, Peru.258 p.

77


Denen, H.; Espínola, N.; Galarreta, V.; Herrera,J.; and Sluimer, P. 1993. Actividadespropuestas para el crecimiento de laproducción de camote mediante laampliación de su utilización: Informe dela misión para formular proyectos dedesarrollo de productos de camoterealizada del 10 al 21 de mayo de 1993,por encargo de la Embajada Real de losPaíses Bajos, en colaboración con laSecretaría Ejecutiva de CooperaciónInternacional del Ministerio de laPresidencia (SECTI/MIPRE), bajo lacoordinación del Centro Internacionalde la Papa (CIP), Lima, Peru. 79 p.(Typescript.)

Espínola, N. 1992. Alimentación animal conbatata (Ipomoea batatas) enLatinoamérica. Turrialba42(1):114-126.

Peralta, P.; Cavero, W.; and Chumbe, V. 1992.Un diagnóstico rápido del pan decamote en el Perú. In: Desarrollo deproductos de raíces y tubérculos enAmérica Latina, vol. II—América Latina:proceedings of an international meetingheld 8-12 April, 1991, at the Institutode Ciencia y Tecnología Agrícolas(ICTA), Villa Nueva, Guatemala. CentroInternacional de la Papa (CIP), Lima,Peru. 375 p.

Ruiz, M. E.; Pezo, D.; and Martínez, L. 1980.The use of sweetpotato (Ipomoeabatatas (L.) Lam) in animal feeding.Trop. Anim. Prod. 5:144-151.

Scott, G. J. 1992. Transformación de loscultivos alimenticios tradicionales:Desarrollo de productos a base deraíces y tubérculos. In: Desarrollo deproductos de raíces y tubérculos,vol. II—América Latina: proceedingsof an international meeting held8-12 April, 1991, at the Instituto deCiencia y Tecnología Agrícolas (ICTA),Villa Nueva, Guatemala. CentroInternacional de la Papa (CIP), Lima,Peru. 375 p.

Timmins, W. H.; Marter, A. D.; Wesby, A.;and Rickard, J. E. 1992. Aspects ofsweetpotato processing in SichuanProvince, People’s Republic of China.In: Product Development for Root andTuber Crops, vol. 1—Asia:proceedings of an internationalmeeting held 22 April-1 May, 1991,at the Visayas State College ofAgriculture (ViSCA), Baybay, Leyte,Philippines. Centro Internacional dela Papa (CIP), Lima, Peru. 384 p.

Vásquez, H. 1994. Procesamiento a bajocosto de la harina de camote. Paperpresented at the second meetingof the Grupo de Camote, held7 January, 1994. CentroInternacional de la Papa (CIP), Lima,Peru. (Typescript.)

Woolfe, J. 1992. The sweet potato, anuntapped food resource. CambridgeUniversity Press, Cambridge, UK.643 p.

78


CHAPTER 11

PROSPECTS FOR CASSAVA STARCH

IN VIETNAM1

Dang Thanh Ha*, Le Cong Tru*, and G. Henry**

analyzing constraints andopportunities for the cassava sector.The 1991 cassava benchmark study(Howeler, 1996) included householdsurveys, focusing on cassavaproduction, on-farm processing, useand consumption, and rural,semiurban, and urban marketing.Also included were processingsurveys, which focused on thetechnical and socioeconomic aspectsof different products processed andmajor marketing channels.

Results from an analysisconducted by Henry et al. (1993) onthe main constraints to cassavaproduction, productivity, processing,and marketing could serve as a basefor strategic research planning inVietnam. Cassava-based productsseem potentially significant for thefuture. Henry et al. (1993) alsoreviewed the products and marketopportunities of cassava in Vietnam,but data were scarce and informationincomplete.

For decisions on future cassavaresearch and development inVietnam, additional in-depth studiesare required to analyze current andfuture potential demand of differentcassava-based products. Research onmarket demand is important becauseconsumer needs (including industry,on-farm use, etc.) should first beassessed and then production,

Introduction

As the third most important crop afterrice and maize, cassava accounts for30% to 40% of secondary foodproduction in Vietnam (Thang, 1993).The total production of cassava was2.47 million tons of fresh roots in1992 (Statistical Yearbook ofVietnam, 1993) planted on277,200 ha. The VietnameseGovernment has shown interest inthis root crop as a cheap raw materialfor further processing.

In 1989, the Vietnamese Root andTuber Research Program was foundedas the first step toward strategicallyreorganizing root crop research inVietnam. In the past, most efforts inagricultural research anddevelopment in Vietnam concentratedon production and little is knownabout consumer and user needs.From 1990 onward, with CIAT’sassistance, a series of cassavaproduction, processing, andmarketing analyses were conductedin Vietnam, aimed at identifying and

* Department of Agricultural Economics,University of Agriculture and Forestry,Thuduc, Ho Chi Minh City, Vietnam.

** Cassava Program, CIAT, Cali, Colombia.


79

Prospects for Cassava Starch in Vietnam

processing, and marketing technologygeared to address specificopportunities.

This study aims, first, to analyzecurrent use of, and relative quantitiesof, starch for different end products;second, to estimate starch demand ofthese products for the future; andthird, to recommend issues for futurecassava research and developmentactivities.

Methodology

Data were collected by interviewingprocessors, traders, personnel fromexport companies, and manufacturerswho use cassava starch as a rawmaterial. With manufacturers, theinterview format included questionson production level, current inclusionratio of cassava starch, productiontechnology, technical requirementsfor cassava starch, its growth rate,and future demand. The time seriesdata available on cassava starchconsumption for each end productand for the manufacture of someproducts are neither reliable norconsistent. Future cassava starchdemand was therefore estimated byusing a simple method based onpopulation and income growth.

Current Cassava Use

Cassava roots have been used fordifferent purposes such as animalfeed (flour), starch production (wetand dry starch), fresh roots forhuman consumption, dried chips forexport, and home-processingpurposes such as maltose and alcohol(Table 1).

Fresh roots (for humanconsumption) and flour for animalfeed (both at the farm and byindustry) account for about 73% ofthe total cassava production inVietnam. Total production of pigswas about 13.9 million head and ofpoultry, about 124.5 million head,representing the use of about1.4 million tons of cassava roots.Cassava starch production is thesecond source of root consumption,representing about 16% of totalproduction. Chips for export accountfor almost 5%, and home-processing(not including dried chips) 6%.

Currently, dried cassava starch isused in food processing and for homeconsumption, exported, and used byseveral industries such as textiles,pharmaceuticals, cardboard,monosodium glutamate (MSG),glucose, maltose, and plywood. The

Table 1. Cassava consumption in Vietnam, 1992.

Use Quantity of fresh roots

(t) (%)

Fresh roots (human consumption) 301,376.60 12Animal feed (by farmers and industry) 1,503,845.90 61Dried chips for export 120,000.00 5

Starch production:Dried starch (80%) 316,062.00Wet starch (20%) 79,015.50Total starch (100%) 395,077.50 16

Home processing (dried chips andstarch not included) 150,000.00 6

Total 2,470,300.00 100

80


total demand for cassava roots forstarch production (both wet anddried) was about 395,077 t in 1992.About 20% went to wet starchproduction, mostly for localprocessing into low quality noodles(Binh et al., 1992).

Cassava Starch Processingand Marketing

During the Vietnamese cassavabenchmark survey in 1991, cassavastarch processing was found to bepracticed in most of the provincessurveyed. But the largest cassavastarch processing areas are in theprovinces of Dongnai and Tayninh,and Ho Chi Minh City. Cassavastarch is produced either dried (about80% of the total starch production) orwet (20%) (Binh et al., 1992).

Most cassava starch production isconducted by the household orvillage. It is constrained bytraditional technology (low conversion

rates), limited and fluctuating rootsupplies, seasonality, restrictedcapital, and poor marketorganization. These lead to lowprofits and fluctuating quality andlevels of supply. Some largeprocessing plants also use oldtechnologies. Currently, investors areinterested in improving cassavastarch processing technologies.

Current Use of CassavaStarch in Vietnamese

Industry

Table 2 summarizes current cassavastarch use. Most dried starch isconsumed at home (about 57%) andby food processing industries (about36%).

Food processing and homeconsumption

Households form the largest group ofconsumers of cassava roots (about60,000 t/year). Cassava starch is

Table 2. Use and quantity of starch in different end products, 1992, and potential demand of cassava byyear 2000.

End product Starch Potential demand in year 2000consumption

(t) (%) (t) (%)

Dried starchFood processing 25,000 35.60 30,000 16.51Home consumption 40,000 56.95 45,000 24.76Textiles 1,550 2.21 2,000 1.10Monosodium glutamate 0 0 90,000 49.53Carton 600 0.85 1,200 0.66Glue (other purposes) 50 0.07 150 0.08Plywood 96 0.14 120 0.07Maltose 40 0.06 100 0.06Glucose 1,800 2.56 3,000 1.65Pharmaceutical products 100 0.14 150 0.08Exports 1,000 1.42 10,000 5.50

Total 70,236 100.00 181,720 100.00

Wet starch(Cakes, noodles, etc.) 17,559 18,000

Total starch consumption 87,795 199,720

Fresh root consumption 395,077 898,740

81


used to bake cakes, fry meat and fish,make soup, and cook other traditionalVietnamese dishes. Cassava starch isbought from retailers, who obtain itfrom wholesalers in urban and localmarkets, who, in their turn, receive itfrom processing centers.

The food processing industry,currently the country’s second largestconsumer, uses about 25,000 t ofhigh quality dried cassava starch peryear. A diverse range of products ismade, including bread, rice chips,and cakes. About 30% of total starchused in rice chips is cassava. Formaking cakes, cassava starch ismixed with other starches fromsoybean, green bean, rice, and wheatflours. To be competitive with otherstarches, cassava starch must becheap and of high quality.

Monosodium glutamate production

The total MSG used in Vietnam iscurrently about 40,000 t/year. Mostis imported from Japan, Taiwan, andSingapore, with only a small amountproduced nationally. In the 1980s,Vietnamese companies producedMSG, using as raw material eithercassava starch (75%) or byproductsfrom the sugar industry (25%). Thesecompanies used old technology withlow conversion rates: 6-6.5 t ofcassava starch produced 1 t of MSG.

Cassava starch was obtained fromprocessing centers throughwholesalers. A starch quality of90%-92% purity was required.Constraints included fluctuatingstarch quality as a result ofprocessors using differenttechnologies; and erratic suppliesbecause root availability depended onharvest seasons. The MSGcompanies had to store starch, butoften lacked good storage facilities,which, with the starch’s variableconsistency and low quality, causedquality losses.

The low conversion rates, poorproduct quality, and high productioncosts made local MSG unable tocompete with imported MSG. Thus,many companies ceased productionor attempted to modernize theirtechnology, sometimes through jointventures with foreign partners. Theproduction of MSG decreased from2,003 t in 1987 to 721 t in 1992(Statistical Yearbook of Vietnam,1993). In 1987, almost 12,000 t ofcassava starch were used by thisindustry. But with moderntechnology, MSG is produced mostlyfrom imported glutamate azide andnot from cassava starch.

Since 1990, several foreignmultinationals have entered the MSGsector. At first, they imported MSG tosell in Vietnam, but after conductingmarket research, they concluded thatproducing MSG in Vietnam was aviable option. Currently, they areproducing MSG from glutamate azideimported from the mother company.At the same time, they areresearching the market potential ofMSG produced from local rawmaterial, availability of raw materials(cassava starch, byproducts from thesugar industry, and other starchsources), possible sites, andproduction organization. Four newMSG factories, with a capacity of35-40 thousand tons/year, are nowbeing planned.

Textile industry

About 1,550 t of cassava starch arecurrently used per year by the textileindustry as size for weaving cottonfabrics. Other possible substitutestarches are maize, wheat flour,potato, and rice. In the past, sometextile factories in northern Vietnamused maize starch, which was morereadily available in the Red RiverDelta than cassava starch. Later, assupplies increased, most factorieschanged to cassava starch, which is

82


technically more suitable and alsocheaper. The Government alsoencouraged the industry to replacestarches from other food crops withthat from cassava.

Starch supplies arrive at thefactories from the processing centersthrough wholesalers. The averageprice of high quality cassava starch,in Vietnamese dôngs (VND), is aboutVND 2,000 to VND 2,200/kg. Starchfor the textile industry must behomogeneous in quality, pure(92%-95%), highly adhesive,white—with no change incolor—unfermented, and resistantto quality loss when stored.

Some textile factories use modernweaving machinery that has a highproduction capacity and high weavingspeed. Such machinery requires highquality glue, which is imported. Thisglue could be made from chemicallymodified cassava starch, a productthat is likely to be used by the textileindustry in the future. But rawcassava starch will still be used insmall weaving factories, and forproducing the currently importedglue.

Glues for cardboard production andother purposes

To produce cardboard and otherpacking materials, starch fromwheat flour, maize, rice, andcassava is used as glue. InVietnam, cassava starch and flourare readily available and relativelycheap. Small cardboard-producingunits with simple technology useboth cassava flour and starch, butlarge modern factories use onlycassava starch, as the flour doesnot reach technical standards.Cassava starch must be highlyadhesive and pure (90%-92%).Whiteness is not so important.Most processing centers can satisfythese criteria.

About 200 kg of cassava starchare needed to produce 5 t ofcardboard. Currently, about 600 t ofcassava starch are used to producean annual 15,000 t of carton.

The estimated consumption ofcassava starch for glues for otherpurposes, such as for offices andpacking materials, is about 50 t/year.

Maltose and glucose production

Maltose production in Vietnamconsumes about 40 t of cassavastarch annually, and that of glucoseabout 1,800 t. To produce 1 t ofmaltose, about 1-1.5 t of cassavastarch is needed, and for 1 t ofglucose syrup, about 300 kg.2 Starchhas to be at least 90% pure for thesetwo products. Maltose and glucoseare used by the pharmaceutical andfood-processing industries, whichthus govern demand.

Plywood industry

Together with urea, formaldehyde,and other chemicals, cassava starchis used to produce industrial glue forplywood production. To produce 1 m2

of plywood, 0.46 kg of industrial glueis used, of which about 30%-35% iscassava starch. Wheat flour couldalso be used, but, because of itsrelatively low price, cassava starch ispreferred. Starch must be pure (lessthan 5% of substance remaining afterburning), with a pH value of 5.5 to 6(in unfermented starch), and containless than 10% cellulose. Whiteness isnot important. Current plywoodproduction is about 700,000 m2,consuming about 96 t of cassavastarch annually.

2. Glucose syrup is a concentrated aqueoussolution of saccharides derived from theoriginal starch. Hence, a greater weight ofsyrup is obtained from a given weight ofstarch.

83


The proportion of cassava starchin glues for plywood is relatively highin Vietnam. In the 1980s, the cost ofglues was about 30%-35% of totalproduction costs. Producers thereforereduced production costs by includinghigh rates of cassava starch (35%),which decreased the product’squality. As chemicals for gluemanufacture became cheaper, theproportion of cassava starch droppedfrom 35% to 30%.

Pharmaceutical industry

At present, the pharmaceuticalindustry uses cassava starch toproduce medicinal tablets and pills.Purity, whiteness, and adhesivenessare the most important criteria, whichtraditional technology usually doesnot meet. Thus, starch bought fromprocessing centers has to undergofurther processing to reach theneeded quality. About 100 t ofcassava starch is consumed annuallyby this industry.

Despite the extra processingneeded, cassava starch is cheaperthan other starches such as rice orpotato, and is in plentiful supply.The technology for using cassavastarch is also more readilyavailable.

One constraint is that, undersome circumstances, theadhesiveness of dried cassava starchis not as good as that of otherstarches. For example, in theprocessing of certain medicines, watercannot be used because it mayinterfere with the medicine’s effects.Research is therefore needed ontechnologies that can directly usedried cassava starch in producingtablets and pills.

Although cassava processingfactories specializing in starch forpharmaceutical use satisfy theindustry’s requirements, the quality

of cassava starch urgently needsimproving to produce high qualitydextrin, maltose dextrin, and glucosefor pharmaceutical use.

Exports

Vietnam exports mostly dried cassavachips and only a small quantity ofdried cassava starch. About 30,000 tof cassava chips is exported annuallyto the European Union (EU), 10,000 tto Asian countries, and only 1,000 tof cassava starch and tapioca pearl toneighboring countries. Cassava chipsexported to the EU cost US$120 toUS$130/t, whereas chips to Asiancountries cost US$70 to US$80/t.Because of the higher prices ofexports to the EU, the local price ofdried chips has increased and exportsto Asian countries have dropped.Asian export companies havetherefore changed to exportingcassava starch or tapioca pearl.

Major constraints to cassavastarch export, however, are poorstarch quality, inefficient processingand marketing system, shortages orpoor storage facilities, relatively hightransport costs, and insufficientsupplies. The current conversionrate from fresh root to dried starchin processing is 5:1 in the wetseason and 4:1 in the dry season.Dry matter content is high in the dryseason but farmers set theircropping calendar so that they canharvest in the wet season when lesslabor is needed, but when bothstarch content and conversion ratesare low and the high humiditymakes drying the starch difficult.Many processors therefore cannotproduce export quality starch. Thesmall-scale processing and lowstarch quality make collectingenough starch for export difficult.These constraints make Vietnamesecassava starch noncompetitive withstarch from China and Thailand.

84


Although labor is cheap andfarmgate prices are low in Vietnam,Thailand has more advantages forproducing cassava starch. These arelower prices for fresh roots (farmgateprice is about US$20/t), large-scaleand highly efficient processing, a veryefficient marketing system with goodstorage facilities, high starch quality,large volumes, and low transportcosts.

Nevertheless, foreign companieswith high production capacity, bettertechnology, and better export facilitieshave invested in cassava starchproduction in Vietnam. Theseinvestments may generate anincreased opportunity for cassavastarch exports in the future.

Cassava Starch MarketPotential

Future domestic demand fordifferent end products

Cassava starch consumption iscurrently important, accounting forabout 20% of the total cassavaproduction. Assessments of thestability of this status revealedconsiderable future potential, withincreasing demand from the MSG andother food processing industries, andhousehold consumption.

The demand for MSG is expectedto grow to 60,000 t by the year 2000.Although MSG production fordomestic consumption is growing,increasing production for export isdifficult because neighboringcountries also produce MSG, and insufficient quantities for their ownconsumption. MSG production inVietnam therefore satisfies mostlydomestic demand.

Currently, most companies areusing imported glutamate azide toproduce the MSG, but the potential

for cassava starch is high, even withbyproducts from the sugar industryas an alternative source of starch.Modern technology and suitablebacteria will help increase theconversion rate from cassava starchto MSG. Despite the increased use ofbyproducts from the sugar industry(because of their lower prices), thedemand for cassava starch will still behigh, probably at about 90,000 t/yearin the year 2000.

However, some constraintsoperate against using cassava starchin this industry, one of which is thelarge volumes of starch needed dailyas raw material. For example, toproduce 10,000 t of MSG per year, acompany needs about 29 t ofcassava starch of 90% purity perday. About 116 t of fresh roots perday are needed to produce thisamount of starch. Hence, obtainingsufficient supplies under currentconditions is difficult. Evencollecting such large volumes iscostly, especially in areas wherecassava production is notconcentrated. Transport costs areacceptable up to a distance of120 km around the plant, butorganizing the collection can be aproblem.

Another problem is the seasonalnature of harvesting cassava,occupying about 5 to 6 months/year,coupled with a lack of adequatestorage facilities. Hence, supplies ofstarch for year-round production areinsufficient.

If these constraints could beresolved, then the demand forcassava starch for MSG productionwould be high.

Cassava starch, as a food forhousehold consumption, is inferiorand its demand declines with increasein consumer income. However,demand grows with population

85


increase. With an expectedpopulation growth of 2.1% and anexpected gross national product (GNP)growth of 8%/year, the demand forcassava starch for householdconsumption, is conservativelyestimated to increase from 60,000 to65,000 t/year by the year 2000.

In the food processing industry,as products diversify, the use ofcassava starch will increase greatly inthe future. Cassava starch will beused in producing higher quality foodproducts and new ones such asdifferent kinds of cakes and snacks.But higher starch quality will beneeded.

Textile industries are expected toexpand substantially. Thus,demand for high quality glues (madefrom chemically modified starch)and modern technology, such ashigh-speed weaving machinery, willincrease demand for modifiedcassava starch, which, in turn, mayencourage the local production ofcassava starch. The potentialdemand for cassava starch in thetextile industry is expected to beabout 2,000 t/year.

Demand for cassava starch asraw material for glues for cardboardproduction is expected to increaseto about 1,200 t and for otherpurposes to 150 t. These demands,however, account for only a smallproportion of the total future starchdemand.

In the plywood industry, higherquality plywood will be needed. Theproportion of cassava starch as rawmaterial for glue production willdecline from 30% to 20%-25% in thefuture. But if plywood productionincreases, about 120 t of cassavawill be used annually.

The demand for cassava starch formaltose production and for the

pharmaceutical industry will notincrease much. The demand inglucose production may be higherwith about 3,000 t/year. Table 2summarizes the estimated futuredemand by industry.

Export potential

By the year 2000, cassava starchexports will have significantlyincreased. At present, foreigncompanies are investing in thecassava processing sector andexporting their products. Theprospects for increased exports ofcassava starch are good, once largeprocessing plants are at full capacityand using modern technology. Betterfacilities and high-yielding varietieswith high starch content wouldimprove the conversion ratio andstarch quality, and lower productioncosts. By the year 2000, exportvolume may reach 10,000 t/year.But once the comparative advantageof cheap labor declines, more starchwill be used for domestic industrialconsumption.

Competition from other starches

Cassava starch has a relatively lowerprice than rice starch and wheat flour(Table 3), a price relationship which is

Table 3. Prices of some products in Ho Chi MinhCity, Vietnam, November 1993.(VND 1,080.00 = US$1.00.)

Product Wholesale price Retail price(VND/kg) (VND/kg)

Cassava starch quality:I 2,200 2,300II 2,000 2,200III 1,800 2,000Rice starch 2,900 3,000Wheat flour 3,000 3,200Cassava flour 950 1,300Cassava noodles 3,200 3,300Monosodium

glutamate 5,500 16,000

86


unlikely to change in the future. Theprice of rice starch will not declinecompared with that of cassavastarch because of technology andbecause the Government will notencourage the use of rice starch inindustry. Wheat flour is lesscompetitive with industrial cassavastarch because it is imported, thususing scarce currency and increasingproduction costs. Other starchessuch as cinnamon and maize aremostly used in the food processingsector, being either too expensive ortoo scarce for industrial use.

The influence of governmentpolicies on market potential

The Vietnamese Governmentemphasizes substitution by cassavaand other roots and tubers to favorrice for domestic humanconsumption and export. TheGovernment also encourages theexport of agricultural products,including cassava-based products,such as dried chips and starch,through a zero export tax. Taxes arealso used to protect domesticproduction by limiting the import ofgoods that can be produced inVietnam. For example, the importtax on MSG and wheat flour is 20%of the c.i.f. (cost, insurance, andfreight) price.

The Government policy alsoencourages investment in thecassava sector. These policies cangreatly affect the business potentialof cassava starch, especially in theMSG industry and cassavaprocessing for export. As a result,several foreign companies haveinvested in these two sectors.

Conclusions

The analysis of current cassavastarch use and proportion of starchin different end products reveals that

food processing, householdconsumption, textiles, and glucoseare the current major cassava starchconsumers. Demand in the MSGindustry is expected to increasegreatly. Little change will occur inthe demand for cassava starch in foodprocessing and householdconsumption, which is expected toremain very high in these two sectors.In the food processing industry,products using cassava starch arehighly diverse, requiring better starchquality. Cassava starch exports willalso increase substantially. Demandfor cassava starch by the year 2000 isexpected to be more than twice thepresent level (Table 4).

The trend of cultivated areas inVietnam (Table 5) shows that the areaplanted to cassava is decreasing andproductivity is only slightlyincreasing. In the future, the cassavaarea is likely to decrease further asother industrial crops of higher valuereplace cassava. With the expectedincrease in future demand for cassavastarch, a gap between supply anddemand will develop. The gap willwiden further with increasingdemand for cassava flour for livestockand poultry feed. If cassavaproductivity is not improved,shortages can be expected in thefuture.

Because cassava area cannot beincreased, extra supplies can beobtained only by intensifying cassavaproduction. Introducing high-yieldingand high starch content varietieswould help solve this problem.

But introducing high-yieldingvarieties requires research onadopting new varieties in differentagronomic regions, transferringtechnology to farmers, and thefarmers’ adopting new technology.Despite the higher yields, improvedvarieties require much more chemicalfertilizers, pesticides, and labor than

87


Table 4. Potential growth of cassava starch by industry, Vietnam, 1992-2000.

Product 1992 demand Estimated demand Growth rate (t) in year 2000 (t) (%)

Dried starchFood processing 25,000 30,000 20Home consumption 40,000 45,000 13Textiles 1,550 2,000 29Monosodium glutamate 0 90,000 very largeCarton 600 1,200 100Glue for other purposes 50 150 200Plywood 96 120 25Maltose 40 100 150Glucose 1,800 3,000 67Pharmaceutical products 100 150 50Exports 1,000 10,000 900

Total 70,236 181,720 159

Wet starch 17,559 18,000 3

Total starch consumption 87,795 199,720 127

Total fresh root consumption 395,077 898,740 127

disposition toward cassava-basedproducts created. Such integratedresearch needs cooperation amongagronomists, plant breeders,processing technologists, andeconomists. Only by these means canpossible losses to society beeliminated. For example, in 1987,farmers were encouraged to producemore cassava even though marketdemand was decreasing. The resultwas an over supply of cassava and asignificant drop in the price of freshroots. The drop was so great thatfarmers did not harvest.

To develop the cassava sector, theGovernment should provide adequatestatistics and make information onprices, demand, and other marketingfeatures widely available to helpfarmers, processors, and otherproducers decide appropriately. TheGovernment should also clearlyindicate its pricing policies.

The technical requirements ofstarch quantity and the quality ofdifferent end products made withcassava starch could be used as

Table 5. Total cultivated area and production ofcassava in Vietnam, 1976-1992.

Year Cultivated area Production(000 ha) (000 t)

1976 243.5 1,843.11980 442.9 3,323.01985 335.0 2,939.81986 314.7 2,882.31987 298.9 2,738.31988 317.7 2,838.31989 284.6 2,585.41990 256.8 2,275.81991 273.2 2,454.91992 277.2 2,470.3

SOURCE: Statistical Year Book of Vietnam, 1993.

do local varieties, and not all farmerscan afford them. Ignorance of newtechnology and lack of credit arefurther constraints to farmers’adopting new technology, requiringincreased extension.

For the cassava sector andagriculture in general to develop inharmony, cassava production,processing, and marketing must becoordinated, and a favorable

88


criteria for cassava production andstarch processing. To satisfy futurerequirements for cassava starchquantity and quality, manyimprovements in production,processing, and marketing should bemade. Traditional processing unitsmust invest in modern processingplant and improve their efficiency,that is, have higher conversion ratiosand better starch quality.

References

Binh, P. T.; Hung, N. M.; Tru, L. C.; andHenry, G. 1993. Socio-economicaspects of cassava production,marketing and rural processing inVietnam. Draft for: Howeler, R. H.(ed.). A benchmark study on cassavaproduction, processing andmarketing in Vietnam: proceedings ofa workshop held in Hanoi, Vietnam,Oct. 29-31, 1992, to present anddiscuss the results of a nation-widesurvey conducted in 1991-1992.Vietnam Ministry of Agriculture andFood Industry (MAFI) and RegionalCassava Program for Asia, CIAT,Bangkok, Thailand. p. 113-158.

Henry, G.; Binh, P. T.; Tru, L. C.; andGottret, M. V. 1993. Cassavaconstraints and opportunities inVietnam: a step toward a commonR&D agenda. Working documentno. 128. CIAT, Cali, Colombia.

Howeler, R. H. (ed.). 1996. A benchmarkstudy on cassava production,processing and marketing inVietnam: proceedings of a workshopheld in Hanoi, Vietnam, Oct. 29-31,1992, to present and discuss theresults of a nation-wide surveyconducted in 1991-1992. VietnamMinistry of Agriculture and FoodIndustry (MAFI) and RegionalCassava Program for Asia, CIAT,Bangkok, Thailand. 284 p.

Thang, N. V. 1993. Cassava in Vietnam: anoverview. Draft for: Howeler, R. H.(ed.). A benchmark study on cassavaproduction, processing andmarketing in Vietnam: proceedings ofa workshop held in Hanoi, Vietnam,Oct. 29-31, 1992, to present anddiscuss the results of a nation-widesurvey conducted in 1991-1992.Vietnam Ministry of Agriculture andFood Industry (MAFI) and RegionalCassava Program for Asia, CIAT,Bangkok, Thailand. p. 12-33.

Statistical Year Book of Vietnam. 1993.Hanoi, Vietnam.

89

Cassava Flour Processing and Marketing in Indonesia

CHAPTER 12

CASSAVA FLOUR PROCESSING AND

MARKETING IN INDONESIA

D. S. Damardjati*, S. Widowati*,T. Bottema**, and G. Henry***

Abstract

In Indonesia, cassava is the fourthmost important food crop after rice,maize, and soybeans. An average of16 million tons of cassava is producedannually, most of which goes to starchextraction or is exported as pellets andchips. But markets are unstable andfarmers have few incentives to producemore cassava.

Processing for cassava flour beganin 1990 to diversify cassava products.The cassava flour agroindustryconsists of three major productionsystems: at the farm (model 1),farmers’ groups (model 2), and the mill,which acts as a nucleus by linkingfarmers and farmers’ groups, throughtheir fresh roots and dried chips, withcassava flour distributors andconsumers (model 3). Processingcapacities of the three systems duringharvesting are about 75 kg of roots perday for model 1, 500 kg for model 2,and 10,000 kg for model 3. Yieldrecovery of 25% to 30% has beenobtained for dried chips, and 24% to29% for cassava flour.

Small farmers and farmer groupsreceive increased added value byproducing cassava chiplets (sawut)instead of gaplek (dried cassavachips). Marketing, however, is still amajor constraint for the cassava flouragroindustry.

A consumer-acceptance study,conducted in the Purwakarta regionand Ponorogo district, showed thatabout 80% of cassava flour wasconsidered acceptable. About 84% ofconsumers thought the flour wasacceptable for householdconsumption, which was estimated at4-7 kg/month per household.Because cassava flour can substitutewheat flour in wheat-based productsby as much as 30%, the entire localproduction of cassava flour can beabsorbed, especially by the foodindustries. PT Mariza, a privatecompany, has begun industrialproduction of cassava flour and isdeveloping a marketing system.

Introduction

Agriculture is an importantcomponent of the Indonesianeconomy, providing 49% of totalemployment, and about 18% of thegross domestic product (GDP). Foodcrops alone represent 62% of the GDPfrom the agricultural sector, that is,12% of the total GDP (1991 figures).

* Bogor Research Institute for Food Crops(BORIF), Bogor, Indonesia

** Center for Research and Development ofCoarse Grains, Pulses, Roots and TuberCrops in the Humid Tropics of Asia and thePacific (CGPRT), Bogor, Indonesia

*** Cassava Program, CIAT, Cali, Colombia.

90


Rice is the chief food crop, followedby soybeans, maize, and cassava.Other secondary crops are peanuts,mung beans, and sweetpotatoes.Minor crops with a potential future infood production are several cereals,legumes, roots, and tubers.

Cassava is grown on about1.4 million hectares throughout thecountry, with an average production of16.3 million tons/year (CBS, 1992).The crop is used as food and forstarch extraction, and is exported asfeed (Damardjati et al., 1992). Thecassava market overall, however, isunstable and tends to discouragefarmers from producing cassava.

Recently, the Government beganpromoting cassava for industrialpurposes and food. A preliminaryeconomic analysis indicated thatcassava flour production and use inprocessed foods would be feasible. Aseries of investigations on cassavaflour production and use was thereforebegun by several research institutesoperating under the auspices of theCentral Research Institute for FoodCrops (CRIFC).

The CRIFC then collaborated withprivate companies to develop acassava-flour agroindustry model atthe village level in several locations inIndonesia (Damardjati et al., 1992).

This paper presents the results ofour study on the development of thecassava-flour production system,consumer acceptance, and themarketing of cassava flour.

Cassava Production,Consumption, and Use

Production

In Indonesia during the past decade,the harvest area has decreased whileboth productivity and the number of

cassava-growing regions haveincreased. (Less than 10 years agoabout 65% of total production camefrom Java alone [Dimyati andManwan, 1992]). Factors causing thereduced harvest area are complex:incentives and sharp pricefluctuations have induced farmers togrow cassava, but factors such asestablishing irrigation facilities andreforestation have reduced plantingarea.

The average yield per hectare ofcassava is rather low at 12 t, but thetrend has been toward a constantincrease in yields. A much higheryield can be obtained throughimproved cultural practices. On theestate of a tapioca plant in Lampung,a yield of 25-30 t/ha of cassava hasbeen continuously attained(Rusastra, 1988) as a result of acassava intensification programstarted by the Government in 1975(Dimyati and Manwan, 1992).Moreover, research findings suggestthat yields can be as much as75 t/ha.

Between 1978 and 1992, cassavaproduction fluctuated, with a peak at17.1 million tons during the late1980s. In 1992, about 16 million tonswere produced (Figure 1).

1978 80 82 84 86 88 90 92

Su

pply

(million

ton

s)

30

25

20

15

10

5

0

Year

Figure 1. Domestic supply of rice ( ), cassava ( ),and wheat ( ) in Indonesia, 1978-1992.(Taken from CBS, 1978-1992.)

91


Domestic demand and consumption

Annual per capita consumption ofcassava as food has been decreasinggradually from about 57 kg in 1983 to51 kg in 1988 (Damardjati et al.,1990). From 1988 to 1992,consumption of cassava and othersecondary crops fluctuated, showing acorrelation with rice consumption(Table 1).

Cassava also became an importantexport crop. Since 1982, Indonesiahas exported a yearly quota of gaplekand pellets to the European Union(EU), as well as to other countries.Since 1989, Indonesia has alsoregularly exported tapioca starch(Table 2).

Cassava availability for domesticfood consumption is related to thetotal export of all cassava products(chips, pellets, and starch). Forexample, domestic consumption ofcassava in 1990 was lower than thatin 1992, because exports were higherin 1990.

The consumption of cassava as afoodstuff is concentrated in Java, thusmaking demand more stable andeasier to estimate. Outside Java, inareas such as Lampung, theinternational market and industrialactivities influence demand.

Processing and use

According to the Indonesian foodbalance sheet for 1991 (CBS, 1992),total cassava production in 1991was almost 16 million tons. Of this,almost 57% was consumed as bothfresh and processed food, 21% wasprocessed into gaplek and pellets, ofwhich 41% were exported and 59%went to industry. Tapioca starchwas produced from about 8% ofharvested cassava, mostly for export,but, if sold on the domestic market,also for making krupuk. The rest isused in other food, textile, paper,glucose, and pharmaceuticalindustries. Postharvest losses arestill relatively high at 13% (Table 3).

Cassava use in Indonesia differsthroughout the country. In Java,where 60% of the population resides,cassava is primarily for humanconsumption. Unnevhr (1990)reported that the rural dwellers,producers, and major consumers ofcassava use about 62% of roots and49% of the gaplek they produce fortheir family needs.

Currently, the demand forcassava in Indonesia seems to havereached a plateau, but expertsanticipate an increase in domesticdemand during the next decade forboth food and industrial purposes.

Table 1. Average per capita consumption of major food crops in Indonesia, 1986-1992.

Crop Consumption

1986 1988 1990 1992

kg/year cal/day kg/year cal/day kg/year cal/day kg/year cal/day

Rice 147.36 1,453 150.18 1,481 150.05 1,480 147.91 1,459Cassava 51.44 154 51.00 154 43.07 129 57.40 172Sweetpotato 11.05 32 10.93 32 9.74 28 10.34 30Wheat 5.96 60 6.59 60 7.54 75 10.36 104Maize 29.25 256 30.75 256 29.68 260 34.63 303Soybean 8.80 80 9.45 80 10.72 97 12.57 114

SOURCES: CBS, 1988; 1990; 1992; 1994.

92


Table 2. International trade in Indonesian cassava (t), 1983-1992.

Year Exports Imports

Chips Starch Chips Starch

1983 358,346 1,602 - 63,8831984 365,161 183 - 31985 343,303 107,000 - 211986 424,600 - 165,000 20,5001987 783,776 116,000 41,750 9,5001988 825,000 - 250,000 23,0001989 834,000 282,000 - -1990 697,000 487,000 - -1991 494,000 317,000 - 12,0001992 372,000 135,000 79,000 34,000

SOURCE: CBS, 1992.

The Indonesian Government isattempting to develop the potential ofcassava flour as a food for domesticconsumption and as a raw material forboth household consumption and thefood industry to complement orsubstitute wheat flour. TheGovernment has recommended thatthe agricultural and industrial sectorsmake special efforts in promotingcassava by diversifying cassavaprocessed products, improving their

Developing of AppropriateTechnology for Cassava Flour

Developing diversified uses for cassavaand the appropriate technology shouldextend the market and strengthenfarmers’ bargaining power in that theywould have other buyers available andcould command better prices. Newalternatives for cassava use should besimple and easy, and give added valuedirectly to the farmers.

Table 3. Trends of cassava production and use in Indonesia (thousands of tons).

Item Fresh cassavaa

1986 1988 1989 1990 1991 Percentageb oftotal production

Total production 13,312 15,471 17,117 15,830 15,954 100Losses 1,572 2,011 2,225 2,058 2,074 13Feed for domestic use 242 309 342 317 319 2Roots for chips 4,281 3,900 3,336 21

Total chips produced (1,540) (1,403) (1,200)Exports (424) (825) (834) (697) (494)

Tapioca starch 1,150 1,881 1,232 8Domestic use (322) (527) (345)Exports (282) (487) (357)

Food consumption 8,573 8,863 9,119 7,674 8,993 56Food industry 5,431 5,781 4,568

a. Values in parentheses indicate dried cassava.b. Percentages are rounded.

SOURCES: CBS, 1988; 1990; 1991; 1992; 1993.

93


Peeling Washing Chippingand soaking

Pressing Slurry

Peel Drying Steeping170 kg

Drying Chips Drying330 kg

Feed Fiber Milling Starch30 kg

Cassava Foodflour industry

310 kg

Packaging

quality, and promoting their useamong the different strata of theIndonesian population.

Several research institutes, thepublic sector, and private companieshave developed machine prototypes forcassava processing and new recipes forpreparing food using cassava products,and have promoted the use ofcomposite cassava-wheat flour inpreparing foodstuffs such as breads,pasta, and cookies.

The use of a model of anagroindustrial system based oncassava flour production wouldsupport efforts to transfer technologyfrom researchers to farmers, who couldthen commercialize the system. Thesystem would then be supported by acontinuous distribution and marketingsystem.

Developing the processing operation

To produce cassava flour, roots arepeeled, washed, chipped, pressed,dried, ground or milled, and then

sieved (Damardjati et al., 1992). Thevillage distribution and processingsystem commonly used for handlingagricultural products involves threetypes of processor groups:(1) individual farmers, (2) farmers’groups, and (3) groups of village unioncooperatives, known as Koperasi UnitDesa (KUDs), processors, millers, andcassava flour producers. Figure 2shows the overall cassava processingsystem, with three tradable productsproduced during the processing offlour from harvesting to marketing:(1) roots, (2) dried cassava chips, and(3) cassava flour.

Models for cassava flour production

The KUD or other entrepreneur groupis appointed as the nucleus processorresponsible for cassava-flourprocessing and marketing. Threemodels of cassava flour production canbe derived from the overall pattern ofcassava distribution and marketingand transaction products. The modelsare individual farmers (model 1),farmers’ groups (model 2), and cassava

Figure 2. Model of cassava processing system in Indonesia. Ovals indicate tradable products.

Cassavaroots

1,000 kg

94


flour plants that act as nuclei that linkfarmers and farmers’ groups, by buyingtheir fresh roots and chips and millingthem, with distributors and consumers(model 3) (Damardjati et al., 1992).

Developing the Agroindustryfor Cassava Flour Processing

With finance from its own projectbudget, the ARM-AARD project, andthe Government Corporation (PT PetroKimia Gresik), SURIF undertook todevelop and establish a project todevelop a cassava-flour agroindustryfor village farmers. PT Petro KimiaGresik acted as the nucleus company.

Site selection

Several sites were chosen for theproject, based on the following criteria:

(1) Desire to participate, andparticipation, on the part offarmers (or farmers’ group) and theKUD or private company asprocessor in the production ofdried chips and cassava flour.

(2) The village or subdistrict selectedmust be a cassava productioncenter at the district or provinciallevel.

(3) Readiness of the KUD or processorto produce cassava flour andcollect dried chips from farmers ortheir groups.

(4) Sufficient infrastructure such astransport and marketing facilities.

Infrastructure needed to implementproject

Implementing the project broughttogether all aspects of production,processing, and marketing ofcassava chips and flour. The mostappropriate available technology wastested in real-life situations. Table 4shows the space and installationsrequired for a village-level processingoperation.

Problems encountered werevariable product quality infarmer-processed chips; inadequatequality control for both chips andflour; inadequate communicationbetween farmers and processors;difficulties in developing handlingsystems for poor quality chips; anddifficulties in establishing the socialorganization needed for processesthat produce marketable products.Operational research directed towardresolving these problems was anessential component of the project.

Supply and sorting area. Freshroots are supplied to this area andhigher quality roots separated forprocessing. This area is mostimportant at the processor level whereit is a part of the processing area. Theroots are weighed, graded, grouped,

Table 4. Infrastructure required by three models of the cassava chip and flour agroindustry, Indonesia.

Infrastructure Model 1 Model 2 Model 3(individual farmer) (farmers’ group) (processor as nucleus)

Supply and sorting area 4-8 10-16 20-30(m²)

Processing area (m²) 10-15 30-40 300-500

Drying area (m²) 20-30 80-100 800-1,000

Storage room for chips 8-12 20-30 200-300(m²)

Storage room for flour - - 1,000(in tons of flour)

95


and sequentially processed accordingto their time of harvest. Ideally, allroots should be processed no laterthan 24 h after harvest.

Processing area. In models 1and 2, roots are peeled, washed,soaked, chipped, and pressed in thisarea. In model 3, milling and packingare also done here. The area musthave a brick or plastered platform withsufficient slope to provide gooddrainage and easy cleaning. In EastJava, the size of the processing area formodel 3 is about 4 x 10 m. In model 1,processing is usually done in thebackyard or kitchen. Some model 2systems have an area set aside,usually in a group leader’s house or“office.”

Drying area. Sun drying is themost appropriate and cheapest methodfor all three models to dry chips. Thedrying area must be completelyexposed to the sun, with a smallshaded area where workers can spreadwet chips on to trays before movingthem into the sun. The area iscompleted with a wooden or bamboorack to hold the trays during drying.Ideally, trays are 0.8 x 2 m and cancarry 7-10 kg of wet chips, dependingon the weather. Model 3 has a dryingarea capacity of 6,000 kg.

Storage room for chips. Oncedried, chips are packed and stored in aroom. The platformed floor is coveredwith wood or bamboo to protect thechips from direct contact with theconcrete or brick floor. In model 3, thestorage room is also used to keepproducts collected from farmers orfarmers’ groups. The chips are storeduntil milled or sold. In models 1 and2, no special area is set aside forstoring chips, which are stored withother field produce in the centralhouse.

Storage room for flour. The floorarea for storing cassava flour is about

the same as for chips. Cassava flouris more compact than chips, andtherefore requires less space. Cassavaflour is stored only for short periodsbefore being sold.

Processing procedures

The procedures followed by the villageplant to process cassava flour are roothandling, peeling, washing andsoaking, chipping, pressing, drying,milling, and packing.

Root handling. Thecharacteristics and quality of theeventual cassava products influencethe way roots are handled by farmers.Root handling includes time andmethods of harvesting, transport fromthe field, and storage. For a goodquality product, roots should beprocessed in less than 24 h afterharvest.

Peeling. Roots are peeledmanually with a knife or traditionalpeeler, usually by women. Peeledcassava yield is about 70%-80%, thatis, 15-20 kg/ha per person.

Washing and soaking. Peeledcassava is washed thoroughly, thensoaked overnight (for high-cyanogencultivars), or for a few minutes(low-cyanogen cultivars) while waitingto be chipped. Soaking should be donein excess water to inhibit browning andreduce cyanogenic (HCN) potentialwhere necessary.

Chipping and pressing. Peeledand soaked roots are then chipped into0.2-0.5 x 1-5 cm chips. The wet chipsare then placed on a tray and pressedwith either a screw or hydraulic press.Pressing reduces moisture, dryingtime, and HCN content, especially forhigh-cyanogen cultivars. It is optionalfor low-cyanogen cultivars.

Drying. Pressed chips are spreadout on a bamboo or aluminum tray,

96


which is put on a rack in directsunlight. Pressed chips take between14 h and 2 days to dry, whereasunpressed chips take 2-3 days. Thefaster the drying, the better the qualityof chips.

Packing and milling. Driedchips are packed in double plasticbags, which are tightly sealed.They can then be stored for about6 months. Chips from models 1 and 2are taken to the processor, whereas inmodel 3 the chips are milled to flour,using a 60-80 µm mesh. Usually theflour is packed into thick plastic bags(0.5-1.0 kg) or into double sacks(25 kg).

Implementing the agroindustrialmodel for cassava flour production

Processing operation system.The agroindustrial model follows a“foster-parent” system in which thebig Government-owned corporationsare appointed as “foster parents.” The“foster-parent” company wassupported with equipment andtechnical skills through collaborationwith research and developmentinstitutes. For example,SURIF/CRIFC and PT Petro AnekaUsaha collaborated in founding anagroindustrial system for cassava flourin Ponorogo district.

At the village level, anagroindustrial model has beendeveloped in which the farmers orfarmers’ groups produce dried chipsas an intermediate product. The“foster parent” is the milling plant,which produces cassava flour. In thissystem, farmers produce dried chipstwo to three times a week, dependingon their capacity and the weather.The farmers or farmers’ groups pooltheir dried chips before sending themto the cassava flour plant.

The plant mixes the chipscollected from this source with those

from other sources at a ratio of 60% to40%, respectively. These mixed chipsare either suitably stored or milled fordistribution, or sent to the distributor.

This system has advantages forboth the plant and the farmers. Theplant guarantees that all chipsproduced by the farmers will beaccepted and bought. Advantages forthe plant are that its equipment,especially the mill, operates at optimalcapacity; it obtains, indirectly, dryingareas from farmers and farmers’groups; labor efficiency is optimized;and plant operational time is longerduring the cassava off-season becauseof its stock of dried chips.

Material, energy, and productioncost analysis. Material conversionvalue in cassava flour processing isinfluenced by root size and peel,cassava variety, and equipment used.Large, easily peeled roots mean higheryields. Table 5 indicates the materialconversion in each processing step(Damardjati et al., 1991). The averageyield of dried chips is 34% and of flour,32%. Screw pressing results in aslightly higher yield than hydraulicpressing. The root cyanogenic potentialstrongly influences yield recovery ofdried starch. Normally, high-cyanogencultivars contain more starch than dolow-cyanogen cultivars.

At the time of the study, laborwages were 2,500 rupiahs/day(exchange rate was Rp 2,126 =US$1.00). Based on theyield recovery of flour (32%), the totalproduction cost of cassava flour was18,725 rupiahs/kg.

Economic analysis

Price determination. Standardprices are an important factor in thecassava flour agroindustry, and areusually higher than those of slicedgaplek, which are unstable and dependon middlemen. Farmers have no

97


bargaining power. Gaplek pricesalways decrease, especially during peakseason. Farmers producing dried chipsalso face the same problem as dogaplek farmers. For example, onefactory buys cassava chips fromfarmers at a higher price (Rp 50/kgmore) than the highest price for gaplek.Another factory, however, uses a tablebased on root prices.

Farmers do not readily accept thesemethods of determining prices. SURIF

and the “foster-parent” factorycollaborated to determine a standardprice for dried chips, which, in 1992,was 270 rupiahs/kg of chips.

Added value for farmer.Traditionally, farmers in Ponorogodistrict processed cassava roots intogaplek. Table 6 compares the addedvalue of chips with that of gaplekfor farmers, showing that theadded value for farmers was2,175 rupiahs/100 kg of roots.

Table 6. Added value of cassava chips compared with that of gaplek for farmers, Ponorogo district,Indonesia, 1992a.

Item Costs in rupiahsb

Dried chips Gaplek

Labor costs:Peeling 1,000 1,000Chipping and drying 1,000

Equipment hire (Rp 10/kg dried chips) 300Total costs 2,300 1,000

Product price 270 125Income from product 8,100 5,625Economic profit 5,000 4,625Economic income + wage 7,800 5,625Added income 2,175 0

a. Calculations based on 100 kg of roots, gaplek yield at 45%, chips yield at 30%, 1992 prices of gaplek atRp 125/kg and sawut (chiplets) at Rp 270/kg, and equipment hire for dried chips at Rp 10/kg.

b. Exchange rate: Rp 2,126 = US$1.00 (January 1994).

Table 5. Yield recovery in cassava flour processing calculated from 500 kg of fresh cassava.

Form of cassava Number of Processing Averagesamples recovery (%) conversion value

Min. Max. (%)

Peeled roots 15 73 83 80Soaked and peeled roots 12 74 88 82Wet chips before pressing 15 70 88 80Pressed chips:

Screw press 6 61 68 65Hydraulic press 6 61 66

Dried chips: 34Screw press 6 29 37Hydraulic press 6 22 37

Flour 15 30 34 32Dried starch (byproduct) 9 2 5

SOURCE: Damardjati et al., 1991.

98


Consumer Acceptanceand Marketing

Survey on consumer acceptance

Consumer acceptance and marketassessment studies were conductedin two locations with differentconsumer characteristics:Purwakarta region and Ponorogodistrict, both in East Java. Theinhabitants of Purwakarta andKarawang districts in thePurwakarta region do not producecassava and eat it infrequently. Incontrast, Ponorogo district is amajor cassava-producing area wherethe inhabitants eat cassava as thesecond staple after rice.

A survey was first carried out on115 families in Purwakarta regionand 124 in Ponorogo district todiscover the acceptability of cassavaflour, its use, and consumerpreferences. In both areas, morethan 80% of respondents did notknow of the product but, when itwas introduced, received it well(more than 84% of respondents).

More than 50% of respondentsfrom both areas used the cassavaflour to make traditional foods.Respondents from Purwakartaregion also tended to like cookiesand cakes (Table 7), whereas thosefrom Ponorogo district tended toprefer traditional foods. One reasonfor the difference may be location:the Purwakarta region is larger andmost respondents were moreeducated and skilled in foodproducts.

Consumers’ use of cassavaflour. Cassava flour can be used asa substitute flour in wheat-basedproducts. For householdconsumption, most respondentspreferred to process it into either(1) traditional foods, (2) cakes,(3) cookies, or (4) krupuk, a

cracker-like product. Table 8 showsthe basic ingredients and processing.

More than 50% of respondentsused cassava flour to bake traditionalfoods because they were simple toprepare, were familiar, and the otheringredients readily available. Therespondents’ different income levelswere reflected in the differentpreferences for food types preparedfrom cassava flour (Table 9). About43% of high-income consumerspreferred to process cassava flourinto cake as compared with 21% ofmedium- and 29% of low-incomeconsumers who tended to prefertraditional foods. Table 10 givesexamples of traditional foods inwhich cassava flour can be used as asubstitute for other flours.

Consumer acceptance forlong-term consumption. The surveyalso assessed consumer acceptanceof and the kind of food productsmade from cassava flour in the longterm. Most respondents used asmuch as 50% cassava flour mixedwith another flour such as wheat,tapioca, or rice for traditional foods(e.g., bala-bale, fried banana, andputu ayu) and cakes. For cookiesand krupuk, relatively little was used(Damardjati et al., 1992), although asmuch as 60% substitution withcassava flour, resulting in goodquality cookies, has been reported(Damardjati et al., 1992). Cassavaflour and tapioca (cassava starch),mixed at a ratio of 1:3, respectively,have been used in krupuk, resultingin an acceptable product (Suismonoand Wheatley, 1991).

In a cooking trial, carried out by115 respondents, cassava flour wasaccepted by about 84% and rejectedby about 15%. The respondents’average demand for cassava flour isabout 5-7 kg/month. The group thatpurchased the most had a mediumincome. Most consumers would

99


Table 7. Consumer preferences for cassava products in Purwakarta region and Ponorogo district (%),Indonesia.

Product Consumers in Consumers inPurwakarta regiona Ponorogo districta

Like Dislike Like Dislike

Traditional foods 57.0 3.3 59.7 12.1Cookies or crackers 13.1 0.9 8.9 0.8Cakes 37.8 1.8 9.7 8.9

a. Number of households surveyed was 115 in Purwakarta and 124 in Ponorogo.

SOURCES: Damardjati et al., 1993; Martini, 1992.

Table 9. Consumer preferences (%) among foods prepared from cassava flour, Purwakarta region andPonorogo district, Indonesia. Samples given during a consumer-preference survey.

Respondent group Processed products

Traditional foods Cookies Cakes Krupuk

Purwakarta regionIncome group:Low (n = 39) 67.6 16.2 29.7 2.7

Medium (n = 46) 75.6 12.2 21.9 2.4

High (n = 30) 53.3 13.3 43.3 -

Ponorogo district:Urban (n = 57) 66.7 17.5 15.8 -Village (n = 67) 76.1 1.8 24.6 -

Table 8. Processing and ingredients of products processed from cassava flour, Indonesia.

Type of ingredient Traditional Cookies Cakes Krupukand process foods

Basic Wheat flour Wheat flour Wheat flour Tapioca flouringredients Rice flour

Additional Margarine Margarine Margarine Cane sugaringredients Eggs Eggs Eggs

Cane sugar Cane sugar Cane sugarVegetablesCoconut milk

Other Salt Leavening Leavening Saltingredients Artificial coloring Flavoring Artificial Spices

flavoring Flavoring

Process Steamed, fried, or Oven-baked Oven-baked Steamed beforeroasted frying

100


Table 10. Traditional food products made with cassava flour as substitute flour, Indonesia.

Local name Cassava Other Brief descriptionflour (%) flour

Bala-bale 50 Wheat Mixture of flour, water, vegetables, and spices. Fried.

Cimplung 50 Wheat Mixture of flour, water, sliced jackfruit, and salt. Fried.

Nagasari 70 Maize Mixture of flour, coconut milk, sugar, salt, vanilla, maize flour, andcooked. Filled with sliced banana and wrapped in banana leaf.Steamed.

Jongkong 50 Rice Dough of flour mixed with coconut milk and salt, and cooked.Filled with sliced palm sugar and thick coconut milk. Wrapped inbanana leaf. Steamed.

Ongol-ongol 65 Wheat Mixture of flour, water, and sugar, and cooked. Formed, cooled, andsliced. Served with grated coconut.

Dodongkal 100 - Cooked flour with water and salt. Dough filled with shredded palmor awug sugar. Served with grated coconut.

Biji salak 100 - Small balls made from flour dough and cooked. Served with sweetcoconut milk and sliced jackfruit.

Bika Ambon 35 Rice Two mixtures of flour, egg, “fermipan,” and coconut water. Onemixture worked into a dough. Other mixture cooked with sugar andcoconut milk until oily. The two mixtures then combined and baked.

500

400

300

200

100

0Jan. Mar. May July Sept. Nov.

Ru

pia

hs

per

kg

250

200

150

100

50

0Jan. Mar. May July Sept. Nov.

Ton

nag

e

Figure 3. Trends in purchasing chips and sellingcassava flour by PT Mariza, Indonesia,1991. ( = purchasing; = selling.)

process cassava flour into traditionalfoods (41.7%) and cakes (21.7%).

Marketing problems. Duringthe several years of establishing anddeveloping the cassava flouragroindustry, marketing was the firstproblem faced. Cassava flour wasunknown and the market had to bedeveloped. Farmers and farmers’groups depend heavily on a mill to actas nucleus for collecting and buyingchips. The mill sells mostly to foodindustries. But the market forprocessed cassava products is small,with the consequence that the plant(nucleus) becomes overstocked inchips and flour. Operationalmanagement also becomes aproblem.

One company which hasexpanded its cassava processingoperations is PT Mariza, a foodcompany that produces snack foodsand cakes. This company hasincreased its monthly output ofcassava flour by over 200% since1991 (Figures 3 and 4).

Figure 4. Trends in prices of cassava roots andcassava flour in transactions carriedout by PT Mariza, Indonesia, 1992.( = flour; = roots.)

101


Future Development of theCassava Flour Agroindustry

Interrelationships amongdeterminants in the agroindustry’sdevelopment

The main interacting participants inthe cassava flour agroindustry are(1) farmers or farmers’ groups,(2) KUD or processor, (3) factoryand distributors, and (4) consumers(Table 11). The four determinantsare (1) policy, (2) infrastructure,(3) participation, and (4) technology.

Farmers or farmers’ groups.Basic policy needs to be constructed atfarmer level. An example isdetermining a suitable floor price forchips, to motivate farmers inprocessing. Small farmers shouldobtain credit from banks throughsimple procedures and flexibleguarantees. They should also receiveextension and training in skills fordeveloping both the cassava flouragroindustry and its processedproducts.

KUD or processor. The KUD orprocessor also expects simpleprocedures and flexible guaranteeswhen obtaining credit from banks. Toensure the agroindustry’s continuity atvillage level, each small industry’sshare of production should beprotected. Chip prices should also besufficiently competitive.

The village KUD or processor has arelatively low management capability,needing guidance in its operationalmanagement. Simple and easilyinstalled equipment with locallyavailable spare parts is to be preferred.The demand for such equipment willprovide opportunities for localworkshops.

Industry and distributor. Tospeed up distribution of cassava flourto consumers the market share of the

flour at distribution level needs to bedetermined through the National FoodAuthority (BULOG). A 10% share fromthe total distributed by BULOG willsuffice to warrant cassava flourmarketing. The increased agroindustrywill speed up the development ofcassava products, the prices of whichwill be heavily influenced by theirquality.

Consumers. Even with BULOG’sintervention in distributing cassavaflour, distributors and retailers areresponsible for making it readilyavailable to consumers. Threemethods are to undercut the prices ofother flours, promote through massmedia, and encourage food and cateringindustries to increase their use ofcassava flour as a raw material inprocessing food.

Supporting activities

To support efforts in developing theagroindustry, certain governmentalpolicies are urgently needed. Thesewould help improve quality; create aproduction environment advantageousto processors, distributors, andfarmers; and change consumerattitudes toward cassava. Thesepolicies are:

(1) Price and distribution policies. Priceand distribution policies for cassavaproducts (cassava flour and chips)of defined levels of quality canencourage increased productionand improved product quality. Theensuring of cassava raw materialsupplies requires an establisheddistribution mechanism.

The continuous distribution ofcassava flour throughout the yearwill encourage farmers to increasecassava production and ensure acontinuous supply for processors.BULOG is expected to play animportant role in the distributionsystem, which will then develop

102

Ca

ssava

Flou

r an

d S

tarch

: Progress in

Resea

rch a

nd

Develop

men

t

Table 11. Matrix correlation of determinants, by participating groups, in the cassava flour agroindustry, Indonesia.

Determinant Subject

Farmers or farmers’ group KUDa or processor Industry Consumer

Policy - Basic price of chips - Credit ensured - Cassava flour marketing at - Good distribution through- Easy credit - Assurance flexibility distributor level, BULOGb BULOGb intervention- Credit ensured - Protection of production share controlled - Ceiling price- Assurance flexibility - Basic price of cassava flour - BULOGb as “foster-parent,” but

no market monopoly

Infrastructure - Extension - Guidance in operation - Facilities for promoting - Promotion through mass- Training management processed cassava products media- Credit for chipping - Credit for equipment and - Facilities for credit - Improvement role of service

equipment operational costs - Products from cassava flour- Market implementation - Market information

Participation - Price expectation - Price expectation - Continuity of chips and cassava - Competitiveness of taste- Processing efficiency - Processing efficiency flour supplies and flour packaging- Added value in processing - Standard quality of chips and - Attractive packaging

flour - Can be mixed with other flour- Export promotion

Technology - Simple and easily installed - Simple and easily installed - Efficient - Serving techniqueequipment - Spare parts available

- Spare parts available - Labor intensive- Relatively cheap

a. Koperasi Unit Desa (village union cooperative).b. BULOG = National Food Authority, Indonesia.

SOURCE: Adnyana et al., 1991.

103


through the self-supportive natureof the existing market.

(2) Support for industrial and exportdiversification. Developing cassavaflour processing plants is the mainstep toward supporting thedevelopment of cassava industriesin general. The consistent demandfor raw material for cassava flourproduction has already increasedat farmer level. To develop, theseindustries need support inproviding an environment that willattract investors.

Quota restrictions have alreadyseverely limited the possibilities ofincreasing the traditional export ofdried cassava (chips and pellets).But export volumes can beexpanded for nontraditionalcommodities such as fructose,sweets, sorbitol, and modifiedstarch. Various food products withpotential as export commoditieshave already been formulated withcassava flour as raw material.

(3) Extension should be aimed atvarious levels of the community:farmers and their families,processors, and other groups.Extension materials should bestructured according to eachtargeted level of the community.

(4) Campaigns and promotion. TheGovernment can help change thecommunity’s attitudes towardcassava through such activities aspromotion, extension, expositions,and cooking festivals.

(5) Community uses of cassava flour.Catering services and bakerieswould be the major consumersof cassava flour, especially bysubstituting for wheat flour intheir products. Other promotersinclude governmental andsemigovernmental organizationssuch as KORPRI, Dharma Wanita,

and Dharma Pertiwi;nongovernment organizations,including social and professionalorganizations; and the mass media.

Conclusions

Developing the cassava flouragroindustry represents for Indonesiaan alternative for diversifying cassavaproducts. It can potentially increasefarmer incomes, extend marketing,support food diversification, reducewheat imports, and contribute to thedevelopment of various chemical andfood industries. Cassava flourprocessing requires the development oftechniques and equipment for peeling,washing, soaking, drying, chipping,pressing, and milling.

The cassava flour agroindustry canbe structured on three models,according to capital, capability,knowledge, and distribution systems ofthe raw material. These models arebased at farmer level (model 1); farmergroup level (model 2); and mill or plantbelonging to a group of privatecompanies or cooperatives as a nucleusin the processing and marketingsystem (model 3). The mills act asprocessors of intermediate products,that is, dried chips, from models 1 and2 to be processed into cassava flour asfinal product. An economic feasibilityanalysis showed that a cassava flouragroindustry is feasible at the villagelevel when it is based on the threemodels being structured into a system.

Cassava flour can be processed intofour groups of food product: traditionalfoods, cookies, cakes, and krupuk. Thehigher the income and education of thehousehold mother, the more likely thatcassava flour will be accepted for use intraditional foods and cakes. As manyas 84% of consumers would acceptcassava flour, and most of these couldbuy 4 to 7 kg of cassava flour permonth. With promotion and improved

104


supplies, the flour therefore has a highpotential to develop a niche in urbanmarkets, especially as supplement tocereal flours.

Marketing is still a majorconstraint to expanding the cassavaflour agroindustry. The PT Mariza’ssuccessful expansion was supportedby its ability to diversify its productsand markets. Governmental supportand policy making is still necessary tocreate a favorable productionenvironment and improve quality atevery step of the production system tomatch market demands.

References

Adnyana, M. O.; Rachim, A.; Damardjati,D. S.; and Basa, I. 1991. Potensi danKendala Pengembangan Agro-industriTepung Kasava dalam SistemUsahatani Terpadu di Lampung.Puslitbangtan, Bogor, Indonesia.

CBS (Central Bureau of Statistics). 1988. Foodbalance sheet in Indonesia, 1986-87.Jakarta, Indonesia.

__________. 1990. Food balance sheet inIndonesia, 1988-89. Jakarta,Indonesia.





Damardjati, D. S.; Seytono, A.; Widowati, S.;Suismono; and Indrasari dan Sutrisno,S. D. 1991. Lap. modelAgro-industri tepung Kasava diPedesan. I. Analisis petensi wilayahpengembangan dan penyajian pilotplant. Bogor, Indonesia.

__________; Widowati, S.; and Dimyati, A.1990. Present status of cassavaprocessing and utilization inIndonesia. In: Howeler, R. H. (ed.).Proceedings of the Third RegionalWorkshop of the Cassava ResearchNetwork in Asia, Oct. 22-27, Malang,Indonesia. CIAT, Cali, Colombia.p. 298-314.

__________; __________; and Rachim, A. 1992.Development of cassavaprocessing at the village level inIndonesia. In: Product developmentfor root and tuber crops. CentroInternacional de la Papa (CIP), Lima,Peru. p. 261-273.

__________; __________; and __________. 1993.Cassava flour production andconsumers’ acceptance at villagelevel in Indonesia. Indones. Agric.Res. Dev. J. 15(1):16-25.

Dimyati, A. and Manwan, I. 1992. Nationalcoordinated research program:cassava and sweet potato. CentralResearch Institute for Food Crops(CRIFC), Bogor, Indonesia. 61 p.

Martini, R. 1992. Study on cassava flour assubstitution ingredient on foodindustry and family level inPonorogo. S1 thesis. BogorAgricultural University, Bogor,Indonesia.

Rusastra, I. W. 1988. Study on aspects ofnational production, consumptionand marketing of cassava. Indones.Agric. Res. Dev. J. 7:57-63.

Suismono and Wheatley, C. 1991.Physico-chemical properties of the“krupuk” product on some of theformulates of cassava compositeflour. In: Suismono (ed.). Cassavaroots: characteristic, utilization andanalysis methods. CIAT, Cali,Colombia. 21 p.

Unnevhr, L. J. 1990. Assessing the impact ofresearch on improving the quality offood commodities. In: Methods fordiagnosing research systemsconstraints and assessing the impactof agricultural research.International Service for NationalAgricultural Research (ISNAR), TheHague, the Netherlands. p. 101-116.

105

World Production and Marketing of Starch

CHAPTER 13

WORLD PRODUCTION AND MARKETING

OF STARCH1

Carlos F. Ostertag*

Main Starch Sources

Starch is extracted from maize,sweetpotatoes, cassava, potatoes,wheat, rice, sorghum, sago palm,arrowroot, and bananas (AVEBE,1989; Jones, 1983). Developedcountries grow most of the world’smaize, potatoes, and wheat, whereasdeveloping countries grow most of thesweetpotatoes and cassava. Forexample, China produces almost 85%of world’s sweetpotatoes (Rhem andEspig, 1991).

These starches differ from eachother in their granule forms and sizes,contents of amylase and amylopectin(the two types of glucose polymerspresent in starches), swellingcapacities (i.e., capacities to absorbwater), and gelatinizationtemperatures (Jones, 1983).

In the early 1980s, 77% of world’sstarch was estimated to derive frommaize (Jones, 1983), mainly because91% of the starch produced in theUSA, the world’s largest producer,was from maize (Farris, 1984). Theincrease in yield per hectare, from2.4 t in 1950 to 7.6 t in 1986,contributed significantly to thiscereal’s importance (Lynam, 1987c).

Table 1 shows the relativeimportance of different starchsources. Two reasons for the

Introduction

Starch production is a major worldagroindustry, with a volume ofaround 33 million tons per year, anda value of US$14 billion (Jones, 1983;Marter and Timmins, 1992;Titapiwatanakun, 1993) (Table 1).Starch is extracted primarily fromcereals and roots through processesthat separate fiber and protein.

Demand for starch is influencedby its versatility. Almost all majorindustries use starch and, as a result,industrialization normally coincideswith a significant increase in thedemand for this raw material (Lynam,1987c).

Three main classes ofstarch-based products exist:unmodified or native starches (UMS),modified starches (MS), andsweeteners. Modified starches arethose in which one or more of theirphysical and chemical properties havebeen changed slightly (Jones, 1983).

* Cassava Program, CIAT, Cali, Colombia.


106


Table 1. Estimated world starch production (1992) (thousand tons)a.

Region or country Raw material Total Percentageof world

Maize Sweet- Cassava Wheat Potato Other productionpotato

North America: 13,450 - - 200 55 20 13,725 41USA 13,200 - - 50 50 20 13,320 40Canada 250 - - 150 5 - 405 1

Latin America 1,000 - 330 - - - 1,330 4

European Union 3,400 - - 1,400 1,200 - 6,000 18

Ex-USSR and 300 - - - 300 - 600 2Eastern Europe

Africa - - 20 - - - 20 <1

Asia: 3,020 4,165 3,442 165 400 30 11,222 34China - 4,000 300 - - - 4,300 13Japan 2,500 120 - 150 400 - 3,170 10Thailand - - 1,800 - - - 1,800 5Indonesia - - 800 - - - 800 2India 200 - 350 - - - 550 2Vietnam - - 90 - - - 90 <1Philippines 75 - 17 - - - 92 <1Malaysia - - 70 - - - 100 <1Taiwan 45 15 15 15 - 30 90 <1South Korea 200 30 - - - - 230 1

Australia 50 - - 300 - - 350 1

Total 21,220 4,165 3,792 2,065 1,955 50 33,247 10064% 13% 6% 6% 6% 0% 100%

a. Includes modified starches and sweeteners.

SOURCES: Estimates based on Jones, 1983; Marter and Timmins, 1992; Titapiwatanakun, 1993.

Holland), Japan, and Eastern Europe.China accounts for almost all of theworld’s production of sweetpotatostarch, whereas the EU, Australia,and Canada dominate wheat starchproduction (Table 1).

The maize starch produced inJapan is derived mainly fromimported U.S. maize, as used to bethe case for the EU. This regionnow locally produces 99% of itsmaize requirements for starchproduction (Leygue, 1993). Localmaize-processing capacity hasdisplaced native starch sources suchas rice, sweetpotatoes, potatoes, andcassava. For example, in Japan, in1962, 80% of the starch produced

decrease in the proportion of maizestarch (64%) are (1) the table includesestimates by Marter and Timmins(1992) of starch production in China,derived mainly from sweetpotato,which Jones excluded in the 1983estimate; and (2) Thailand has greatlyexpanded cassava starch productionrecently (Titapiwatanakun, 1993).

Production of maize starch isconcentrated in the USA, Japan, andthe European Union (EU). Asia is thechief cassava starch producer,primarily Thailand, Indonesia, China,and India; with Brazil, in LatinAmerica, also an important producer.Production of potato starch iscentered in the EU (especially

107


was derived from sweetpotatoes andpotatoes. But this share fell to 20%by 1982, displaced by U.S. maize.The main reason was that Japanesestarch was used mostly for producingsweeteners, a category for which thetechnology for maize wet-milling isvery advanced (Lynam, 1987c).

Simplified Classification ofStarch-Based Products

Starch is a versatile raw materialcompared with other carbohydrates.Native starch can be modified orchemical derivatives obtained from itby using relatively simple processes.Starch is dispersible in cold waterand has a higher reactivity than thehighly polymeric cellulose. Starch isalso highly susceptible to partial ortotal hydrolytic degradation by acidsor enzymes, yielding oligomeric ormonomeric products, which, in turn,can be further modified or used to

obtain chemical derivatives notpossible from cellulose or sucrose.Starch can also be separated intoamylose and amylopectin, and can beused in solvolysis with alcohols (Kochand Roper, 1988) (Figure 1).

A simple way to classifystarch-based products is as follows:UMS, MS (e.g., dextrins,pregelatinized starches, and oxidizedstarches), starch derivatives (e.g.,esters, ethers, and cross-linkedstarches), and sweeteners (glucosesyrups, high fructose syrups,dextrose, and maltodextrins) (Jones,1983; Koch and Roper, 1988). Starchderivatives and sweeteners are usedprimarily in the food industry.

Native starches are marketed dry,under different grades for human andindustrial consumption. Mostdeveloping countries produce onlythis type of starch, except for thosewith Corn Products Corporation (CPC)

Process Product

Separation

Mechanical/thermaltreatment

Esterification Native Etherification starch Cross-linking

Transglicosilation(solvolysis)

Hydrolysis

Figure 1. Starch-based products. (After Koch and Roper, 1988.)

* Amylose

* Amylopectin

* Modified starches

* Pregelatinized starches

* Dextrins

* Starch derivatives

* Glucosides

* Monosaccharides

* Disaccharides

* Oligosaccharides

* Maltodextrins

108


subsidiaries, which also produce MS(Jones, 1983).

Modified starches are those thathave been changed slightly in one ormore of their physical and chemicalproperties. Modification aims toadjust the product to the particularneeds of a client or to imitate acompetitive product. The range ofmodifications and processes is vastand complex but can besummarized as follows:

(1) Pregelatinized type by feeding astarch suspension on to a heatedrotating drum.

(2) Dextrin type from dry chemicaltreatment.

(3) Wet chemical treatment,including thin boiling andoxidized starches.

(4) Other modifications, includingthe use of catalysts andcross-linking and etherifyingagents (Jones, 1983).

The term “sweetener” refers toproducts such as glucose syrup,high fructose syrup (HFS), anddextrose. Sweetener production isbased on acid or enzymatichydrolysis of starch. Chemically,glucose and dextrose aresynonymous, but, commercially,“dextrose” is used to describe thepure crystalline product and“glucose syrup” products obtainedfrom incomplete starch hydrolysis(Jones, 1983).

By subsequent complexprocessing, based on enzyme action,HFS can be obtained. This producthas grown substantially inimportance, particularly in the USA,where it was introduced in 1968.Usually based on maize, it is knownas high fructose corn syrup (HFCS)in the USA and as isoglucose inEurope (Jones, 1983).

Main Starch Producers

Table 1 shows that starch productionin the USA, almost exclusively maizestarch, accounts for 41% of worldproduction. Asia has become animportant starch producer,contributing to more than one-third ofworld production. The majorproducers are China (sweetpotato),Japan (maize), and Thailand(cassava). Asian starch is extractedmainly from sweetpotatoes (38%),cassava (31%), and maize (28%).Although maize starch is from U.S.maize, the other starches areobtained from local raw materials.

No exact records of sweetpotatostarch production in China exist but,according to Marter and Timmins(1992), the volume could be about4 million tons/year. Processing is athousehold level in villages and thestarch is used primarily for makingnoodles, a traditional oriental food.

Starch consumed in Japan can becategorized as (1) starch obtainedfrom local crops such as potatoes andsweetpotatoes, (2) starch derived fromimported raw materials such as maizeand wheat, and (3) imported starches,such as cassava, sago, and potatostarches (Jones, 1983).

The EU produces nearly 18% ofworld starch, principally from maize,wheat, and potato. France aloneaccounts for one quarter of thisvolume (Leygue, 1993). Obtainingstarch from wheat flour throughwheat-washing technology (glutenseparation) has increased since 1983,with an annual growth rate of 15%.This source has displaced maizestarch to a significant extent (Leuch,1990). Starch production in the EUhas grown at an annual compoundrate of 4.4% during 1981-1990 (Kochet al., 1993).

109


Latin American starch productionaccounts for 4% of world productionand consists of maize and somecassava starches. Brazil dominatesthe production of both starches.

The starch industry mostlysupplies its own domestic markets. In1980, excluding internal EU trade,only 4% of world production(600,000 t/year) was estimated to beexported (Jones, 1983). Since then,starch production has expanded tosuch an extent in Thailand and China,that exports have more than doubled,to almost 1.5 million tons. Almost70% of Thai starch is exported, mainlyto the USA, Japan, and Taiwan(Titapiwatanakun, 1993).

International trade is concentratedin UMS, mainly from Thailand, China,Indonesia, and Brazil, and consistschiefly of cassava and sweetpotatostarches (Jones, 1983; Shuren andHenry, 1993; Titapiwatanakun, 1993).The major markets for exported starchare Japan, Taiwan, the USA, and theEU (Jones, 1983).

The recent high growth rate of thestarch industry in Thailand is worthstudying. The Thai starch industryhas confronted two limitations: first,the high tariffs for starch imports inalmost all nations except the USA; andthe second, competition with the pelletand chip export market for rawmaterial. The domestic EU price forcereals determines local root prices,which is why roots for the starchindustry are so expensive. The starchindustry, in turn, has to compete withthe low international maize prices(Lynam, 1987c).

But the establishment of exportquotas for Thai pellets to the EU in theearly 1980s lowered domestic prices ofcassava roots and led to a doubling ofcassava starch exports (Lynam,1987c). Titapiwatanakun’s study(1993) of the impact of the recent 29%

reduction in the domestic price ofcereals in the EU indicates that thefarm price of local roots will decreasein Thailand, liberating raw materialfor producing cassava starch. Thaicassava starch exports have increasedat an annual growth rate of about14% since 1975 (Atthasampunna,1990; Lynam, 1987b;Titapiwatanakun, 1993); for example:

Year Exports in tons

1975 145,0001980 248,0001985 497,0001989 646,0001991 1 million

Current Starch Markets

In most countries, starchconsumption is highly correlated withproduction. Exceptions are thosecountries where starch production isprimarily export-oriented (e.g.,Thailand and China) or where starchis imported (e.g., Taiwan and, to alesser extent, Japan).

General uses

Starch has one of the widest ranges ofend-uses of any product derived fromvegetable sources. It is a good sourceof carbohydrate but in the foodindustry it is used mainly as athickener, filler, binder, stabilizer, ortexture improver. Some examples ofthese uses are in canned andpowdered soups, instant desserts,custard powder, sausages andprocessed meat, sauces, bakeryproducts, confectionery, and icecream. Sweeteners such as syrupsare used for soft drinks, pastries, andcanned goods; this segment hasshown the most growth in the last25 years. Edible starch is also usedin the pharmaceutical and brewingindustries (Jones, 1983).

110


The industrial uses of starchand starch products are numerous.Among the most important are inthe paper and board industry(printing papers, coated papers,corrugated board), adhesives (labels,laminating, gummed paper, tape),textiles (sizing, finishing), oil-welldrilling (drilling “mud”), dye stuffs,and the building, metal, andchemical industries (Jones, 1983).

Uses in the USA

More than 95% of the starchcurrently used in the USA isobtained by wet milling maize. In1992, 48% of the wet milling outputwas destined for HFCS production,25% for glucose and dextrose, and27% for actual maize starch (USDA,1993a). These figures excludeethanol production based on the wetmilling technology but it should benoted that more than 10 milliontons of maize were used for thispurpose in the USA in 1992 (USDA,1993a).

Table 2 shows starch productionand relative weight for 1980 and1992 of the main starch-derivedproducts in the USA. The end-useas sweetener is prominent,representing more than 70% of thetotal. The high growth of the HFCSsegment can also be noted for theabsolute domination of the softdrink market for sweeteners since1985 (Claassen and Brenner, 1991).HFCS production is divided intoHFCS-55 (containing 55% fructose)with a market share of 58% andHFCS-42 (containing 42%) with42%.

The use of maize-basedsweeteners, especially HFCS, hasgrown dramatically because of theirexcellent quality, their usefulness asfunctional agents in foods, and theirlower cost versus sugar (Long,1985). HFCS is a direct substitute

for sugar in every area except drymixes or wherever a nonhygroscopicsweetener is required, as is the casefor hard candy and table sugar (Long,1985).

The end-uses for UMS and MS inthe USA (1980 data from Jones, 1983)include:

(1) Paper industry (60% of UMS and50% of MS), including for sizing,coating, and corrugation.

(2) Food industry (20% of UMS and20% of MS), including foringredients in cookies andconvenience foods (e.g., instantsoups, desserts, and frozendinners).

(3) Other important industrial usersare the brewing, pharmaceutical,and adhesive industries (20% ofUMS) and the textile industry(30% of MS).

Table 2. Production of the principal starchproducts in the USA in millions of tons(mill. t)a.

Product 1980 1992

(mill. t) (%) (mill. t) (%)

Maize starchHFCSb 1.91 31 6.00 45Glucose syrup 1.86 30 2.90 22Dextrose 0.41 7 0.60 4

Subtotalsweeteners 4.18 68 9.50 71

Unmodified 1.18 19 2.20c 16

Modified 0.68 11 1.40c 11

Other 0.06 1 0.10c 1

Other starches(e.g., wheat,potato) 0.07 1 0.12 1

Total 6.17 100 13.32 100

a. Excludes ethanol production derived from wetmilling maize.

b. HFCS = High fructose corn syrup.c. C. F. Ostertag, 1993, unpublished data.

SOURCES: 1980 data: Jones, 1983.

1992 data: Farris, 1984; Kirby, 1990;USDA, 1993a.

111


Almost 100% of the starch-basedproduction of sweeteners is destinedfor the food industry. More than70% of the HFCS production is usedby companies producing carbonatedsoft drinks; 90% of HFCS-55 is usedfor carbonated soft drinks (USDA,1993b). Other uses for sweetenersinclude pastries, canned fruit,dessert dairy products, anddressings and ketchup. Apart fromtheir sweetening properties, they arealso useful for controllinghygroscopicity, texture, freezingtemperature, and viscosity (Long,1985).

Uses in the EU

Of the 6 million tons of starchproduced in the EU, 54% is used bythe food industry and the remaining46% by the nonfood sector, includingpaper (19%), chemicals andfermentation (13%), corrugation (7%),and others (7%) (Koch et al., 1993).Other relatively new uses in the EUare for the production of ethanol,plastics, and polymers (Agra Europe,1990). The annual growth rate ofnonfood markets (4.8%) has beengreater than the global market (4.4%)(Koch et al., 1993).

Starch is consumed in its native(29%), modified (17%), and hydrolyzedforms (54%). Hydrolyzed starch isused in sweets, beverages, fruitpreparations, and pastries (Koch etal., 1993).

Of the 1 million tons of starchemployed by industry in the thenWest Germany (1987), 41% was usedfor the following product categories:adhesives, pharmaceuticals, paperand cardboard, soap, chemicals, dyes,paints, building materials, and rubberproducts (Christmann, 1991).

The use of wheat flour in nonfoodindustries in West Germany increasedfrom 1,000 to 90,000 t during

1980-1987 because of its specialchemical properties. At the sametime, the consumption of potatostarch also increased, whereas that ofmaize starch decreased (Christmann,1991). The demand for starch overallin Germany by nonfood industries inthe year 2000 is predicted to bebetween 600,000 and 800,000 t,and for the EU, between 2.5 and3.0 million tons (Christmann, 1991).

Uses of potato starch

The chief world markets for potatostarch, mostly located in the EU, arethe following industries: food, paper,textiles, and mineral oil (additives foroil-well drilling) (AVEBE, 1989). Inthe early 1980s, these uses weredistributed in the USA as follows:33% for paper, as a cationicderivative; 30% in the food industryin native or modified form forpreparing soup mixes, puddings, andsweets in general; 19% for adhesives,preferably in dextrin form; and 15%in pregelatinized form as an additivein oil-well drilling. The latter segmentexhibits the highest growth rate(Mitch, 1984).

Uses in Japan

In 1980, almost 60% of the starchproduced in Japan was forsweeteners, mainly HFCS, derivedprimarily from U.S. maize, but alsofrom sweetpotato and potato starches.Nearly 15% of the starch was destinedfor MS production, principally basedon imported maize; the main MSproduced was oxidized starch. Otherimportant uses were for paper,cardboard, and textiles (7%); fishproducts such as kamaboko (7%);beer (3%); and monosodiumglutamate (MSG), a popular flavorenhancer in Asian cuisine (1%). Therest of the starch (12%) was usedchiefly for food products (Jones,1983). Currently, Japan producesalmost 2 million tons of HFCS.

112


Uses in Taiwan

Taiwan is a major importer of starch,principally cassava starch, which, in1983, was used mainly for preparingmaltose for bakery products andsweets and alpha starch for eel feed.Other uses for MS were adhesives forcorrugated cardboard, dextrins,ingredients for food products such asnoodles, and other uses in the textileand paper industries. Potentially,MSG can be the largest consumer ofstarch but molasses are normallyused. When the price of molassesgoes up, cassava starch is preferred(Jones, 1983).

Other Asian markets

Asians tend to use cassava andsweetpotato starch more for industrialuses than for human consumption.For example, the percentage ofcassava starch destined for humanconsumption fell from 65% to 50%between 1966 and 1980 (Ghosh,1988). In Indonesia, 3% of cassavaroots in 1983 were directed towardstarch production, whereas by 1988this percentage had increased to 10%(Damardjati et al., 1990).

Sweetpotato and cassavastarches in China have traditionallybeen used to prepare noodles andMSG. Almost half the starchproduction is directed to noodles. Ofthe 200,000 t of MSG producedannually, 90% are prepared fromsweetpotato and 10% from cassavastarch (Shuren, 1990).

Other industrial uses of starch inChina include sweeteners such asglucose syrup (100,000 t/year),medical glucose, maltose, and HFS.Production of HFS is low because itcannot compete with thesophisticated sugar industry (Shuren,1990). China has also pioneered theproduction of sophisticated chemicalproducts.

In Thailand, for the 510,000 t ofstarch consumed domestically, themain markets in 1991 werehousehold use and food (e.g.,noodles), 33%; MSG and lysine, 19%;glucose syrup, 15%; paper industry,9%; textiles, 3%; plywood, 1%; andothers, 13% (Titapiwatanakun, 1993).When comparing these figures with1983 data (Lynam, 1987b), the foodand glucose markets present thehighest rates of growth—the glucosemarket did not even exist in 1983.The markets for MSG, lysine, andend-uses in the paper industry, incontrast, have decreased significantly.

Predictions for the year 2001suggest that the consumption by thefood industry will fall to 18%,whereas the share held by MSG andlysine will rise to 27%, and that ofthe paper industry to 15%(Titapiwatanakun, 1993). Thailandexports MS mostly to Japan.

About 10% of Indonesiancassava production is processed toobtain starch, which is mostly used(65%) to make krupuk, a crunchynative food. Another 15% of cassavastarch is used for other foods, 10%for textiles, and 3% for glucose(Damardjati et al., 1990; Lynam,1987a).

Cassava starch in India ismainly used for householdconsumption and to prepare glucoseand dextrins. In the northernstates, it is also used in the textileindustry (Padmaja et al., 1990).

Cassava starch production in thePhilippines is destined chiefly for thefood industry and for glucose. Otherminor markets include thepharmaceutical, paper, textile, andadhesive industries (van Den et al.,1990).

113


Uses in Latin America

Starch in Brazil is used for householdconsumption, and in the food (e.g., asa thickener, stabilizer in processedmeat, base for colors and aromas,and in bread making) andpharmaceutical industries. Innonfood industries, it is employed tomanufacture adhesives, paper,explosives, and biodegradable plastics(Cereda, 1991; I. C. Takitane, 1992,personal communication).

The market for sweeteners

Of the market segments for starch,that of sweeteners deserves specialattention because it has displayed thehighest growth in the last 25 years.The birth of enzyme engineeringallowed the low-cost conversion ofstarch to D-glucose and then to amixture in equilibrium of D-glucoseand D-fructose (HFCS) exhibiting thesame degree of sweetness as invertsugar from sugar cane or sugar beet.

High fructose corn syrup. HFCScontains from 55% to 90% fructose (amedian of 60%) but the most commonin the U.S. market are HFCS-55 andHFCS-42. The HFCS-55 is slightlysweeter than sucrose. A fructose witha 97% purity can be obtained fromthese syrups (Sasson, 1990).

When HFCS was launched in1968, it immediately captured 30% ofthe market for sugar in the USA anddoubled the amount of starchproduced by the maize wet-millingindustry (Whistler, 1984). Thesuccess of the HFCS resulted from theprotection of the domestic sugarindustry, reflected by high internalprices, and the lowered price of maizedue to yield increases (Lynam,1987c).

In 1984, 3.5 million tons of HFCSwere produced in the USA, almost1 million in Japan, and 200,000 in

the EU (Sasson, 1990). The followingyear, the main soft drinkmanufacturers in the USA decided toincrease the proportion of fructosesyrup from 50% to 100% (Sasson,1990). Per capita consumption ofHFCS rose almost three-fold between1980 and 1988 (Table 3). The growthrate of the HFCS began to fall in 1985(The advance of..., 1991) afterconquering the soft drink market, themain market for sweeteners in the USA(Claassen and Brenner, 1991).

The annual production of HFCS inthe USA in 1992 was 6 million tons,58% of which corresponded toHFCS-55 and 42% to HFCS-42 (USDA,1993b). This volume represents 70%of world production, followed by Japan(The advance of..., 1991).

The current world production ofHFS, about 8.5 million tons, isconcentrated in developed countries.In the EU, the annual HFS production,about 500,000 t, has been voluntarilyrestricted to protect the sugar beetindustry (Coutouly, 1991).

But an increasing proportion ofgrowth is expected in developingcountries, which have alternativesources such as cassava, rice, wheat,and sorghum. Currently,starch-based sweetener production is

Table 3. Growth of annual consumption (kg percapita) of caloric sweeteners in the USA(1965-1992).

Year Sweetener

Glucose High Dextrose Sucrosesyrup fructose

syrup

1965 5.6 .0 1.9 44.01975 7.4 3.1 2.3 41.61980 8.3 7.0 1.6 38.61985 8.9 20.0 1.8 29.11992 9.6 23.5 2.0 29.3

SOURCES: Farris, 1984; Higley and White, 1991;USDA, 1993b; Whistler, 1984.

114


growing faster in Asia and LatinAmerica (Claassen and Brenner,1991). Some countries, like Chinaand Vietnam, already have small HFSindustries. But the internationalsweetener trade is small, with theUSA as a net importer—it purchasedalmost 200,000 t of HFCS fromCanada in 1992 (USDA, 1993b).

Glucose syrup. In contrast,processing for glucose syrup issimpler. Hence, its production iswidespread, even in the developingcountries of Asia (e.g., Thailand,China, India, the Philippines, andIndonesia) and Latin America (incountries with CPC subsidiaries).

Marketing factors. Table 3shows the growth of per capitaconsumption of sweeteners in theUSA from 1965 to 1992. By 1985,the combined per capita consumptionof sweeteners surpassed that ofsugar.

The world sweetener market,however, is divided into caloric (e.g.,sugar and HFS) and noncaloric (e.g.,saccharine and aspartame)sweeteners. Total consumption ofnoncaloric sweeteners is equivalent,in sweetness, to 8 million tons ofsugar, similar to the equivalent HFSproduction. As a reference, totalsugar production exceeds 110 milliontons (The advance of..., 1991).

Prices of the differentstarch-derived sweeteners in the U.S.market in July 1993 were HFCS-55,US$0.52/kg; HFCS-42, $0.47;glucose, $0.33; and dextrose, $0.54.These prices were considered to behigh and a reduction was expected inthe fourth quarter (USDA, 1993b).Prices respond to the cost of maizeand other inputs and to demand,which is high in summer and declinesin other seasons. Dextrose ischaracterized by a high but stableprice (USDA, 1993b).

Marketing Opportunities forDeveloping Countries

According to Jones (1983), tradebarriers such as duties, levies, andquotas limit export opportunities,particularly to the EU and Japan.Price competitiveness is the othermajor factor affecting marketprospects.

“Mass” and “specialized” marketsmust be distinguished: in the massmarket most UMS must compete withone another, with the result that thecheapest, assuming acceptablequality levels, enjoy market success.

In contrast, the specializedmarket is a small segment of the UMSmarket, where end-users requirespecific characteristics, such as acertain granule size or pastingtemperature, that can only besupplied by one or two particularstarches. But, because otherstarches can be profitably modified toreproduce the desired properties,these modified starches (usuallymaize or sweetpotato) may graduallyreduce the UMS’ share in thespecialized market. For example,after World War II, much of the USA’scassava starch imports weresubstituted with cheaper maizestarch (Lynam, 1987c).

Future prospects for starchexporters in developing countries varygreatly with regard to the massmarket. In the EU, for example, withthe current restrictions, the outlookfor import growth is nil. In Japan,import growth prospects are limited toachieving the full quota. If theexporter can compete with U.S. maizeprices, then considerable potentialexists in the USA.

International trade in MS andsweeteners is small. Sweetenerexports are limited. Most MS exportscorrespond to dextrins. Traditionally,

115


developing countries participate littlein this trade. The main potential liesin increasing the exports of thealready modified cassava starchesbeing produced in the USA, Japan,and the EU from imported cassavastarch.

For developing countries, modifiedstarch exports would earn addedvalue and are likely to have easieraccess to foreign markets. But a highlevel of technical expertise and a closerelationship between modifier andend-user are desirable. Thailand andChina are currently increasingexports of MS to Japan.

In view of the uncertain exportopportunities, prospective starchproducers in developing countries arerecommended to concentrate first ontheir domestic markets, then expandto neighboring markets, before sellingon the international market.

The domestic markets for UMSwill grow if the UMS-using industries(food, textile, paper and cardboard)develop. Demand for MS, sweeteners,and, in countries where sugar isscarce or expensive, HFS may arise.Other uses for starches are forcomposite flours and biotechnologicalapplications.

New Market Perspectivesfor Starch

Because of starch’s versatility, newuses arise every year. It is also anexcellent example of a renewable rawmaterial. It should be used morewidely in the medium and long term,taking into account three aims: (1) topreserve natural resources, (2) toproduce biodegradable products thatare environmentally friendly, and(3) to reduce agricultural surpluses(Koch and Roper, 1988).Starch-based derivatives cansubstitute to some extent for

petrochemical products, but focusshould be given to preparing andsynthesizing new compounds withspecific and improved properties(Koch and Roper, 1988).

Ethanol

One new compound of high potentialis ethanol. In the USA, since the late1960s, 95% of ethanol has beenproduced from wet milling maize.Initially, ethanol was used to blendwith gasoline at a 1:9 ratio to produce“gasohol.” Currently, this blendrepresents 8% of gasolines sold in theUSA. New uses later emerged: as anoctane enhancer in gasoline, and asan oxygenate to reduce theenvironmental pollution ofautomobiles. In 1992, ethanolproduction reached almost 1 billiongallons. Ethanol is now not onlyfeasible but will be the predominantenergy alternative in the future. Themain oxygenate competitor forethanol is methyl tertiary butyl ether(MTBE) (Hiunok, 1993; Russo, 1993).

Germany and the EU are studyingthe feasibility of replacingpetroleum-derived products withrenewable raw materials such asstarch, among others. Analyses arebeing conducted on substituting 10%of diesel and heating oils and 5% ofpetroleum with renewable rawmaterials such as starch, and on thegreater use of renewable rawmaterials as fuel (Schmitt, 1988).

Biodegradable polymers

The USA is increasingly concernedwith environmental pollution causedby its production processes and thedisposal of its end products. This hasforced the plastic industry to look foralternative raw materials and to makeits products more recyclable andbiodegradable (Beach and Price,1993).

116


The current use of biodegradablepolymers, mostly based on maize, inthe USA has been for items wheredisintegration after use is of directbenefit. Some examples areagricultural mulch films, plantingcontainers and protectors, hay twine,surgical stitching, medicine capsules,and compost bags. Agriculturalpesticide firms are also examining theuse of starch-based polymers toencapsulate products (Beach andPrice, 1993).

Markets for biodegradablepolymers in the USA are foodpackaging, nonfood packaging,personal care and medical products,and other disposable products. Butbecause the U.S. Government has notregulated the use of biodegradablepackaging for food, the key market inthe near future will be for nonfoodpackaging. By 1992, biodegradablepolymeric resins had captured 0.06%of the market for plastic resins usedby the nonfood sector, representing2.3 million kg of a total of3.6 billion kg (USDA, 1993a).

Biodegradable polymers competein the market for plastic materialsand resins, whose composition in theUSA (1992) was low-densitypolyethylene (19%), high-densitypolyethylene (16%), polyvinyl chlorideor PVC (15%), polypropylene (13%),polystyrene (8%), and more than 18additional materials account for theremaining 29% (Beach and Price,1993).

Vegetable adhesives

Environmental concerns in the USAover synthetic adhesives have spurrednew starch-using technologiesdestined mainly for the packagingmarket. Starch-based adhesives areusually less expensive than syntheticadhesives and are free of theunpleasant odors of some animalglues. In 1990, the USA consumed

about 4.5 million tons of adhesives, ofwhich 40% were natural. Maizestarch dominates the market fornatural adhesives with an annualconsumption of 1.6 million tons(USDA, 1993a).

Organic chemicals

Agricultural products are increasinglyconsidered as alternative rawmaterials for organic chemicals.Within the common ground shared byagricultural and chemical industries,a new industry, labeled “greenrefinery,” may result. The starchindustry competes with that ofglucose syrup to supply fermentationsubstrata. Of the possiblebiotechnological processes, the mostviable in the short and medium termsinvolve producing energy,fermentation products such as aminoacids and other organic acids,biodegradable plastics andsurfactants, antibiotics, andbiocatalysts (Malerbe, 1990).

China has pioneered thedevelopment of refined chemicalproducts such as sorbitol, mannitol,oxalic acid, gluconic acid, acetic acid,and ethylene. Sorbitol is used tomanufacture vitamin C, fortoothpaste, cosmetics, and paints.Mannitol is used to produce polyesterand plastic foam, and is also usedmedically as a plasma expander(Ghosh, 1988; Shuren, 1990).

Large-scale, bioprocess technologyalready has had a significant impacton the use of starch for producingcitric and lactic acids. Improvedfermenting processes have beendeveloped for producing butyric,succinic, and propionic acids (Zeikus,1993).

Food

Previously, starch was appreciatedmainly for its nutritional value as a

117


source of calories. Currently, it is alsovalued for its functional propertiesrelated to human health. For example,resistant starch, as does fiber, helpsstimulate digestion, and plays a role inthe prevention of colon cancer. Fatsubstitutes such as maltodextrins andcomplex carbohydrates help reducecaloric intake (Koch et al., 1993).

Other novel starch-based productsinclude solid glucose and fructose,new combinations of starch and fiber,and modified maize and cassavastarches used to replace lactic proteinin processed meat and yogurt.

Other uses

Other examples of new or expandedmarkets include systems forcontrolled release of chemicals,coating agents, surfactants,plasticizers, and sequestrantsbased on starch (Doane, 1993).

Figure 2 divides the potentialindustrial (nonfood) use of a rangeof new starch-derived products intofive categories. Figure 3summarizes both current andfuture uses of starch-basedproducts.

* Binders and glues

* ThickenersAuxiliary material Process aids * Texturizers

* Formulation aids

* Protective colloids

* Polyols

Biotechnology* Organic acids

Raw materials(fermentation)

* Amino acids

* Polysaccharides

* Enzymes

* Polyethylene and polypropylene

FunctionalProcessing of * Polystyrene

Starchadditives

synthetic * Polyvinyl chloride (PVC)polymers * Polyurethane foam

* Styrene and butadien lattices

* Polyesters and alkyd resins

Incorporation into* Polyurethanes

Componentsynthetic polymers

* U/F or M/F resins

* Phenolic resins

* Grafted polymers

* Surfactants and detergentsEnd product or * Sequestrants

Active materials intermediate * Buildersproduct * Perborate boosters

* Chiral building blocks

Figure 2. New categories of nonfood industrial uses for starch. (After Koch and Roper, 1988.)

118


Ethanol Oxygenated motor fuels/solvents/cosmetics/pharmaceuticals

Lactic acid Biodegradable plastics, mold retardant, acidulants

Itaconic acid Plastic objects

Fermentation processes Citric, gluconic acids Cleaners, control of metal contaminants

Enzymes Catalysts HFCS, denim, ethanol, detergents

Pharmaceuticals Antibiotics, vitamin B12, etc.Acetic acidAcetonen-ButanolGlycerolFumaric acidMalic acidSuccinic acid

Commercialproducts fromagriculturalmaterials:

Xanthan gum Oil, textiles

Fermentation byproducts Dry ice, fuel gas, animal feed

Oxidation/dehydration Polyester plastics

Starch Hydrolysis GlucoseSorbitol/MannitolHydrogenation

Plastics/alkydpaints, coatings

Depolymerization with heat,acids, oxidizing agents

Economical starches withimproved performance

Paper goods, textiles,adhesives

CrosslinkingStarch with improved resistance

to shear and temperaturePowder for surgical gloves, other anti-stick products

Reaction with ethylene oxide Hydroxyethyl starch Coated papers

Reaction withreactive amines

Cationicstarch

Floculants/fibermodification

Water treatment, papergoods, textiles

Acrylonitrile polymerizationand hydrolysis

Adsorbentmaterial

Diapers, adsorbent pads, seed coats, filters

Figure 3. Current and future starch-based products (with processes in italics). (After Parker, 1993,personal communication.)

References

The advance of sugar’s competitors inperspective. 1991. Int. SugarSweetener Rep. 123:22, 355-358.

Agra Europe. 1990. Prospects for alternativeuses of cereals and other crops.Agra-Briefing, no. 23. Kent, UK. 35 p.

Atthasampunna, P. 1990. Cassava processingand utilization in Thailand. In:Howeler, R. H. (ed.). Proceedings ofthe Third Regional Workshop of theCassava Research Network in Asia,held Oct. 22-27, 1990, Malang,Indonesia. CIAT, Cali, Colombia. p.315-326.

AVEBE. 1989. Potato starch. Foxhol, theNetherlands. 17 p.

Beach, E. D. and Price, J. M. 1993. The effectsof expanding biodegradable polymerproduction on the farm sector. In:Industrial uses of agricultural materials.Situation and outlook report, no. 6.Economic Research Service, UnitedStates Department of Agriculture (ERS/USDA), Washington, DC, USA. p. 41-48.

Cereda, M. P. 1991. General viewpoint ofcassava starch industries in Brazil.Paper presented at the Cassava StarchWorkshop, 17-20 June, CentroInternacional de Agricultura Tropical(CIAT), Cali, Colombia. CIAT, Cali,Colombia. (Abstr.)

119


Kirby, K. W. 1990. Specialty starches: use inthe paper industry. In: Glass, J. E.and Swift, G. (eds.). Proceedings ofthe American Chemical Society (ACS)Symposium. ACS, Washington, DC,USA. p. 274-287.

Koch, H. and Roper, H. 1988. Newindustrial products from starch.Starch/Stärke 4:121-131.

___________; __________; and Hopcke, R.1993. New industrial uses of starch.In: Meuser, F.; Manners, D. J.; andSeibel, W. (eds.). Plant polymericcarbohydrates. Royal Society ofChemistry, Cambridge, UK.p. 157-179.

Leuch, D. J. 1990. The effects of theCommon Industrial Policy on theEuropean Community wheat-washingindustry and grain trade. Staff report,no. AGES 9023. Economic ResearchService, United States Department ofAgriculture (ERS/USDA),Washington, DC, USA. 26 p.

Leygue, J. P. 1993. Débouchés industrielsdes céréales. Institut technique descéréales et des fourrages (ITCF),Céréaliers du France, Paris, France.32 p.

Long, J. E. 1985. United States markets forstarch-based products. In: vanBeynum, G. M. A. and Roels, J. A.(eds.). Starch conversion technology.Marcel Dekker, Delft, theNetherlands. p. 335-347.

Lynam, J. 1987a. Indonesia, a multi-marketcassava economy. In: Lynam, J. Thecassava economy of Asia: adapting toeconomic change. Section 4, draftversion. CIAT, Cali, Colombia. 55 p.

__________. 1987b. Thailand, rapid growthdriven by export markets. In: Lynam,J. The cassava economy of Asia:adapting to economic change. Section7, draft version. CIAT, Cali, Colombia.55 p.

__________. 1987c. World and Asian marketsfor cassava products. In: Lynam, J.The cassava economy of Asia:adapting to economic change. Section8, draft version. CIAT, Cali, Colombia.49 p.

Christmann, V. 1991. Price formation andthe use of starches in the non-foodarea. In: The production andalternative uses of renewable rawmaterials from agriculture andforestry. Research documentprepared for the GermanGovernment, Sonderheft, Germany.p. 111-115. (Typescript.)

Claassen, T. L. and Brenner, K. 1991. A‘new world order’ for sweeteners?Sugar y Azúcar 86:10, 22-24, 26.

Coutouly, G. 1991. Genie enzymatique.Masson et Doin, Paris, France.

Damardjati, S. D.; Widowati, S.; and Dimyati,A. 1990. Present status of cassavaprocessing and utilization inIndonesia. In: Howeler, R. H. (ed.).Proceedings of the Third RegionalWorkshop of the Cassava ResearchNetwork in Asia, Oct. 22-27, Malang,Indonesia. CIAT, Cali, Colombia.p. 298-314.

Doane, W. M. 1993. Starch: opportunities fornew industrial uses. Cereal FoodsWorld 38(8):613.

Farris, P. L. 1984. Economics and future ofthe starch industry. In:Whistler, R. L. and Paschall, E. F.(eds.). Starch: chemistry andtechnology. Academic Press, Orlando,FL, USA. p. 11-24.

Ghosh, S. P. 1988. Tuber crops. Oxford andIBH Publishing, New Delhi, India.

Higley, N. A. and White, J. S. 1991. Trends infructose availability andconsumption in the United States.Food Technol. 10:118-122.

Hiunok, L. 1993. Ethanol’s evolving role inthe U.S. automobile fuel market.In: Industrial uses of agriculturalmaterials. Situation and outlookreport, no. 6. Economic ResearchService, United States Department ofAgriculture (ERS/USDA),Washington, DC, USA. p. 49-54.

Jones, S. F. 1983. The world market forstarch and starch products withparticular reference to cassava(tapioca) starch. Report no. G173.Tropical Development and ResearchInstitute (TDRI), London, UK. 98 p.

120


Malerbe, A. 1990. La chimie verte: quellesstrategies pour les industries dusucre et de l’amidon. Economie etSociologie Rurales, no. 34. Grignon,France. 101 p.

Marter, A. D. and Timmins, W. H. 1992.Small-scale processing of sweetpotato in Sichuan Province, People’sRepublic of China. Trop. Sci.32:241-250.

Mitch, E. L. 1984. Potato starch: productionand uses. In: Whistler, R. L. andPaschall, E. F. (eds.). Starch:chemistry and technology. AcademicPress, Orlando, FL, USA. p. 479-489.

Padmaja, G.; Balagopalan, C.; Kurup, G. T.;Moorthy, S. N.; and Nanda, S. K.1990. Cassava processing, marketingand utilization in India. In: Howeler,R. H. (ed.). Proceedings of the ThirdRegional Workshop of the CassavaResearch Network in Asia, Oct.22-27, Malang, Indonesia. CIAT, Cali,Colombia. p. 327-338.

Rhem, S. and Espig, G. 1991. The cultivatedplants of the tropics and subtropics.Margraf, Germany. 552 p.

Russo, L. J. 1993. The evolution oftechnology in the fuel ethanolindustry. Cereal Foods World38(8):636.

Sasson, A. 1990. Feeding tomorrow’s world.United Nations Education,Scientific, and Cultural Organization(UNESCO) and Editorial Reverté,Barcelona, Spain. 807 p.

Schmitt, H. 1988. Renewable raw materials:effects on agricultural markets.Politische Studien 301:39, 609-618.

Shuren, J. 1990. Cassava processing andutilization in China. In: Howeler,R. H. (ed.). Proceedings of the ThirdRegional Workshop of the CassavaResearch Network in Asia, Oct.22-27, Malang, Indonesia. CIAT,Cali, Colombia. p. 355-362.

__________ and Henry, G. 1993. The changingrole of cassava in South China’sagro-industrial development:problems and opportunities. Paperpresented at the regional seminar on“Upland Agriculture in Asia”, April6-8, Regional Coordination Centre forResearch and Development of CoarseGrains, Pulses, Roots, and TuberCrops in the Humid Tropics of Asiaand the Pacific (CGPRT), Bogor,Indonesia.

Titapiwatanakun, B. 1993. Thai cassavastarch industry: current and futureutilization. Paper presented at theInternational Meeting on CassavaFlour and Starch, Jan. 11-15, 1994,Cali, Colombia. CIAT, Cali, Colombia.(Abstr.)

USDA (United States Department ofAgriculture). 1993a. Industrial usesof agricultural materials. Situationand outlook report, no. 6. EconomicResearch Service (ERS), USDA,Washington, DC, USA. 71 p.

__________. 1993b. Sugar and sweetener.Situation and outlook report, no. 9.Economic Research Service (ERS),USDA, Washington, DC, USA. 57 p.

van Den, T.; Palomar, L. S.; and Amestos,F. J. 1990. Cassava processing andutilization in the Philippines. In:Howeler, R. H. (ed.). Proceedings ofthe Third Regional Workshop of theCassava Research Network in Asia,Oct. 22-27, Malang, Indonesia. CIAT,Cali, Colombia. p. 339-354.

Whistler, R. L. 1984. History and futureexpectation of starch use. In:Whistler, R. L. and Paschall, E. F.(eds.). Starch: chemistry andtechnology. Academic Press, Orlando,FL, USA. p. 1-9.

Zeikus, J. G. 1993. Production of organicacids from fermentation of starch.Cereal Foods World 38(8):609.

SSSSSESSIONESSIONESSIONESSIONESSION 3: 3: 3: 3: 3:

PPPPPHYSICOCHEMICALHYSICOCHEMICALHYSICOCHEMICALHYSICOCHEMICALHYSICOCHEMICAL S S S S STUDIESTUDIESTUDIESTUDIESTUDIES OFOFOFOFOF

FFFFFLOURSLOURSLOURSLOURSLOURS ANDANDANDANDAND S S S S STTTTTARCHESARCHESARCHESARCHESARCHES

123

The Role of Common Salt in Maintaining Hot-Paste Viscosity...

Abstract

The amylographs of starch and flourfrom three cassava varieties weredetermined in salt (NaCl) solutionsof 0%, 2.5%, 5%, and 7.5%concentrations. The salt increasedthe pasting and peak viscositytemperatures. Peak viscosity differedwith variety, and increased with salt insome cases, but was reduced to belowthat of the control in others. Salt alsoreduced the extent of retrogradation ofstarch, compared with the control.

Introduction

The average Ghanaian housewifeknows that, if a family member is latefor the evening meal of fufu, she mustadd salt, pounding it into the cassavapaste (or plantain or cocoyam). Thispractice helps prolong the elasticity ofthe pounded paste, which otherwisewill harden, and, in some cases,become watery.

For industrial starches, certainadditives are often used to modifystarch properties to make themsuitable for particular end products.

CHAPTER 14

THE ROLE OF COMMON SALT INMAINTAINING HOT-PASTE VISCOSITY OF

CASSAVA STARCH

O. Safo-Kantanka* and Rita Acquistucci.**

Tipples (1982) pointed out that, inwheat starch, these additives affect itsgelatinization properties. Additivesinclude sugars, syrups, various ions,and some bread ingredients. He citedthe study of Hester et al. (1956) on theeffect of sucrose on the pastingcharacteristics of several starches.They reported that:

(1) The temperature of the initial risein paste viscosity increased formost starches, indicating adelayed swelling of granules.

(2) The temperature of maximumviscosity of starch pastes waslower than, or did not reach,95 °C, indicating less swelling ofgranules.

(3) Granules disintegrated less.(4) The amount of soluble material

diffusing from the granules wasless.

(5) Starch gels became less rigid, and,when high sucrose concentrationswere used, gels did not form.

Bean and Osman (1959)investigated the effect of 10 differentsugars and syrups on hot-pasteviscosity curves and gel strength of 5%maize-starch paste. The maximumhot-paste viscosity increased slightlyduring gelatinization withconcentrations of sugar as high as20%, but decreased with higherconcentrations. Tipples (1982) alsocites Medcalf and Gilles’s 1966 study

* Crop Science Department, University ofScience and Technology (UST), Kumasi,Ghana.

** National Nutrition Institute, Rome, Italy.

124


on the effect of different salts onwheat-starch amylograms. They foundthat pasting temperature and peaktemperature progressively increased,according to the effects of the classicallyotropic anion series (SCN-, I-, NO-

3,Br-, Cl-, F-, SO-

4). Except for Na2SO4

and NaF, the salts studied gavemarkedly increased starch peakviscosities. Even NaCl concentrationsas low as 0.05 M resulted in asignificant increase in wheat-starchpeak viscosity.

Ganz (1965) found that when asuspension of wheat starch was heatedin 2.5% (0.43 M) NaCl solution in aBrabender viscoamylograph, peakviscosity markedly increased. Thisincrease was believed to be a result ofan enhanced maintenance of “granuleintegrity.” That is, the granule swells,or remains intact, over a longer timebefore fragmenting. The use of saltswas therefore suggested as a way ofregulating starch swelling.

Our study accordingly aimed toverify the Ghanaian housewife’spractice, and to find agreement withthe observed effects of additives onstarch gelatinization reported in theliterature.

Materials and Methods

A previous study had alreadyexamined the swelling power andsolubility of three cassava varietiesthat differed in cooking quality(i.e., mealiness of the cooked root, andelasticity and smoothness of thepounded paste).

Results showed that the threevarieties differed in granule swellingcharacteristics. ‘Ankra’, a goodcooking variety, showed a two-stepgradual swelling of granules. Butneither the variety 91934, whichshowed a two-step rapid swelling ofgranules, nor the variety 30474, which

showed a one-step slow swelling, hadgood cooking qualities. The threevarieties differed in the strength of thebonding forces between granules(Figures 1 and 2). These three varietieswere used in the present investigation.

The concentrations of NaClsolutions used by Ganz (1965) forwheat were adopted. They were0.43 M, 0.86 M, and 1.29 M, thusgiving 2.5%, 5%, and 7.5% saltsolutions, respectively. One sample of35 g (dry basis) of cassava starch was

60 65 70 75 80 85 90 95

50

40

30

20

10

0

Sol

ubilit

y (%

)

Temperature (°C)

70

60

50

40

30

20

10

0

Sw

ellin

g pow

er

60 65 70 75 80 85 90 95

Temperature (°C)

Figure 1. The effect of temperature on theswelling power of starch from threecassava varieties. ( = cv. 91934; = cv. Ankra; = cv. 30474.)

Figure 2. The effect of temperature on thesolubility of starch from three cassavavarieties. ( = cv. 91934; = cv. Ankra; = cv. 30474.)

125


dissolved in each solution and thestandard Brabender procedurefollowed. The flours were also studiedin the same way.

Results and Discussion

The pasting cycles of the starches andflours are presented in Tables 1 and 2

and in Figures 3 and 4, and arediscussed below.

Pasting temperature

The addition of salt increased pastingtemperature, although the degree ofincrease varied with variety. Theseresults agree with the findings ofHester et al. (1956) (see p. 123). This

Table 2. Viscosity changes (in Brabender viscosity units = BU) in flour during gelatinization in thepresence of common salt (NaCl).

Variety Salt Pasting Peak temp. Peak Visc. at Visc. Visc. at Visc.conc’n temp. (°C) (°C) visc. 95 °C after 1 h 50 °C after 1 h

at 95 °C at 50 °C

Ankra 0 M 66.5 75.5 480 480 160 240 2500.43 M 72.5 80.0 500 380 70 120 900.86 M 74.0 81.5 500 400 80 120 1001.29 M 78.5 86.0 420 400 120 180 140

91934 0 M 66.5 75.5 380 40 0 0 00.43 M 69.5 77.0 390 100 0 0 00.86 M 74.0 80.0 420 180 0 0 01.29 M 75.5 81.5 400 220 0 0 0

30474 0 M 71.0 78.5 90 60 30 50 500.43 M 75.5 87.5 90 80 10 10 100.86 M 75.5 84.5 290 170 10 10 101.29 M 75.5 86.0 240 190 20 20 20

Table 1. Viscosity changes (using Brabender viscosity units = BU) in starch during gelatinization in thepresence of common salt (NaCl).

Variety Salt Pasting Peak temp. Peak Visc. at Visc. Visc. at Visc.conc’n temp. (°C) (°C) visc. 95 °C after 1 h 50 °C after 1 h

at 95 °C at 50 °C

Ankra 0 M 74.0 82.0 560 460 260 480 4200.43 M 75.5 89.0 360 340 150 220 1800.86 M 74.0 92.0 300 300 160 200 1601.29 M 79.3 93.5 360 360 200 300 260

91934 0 M 74.0 77.0 500 380 145 280 2400.43 M 75.5 81.5 460 380 30 50 400.86 M 77.0 84.5 500 440 60 90 801.29 M 77.8 87.5 560 500 100 140 100

30474 0 M 71.0 85.0 340 290 140 280 2600.43 M 78.5 92.0 340 270 110 160 1200.86 M 81.5 95.0 380 380 180 260 2001.29 M 81.5 95.0 380 380 260 340 280

126


found similar results when theyexamined the effect of different saltson wheat starch amylographs.

Peak viscosity

The effect of salt concentrations onpeak viscosity varied with variety. Inthe flours and starches of ‘30474’ and‘91934’, salt concentrations of 5% and

means that the salt caused a delay ingranule swelling.

In all three varieties, adding saltalso increased the temperature atwhich peak viscosity was attained.For example, in ‘30474’, peakviscosity of its starch in 5% and 7.5%salt solutions occurred at 95 °C. Thecited Medcalf and Gilles study (1966)

600

500

400

300

200

100

0

0 14 15 30 45 60 90 120 150 180

600

500

400

300

200

100

0Time(in minutes)

Temp. 50 °C Held at 95 °C Held at 50 °C

600

500

400

300

200

100

0

Vis

cosi

ty (B

U)

cv. 30474

cv. 91934

cv. Ankra

0 13 14 15 30 60 90 120 150 180

300

250

200

150

100

50

0

500

400

300

200

100

00 14 15 30 45 60 90 120 150 180

600

500

400

300

200

100

00 14 15 30 45 60 90 120 150 180

Figure 4. Effect of different salt concentrationson hot-paste viscosity of flour fromthree cassava varieties. (Saltconcentrations: = control; = 7.5%; = 5%; = 2.5%.

Figure 3. The effect of different salt concentrationson hot-paste viscosity of starch fromthree cassava varieties. (Saltconcentrations: = control; = 7.5%; = 5%; = 2.5%.)


(in minutes)Time

Cooking temperature and time

Hot

-past

e vi

scos

ity

(BU

)


cv. 30474

cv. 91934

cv. Ankra

Cooking temperature and time

127


7.5% resulted in increases in peakviscosity. These results agree withfindings by Medcalf and Gilles (1966)and Ganz (1965), who found thatNaCl concentrations of 0.05 M and0.43 M significantly increased peakviscosity in starches, for which Ganz(1965) postulated the “granuleintegrity” hypothesis (p. 124). In‘Ankra’, results were slightly different:peak viscosity of untreated starchwas far higher than those with saltadded, even though peak viscosityincreased with salt concentration.

For flour made from ‘Ankra’, thehighest salt concentration of 7.5%had the lowest peak viscosity. Again,this contrasted markedly with thebehavior of the other two varieties.

Our results seem to demonstratethat, if starch granules are fragilewhen swollen, as for variety 91934,adding NaCl may reduce the fragility,but in other cases, as with ‘Ankra’,salt may inhibit granule swelling.

Salt also affected thetemperatures at which peak viscositycould be maintained: at temperatureshigher than 95 °C peak viscositywould begin to drop. For all threevarieties, even though adding salt didnot increase peak viscosity, comparedwith the control, no differencesexisted between peak viscosity andviscosity at 95 °C. The reason may bethat either the salt increased thetemperature at which peak viscositywas attained, or it enabled theswollen granules to remain intact fora longer time before fragmenting.

Retrogradation

“Retrogradation” is an increasedrigidity in the starch gel that occursas starch granules re-associateduring cooling, sometimes leading to

syneresis, or the release of water.Retrogradation is heavily influencedby the amylose content of the starch.It declines when salt is added.Hence, the Ghanaian housewife, bypounding salt into the pounded paste,is reducing its tendency to retrograde,thus extending its “table-life.”

For the three varieties, the extentof retrogradation in pure starch andflour tended to increase as the saltconcentration increased, but, exceptfor starch from cv. 30474, the extentof retrogradation was always lessthan the control. The cited study byHester et al. (1956) also found thatstarch gels became less rigid whensucrose was added and that, whenhigh sucrose concentrations wereused, gels did not form at all.

Acknowledgments

This work was undertaken as partof a contract (Research ContractNumber GHA 5416) with theInternational Atomic Energy Agency(IAEA). The award of a six-monthfellowship, which enabled the seniorauthor to travel to Rome to carry outthis research, is gratefullyacknowledged.

References

Bean, M. M. and Osman, E. M. 1959.Behaviour of starch during foodpreparation. II. Effects of differentsugars on the viscosity and gelstrength of starch pastes. Food Res.24:665.

Ganz, A. J. 1965. Effect of sodium chlorideon the pasting of wheat starchgranules. Cereal Chem. 42:429.

Tipples, K. H. 1982. Uses and applications.Brabender.viscoamylographhandbook, 1982.

128


CHAPTER 15

AMYLOGRAPHIC PERFORMANCE OF

CASSAVA STARCH SUBJECTED TO

EXTRUSION COOKING1

Z. González and E. Pérez*


A commercial cassava starch wassubmitted to extrusion cooking in twoRheocord Torque Rheometers. Model104 had a single rotating screw thatoperated at 90 rpm at a temperatureof 150 °C, with a 25% samplemoisture content. Model 3000 haddouble, co-rotating screws that alsooperated at 90 rpm, but attemperatures of 100 and 150 °C, andwith 10%, 21%, and 25% samplemoisture contents.

Starch suspensions at 6.88%(w/w dry basis) were prepared. Thesewere heated at a rate of 1.5 °C perminute in the bowl of a Brabenderamylograph (model A.V. 40, 60-cycle),from 30 until 90 °C. Suspensionswere maintained at this temperaturefor 30 min and then cooled at thesame rate to 50 °C, at which theywere maintained for 30 min more.Water absorption capacity, solubility,and swelling power of both extrudedand native starches were determinedby Schoch’s method (1964).

Results

Table 1 shows the most importantparameters of the amylogramsobtained. The initial gelatinizationtemperature of starches extruded at25% moisture content (61.5 °C) and

Introduction

In Venezuela, cassava (Manihotesculenta Crantz) is consumedpreferably fresh but also in otherforms. During 1991, 381,069 t ofcassava were produced, of which183,913 t were used for humanconsumption, including 19,549 t forproducing “casabe,” a type of cassavabread. Another 38,107 t were usedfor animal feed, 244 t for export.About 152,428 t were estimated aslost (INN, 1991).

Because starch has multiple usesin the food, pharmaceutical, oil, andtextile industries, great interest hasarisen in the use of alternative,low-cost sources of thispolysaccharide. Starch’s functionalproperties can be modified bydifferent methods. Our studyevaluated the effect of extrusioncooking on the amylographicperformance of cassava starch.

* Instituto de Ciencia y Tecnología deAlimentos, Facultad de Ciencias, UniversidadCentral de Venezuela (UCV), Caracas,Venezuela.


129

Amylographic Performance of Cassava Starch Under Extrusion Cooking

Table 1. Effect of extrusion cooking on the most important parameters used in cassava starchamylography.

Parametera Sample

Native Extruded Extruded Extrudedstarch starchb starchc starchd

Initial gelatinization temperature (°C) 60.8 61.5 55.5 61.5Final gelatinization temperature (°C) 70.5 77.5 67.5 78.0Maximum viscosity (Vmax); (BU) 900 740 860 700V

max temperature (°C) 69.8 77.3 66.8 76.5

Viscosity at 90 °C (BU) 380 520 410 530Viscosity after 30 min at 90 °C (V90/30); (BU) 40 300 260 290Viscosity at 50 °C (V50); (BU) 400 480 420 320Viscosity after 30 min at 50 °C (BU) 420 480 430 380Stability (Vmax - V90/30); (BU) 660 440 600 410Sedimentation (V50 - V90/30); (BU) 160 180 160 30Consistency (V50 - V

max); (BU) -500 -260 -440 -380

a. BU = Brabender viscosity units.b. Single-screw, 150 °C, 25% moisture content, and 90 rpm.c. Double-screw, 100 °C, 10.21% moisture content, and 90 rpm.d. Double-screw, 150 °C, 25% moisture content, and 90 rpm.

that for native starch (60.8 °C) did notdiffer significantly. However, initialgelatinization temperature for thesample extruded at 10.21% moisturecontent by the double-screw extruderwas 55.5 °C. Apparently, theconditions under which this lastoperation was performed favored theaccess of water to the amorphouszones of the starch granules, causingthem to swell faster. Gelatinizationtherefore began at a lowertemperature.

The interval between initial andfinal gelatinization temperatures wasgreater in starches extruded at 25%moisture content by both single- anddouble-screw extruders (about 16 °C)than those corresponding to starchprocessed by the double-screwextruder at 10.21% moisture content(12 °C) and to native starch (9.7 °C)(Table 1). These results indicate thatextrusion partially transformed thestarch granule structure and affectedthe macromolecules. This caused agreater temperature interval inextruded products than in nativestarches.

All processed samples had lowermaximum viscosity (Vmax) values(Table 1) compared with nativestarches (900 Brabender viscosityunits [BU]). Starches extruded at25% moisture content by single-screwextruders showed Vmax of 740 BU andby double-screw extruder, 700 BU.These values were lower than that forstarch extruded at 10.21% moisturecontent by a double-screw extruder,which, in its turn, differed by 40 BUfrom that of native starch (860 BU).

Maximum viscosities of extrudedstarches, at 25% moisture content,were obtained at 77.3 (single-screwextruder) and 76.5 °C (double-screwextruder), higher than thatcorresponding to native starch(69.8 °C) (Table 1). Starch extrudedat 10.21% moisture content by thedouble-screw extruder not onlyshowed the lowest temperature forVmax (66.8 °C), but also the lowestgelatinization temperature range(55.5-67.5 °C). This finding probablyindicates that, because the extrusionprocess makes more water availableto the amorphous zones of starch

130


granules, gelatinization advancedmore rapidly and so reached Vmax at alower temperature.

Because no Vmax value similar tothat of native starch was obtained inthe samples processed, a certaindegree of macromolecule ruptureand/or reorganization can beinferred. Although native starchpresented the highest Vmax, itsswelling power was not the highest(Table 2). Also, at about 70 °C, thetemperature at which Vmax of nativestarch was obtained, the highestvalue of water absorption (24.31 g/gstarch) and of swelling power(2.34 g/g starch) corresponded toextruded starch at 10.21% moisturecontent by the double-screw extruder(Table 2).

These findings suggest that theexpansion corresponding to Vmax ofextruded starches was the result ofvarious factors acting together,principally swelling power andsolubility. The greatest value of Vmax

of extruded samples thuscorresponded to processed starch at10.21% moisture content in thedouble-screw extruder, whoseswelling capacity was the highest ofthe starches tested. Extrudedstarch at 25% moisture content inthe single-screw extruder had,overall, the lowest solubility valuesof the starches, even though ittended to swell less than thestarches processed by thedouble-screw extruder.

Table 2. Effect of extrusion cooking on water absorption, solubility, and swelling power of cassavastarches.

Temperature (°C) Sample

Native Extruded Extruded Extrudedstarch starcha starchb starchc

Water absorption (g/g starch)65 2.40 7.40 0.90 0.8970 15.04 14.46 24.31 19.0975 23.22 17.74 23.14 -80 29.47 21.09 29.41 24.2785 31.77 25.17 37.95 28.4990 43.41 27.86 38.93 30.41

Solubility (%)65 1.90 5.44 9.21 9.1170 11.48 11.09 17.42 17.5075 21.57 13.52 16.23 -80 22.11 16.88 22.20 21.7885 25.21 19.52 39.79 26.3390 33.09 21.79 59.87 26.54

Swelling power65 0.50 0.83 0.90 0.8670 0.83 1.54 2.34 1.9075 1.15 1.80 2.24 -80 1.29 2.19 2.78 2.3285 1.53 2.62 3.48 2.6690 2.04 2.93 3.28 2.55

a. Single-screw, 150 °C, 25% moisture content, and 90 rpm.b. Double-screw, 100 °C, 10.21% moisture content, and 90 rpm.c. Double-screw, 150 °C, 25% moisture content, and 90 rpm.

131

Amylographic Performance of Cassava Starch Under Extrusion Cooking

Starch modified at 25% moisturecontent in the double-screw extruder,despite showing an intermediateswelling ability compared with therest of the processed samples, hadthe lowest Vmax. The reason may havebeen that its solubility was usuallyhigher at the same moisture contentthan that of the single-screw extruder(Table 2).

In summary, extrusion tended toreduce water absorption capacity andsolubility of samples processed at25% moisture content by single- anddouble-screw equipment, whereas theswelling power of all extrudedstarches increased. However, instarch processed at 10.21% moisturecontent by the double-screw extruder,solubility tended to increase.

Without exception, all starcheshad reduced viscosity values at 90 °Cin relation to Vmax and, after 30 min at90 °C, in relation to the initialviscosity (Table 1). The differentstarch suspensions showed lowstability during cooking, that is,granules were highly susceptible toshearing stress. This was reflected inthe positive values of the stabilityindex, which is defined as thedifference between Vmax and viscosityafter 30 min at 90 °C (Rasper, 1980).Native starch was the least stableduring cooking (660 BU), followed bystarch extruded at 10.21% moisturecontent in the double-screw extruder(600 BU), starch processed at 25%moisture content in single-screwextruder (440 BU), and starchprocessed in the double-screwextruder (410 BU).

Viscosity values at 50 °C of allstarches were higher than thecorresponding viscosity values after30 min at 90 °C. This findingsuggests that a certain degree ofretrogradation occurred in thesestarches, which could be quantifiedas a sedimentation index, or the

difference between the viscosity at50 °C and that after 30 min at 90 °C(Rasper, 1980). Native starch andstarch extruded at 10.21% moisturecontent by the double-screw extruderpresented the same retrogradationtendency (160 BU), and starchprocessed by the single-screwextruder at 25% moisture contentshowed the greatest sedimentation(180 BU). The lowest value for thisindex corresponded to starchextruded at 25% moisture content bythe double-screw extruder (30 BU).

Viscosity after 30 min at 50 °Cwas higher than viscosity at 50 °C,except for starch extruded by thesingle-screw extruder, whose valueremained constant (480 BU). Ingeneral terms, all starches showedstability during cooking at 50 °C.

Consistency (the differencebetween viscosity at 50 °C and Vmax;Rasper, 1980) increased as aconsequence of the extrusion process.Native starch showed a value of-500 BU, while extruded starchesshowed values of -440 BU (starchextruded at 10.21% moisture contentby a double-screw extruder), -260 BU(starch extruded by a single-screwextruder), and -380 BU (starchextruded at 25% moisture content bya double-screw extruder).

Conclusions

Extrusion cooking of cassava starchcaused a series of modifications in thestarch structure, depending oncooking conditions. The amorphouszones of starch extruded at 10.21%moisture content by the double-screwextruder apparently had greateraccess to water. This translated intoquicker swelling of starch granulesand the start of gelatinization for allextruded samples at a lowertemperature, although the intervals ofgelatinization temperature increased

132


due to the process. Swelling power,stability, and consistency of extrudedstarches also increased, while Vmax

decreased. This appeared to dependprincipally on swelling power andsolubility, among other factors. Inthe sample processed at 25%moisture content by the double-screwextruder, the tendency of starchretrogradation was notably reduced.

References

INN (Instituto Nacional de Nutrición). 1991.Hoja de balance de alimentos, versiónpreliminar. Caracas, Venezuela.

Rasper, V. 1980. Theoretical aspects ofamylographology. In: Shuey, W. C.and Tipples, K. H. (eds.). Theamylograph handbook. AmericanAssociation of Cereal Chemists, St.Paul, MN, USA. p. 1-6.

Schoch, T. J. 1964. Swelling power andsolubility of granular starches. In:Whistler, R. L. (ed.). Methods incarbohydrate chemistry, vol. 4.Academic Press, New York, USA.p. 106-108.

133

Improving the Bread-Making Potential of Cassava Sour Starch

CHAPTER 16

IMPROVING THE BREAD-MAKING

POTENTIAL OF CASSAVA SOUR STARCH

D. Dufour*, S. Larsonneur*, F. Alarcón**, C. Brabet*, and G. Chuzel***

* CIRAD/SAR, stationed at the CassavaUtilization Section, CIAT, Cali, Colombia.

** CIRAD/SAR, Montpellier, France.*** CIRAD/SAR, stationed at the Faculdade de


Abstract

Cassava sour starch, fermented foruse in bread making, is traditionallysun-dried. Changes in thephysicochemical and functionalproperties of the starch duringsun-drying were examined forcorrelations between these changesand the starch’s bread-makingpotential. Starch samples collectedafter fermentation and drying wereanalyzed for their pH, total acidity,and lactic acid. Viscoamylogramswere plotted and bread-makingpotential determined. Resultsindicated that exposure to sunlightconsiderably changes thephysicochemical and rheologicalproperties of cassava sour starch,correlating directly withbread-making potential. Duringoven-drying, the lactic acid contentremained steady, whereas sun-dryingat a similar temperature greatlyreduced it, thus augmenting thecassava sour starch’s bread-makingpotential.

Introduction

Cassava sour starch is a product oftraditional rural industry in LatinAmerica. It is used for making breadssuch as pandebono and pan de yucain Colombia, and pão de queijo inBrazil; and for industrially processedsnack foods (Cereda, 1973, 1991;Cereda and Nuñes, 1992; Chuzel,1990). Urban markets for sour starchare growing in Brazil (where it isknown as polvilho azedo) and inColombia (almidón agrio).

Bakers and manufacturersregard swelling power as the maincriterion of quality, but this is oftenunpredictable. Our study aimed tounderstand how bread-makingpotential is increased duringtraditional processing, so we couldsuggest ways of achieving a betterquality sour starch (Chuzel andMuchnik, 1993).

The traditional method consists ofwet-process extraction of starch fromcassava roots (Pinto, 1978; Ruiz,1988, 1991). The starch is thenstored in 0.5 to 5-t capacity tanks andfermented for 20 to 60 days, accordingto climatic conditions (temperaturesmay range from 15 to 25 °C) (Jory,1989). Lactic fermentation takes placeand the starch pH drops to about3.5-4.0 (Cárdenas and de Buckle,

134


1980). It is then sun-dried ondrying tables (Brazil) or on blackplastic sheeting laid on the ground(Colombia).

Both fermentation andsun-drying give the cassava starchits bread-making potential (Chuzel,1992). Fermentation also causessubstantial modifications to thestarch’s organoleptic andphysicochemical characteristics(Camargo et al., 1988; Cereda,1985; Nakamura and Park, 1975).

Larsonneur (1993) achievedoptimal swelling power by exposingthin layers (0.5 to 1 cm) of sourstarch to the sun (solar radiationintensity ≈1,200 W/m2) on sheets ofblack plastic, and shaking thesheets frequently. In confirmation,Colombian sour starch producersmaintain that drying in low levels ofsunlight results in poor-quality sourstarch with low bread-makingpotential. Brazilian large-scalemanufacturers prefer drying sourstarch in the sun (sometimes using12 km of drying tables) to thevarious types of driers (e.g., hot air,flash driers, and drum driers) usedfor producing unfermented cassavastarch. Industrial trials have shownthat sour starch dried artificiallyhas no significant swelling power.

Sour starch is the mainingredient (mixed with fats orcheese, eggs, and salt) intraditional, high-swelling breads.Such breads contain no wheat flour,nor do they undergo yeastfermentation before baking.Additives are not used and thedough is baked immediately afterkneading, with no rising or“proofing” time. Rising, therefore,does not involve a protein-glutennetwork nor the production ofcarbon dioxide by yeasts as seen in,for example, the making of Frenchbread (Godon, 1981).

Our study examines how sunlightchanges the bread-making potential ofcassava sour starch by changing thefollowing physicochemical andrheological properties: pH, totalacidity, lactic acid content, andBrabender viscosity. The importantrole of lactic acid is alsodemonstrated.


Preparing samples

Starch samples were collected fromproduction units at Santander deQuilichao (Department of Cauca,Colombia). Three local cassavacultivars were used: ‘Amarga’ (referredto as starch A), ‘CMC 40’ (B), and‘Algodona’ (C).

Extracted starch was left toferment in tiled tanks (0.95 x 0.82 x0.79 m, capacity 0.5 m3). The averagetemperature in the zone was 20 °C,with a small day-night variability of18 to 22 °C. For starch A, a sample ofstarch milk (unfermented starchsuspension in water) was collectedimmediately after extraction. Asample of fermented starch was takenafter 30 days of fermentation, justbefore farmers typically initiatesun-drying. The samples were thentransferred, in an insulated box, to theCIAT laboratory, frozen at -20 °C, andsubjected to drying tests and analysis.

Drying conditions

Sun-drying, on black plastic sheets for8 h, was similar to traditional sourstarch-drying conditions in Colombia(layer 1 to 1.5 cm thick, with agitationevery 2 h).

Exposure to sun

Starch samples were sun-dried fordifferent lengths of time (2, 4, 6, and8 h) and then oven-dried at 40 °C to a

135


final moisture content of about 11%.A control sample was oven-dried only,at 40 °C. This temperature waschosen because it does not causegelatinization, but is representative ofan average daytime temperature instrong sunlight. The sampling planwas as follows:

Moist starch Dry starch

0 2 4 6 8

hours of drying

S______________________________ S

S______________________S

S______________ S

S______ S

S

where:

S = sampling forsubsequent analysis

= sun-drying= oven-drying

Rheological and physicochemicalanalyses

Each of the following analyses wasperformed in duplicate on starches A,B, and C:

Viscoamylograms. Therheological properties were determinedby using a Brabenderviscoamylograph. The sour starchwas first ground and sieved through a65/cm ( ≈150 µm) mesh. An aliquot(500 ml) of an aqueous sour-starchsuspension (5% dry matter) was usedto plot the viscoamylogram. Theanalysis unit rotated at 75 rpm; thetemperature of the reaction mixtureincreased steadily at 1.5 °C/min from25 to 90 °C. The mixture was keptat 90 °C for 20 min, then steadilycooled at 1.5 °C/min to 50 °C; this

temperature was kept constant for10 min. Viscosities are expressed inBrabender units (BU).

Measurement of pH. A 10% (w/v)aqueous suspension was agitated atambient temperature (20 ± 2 °C) for30 min and then centrifuged at15,000 g for 15 min. Supernatant pHwas measured.

Assay of total acidity. Totalacidity was assayed in 50 ml ofthe supernatant described aboveby titration of a NaOH 0.1 Nsolution in the presence of 1%phenolphthalein-alcohol solution. Theresults were measured in moles of acidper gram of dry weight of sour starch.

Assay of lactic acid. Cassavasour starch (10 g) was added to15 ml H2S04 (0.006 M) and agitatedfor 1 min. The suspension washomogenized for 1 min at 24,000 rpmin an Ultraturrax blender, agitatedin a vortex mixer for 1 min, andcentrifuged at 9,800 g for 25 min.The supernatant was passed througha 0.45 m filter and analyzed byhigh-performance liquidchromatography (HPLC) as follows: ofthe filtrate, 20 µl was injected into anAminex HPX87H column (Biorad),which was controlled thermostaticallyat 65 °C. Column separation wasbased on a combination of ionexchange, molecular screening, andhydrophobic exchange. A solution ofH2S04 (0.006 M) was used as eluant ata flow rate of 0.8 ml/min (Giraud andRaimbault, 1991). The lactic acidpeak was detected under ultravioletlight at 210 nm. The results wereexpressed in grams of lactic acid per100 g of initial sour-starch dry weight.

Measuring bread-makingpotential. Procedures for breadmaking with sour starch andevaluation of swelling power weredeveloped by Escobar and Molinari(1990) and modified by Laurent (1992)

________

136


and Larsonneur (1993). Sour starchwas ground in a mortar and sieved for10 min through a 65/cm (≈150 µm)mesh. Of this fraction, 85 g (dryweight) was mixed with 100 g ofColombian cheese (Campesino, brand“Alpina”) in a Hobart kneadingmachine operated at low speed(165 rpm) for 1 min. Water was addedto obtain a total of 65 ml of water inthe dough. It was then kneaded atmedium speed (300 rpm) for 2 min.Six 30-g rings of dough with an insidediameter of 2 cm were prepared. Thesewere baked at 280 °C for 17 min andcooled for 2 h at ambient temperature.Each loaf was weighed and its volumemeasured with a volumeter accordingto Vanhamel et al. (1991). The specificvolume of the bread was thenexpressed in cm3/g.


Previous tests had shown that freezinghad no effect on the viscoelasticproperties of starch or on thebread-making potential of sour starch

(Larsonneur, 1993). The samplestaken after fermentation were thereforefrozen to be sun-dried the followingday.

The viscoamylograms in Figure 1show the performances of twosubsamples taken from the originalstarch sample A. One subsample wasfrozen and the other fermented for33 days. They were then sun-driedunder identical conditions. Theviscoamylograms reveal considerablemodification of the rheologicalproperties of the fermented sour starch.

Pasting temperature (62.5 °C) andthat of maximum viscosity (70 °C)were identical in all the samples. Thetendency toward retrogradation (adecrease in viscosity after the peak)increased relative to fermentation, afinding which agrees with those ofNakamura and Park (1975). Peakviscosity decreased in relation tothe time allowed for thefermentation—bacterial amylases tobreak down the large starch molecules(Camargo et al., 1988).

Figure 1. Changes in the rheological properties of starch, extracted from cassava variety ‘Amarga’, duringfermentation. The samples were sun-dried for 8 h before analysis. ( = unfermented starch; = starch fermented 33 days.)

Analysis time (minutes)

62.5 °C

70 °C

Vis

cosi

ty (B

U)

400

300

200

100

00 20 40 60 80 100

137


The specific volume of the loavesincreased from 3.5 to 6.5 cm3/g forstarch A, from 2.0 to 5.8 cm3/g for B,and from 1.9 to 5.2 cm3/g for C.

The effect of sunlight

Direct exposure to sunlight (8 h underequatorial conditions) causedsubstantial changes in the rheologicalproperties of fermented starch A. Theviscoamylogram (Figure 2) differswidely from those of the same starchanalyzed before drying (wet starch) andafter oven-drying at 40 °C. In addition,

the two latter viscoamylograms aresimilar, indicating that oven-dryingbarely affects the physicochemicalproperties of sour starch. Thesun-dried starch shows a strongretrogradation tendency and a notabledecrease in maximum viscosity (from320 to 220 BU).

Analysis of pasting properties ofstarch A after 0, 2, 4, 6, and 8 h ofsun-drying reveals a rapid increase inretrogradation tendency after about3 h of exposure to sunlight (Figure 3).In contrast, the decrease in maximum

0 20 40 60 80 100

0 20 40 60 80 100

400

300

200

100

0

Vis

cosi

ty (B

U)

400

300

200

100

0

.

Figure 2. Influence of type of drying on the rheological properties of starch extracted from cassavavariety ‘Amarga’. Samples were taken after 33 days of fermentation. ( = wet starch; = starch oven-dried 8 h; = starch sun-dried 8 h.).

Vis

cosi

ty (B

U)


Figure 3. Influence of sun-drying time on the rheological properties of starch extracted from cassavavariety ‘Amarga’. Samples were taken after 33 days of fermentation. ( = wet starch; = starch sun-dried 2 h; = starch sun-dried 4 h; = starch sun-dried 6 h; = starch sun-dried 8 h.)


138


viscosity is linear (r2 = 0.934) againsttime of exposure to sunlight within the0 to 8-h range (Figure 4). Becausemaximum bread-making potential wasattained after 3 h of sun-drying(Figure 5), bread-making potentialappears to relate more directly to theincrease in retrogradation tendency ofstarch.

In addition, when oven-dried, thesame starch showed no increase inbread-making potential. Sun-dryingkinetics observed for other cassavacultivars (starches B and C) are shownin Figure 6. They confirm thatbread-making potential is acquiredduring exposure to solar radiation andnot after oven-drying.

Significantly, the pH of starch A(sampled after 33 days offermentation) rose from 3.45 to 3.70after sun-drying and increased to only3.50 when oven-dried (Figure 7).Starches B and C similarly increasedin pH during sun-drying from 3.48 to3.55 (B) and from 3.45 to 3.55 (C)(Table 1). Because the pre-drying pHof 3.45 corresponded to the pKa oflactic acid, the medium would havebeen strongly buffered, with lactic andlactate forms in equal proportions.This small increase in pH duringsun-drying (0.25 unit, from 3.45 to

340

320

300

280

260

240

2200 1 2 3 4 5 6 7 8

Sun-drying time (hours)

y = 316.60 - 11.60 x r2 = 0.934

Maxi

mu

m v

isco

sity

(BU

)

Sun-drying time (hours)

Spec

ific

vol

um

e (cm

3/g)

7

6

5

4

3

2

10 1 2 3 4 5 6 7 8 9 10

Spec

ific

vol

um

e (cm

3/g)

0 1 2 3 4 5 6 7 8 9 10

7

6

5

4

3

2

Drying time (hours)

Figure 5. Changes in bread-making potential ofstarch extracted from cassava variety‘Amarga’ in relation to duration ofsun-drying.

Figure 6. Changes in bread-making potentialof starches extracted from cassavavarieties ‘Amarga’ ( ) ‘CMC 40’ ( ),and ‘Algodona’ ( ) in relation toduration of sun-drying. All threevarieties were also oven-dried( ; oven-dried ‘CMC 40’ ).

3.80

3.75

3.70

3.65

3.60

3.55

3.50

3.45

3.40

pH

0 1 2 3 4 5 6 7 8

Drying time (hours)

Figure 7. Changes in the pH of starch extractedfrom cassava variety ‘Amarga’ inrelation to drying time ( = sun-drying; = oven-drying).

Figure 4. Changes in maximum viscosity ofstarch extracted from cassava variety‘Amarga’ in relation to duration ofsun-drying. Samples were taken after33 days of fermentation.

139


Table 1. Acquisition of bread-making potential during fermentation.

Starch Cassava varietya

A B C

pH of wet sweet starch 6.8 - -

pH of sour starch before sun-drying 3.45 3.48 3.45

pH of sour starch after sun-drying 3.70 3.55 3.55

Bread specific volume(ml/g) of oven-dried starch 3.5 2 1.9

Bread specific volume(ml/g) of sun-dried starch 6.5 5.8 4.7

Total acidity before sun-drying(10-5 mol/g dry matter) 10.5 10.4 7.7

Total acidity after sun-drying(10-5 mol/g dry matter) 6.8 9.4 6.7

Lactic acid before sun-drying(10-6 mol/g of dry matter) 105 105 76

Lactic acid after sun-drying(10-6 mol/g of dry matter) 68 94 66

a. A = ‘Amarga’; B = ‘CMC 40’; C = ‘Algodona’.

0 1 2 3 4 5 6 7 8

Lact

ic a

cid +

lact

ate

con

ten

t (1

0-6 m

ol/g)

110

100

90

80

70

60

3.70, starch A) therefore suggestsconsiderable variation in theproportions of lactic acid and lactate,considering the chemical equation ofthe buffer solutions (pH = pKa + logbase/acid).

This variation during sun-dryingcan be interpreted either by thetransformation of lactic acid intolactate or by the disappearance of thelactic form. The lactic acid assay offermented starch sample A (Figure 8),sun-dried and oven-dried, indicatesthat the initial (lactic acid + lactate)content was 10.5 x 10-5 mol/g dryweight, which corresponds to theconversion of 1% of the initial starchinto lactic acid during fermentation.This decreased to 6.8 x 10-5 mol/g dryweight (a decrease of 35%) duringsun-drying but remained unchangedduring oven-drying for 8 h. Becausethe oven-drying temperature wassimilar to that of sun-drying, thedisappearance of lactic acid cannot beascribed to volatilization. Starches Band C similarly produced decreases of

10.4 to 9.4 x 10-5 and 7.7 to6.7 x 10-5 mol/g dry weight,respectively. The greater fall in thelactic acid content of starch A (35%against 10% and 13%, respectively)and its higher bread-making potential(6.5 against 5.8 and 4.7, respectively)correlate well (Table 1). Thedifferences observed between starchesA, B, and C may be heightened by thediversity of cultivars used, a finding

Drying time (hours)

Figure 8. Changes in the lactic acid contentof starch extracted from cassavavariety ‘Amarga’ in relation to dryingtime ( = sun-drying; = oven-drying).

140


which agrees with that of Chuzel(1992).

The pH increase suggests thatlactic acid is consumed in a chemicalreaction during sun-drying. TheHPLC method used did not permitassay of combined forms of lactic acidand thus did not detect polymerizedforms or covalent bonds which mighthave formed during drying. Detectionwould be possible after totalhydrolysis of the starch.

In classic bread making, wheatgluten forms a three-dimensionalnetwork that retains gas bubblesduring baking. Baking additives(xanthan gums) are added tonon-panifiable flours to increase theirbread-making potential (Eggleston,1992; Godon, 1981). Because starchis the only significant macromoleculein sour starch (no protein or celluloseis present), a three-dimensionalnetwork may be formed by aphotochemical reaction involvinglactic acid and fermented starch.Such a network may account for theacquired bread-making potential ofsun-dried cassava sour starch.

Conclusions and Prospects

Fermentation and sun-drying clearlyplayed a role in obtaining sour starchwith high swelling power anddesirable organoleptic characteristics.At the end of fermentation, the pHwas 3.45, following conversion ofabout 1% of the initial starch to lacticacid. Fermentation gave the starchthe necessary physicochemicalproperties required to later achievebread-making potential throughexposure to sunlight. Fermentationand sun-drying modified therheological properties of the starchand produced a more markedretrogradation and lower maximumviscosity, together with an increasedswelling power. The lactic acid

content decreased by as much as35% during sun-drying only,suggesting a photochemical reactioninvolving the starch, resulting in theformation of a three-dimensionalnetwork that retains gas bubblesduring baking and, hence,accounting for the acquisition ofbread-making potential by sourstarch.

Characterization of the solarradiation involved in this processshould lead to the design of dryingapparatus that would combine bothair-drying and radiation. Such asystem would overcome dependenceon climate, reduce labor costs anddrying space, and reduce losses towind, poor handling, and externalcontamination. Brazilianindustrialists favor developing anartificial drier to manufacturehigh-quality industrial sour starch.

A better understanding of thephenomena described in this studyshould permit the development ofmodified cassava starch with a highbread-making potential. Suchmodified starch could be used as anadditive (as are xanthan gums) toimprove the bread-making capacity offlours which expand little.Furthermore, modified cassava starchcould have great potential for thedevelopment of gluten-free bread.

Good fermentation practice andsolar drying, combined with the useof cassava cultivars specificallychosen for sour starch production,should facilitate the production ofhigh-quality cassava sour starch forwhich demand exists in breadmaking and various industries.

Acknowledgments

The authors thank Marney PascoliCereda (UNESP, Botucatu, Brazil) forher advice, hospitality, and access to

141


her extensive knowledge of cassavasour starch.

The research used facilities atCIAT’s Cassava Quality andUtilization Section, and was fundedby CIAT, CIRAD/SAR, and theEuropean Unión.

Special thanks go to Alba LucíaChávez and Jorge Mayer, of CIAT’sBiotechnology Research Unit, andLuc Laurent, of UniversitéTechnologique de Compiègne,Compiègne, France, for their activeparticipation in the trials.

References

Camargo, C.; Colonna, P.; Buleon, A.; andRichard-Molard, D. 1988.Functional properties of sourcassava (Manihot utilissima).starch:polvilho azedo. J. Sci. Food Agric.45:273-289.

Cárdenas, O. S. and de Buckle, T. S. 1980.Sour cassava starch production: apreliminary study. J. Food Sci.45:1509-1512, 1528.

Cereda, M. P. 1973. Alguns aspectos sobre afermentação da fécula demandioca. Ph.D. dissertation.Faculdade de Ciências Médicas eBiológicas, Universidade EstadualPaulista, Botucatu, SP, Brazil. 89 p.

__________. 1985. Avaliação da qualidade dafécula fermentada comercial demandioca (polvilho azedo).I. Características viscográficas eabsorção de agua. Rev. Bras. Med.3(2):7-13.

__________. 1991. Technology and quality ofsour starch. In: [Proceedings of theworkshop on] “Avances sobreAlmidón de Yuca” held at CIAT, Cali,Colombia, 17-20 June 1991. (Abstr.)

__________ and Nuñes, O. L. S. 1992.Brazilian fermented cassava starch.I. Production and use. In: XVIthInternational CarbohydrateSymposium held in Paris, France,July 5-10, 1992. (Abstr.)

Chuzel, G. 1990. Cassava starch: currentand potential use in LatinAmerica. Cassava Newsl. (Cent.Int. Agric. Trop.) 15(1):9-11.

__________. 1992. Amélioration techniqueet économique du procédé defabrication de l’amidon aigre demanioc. In: D. Dufour andD. Griffon (eds.). Amélioration dela qualité des aliments fermentésà base de manioc: Rapportfinal du contrat CEE/STD2TS2A-0225. CIRAD, Montpellier,France.

__________ and Muchnik, J. 1993. Lavalorisation des ressourcestechniques locales: L’amidonaigre de manioc en Colombie. In:J. Muchnick (ed.). Alimentation:Techniques et innovations dansles régions tropicales. EditionsL’Harmattan, Paris, France.p. 307-337.

Eggleston, G. 1992. Can we make amarketable cassava breadwithout wheat? Cassava Newsl.(Cent. Int. Agric. Trop.) 16(1):7-8.

Escobar, C. A. and Molinari, J. E. 1990.Obtención de parámetros para laevaluación de la calidad de unalmidón agrio de yuca. B.S.thesis. Plan de Estudios deIngeniería Química, Universidaddel Valle, Cali, Colombia. 75 p.

Giraud, E. and Raimbault, M. 1991.Utilización de la cromatografíalíquida de alta resolución (HPLC)para la caracterizaciónbioquímica de la fermentación delalmidón de yuca. In: [Proceedingsof the workshop on] “Avancessobre Almidón de Yuca” held atCIAT, Cali, Colombia, 17-20 June1991. (Abstr.)

Godon, B. 1981. Le pain. Pour la Science50:74-87.

Jory, M. 1989. Contribution à l’étude dedeux processus de transformationdu manioc comportant une phasede fermentation: Le gari au Togo,l’amidon aigre en Colombie.Mémoire de mastère entechnologie alimentaire régionschaudes. Ecole nationalesupérieure des industriesagricoles et alimentaires (ENSIA)and CIRAD, Montpellier, France.45 p.

142


Larsonneur, S. 1993. Influence du séchagesolaire sur la qualité de l’amidonaigre de manioc. Mémoire ingénieur.Université Tecnologique de Compiègneand CIAT, Cali, Colombia. 114 p.

Laurent, L. 1992. Qualité de l’amidon aigre demanioc: Validation d’une méthoded’évaluation du pouvoir depanification et mise en place d’uneépreuve descriptive d’analysesensorielle. Mémoire ingénieur.Université Tecnologique deCompiègne and CIAT, Cali, Colombia.88 p.

Nakamura, I. M. and Park, Y. K. 1975. Somephysico-chemical properties offermented cassava starch (polvilhoazedo). Starch/Stärke 27(9):295-297.

Pinto, R. 1978. Extracción de almidón deyuca en rallanderías. ICA (Inst.Colomb. Agropecu.) Informa 12(9):3-6.

Ruiz, R. 1988. Informe de actividad:Programa de apoyo a las empresasproductoras de almidón de yuca enel norte del Cauca. Corporaciónpara Estudios Interdisciplinarios yAsesorías Técnicas (CETEC) andServicio de Desarrollo y Consultoríapara el Sector Cooperativo y deMicro-Empresas (SEDECOM), Cali,Colombia.

__________. 1991. Agroindustria dealmidón agrio en el norte delCauca. In: [Proceedings of theworkshop on] “Avances sobreAlmidón de Yuca” held at CIAT,Cali, Colombia, 17-20 June 1991.(Abstr.)

Vanhamel, S.; Van den Ende, L.; Darius,P. L.; and Delcour, J. A. 1991. Avolumeter for breads prepared from10 grams of flour. Cereal Chem.68(2):170-172.

143

Physicochemical Properties of Cassava Sour Starch

CHAPTER 17

PHYSICOCHEMICAL PROPERTIES OF

CASSAVA SOUR STARCH1

C. Mestres*, X. Rouau**,N. Zakhia***, and C. Brabet†

* Département des cultures annuelles (CA),CIRAD, Montpellier, France.

** Institut national de recherche agronomique(INRA), Montpellier, France.

*** CIRAD/SAR, Montpellier, France.† CIRAD/SAR, stationed at the Cassava

Utilization Section, CIAT, Cali, Colombia.


Introduction

This chapter describes a collaborativestudy involving CIRAD, the Institutnational de recherche agronomique(INRA), and CIAT. The studyinvestigated the physicochemicalbases of baking pandebono, aColombian traditional bread, usingcassava sour starch samples providedby CIAT. We evaluated thephysicochemical modifications of sour(i.e., fermented) starch duringfermentation and drying and tried torelate them with the starch’sexpansion property and potential formaking pandebono.

Pandebono dough, a mixture ofsour starch, water, and cheese,expands during cooking. This impliesthat gas is produced, which expands,thus increasing the product’s volume(Figure 1). If sweet (i.e., unfermented)starch is used, expansion does notoccur, because either no gas isproduced, or it escapes from thedough.

If pandebono is to expand, as inwheat bread, gas must form and beretained. Retention supposes thatthe dough has viscoelastic propertiesthat make it gas-tight. Viscoelasticmaterials in food are generallypolymers and polymer networks.They can be proteins such as glutenin wheat bread, nonstarchpolysaccharides such as pentosans inrye bread, or exopolysaccharides frommicroorganisms often added innonwheat bread recipes (e.g.,dextran). Pentosans need oxidativereticulation, and starchgelatinization, to improve theirrheological properties. Weinvestigated all these possibilities toexplain gas retention in cassava sourstarch.

Determining the Presenceof Polymers in

Pandebono Dough

Dufour et al. (Chapter 16, thisvolume) describe collecting two sets ofsamples: (1) at different stages offermentation (0 to 33 days), followedby sun-drying for 8 h, and (2) aftercomplete fermentation (33 days) andat different stages of sun-drying (0 to8 h). The resulting loaf volumes areshown in Figure 2.

We first determined the nitrogencontent of the starch samples, which

144


In unfermented starch,gas bubbles escape

In sour starch, gasbubbles are retained

was very low, ranging between 0.3and 0.6 g/kg and implying about0.2% as protein (Figure 3).Fermentation even decreased nitrogencontent slightly. Proteins are,therefore, highly unlikely to directlyinfluence the viscoelastic properties ofsour starch.

The samples’ pentosan contentranged from 6 to 7 g/kg and did notchange significantly with the durationof fermentation (Figure 4). Thesepolysaccharides originated fromresidual fibers of cassava roots,rather than from production ofpentosan-like polymers bymicroorganisms during fermentation.

The molecules of phenoliccompounds are efficient oxidativereticulation agents of pentosans,improving their functional properties.They can also absorb ultraviolet light,which greatly increases their activityas oxidative agents. However, wefound only traces of ferulic acid,which probably originated fromresidual fibers. We could notdetermine whether sour starchcontains dextran because this

Loa

f vo

lum

e (c

m3/g)

Figure 2. Loaf volumes of pandebono (traditionalColombian cheesebread) obtained intwo sets of samples. Data fromDufour et al., Chapter 16, thisvolume.

8

6

4

2

00 2 4 6 8

Sun-drying (hours)

After 33 days of fermentation

8

6

4

2

00 10 20 30 40

Duration of fermentation (days)

After 8 h of sun-drying

Figure 1. Presumed mechanism of expansion in pandebono dough during cooking. Pandebono is atraditional cheesebread eaten in Colombia.

Mixing CookingShaping

Starch

Water

Cheese

145


molecule is very similar to starch andcannot be separated from it.

Protein and pentosan contents ofpandebono dough are too low to havea significant influence on theviscoelastic properties of sour starchin general, and on gas retention inparticular.

The Role of Gelatinizationin Gas Retention

Apart from cheese, starch remains themain component of pandebono(95%-98% dry matter). We thereforestudied the gelatinization andrheological properties of starch todetermine whether fermentation anddrying modify it in a way that wouldexplain sour starch’s ability to expandand retain gas.

We determined starch’s thermalproperties by using differentialscanning calorimetry (DSC). Weheated starch at a constant rate of10 °C/min and measured theheat-flux between 35 to 140 °C. Thisway we could determine thegelatinization onset temperature (theintercept of base line and tangent tothe energy change) and the enthalpychange (the area of heat flux duringthe gelatinization transition)(Figure 5).

Table 1 gives the results for themost significant samples: unfermentedstarch, oven-dried sour starch, andsun-dried sour starch. Only the lastsample expanded well. The cassavasamples did not differ markedly intheir thermal properties: for all,gelatinization temperature was close to60 °C and enthalpy change to 16 J/g.Fermentation and drying did notsignificantly modify the thermalproperties of starch crystallites, thusthe specific expansion property of sourstarch cannot be explained by changesin crystallites.

Figure 3. Nitrogen content (g/kg) of cassavasour starch samples.

Nit

roge

n (g

/kg)

0.6

0.5

0.4

0.30 10 20 30 40



0.6

0.5

0.4

0.3

Sun-drying (hours)


0 2 4 6 8

123451234512345123451234512345

12341234

1234123412341234123412341234

123412341234

12345123451234512345123451234512345

1234123412341234 1234

123412341234123412341234

123412341234 1234

123412341234123412341234

1234123412341234

123451234512345123451234512345

1234123412341234 1234

12341234123412341234

1234123412341234 12345

123451234512345123451234512345

1234123412341234

Con

ten

t of

pen

tose

s (%

dry

basi

s)

8

6

4

2

00 3 7 9 13 19 26 33


Figure 4. Sugar content of cassava sour starchsamples after total acid hydrolysis(glucose is not reported). ( = xylose;121212 = arabinose;

123123123 = ribose;

= rhamnose.)

146


Table 1. Thermal propertiesa of cassava starch samples.

Sample Fermentation Sun-drying In pH 4.0 In water In pH 7.0of starch (days) (h) buffer buffer

GT EC GT EC GT EC GT EC GT EC

Maize 0 0 0 0 66.2 14.3 - - 69.0 14.7Cassava 1b 0 0 8 8 60.5 16.6 59.8 16.7 63.9 17.4Cassava 2c 33 33 8 8 60.6 15.5 60.2 16.1 63.5 17.3Cassava 3d 33 33 0 0 60.5 16.1 60.3 16.3 64.0 16.7

a. GT = gelatinization temperature (° C); EC = enthalpy change (J/g dry basis).b. Unfermented cassava starch.c. Sour, oven-dried cassava starch.d. Sour, sun-dried cassava starch.

19

17

15

13

Hea

t fl

ux

(mW

)

40 60 80 100

Temperature (°C)

Figure 6. Viscosity profile of cassava sour starch observed by using a Rapid Visco Analyzer (RVA).( = viscosity; ...... = temperature.)

Figure 5. Enthalpy change (EC) andgelatinization temperature (GT) ofcassava starch observed withdifferential scanning calorimetry.

200

160

120

80

40

0

90

70

50

300 5 10 15 20 25

Time (minutes)

We then characterized therheological properties of cassavastarches. We madeviscoamylographic determinationswith the Rapid Visco Analyzer, asimilar device to the Brabenderviscoamylograph. We measuredpasting temperature, maximumviscosity, and gelification index(Figure 6).

Our results (Figure 7) confirmthose obtained at CIAT (Dufour et al.,Chapter 16, this volume):

(1) The pasting temperature is similarfor all samples (which matchesthe DSC measurements).

Pasting temperature

Vis

cosi

ty (R

VA

un

its)

Gelification

Maximum viscosity

Tem

per

atu

re (°C

)

EC

GT

147


80

60

40

20

0

Past

ing

tem

per

atu

re (°C

)

80

60

40

20

0

0 10 20 30 40

0 2 4 6 8 10

Duration of fermentation after8 h of sun-drying (days)

Sun-drying after 33 daysof fermentation (hours)

250

200

150

100

50

0

250

200

150

100

50

0

Figure 7. Variation of pasting temperature (- -)and maximum viscosity (- -) for twosets of samples of cassava sour starchin phosphate buffer at pH 7.0.(RVA = Rapid Visco Analyzer.)

(2) The maximum viscosity decreasedwith increased duration offermentation and sun-drying.This figure seemed related to theloaf volume of pandebono: thelower the maximum viscosity, thehigher the loaf volume.

These observations were madewith samples in a pH 7.0 bufferwithout amylase inhibitor. However,pH did have a significant effect(Figure 8): in fermented and sun-driedsour starch, which had the bestexpansion property, maximumviscosity continuously decreased aspH increased from 4 to 10. Thisphenomenon did not occur forstarches unsuitable for pandebonomaking, such as unfermented starchor oven-dried sour starch.

Consequently, the maximumviscosity is similar for all samples inan acid medium. We hypothesizedthat the sour starch with the bestexpansion property may contain anamylase that hydrolyzes the productduring measurement, lowering theviscosity of the medium. Because thisamylase should be active in neutraland basic pH, we tried to determineamylase activity within this sample byestablishing the presence of reducingsugars and starch solubility. That is,if exo-amylase activity exists,reducing sugars should be releasedwith time, but if endo-amylaseactivity exists, then starch solubilitywould increase with time.

In fact, we did not find change ineither of these two parameters. Thisindicates that amylase activity iseither very low or nonexistent.

We then investigated themacromolecular structure of starch,determining intrinsic viscosity bymaking the starch soluble with alkali(pH 13). Intrinsic viscosity representsthe hydrodynamic volume of themolecules (polymers) and depends ontwo factors: first, the molecularweight of the polymer—the higher themolecular weight, the higher the

Maxi

mu

m v

isco

sity

(R

VA

un

its)

Maxi

mu

m v

isco

sity

(RV

A u

nit

s)

200

160

120

80

40

04 6 8 10

pH

Figure 8. Influence of pH on the maximumviscosity of cassava sour starchsamples. (... ... = unfermented,sun-dried; o = fermented 33 days,artificially dried; -- -- = fermented33 days, sun-dried; RVA = Rapid ViscoAnalyzer.)

148


intrinsic viscosity—and, second, theconformation of the molecules—theless “folded” they are in solution, thehigher the intrinsic viscosity.

The intrinsic viscosity follows thesame pattern as the maximumviscosity observed on the amylograph(Figure 9): it decreases within the firstdays of fermentation and within thefirst hours of sun-drying.

A marked relationship thereforeexists between intrinsic viscosity andthe pandebono’s loaf volume: thelower the viscosity, the higher the loafvolume (Figure 10). We confirmedthis relationship with another set ofsamples from a different cassavavariety (CMC 40). How can thelowering of intrinsic viscosity beexplained?

Conclusions

We did not see any differences in thecrystalline structure of starch, but weshowed that viscosity of solubilized(intrinsic viscosity) or dispersed(viscoamylograph) starch decreaseswith increased fermentation andsun-drying. Such reduction in

(1) A decrease of molecular weight—unlikely, because of the lack ofamylase activity.

(2) A change in macromolecularconformation and increasedconvolution through interactionwith other molecules. Suchinteraction facilitates polymerfolding.

Intr

insi

c vi

scos

ity

(ml/

g)

180

160

140

120

100

800 10 20 30 40


200

170

140

110

80

500 2 4 6 8 10



Figure 9. The intrinsic viscosity of two sets ofsamples of cassava sour starch.

Sun-drying (hours)

Loa

f vo

lum

e (c

m3/g)

100 120 140 160 180 200

6

4

2

0

8

6

4

2

0

Intrinsic viscosity (ml/g)

Figure 10. Relationship between intrinsicviscosity of sour starch from twocassava varieties and pandebono loafvolumes. ( = variety ‘Amarga’; = variety ‘CMC 40’; pandebono = a

traditional cheesebread eaten inColombia.)

149


viscosity seems related to thepandebono’s loaf volume, but isobserved only at neutral and basicpH, and after 2 h of sun-drying butnot after 33 days of fermentation. Itis not linked to an amylase or acidicdegradation of starch.

We can only propose somehypotheses to explain our results.The starch may have undergone anoxidative degradation (possible inoven-drying). Or interactions may

have occurred with other molecules,either lactates or derivatives of lacticacid. Lactates cause starch toplasticize, and the effect is so notablethat a patent has recently been takenout. Through lactates starchderivatives can be obtained. Volatilederivatives of lactic acid maycontribute to gas production, and tothe flavor and smell of sour starch. Ifflammable, their flame colors couldindicate the quality of a sour starch.

150


CHAPTER 18

INFLUENCE OF GELATINIZATION

CHARACTERISTICS OF CASSAVA STARCH

AND FLOUR ON THE TEXTURAL

PROPERTIES OF SOME FOOD PRODUCTS

S. N. Moorthy*, J. Rickard**, and J. M. V. Blanshard***

Abstract

Cassava flour contains fiber, sugars,and smaller quantities of lipids andother components. It exhibitsproperties different from those ofcassava starch, which cooks to a morecohesive paste. The gelatinizationcharacteristics of starch and flour,extracted from selected cassavacultivars, were examined. The peakviscosity of flour was generally lowerthan that of starch, although morestable. Swelling volumes were alsocorrespondingly lower. Thegelatinization temperature of flour,whether ascertained by differentialscanning calorimetry or viscography,was consistently several degreeshigher than that of starch. Thelower peak viscosity and highergelatinization temperature probablycontribute significantly to the texturaldifferences between flour and starch.Defatting and ethanol extraction hadlittle influence on the gelatinizationcharacteristics of either starch orflour, indicating that fiber, rather than

lipids or sugars, probably makes themost important contribution to flourtexture. The importance of thesefindings to the texture of food productsmade from cassava flour and starch isdiscussed.

Introduction

Cassava is an important root crop inmany tropical countries, where thestarchy and tuberous roots are eatenin various forms, including as starchand flour. The starch is extracted by awet process and the flour obtained bymilling dried chips.

The texture of cooked roots differswidely between cultivars, andconsiderable work has been carried outto identify reasons for this variability(Asaoka et al., 1992; Kawano et al.,1987; Moorthy et al., 1993a;Safo-Katanka and Owusu-Nipah,1992; Wheatley et al., 1993). Fewconclusions have been reached but thequantity and quality of the starch inthe root and the presence of variousnonstarchy polysaccharides areconsidered important.

Several differences also exist inthe rheological and functionalproperties between starch and flour.We attempted to identify the reasonsfor these differences.

* Division of Postharvest Technology, CentralTuber Crops Research Institute (CTCRI),Trivandrum, India.

** Natural Resources Institute (NRI), Chatham,UK.

*** University of Nottingham, UK.

151

...Gelatinization Characteristics of Cassava Starch and Flour...


Starch and flour were obtained fromfive cultivars of freshly harvestedcassava roots (M-4, H-165, H-1687,S-856, and H-97), each havingdifferent cooking qualities. The mainconstituents of the samples—starch,fiber, lipids, and sugars—weredetermined by standard procedures.To assess the influence of lipids,samples were defatted by extraction(Soxhlet), using petroleum ether(40-60 °C). Ethanol extraction wassimilarly undertaken, with 80%ethanol (Soxhlet, 6 h), to examinethe influence of sugars andethanol-soluble components.

Differential scanning calorimetry(DSC) data were obtained by usingPerkin Elmer DSC-2 equipment withIndium as a standard (temperaturerange 25-100 °C, at a heating rate of10 °C/min). Gelatinization profiles ofthe samples (5%) were obtained on aBrabender Viscoamylograph (350 cmg[torque] sensitivity cartridge, heatingrate 1.5 °C/min). Swelling volumeswere determined at 95 °C (Schoch,1964).


Table 1 presents the results of thechemical analyses of starch and flourfrom the five cassava varieties.Starch content on a dry weight basiswas 98% or more in all the starchsamples and between 79.1% and86.0% in the flour samples. Crudefiber content was 0.13% or less in thestarch samples, whereas it rangedfrom 1.50% to 2.98% in the flour.Earlier studies show similar starchand fiber compositions (Abraham etal., 1979).

The lipid content, by nature muchlower than that of cereals, varied from0.11% to 0.22% in the starches andfrom 0.25% to 0.56% in the flours.

Lipids, in common with manysurfactants, significantly affect starchby complexing strongly with amyloseand amylopectin side chains,rendering these less labile (Krog,1973). This capacity has beenexploited for reducing thecohesiveness of potato starch products(Hoover and Hadziyev, 1981, 1982).Low levels of surfactants can have aprofound effect on cassava starch(Moorthy, 1985). Ethanol-solubleconstituents in the flours ranged from2.5% to 3.7% and only 0.9% to 1.3%in the starch samples (data notshown). The predominant sugar incassava flour has been identified assucrose.

The recorded gelatinizationtemperatures (Table 2) reveal aconsistent difference between the flourand starch samples. Comparingvalues for initial, maximum, and endtemperatures, the results for flourare each 2-3 °C higher than for thecorresponding starches. Componentswithin the flour, by restricting accessof water into the starch granules, candelay gelatinization. Surfactantsand lipids, by forming complexes, areknown to raise gelatinizationtemperatures (Osman, 1967).However, the defatted andethanol-extracted flours had the samevalues as native flour, indicating thatneither lipids nor sugars wereresponsible for enhancedgelatinization temperatures. Recentexperiments on cassava starch showcorrelation between higher fibercontent and higher gelatinizationtemperatures (Moorthy et al., 1993a).Thus, the elevation in gelatinizationtemperature of flour may be attributedto the fiber.

The DSC peak patterns of starchand flour from the same cultivar weresimilar. ‘H-97’ starch and flour had acharacteristic shoulder in their peaks,whereas ‘M-4’ starch and flour hadtypically broad peaks. The peak

152


Table 1. Biochemical constituents of starch and flour made from five cassava varieties.

Variety Product Starch Sugar Lipids Crude fiber(%) (%) (%) (%)

M-4 Starch 98.1 - 0.11 0.11Flour 86.0 2.20 0.45 1.50

H-165 Starch 98.0 - 0.22 0.13Flour 79.1 3.49 0.27 2.98

H-1687 Starch 98.2 - 0.18 0.15Flour 80.5 2.72 0.29 2.23

S-856 Starch 98.5 - 0.20 0.12Flour 81.2 3.23 0.56 2.56

H-97 Starch 98.3 - 0.20 0.11Flour 82.7 3.05 0.25 2.70

Table 2. Data from differential scanning calorimetry (DSC) of cassava starch and flour.

Variety Product Temperature (°C) ∆Ha (cal/g)

Initial Maximum End

M-4 Starch 68.10 73.24 78.54 2.95Flour 71.11 75.67 81.29 2.02

H-165 Starch 65.35 69.22 74.86 3.27Flour 68.65 72.02 77.19 2.14

H-1687 Starch 67.12 71.45 75.39 2.15Flour 70.02 73.90 79.11 2.22

S-856 Starch 65.62 70.14 74.94 2.65Flour 68.72 72.92 76.95 2.09

H-97 Starch 69.36 72.29 77.13 3.43Flour 71.82 75.02 79.92 2.27

a. ∆H = Enthalpy change.

patterns were not modified bydefatting or by ethanol extraction,indicating dependence on the starchgranular structure. The enthalpy ofgelatinization of flour was lower thanthat for starch for every variety andneither defatting nor ethanolextraction affected the values to anynoticeable extent. However, the lowerenthalpy for the flour can beattributed in part to the lower starchcontent of the samples. The enhanced

gelatinization temperatures in theDSC results for the flour samplesare supported by the Brabenderviscographic data (Table 3), whichshow that pasting temperatures forflours were 3-5 °C higher than for therespective starches.

The peak viscosity and viscositybreakdown for each flour weredifferent from those of thecorresponding starch, and most

153


Table 3. Viscosity and swelling properties of cassava starch and flour.

Variety Producta Viscosity (BU)b Break-down Pasting Swellingtemp. (°C) vol. (ml/g)

V97 VH

M-4 Starch 540 380 160 68-73 32.0Flour 380 320 60 71-74 28.0Starch (d.) 580 440 140 70-76 33.5Flour (d.) 380 310 70 72-75 29.5Starch (e.) 560 420 140 70-75 33.0Flour (e.) 410 380 30 72-76 29.0

H-165 Starch 940 480 460 66-78 38.5Flour 460 380 80 71-75 32.0Starch (d.) 1,000 580 420 69-82 39.5Flour (d.) 440 360 80 72-76 33.5Starch (e.) 1,000 520 480 69-83 39.5Flour (e.) 470 380 90 71-78 33.0

H-1687 Starch 540 480 60 70-81 33.5Flour 460 440 20 71-83 29.5Starch (d.) 570 510 60 71-82 33.5Flour (d.) 440 390 50 70-80 29.0Starch (e.) 520 480 40 70-83 34.0Flour (e.) 440 400 40 71-83 29.5

S-856 Starch 500 340 160 67-80 33.0Flour 440 360 80 71-86 29.0Starch (d.) 580 390 190 69-75 33.5Flour (d.) 470 380 90 70-85 29.5Starch (e.) 490 360 130 69-75 34.0Flour (e.) 740 390 80 70-89 30.0

a. (d.) = after defatting; (e.) = after ethanol extraction.b. V97 = viscosity at 97 °C; VH = viscosity after holding at 97 °C.

affecting starch granule expansion andbreakdown.

Defatting and ethanol extractionslightly enhanced paste viscosity instarch, whereas flour samples remainedunaffected. In contrast, the slightincrease in the flours’ pastingtemperature was probably due to thecontinuing presence of fiber in thedefatted ethanol-extracted samples.Similar results have been obtained in afiber-rich starchy flour extracted fromfermented roots (Moorthy et al., 1993b).According to Osman (1967), high levelsof sugars are needed to bring aboutperceptible changes in the viscosity ofstarch. The absence of significantchanges in the peak viscosity of floursand starches thus indicates that sugars

pronounced in the comparativelymuch lower peak viscosity of flour from‘H-165’. Again, the lower starchcontent in the flour samples canaccount in part for the low readings.However, while the viscosity was lower,it was more consistent throughout thetemperature program. Stabilizationoccurs through the presence ofnonstarchy components in the flour.Lipids, although known to stabilizestarch viscosity (Krog, 1973), had littleeffect here. The absence ofstabilization in the ethanol-extractedsamples indicates that sugars were notinvolved either. The reduced viscositynoted in all varieties was mostpronounced in ‘H-165’, which had thehighest crude fiber content. Stabilitymay therefore result from the fiber

154


do not greatly influence viscosity incassava.

Neither defatting nor ethanolextraction affected swelling volumes ofthe starches and flours, furtherindicating the influence of fiber inmodifying starch rheological propertiesin the flour.

Cassava starch, cooked in water,generally gives a cohesive, long paste,whereas flour texture is less cohesive.Cohesiveness is attributed to thebreakdown of starch molecules duringheating and stirring. Earlygelatinization can render starch moresusceptible to breakdown because itundergoes a longer period of shear.High swelling necessitates theweakening of associative forces andthus easier breakdown of starch. Thefiber may act as a barrier to earliergelatinization and to higher swelling,reducing the cohesiveness of the paste.

Starch can act, for example, as abinder, thickener, or glazing agent indifferent foods (Smith, 1982[?]). Inproducts where a cohesive texture isdesired, such as gravies and puddings,starch would be favored, whereas inproducts where a nonsticky consistencyis sought, flour would be more suitable.These mirror the findings ofcomparative studies conducted at theCentral Tuber Crops Research Institute(CTCRI) for starch and flour in locallyproduced extruded products (whenstarch was used, the product tended tobe hard and oily; with flour, the sameproduct was crisp and nonsticky).Similarly, food items prepared fromstarchy flour made from fermentedroots had a higher fiber content andbetter texture.

The study thus clearly indicatesthat fiber is a significant determinantof the characteristics and functionalproperties of cassava starches andflours. Future work should examinespecific fiber components (e.g., cellulose

and hemicellulose) and their interactionwith starch. It should also focus on howrheological characteristics can lead todifferent functional properties in starchand flour.

Acknowledgments

We thank the Director of the CTCRIand Dr. C. Balagopalan, Head of thePHT Division, CTCRI, for their kindhelp and encouragement.

References

Abraham, T. E.; Raja, K. C. M.; Sreedharan,V. P.; and Sreemulanathan, H. 1979.Some quality aspects of a few varieties ofcassava. J. Food Sci. Tech. 16:237-239.

Asaoka, M.; Blanshard, J. M. V.; and Rickard,J. E. 1992. Effects of cultivars andgrowth on the gelatinization propertiesof cassava (Manihot esculenta Crantz)starch. J. Sci. Food Agric. 59:53-58.

Hoover, R. and Hadziyev, D. 1981. The effectof monoglycerides on amylosecomplexing during a potato granuleprocess. Starch/Stärke 33:346-355.

_______ and _______. 1982. Effect ofmonoglyceride on some rehydrationproperties of potato granules.Starch/Stärke 34:152-158.

Kawano, K.; Fukuda, W. M. G.; and Cenpukdee,U. 1987. Genetic and environmentaleffects on dry matter content of cassavaroot. Crop. Sci. 27:69-74.

Krog, N. 1973. Amylose complexing effect offood-grade emulsifiers. Starch/Stärke23:206-210.

Moorthy, S. N. 1985. Effect of different typesof surfactants on cassava starchproperties. J. Agric. Food Chem.33:1227-1232.

_______; Blanshard, J. M. V.; and Rickard, J. E.1993a. Starch properties in relation tocooking quality of cassava. In: Roca,W. M. and Thro, A. M. (eds.). Proceedingsof the First International ScientificMeeting, Cassava BiotechnologyNetwork, Cartagena de Indias, Colombia,25-28 August 1992. Working documentno. 123. CIAT, Cali, Colombia.p. 265-269.

155


_______; George, M.; and Padmaja, G.1993b. Functional properties ofthe starchy flour extracted fromcassava on fermentation with amixed-culture inoculum. J. Sci.Food Agric. 61:442-447.

Osman, E. M. 1967. Starch in the foodindustry. In: Whistler, R. L. andPaschall, E. F. (eds.). Starchchemistry and technology, vol. 2.Academic Press, New York, NY,USA. p. 163-215.

Safo-Katanka, O. and Owusu-Nipah, J.1992. Cassava varietal screeningfor cooking quality: relationshipbetween dry matter, starchcontent, mealiness and certainmicroscopic observations of theraw and cooked tuber. J. Sci. FoodAgric. 60:99-104.

Schoch, T. J. 1964. Swelling power andsolubility of granular starches. In:Whistler, R. L. (ed.). Methods incarbohydrate chemistry, vol. 4. AcademicPress, New York, NY, USA. p. 106-108.

Smith, P. S. 1982[?]. Starchy derivatives andtheir use in foods. In: Linebeck, D. R.and Inglett, G. E. (eds.). Foodcarbohydrates. AVI Publications,Westport, CT, USA. p. 237-258.

Wheatley, C. C.; Orrego, J. I.; Sánchez, T.; andGranados, E. 1993. Quality evaluationof the cassava core collection at CIAT.In: Roca, W. M. and Thro, A. M. (eds.).Proceedings of the First InternationalScientific Meeting, CassavaBiotechnology Network, Cartagena deIndias, Colombia, 25-28 August 1992.Working document no. 123. CIAT, Cali,Colombia. p. 255-264.

156


CHAPTER 19

TWO RAPID ASSAYS FOR CYANOGENS

IN CASSAVA:THEIR EVALUATION, MODIFICATION,AND COMPARISON

G. M. O’Brien* and C. C. Wheatley**

in 68% of cases compared with 66%by the T.B. assay. The T.B. assay,however, performed more reliably withlow cyanogen samples, whereas thepicrate assay was more reliable withintermediate cyanogen samples. Thesampling protocol used at CIAT forthe rapid assay of cyanogen contentsof cassava clones was also evaluated.

Introduction

Cassava (Manihot esculenta Crantz) isthe fourth most important food cropof the tropics (Cock, 1985). Anefficient source of low-costcarbohydrates, cassava is importantfor food security, particularly inAfrica, and as an industrial rawmaterial, especially in Asia and LatinAmerica. Cyanogens have long beenrecognized as a toxic component ofcassava’s edible roots and leaves.The cyanogenic contents of the rootscan vary from less than 10 to morethan 500 mg/kg, measured ashydrogen cyanide (HCN), on a freshweight basis (fwb).

Guignard first developed thealkaline picrate assay for cyanide in1906. It was introduced in asemiquantitative, rapid format toCIAT and the International Instituteof Tropical Agriculture (IITA), Nigeria,during the 1970s as a routine assayin cassava breeding programs. The

Abstract

Two rapid, semiquantitative assaysfor total cyanogens in cassava wereevaluated. These were the rapidpicrate paper test, now well-known,and a recently proposed, rapid, papertest that involves the reagent tetrabase (T.B.; 4,4'-methylenebis-[N,N-dimethylaniline]). A precisecolorimetric assay was used ascontrol. After preliminary evaluation,both assay methods underwentmodification to improve accuracy ofscoring. As a result, the reliability ofthe picrate assay was greatlyimproved. The T.B. assay wasmodified in the interests of safety.Evaluation of the latter assay over arange of temperatures from 20 to35 °C showed no significant effects oftemperature on performance whenthe new scoring system was used.The level of endogenous linamaraseactivity in each sample was aninfluential factor in rapid assayperformance. In a series ofcomparative trials in three distinctecosystems, the newly modifiedpicrate assay produced correct results

* Natural Resources Institute (NRI)/CIAT,currently Research Fellow at CIAT.

** Centro Internacional de la Papa (CIP),stationed at Bogor, Indonesia.

157

Two Rapid Assays for Cyanogens in Cassava:...

rapid picrate assay is based onproducing a red-brown color whoseintensity increases with the quantityof HCN liberated from the sample.The HCN is liberated by autolysisand results from the hydrolyticaction of the endogenous enzyme,linamarase, on cyanogenicglucosides in the sample. Thepicrate system, when usedquantitatively, has been observed toproduce very high results fromcassava compared with otherquantitative methods(Izomkun-Etiobhio and Ugochukwu,1984; Mendoza et al., 1984). Theassay has been criticized for falselydetecting cyanogens in acyanogenicsamples (Nahrstedt, 1980). Therapid picrate assay originally usedwith a 9-point scale at CIAT hasbeen criticized for giving poorcorrelation between cyanogencontent and result (CIAT, 1993).

More recently, an alternative,rapid semiquantitative assay wasdeveloped, based on the reagenttetra base (T.B.; [4,4'-methylenebis-[N,N-dimethylaniline]). Also knownas Michler’s reagent, T.B. wasoriginally used for qualitative assayof cyanogens (Feigl and Anger,1966). The modified T.B. methodwas proposed as more sensitive andrapid than picrate (Bradbury andEgan, 1992). In this test, ablue-violet color forms, whichincreases in intensity and in violethue the higher the sample’scyanogen content.

We have evaluated both thepicrate and the T.B. rapid methods,using a reliable quantitativecolorimetric assay (Cooke, 1978;O’Brien et al., 1991) as control. Theassay method(s) selected after thisevaluation would be expected toperform well in any cassava-growingenvironment. Because ambienttemperatures in the tropics andsubtropics vary considerably, we

have carried out controlledevaluation and comparison of thetwo assay methods in three distincttropical environments in Colombiaunder field conditions.

Reagents

Picric acid (99%), copper (II) acetate(99%), toluene (99.5%), sodiumcarbonate (99.5%, anhydrous)(E. Merck, Darmstadt, Germany),and T.B. (Sigma Chemical Company,St. Louis, USA) were used for therapid methods. Reagents used forthe colorimetric quantitativecyanogen assay were as describedby O’Brien et al. (1991).

Alkaline picrate mixture wascomposed of picric acid (5 g) andsodium carbonate (25 g, anhydrous)dissolved in water and made up to1 liter.

The T.B. mixture was asdescribed by Bradbury and Egan(1992), a 1:1 (v/v) mixture producedon a daily basis, using two reagents:

(1) Copper (II) acetate, 3 g/L in 15%acetic acid.

(2) Tetra base, 3 g/L in acetone.

According to Bradbury and Egan(1992), the two reagent solutionsshould be stable for several months.But we noted a slight darkening ofthe reagent mixture made fromreagents 3 weeks old and older.Reagents were therefore freshlymade every 3 weeks.

Because of the highly toxic andcarcinogenic nature of T.B., nitrilerubber gloves were used in handling.The T.B. reagent was prepared andstored in a fume-cupboard.

Reagents for assay of linamaraseactivity were as described by Cooke(1979).

158


The Tetra Base Assay

The T.B. rapid assay of Bradburyand Egan (1992) represented a newdevelopment in cassava cyanogenassay, which had been tested, usingonly low-cyanogen cultivars. Theassay was also reported to be alittle more rapid at higher ambienttemperatures, suggesting thatchanges in temperature may affectthe endogenous linamarase in agiven sample (J. H. Bradbury, 1992,personal communication).

This reference to linamaraseenzyme also prompted interest inthe relationship between theamount of endogenous linamarasein a given sample, its activity undergiven environmental conditions,and, consequently, the reliability ofautolytic assays. Thus, evaluatingthe T.B. assay in some detailbecame necessary. At CIAT,Palmira, the T.B. assay was run atdifferent temperatures, with anassay of endogenous linamaraseactivity in parenchyma surroundingthe sample taken for the T.B. assay.The quantitative, colorimetric assayof cyanogen content was used ascontrol. The duration of the assaywas determined after constantlyobserving samples under assay for3 h and after an overnight period.The testing of the method resultedin a 1-h assay.

Conducting the assay

The T.B. assay was carried out inquadruplicate. A central disc wassliced out crosswise from a cassavaroot and parenchymal plugsremoved (Figure 1). The plugs weretrimmed with a scalpel to 0.5 cmthick, placed individually in smallglass vials (2 x 5 cm), and sealedwith tightly fitting, plastic stoppers.The plugs were maintained in thevials for 1 h before assay, to allow

buildup of HCN in the vial. The assaywas then carried out.

To start the assay, the stopper inthe sample vial was replaced by asimilar stopper with a T.B. test-paperattached, so that the paper wassuspended inside the vial, 1 cm abovethe sample. The blue-violet colorproduced at the bottom end of thepaper was recorded after 10 and60 min. The result was interpreted interms of cyanogen content.

The T.B. test-paper was made asfollows: paper strips (Whatman’sno. 1 filter paper, 4 x 1 cm) wereattached to clean vial stoppers withadhesive tape. A 1-cm portion at oneend of the paper was attached to thestopper, leaving a length of 3 cm toact as support for the T.B. mixture.The stoppers and papers were placedwithin the fume cupboard. One dropof T.B. mixture was placed on eachpaper at the end away from thestopper. The drops of mixture wereleft to soak through before each paperwas sealed in an empty vial for safety,before assay.

To evaluate the T.B. assaymethod, assays were made of 72cassava roots from 10 varieties,ranging between 10 and 456 mg/kgtotal cyanogens (as HCN, fwb). Ineach case, the root was first sampledfor T.B. assay, then the rest of thepeeled root was assayed for both totalcyanogens and endogenouslinamarase activity (Figure 1) byquantitative colorimetric assay(Cooke, 1978; O’Brien et al., 1991).

Tetra base assays were carriedout at 20, 25, 30, and 35 °C, using anincubator. At least one root fromeach of the 10 cassava clones used inthe experiment was assayed at eachtemperature. The samples werestored in the incubator during the 1-hpreassay period and during the assay

159


Figure 1. Steps in sampling a cassava root for rapid and quantitative cyanogen assays. (A) The root isfirst measured longitudinally and a disc, 1 cm thick, is removed from the center. (B; i) Thegeometric radius of the disc is measured from the center to the inner edge of the peel. (ii) Halfway along a radius, a 1-cm round plug is removed with a borer. The plug is used for the tetrabase assay. The sampling is replicated four times. (iii) From the space between two plugs,again half way along a radius, a 1-cm cube is cut out with a scalpel. The cube is used for thepicrate assay. (iv) The rest of the root is peeled and the parenchyma chopped into cubes ofabout 1 cm3. From these, a random sample of 50 g is quantitatively assayed for totalcyanogens. Where required, another 50-g sample is taken for a linamarase enzyme assay.

Table 1. Tetra base rapid assay, groupingformat.

Range Score Total cyanogens(mg/kg as HCN, fwb)

1 <8.5 0-502 <8.5 after 10 min,

>8.5 after 1 h 50-1003 >8.5 after 10 min >100

itself. A 10-point numerical scale wasdevised, using the “Munsell colorguide.” It was based on the intensityof color attained, differing from thatused by Bradbury and Egan (1992).The scale ran from very pale blue todeep violet.

Figure 2 shows the scoresobtained by 72 roots in a 1-h test oftheir cyanogen contents. Therelationship between total cyanogensand T.B. score in the 1-h test waslinear only to about 50 mg/kg (asHCN, fwb). Most samples withcyanogen contents greater than50 mg/kg produced a score between8.5 and 10. Roots containing morethan 100 mg/kg, almost withoutexception, gave scores between

8.5 and 10 within 10 min.Accordingly, a system of scoring(Table 1) for total cyanogens incassava parenchyma was devised.The maximum permitted error in thisgrouping method was ± 1 mg/kg. Asample with a cyanogen content of50.9 mg/kg could therefore be classedeither in range 1 or range 2.

A

B

160


sample is neither grated normacerated in buffer and where asignificant degree of autolyticbreakdown of cyanogens is soughtwithin 1 h. Cooke and De la Cruz(1982) found that 24 h in excessbuffer was needed for a completeautolytic breakdown of cyanogens ina cassava sample. Yet, 84.7% of allresults in this experiment werecorrectly assayed, regardless ofendogenous linamarase activity andof temperature within the rangestated (20 to 35 °C).

This experiment and its resultsare described in greater detailelsewhere (O’Brien et al., 1993).

Picrate Assay

Plant breeding programs have beenroutinely using the alkaline picraterapid assay for about 20 years,unlike the relatively untested T.B.assay. The original scoring format,used for the picrate assay, assigneda cyanogen-content range to eachof the nine points on thecolor scale:

Of the 72 assays carried out inthis experiment, 61 results proved tobe correct within the groupingformat shown in Table 1. Thus,84.7% of all results were correctlyassayed.

Influence of endogenouslinamarase activity

During the T.B. experimental workat CIAT, the endogenous linamaraseactivity of roots being assayed wasmeasured colorimetrically. Activityranged between 0.03 and 0.63enzyme units per gram (EU/g, fwb).The T.B. score and colorimetricassay correlated better when datapoints from samples with less than0.2 EU/g were removed, reducingthe total from 72 to 43 cases. TheSpearman correlation of linearityincreased from 0.77 to 0.85. Hence,T.B. performance in roots withlinamarase activity below 0.2 EU/gwas negatively affected.

It is unsurprising that asample’s endogenous linamaraseactivity should affect results in asystem where the parenchymal

12

10

8

6

4

2

00 50 100 150 200 250 300 350 400 450 500

Tet

ra b

ase

sco

re

Figure 2. Evaluation of the tetra base assay used at CIAT, Palmira, to determine the cyanogen contentof 72 roots from 10 cassava varieties.

Total cyanogens (as HCN, mg/kg, fwb)

161


toluene were placed on to the sample.The tube was tightly sealed with arubber stopper, entrapping a strip ofpaper (Whatman’s no. 1, 6 x 1 cm),saturated in alkaline picrate mixtureand suspended above the sample.After 10 h (Pivijay) or 12 h (Palmiraand Cajibío), the resultant colorchange was noted and interpreted interms of cyanogen content.

Field-Based Comparison ofthe Two Rapid Methods,

with Colorimetric Control

Three sites were selected for thiswork:

(1) Cajibío (near Popayán,southwestern Colombia): ahighland area, 2,000 m above sealevel, temperatures were 19 to26 °C during trials.

(2) CIAT (Palmira, southwesternColombia): mid-altitude, 1,000 mabove sea level, temperatures were26 to 33 °C during trials.

(3) Pivijay (North Coast, Colombia): atsea level, temperatures of 30 to34 °C during trials.

In this experiment, 100 rootsfrom 12 different clones were assayed.Each root was sampled for boththe rapid T.B. assay and the rapidpicrate assay (Figure 1). The restof the root parenchyma thenunderwent quantitative colorimetricassay.

Table 2. Scoring ranges and levels of accuracy for tetra base and picrate rapid assays.

Group Scoring ranges Assay accuracy (%)

1 2 3 TB Picrate Q. color.

A 0-50 50-100 100+ 66 68 100

B 0-40 40-90 90+ 61 73 100

C 0-60 60-110 110+ 68 65 100

Score Total cyanogens(mg/kg, as HCN, fwb)

1 = pale yellow <102 10 - 153 15 - 254 25 - 405 40 - 606 60 - 857 85 - 1158 115 - 1509 = dark brown >150

Some time before this projectbegan, CIAT had made a limitedevaluation of the picrate rapid assay.The original scoring format, with its9-point scale, was regarded asunworkable because of errors (CIAT,1993). The method of scoring wasthus modified into anecdotal ranges:

Score Cyanogen contents

1-4 Low

5-7 Medium

8-9 High

With the picrate assay, weevaluated only this “anecdotal”scoring system, assigning specificquantities to the anecdotal rangeslisted above. The assay wascompared with the evaluated andmodified T.B. assay, under fieldconditions, using the quantitativecolorimetric cyanogen assay as acontrol.

For the picrate assay, a cube wascut from the parenchymal disc takenfrom the root (Figure 1) and placedin a 12-cm test tube. Five drops of

162


Using the results of the 100analyses carried out, three rangegroups of total cyanogen content(measured in mg/kg of HCN, fwb)were considered for scoring in eachrapid assay and their levels ofaccuracy evaluated (Table 2). GroupB gave 73% correct results for picratebut only 61% for T.B. Group C gave65% for picrate and 68% for T.B.Group A gave the most favorableproportion of correct results for boththe T.B. (66%) and picrate (68%)methods.

Of the three groups, group A wasadopted. Hence, the picrate rapidassay gave slightly but notsignificantly better results than theT.B. rapid assay. Also, the picratescoring format was slightly modified:the score 7 was reassigned to highinstead of intermediate cyanogencontent. The scores were thereforegrouped as shown in Table 3.Figure 3 shows the results ofcomparing the two assays.

1 2 3

1 9 3 4

2 2 3

3 2 17

1 2 3

1 1 1

2 7 8 13

3 10

1 2 3

1 2 7

2 7 3

3 2 9

1 2 3

1

2 2 13 2

3 3 10

1 2 3

1 1 1

2 9 7

3 2 10

1 2 3

1

2 1 12 3

3 14

Picrate assay score Tetra base assay score

Figure 3. Performance of picrate and tetra baseassays in measuring total cyanogensin cassava grown in three ecosystems:A = CIAT, Palmira, mid-altitudes,southwestern Colombia; B = Cajibío,highlands, southwestern Colombia;C = Pivijay, sea level, North Coast,Colombia. Values indicated on thegraphs’ axes correspond to cyanogencontent ranges (mg/kg, as HCN, fwb):1 = 0-50; 2 = 50-100; 3 = > 100.Correct results are reported in shadedboxes. Maximum error permitted is± 1 mg/kg.

A

B

C

intermediate (range 2) and even somehigh (range 3) cyanogen samples. Thepicrate assay wrongly classified anumber of high and some low cyanogensamples as “intermediate.” Forsamples classified as “high” the IQRswere almost identical. Despite a smallnumber of intermediate resultswrongly classified as “high,” nearly allsamples with cyanogen contents of100+ mg/kg or more were correctly

Sensitivity to cyanogen content

A statistical study was carried out onthe data from this experiment, usinga box plot model (SAS statisticalanalysis package). It was found thatthe methods produced differentinterquartile ranges (IQR,representing the central 50% ofresults) for cyanogen content.Range 1 denoted low cyanogencontent and range 2 intermediate.The T.B. assay wrongly classified as“low” (range 1) a number of

Table 3. Picrate rapid assay, revised groupformat.

Range Score Total cyanogens(mg/kg as HCN, fwb)

1 1-4 0-502 5-6 50-1003 7-9 >100

163


placed within range 3 in bothmethods.

The findings of the three trialssuggest that the picrate assayperforms more reliably than the T.B.with roots containing 50-100 mg/kgas HCN, whereas the T.B. assay ismore reliable than picrate with rootscontaining less than 50 mg/kg.

Comparing the two assays acrossecosystems

Figure 3 shows that the picrate assayperformed less well than the T.B. atCIAT, Palmira (almost 48% correct,compared with 70%, respectively,Figure 3A). But, at Cajibío, thepicrate was better than the T.B. assay(87% compared with 67%,respectively, Figure 3B). At Pivijay,results were 77% for picratecompared with 60% for T.B.(Figure 3C).

This apparent irregularityprobably had more to do withsamples’ cyanogen content than withother factors (temperatures weresimilar at CIAT and Pivijay). At CIAT,of the eight clones used, twoproduced roots whose cyanogencontents belonged to range 1. Theseroots were all erroneously classifiedunder range 2 in the picrate assay,whereas the T.B. assay correctlyclassified them under range 1. In theother two trials, no low-cyanogenclone was used and the erroneouspicrate results therefore did not recur.

Interpreting the results of therespective assays by means of thesystem chosen, no significantdifference in overall performance wasfound between the two methods, withoverall success rates of 68% (picrate)and 66% (T.B.).

That both methods producedresults of which more than 30% wereerroneous suggests that neither is

particularly good for rapid screening.But variations in sample linamaraseactivity constitute an inbuilt source oferror for any rapid autolytic assay ofcyanogens in cassava. Nonautolyticmethods (for example, destruction ofthe endogenous linamarase of thesample, followed by adding an excessquantity of exogenous linamarase) aremore effective in this respect, but theyare much slower for mass-screeningsamples. Autolysis is therefore stillthe only practicable system.

Evaluation of Field SamplingMethod

In the comparison trials undertakenat Cajibío and Pivijay, the two rapidassays were compared, using thesame root samples. The CIAT plantbreeders’ sampling method itself wasalso tested for its representativeness,using the quantitative colorimetricassay. CIAT usually samples clonesfor mass screening from a plot of25 plants (5 x 5), selecting one plantnear the center. From this plant, oneroot is selected for rapid assay of totalcyanogens. The result obtained fromthis root is treated as representativeof the entire plot.

At Cajibío, plots containing25 plants of each clone, 13 monthsold, were used. At Pivijay, rows of sixplants per clone, 8.5 months old,were used. In each case, one plantwas harvested as representing its plotor row. From this plant, one root waschosen to represent the whole plant(and therefore the whole plot or row).At the same time, a further four rootswere taken from the same plant: theirparenchyma was pooled and triplicateextracts made. Their assay served ascontrol to show whether the chosenroot represented the whole plant.

A further four plants wereharvested from the same plot or row.The parenchyma of their roots was

164


thoroughly chopped and pooled.Triplicate assays were taken to showif the chosen plant properlyrepresented the entire plot or row,and further, if the chosen root fromthat plant properly represented theentire plot or row.

The semiquantitative, rapidassays operate with three ranges ofcyanogen content. To berepresentative, the sample root takenfrom a given plot or row had to give aresult within the same range as themean of the rest of the plot or row.Analysis of the data from the trial forrepresentativeness shows that, in 47cases out of 60 (i.e., in 78% of cases),the selected root was found to berepresentative both of the plant fromwhich it was taken and of the groupof five plants taken to represent theentire plot or row. Thus, thesampling method appearssatisfactory, although it would bedesirable to continue and expand thisinvestigation, extending it to othercassava clones.

Conclusions

In comparing the Bradbury and Egan(1992) picrate assay with a newlymodified T.B. assay, the success ratesof both were very similar—68% forpicrate and 66% for T.B.

The high toxicity and carcinogenicnature of the reagent tetra baserequires a comprehensive andcarefully controlled methodology. Interms of reagent costs, the T.B. assayis potentially less expensive than thepicrate assay, although costsassociated with safety precautionsand equipment are considerablyhigher. In view of these findings, thefollowing recommendations are made:

(1) The newly modified, picrate assayshould be used for rapid screening

of cassava cyanogens in all cases,except:

(a) where a significantly highproportion of low-cyanogenclones (0-50 mg/kg, fwb) areused (i.e., where the risk ofrejecting low-cyanogenmaterial would be high); and

(b) where a very rapid result isrequired.

(2) Under circumstances described in(1a) and (1b), the tetra base assayshould be used but only if:

(a) a well-maintained fumecupboard and good disposalfacilities are available; and

(b) workers have been trainedand are willing to applystrict safety procedures.

References

Bradbury, J. H. and Egan, S. V. 1992. Rapidscreening assay of cyanide content ofcassava. Phytochem. Anal. 3:91-94.

CIAT, Cassava Program. 1993. Activitiesduring 1989: utilization. In:Cassava Program report, 1987-1989.Working document no. 91. Cali,Colombia. p. 567-568.

Cock, J. 1985. Cassava: new potential for aneglected crop. Westview Press,Boulder, CO, USA. 191 p.

Cooke, R. D. 1978. An enzymatic assay forthe total cyanide content of cassava(Manihot esculenta Crantz). J. Sci.Food Agric. 29:345-352.

__________. 1979. Enzymatic assay fordetermining the cyanide content ofcassava and cassava products. CIAT,Cali, Colombia. 14 p.

__________ and De la Cruz, E. M. 1982. Anevaluation of enzymic andautolytic assays for cyanide incassava (Manihot esculenta Crantz).J. Sci. Food Agric. 33:1001-1009.

165


Feigl, F. and Anger, V. 1966. Replacement ofbenzidine by copperethylacetoacetate and tetra base asspot-test reagent for hydrogencyanide and cyanogen. Analyst91:282-284.

Izomkun-Etiobhio, B. O. and Ugochukwu,E. N. 1984. Comparison of analkaline picrate and apyridine-pyrazolone method for thedetermination of hydrogen cyanide incassava and in its products. J. Sci.Food Agric. 35:1-4.

Mendoza, E. M. T.; Kojima, M.; Iwatsuki, N.;Fukuba, H.; and Uritani, I. 1984.Evaluation of some methods for theanalysis of cyanide in cassava. In:Uritani, I. and Reyes, E. D. (eds.).Tropical root crops: postharvestphysiology and processing. JapaneseScientific Society Press, Tokyo,Japan. p. 235-242.

Munsell color guide. Kollmorgen InstrumentCorporation, Baltimore, MD, USA.

Nahrstedt, A. 1980. Absence of cyanogenesisfrom Droseraceae. Phytochemistry19:2757-2758.

O’Brien, G. M.; Taylor, A. J.; and Poulter,N. H. 1991. Improved enzymic assayfor cyanogens in fresh and processedcassava. J. Sci. Food Agric.56:277-289.

__________; Wheatley, C. C.; and Poulter,N. H. 1993. Evaluation of a rapidsemi-quantitative assay forcyanogensis in cassava. In: Roca,W. M. and Thro, A. M. (eds.).Proceedings of the First InternationalScientific Meeting, CassavaBiotechnology Network, Cartagena deIndias, Colombia, 25-28 August,1992. Working document no. 123.CIAT, Cali, Colombia. p. 390-399.

166


CHAPTER 20

ACUTE POISONING IN TANZANIA:THE ROLE OF INSUFFICIENTLY

PROCESSED CASSAVA ROOTS

N. L. V. Mlingi*

cyanogenic glucosides andcyanohydrins from the roots andprevent poisoning are urgently neededin this area. An intervention programhas been established to develop anextension package for cassavaprocessing, and to make thepopulation aware of the problem andadopt more efficient processingmethods.

Introduction

The advantages of cassava as a foodsecurity crop in sub-Saharan Africausually outweigh the nutritionaldrawbacks that sometimes makecassava appear an inferior food.Drawbacks include low proteincontent of the roots, low energydensity, and potential toxicity from thepresence of the cyanogenic glucosideslinamarin and lotaustralin (Rosling,1987).

The amount of glucosides, mainlyconsisting of linamarin (90%), canreach 1,500 mg CN equivalent per kgdry weight in fresh roots, particularlyin those of bitter varieties grown fortheir higher yields (Sunderesan et al.,1987). Environmental factors such asdrought, pests, and diseases mayincrease the glucoside content(Gondwe, 1974). If the food securityprovided by cassava is to have anoptimal impact on community health,

Abstract

In 1988, an outbreak of acutepoisoning occurred in adrought-stricken district in southernTanzania. Studies carried out in thearea revealed that the victims hadhigh levels of thiocyanate, a cyanidemetabolite found in the body’s plasmaand urine. The high dietary cyanidecame from consuming insufficientlyprocessed roots of cassava, the onlycrop to survive the prolonged drought.Because of food scarcity, thecustomary, but lengthy, sun-drying ofpeeled cassava roots was replaced bya repeated pounding and sun-dryingof peeled roots to obtain flour forconsumption the same day. Anexperiment in one village showed thatthe principal source of dietary cyanidecomprised the high residual levels ofcyanohydrin (the intermediatebreakdown product), which rangedfrom 16 to 20 mg CN equivalent perkg dry weight. The shortenedprocessing method adopted during thedrought resulted in high glucosidelevels ranging from 3 to 879 mg CNequivalent per kg dry weight in thefinal products. Rapid, but moreeffective, tissue disintegration anddrying techniques that easily remove

* Tanzania Food and Nutrition Center, Dar esSalaam, Tanzania.

167

Acute Poisoning in Tanzania:...

then the nutritional drawbacks mustbe avoided or balanced.Supplementing cassava-based mealswith various protein sources canbalance low protein content and lowenergy density, and efficientprocessing can solve toxicityproblems.

During processing, disintegratingthe root tissue releases anendogenous enzyme—linamarase—which hydrolyzes the glucosides to thecorresponding intermediate products:cyanohydrins (Figure 1). Theintermediate products at pH >6spontaneously decompose to volatilehydrogen cyanide (HCN), whichrapidly evaporates into the air ordissolves in water (Figure 1) (Cooke,1978).

The glucosides, cyanohydrins, andHCN are collectively known ascyanogens, and efficient processingcan reduce them all to negligiblelevels. If insufficiently processedroots are consumed, cyanide exposurecan occur from glucosides orcyanohydrins breaking down in thegut. The human body detoxifiescyanide by enzymatically converting itto the less toxic thiocyanate (or SCN),using sulfur as a substrate. Sulfur isobtained from sulfur-containingamino acids in the diet (Banea et al.,1992). The thiocyanate is found inserum and urine, through which it isexcreted. Thus, cyanide exposure inhumans can be estimated bydetermining thiocyanate in serum andurine.

Despite cassava being extensivelyused as food, reports of acutepoisonings are rare. Most consumersare aware of the potential toxicity andknow how to detoxify the roots. Wherecases of acute poisoning and othereffects have occurred, they were mainlyin populations suffering severe foodshortages (Rosling, 1987). Lack ofscientific attention to such populationsmay also partly explain the scarcity ofreporting.

A prolonged drought occurred inMasasi District, southern Tanzania, in1987/88, causing a severe foodshortage that motivated people todeviate from traditional cassavaprocessing methods. The first visit tothe District, in September 1988, wasduring a “situation of emergency,”declared by authorities who hadreceived reports from many villages ofacute poisoning after consumption ofcassava-based meals.


The study area

In 1988, Masasi District, Mtwararegion, southern Tanzania, had apopulation of 335,000, correspondingto 38 inhabitants per square km. Mostof the soil in the District is a red loam,suitable for cassava and maizecultivation. The District has aunimodal rainfall that usually starts inlate November and ends in May.Average annual rainfall is 940 mm andtemperatures vary from as low as 18 °C

Figure 1. The breakdown of cyanogenic glucosides in cassava to produce hydrogen cyanide. R1 = CH3 forlinamarin; R2 = C2H5 for lotaustralin.

O C C≡NGlucose HO C C≡N

Dissolves inwater and/orevaporatesinto the air

Mainlynonenzymatic

Slow at pH < 6-7EnzymaticCH3CH3

R1 or 2

R1 or 2

> HC≡N

168


in July to as high as 35 °C inDecember.

Cassava is a secondary staple inMasasi but in difficult years becomesthe primary one. In neighboringNewala District, cassava is theprimary staple and people fromMasasi often beg or barter cassavaroots and seedlings from Newaladuring periods of food shortage. Anearlier rapid rural appraisal revealedthat prolonged sun-drying of peeledroots was the main cassavaprocessing method in Masasi(Seenappa and Mlingi, 1988).

Key informants, focus-groupinterviews, household surveys, andsample analysis

Local authorities and other keyinformants of Masasi District providedgeneral information on theagricultural and dietary situation inthe District before and during thedrought in 1988. The districtauthorities selected, for the study, thevillages of Mtandi, Chanikanguo, andMumbaka, all 5 to 15 km from Masasitown and which had the highestnumbers of reported cases of acutepoisoning.

In each village, the leaders,elders, and women were gathered forfocus-group interviews (Scrimshawand Hurtado, 1987) on agriculturaland dietary practices. One 10-cellunit (an administrative structure withabout 10 households) was selected ineach village. In the 35 householdsfound in those units, the husbandand/or wife were interviewed througha formulated open-endedquestionnaire. This covered foodconsumption over 24 h (Table 1),cassava cultivation and processing,and occurrence of acute poisoning.To determine thiocyanate, anindicator of cyanide exposure, plasmaand urine specimens were collected

Table 1. Dietary practices during food shortagesin 35 interviewed households, Masasidistrict, southern Tanzania, 1988.

Foods consumed in Householdslast 24 h

(no.) (%)

Cassava roots 35 100

Wild vegetables 30 86

Fruit (mainly mangoes) 29 83

Cassava leaves 19 54

Maize 10 29

Dried small fish 7 20

Legumes 6 17

from 28 men, 37 women, and 30children aged 5 to 14 years—a total of95 subjects in the interviewedhouseholds.

A year later, Mtandi andChanikanguo villages were revisited.Of the previously studied households,12 volunteered to be brieflyinterviewed on diet and give urinespecimens, that is, 32 subjectsmaking up 9 men, 13 women, and10 children. At the same time,specimens of cassava flour used in theinterviewed households were collected.Clinical records of the Ndanda MissionHospital and Masasi District Hospitalwere reviewed for cases of cassavapoisoning. In Mumbaka, sixhouseholds that had reported acutepoisoning in 1988 were extensivelyreinterviewed on the symptomsexperienced.

As a reference, thiocyanate wasdetermined in plasma and urinespecimens collected from 201 adults ofboth sexes, randomly selected, from avillage in the Kilimanjaro region,northern Tanzania, where the dietarystaple was banana, and cassavaconsumption was rare.

A cassava-processing experimentwas conducted in Mtandi villageduring a follow-up study in 1988.

169


Cyanogen levels were determined incassava flour obtained by twoshortened processing methods usedduring the food shortage. Roots from20 plants of a bitter variety, chimaje,were harvested from the same field.Roots were peeled and split lengthwiseto form 19 identical pairs of batches.Under supervision, an elderly womanprocessed one set of batches intochinyanya flour and another set intosmall makopa (or dried root pieces),which were later pounded into flour, aspracticed during the food shortage.

All specimens collected were keptfrozen before analysis and thiocyanatewas determined according to Lundquistet al., (1979, 1983). Cyanogens inflour samples were determined by anenzymic assay method modified byO’Brien et al. (1991), permittingseparate quantification of glucosides,cyanohydrins, and HCN.

Results

Drought and food shortage

In normal years, cassava, maize, andsorghum dominate production inhigher areas while rice is restricted tosome lower swampy areas. Other foodcrops in the district includesweetpotatoes, cowpeas, and pigeonpeas. Cashews, groundnuts, andbambara nuts are cultivated as cashcrops. The area’s rainfall patternshows that rainfall was halved in theagricultural year 1987/88, the periodbefore the food shortage. Interviewswith key informants and focus groupsconfirmed that the drought from June1987 to September 1988 caused theworst food shortages ever experiencedin the district since 1966. Cassavawas the only crop which survived;maize, rice, sorghum, and millet allfailed.

The focus groups revealed thatmany families grow both sweet and

bitter varieties of cassava but that,during the drought, roots from bothvarieties tasted more bitter than innormal years. Among the interviewedhouseholds, 71% cultivated only bittervarieties, while the rest cultivated bothbitter and sweet. Of the six mostcommonly grown varieties, two,chimaje and limbanga, were identifiedas bitter; three—liumbukwa, kigoma,and mba safi—as sweet; and the sixthvariety, mreteta, as either sweet (by58% of households) or as both sweetand bitter (23%).

During the drought, the PrimeMinister’s Office (PMO) promptlyreacted to news of the food shortage bydistributing about 400 t/month ofrelief food, thus alleviating the threatof famine. But several households didnot receive relief because it wasinsufficient, and used mainly to inducepeople to cultivate communal fields inthe most affected villages. Old anddisabled people received free relieffood.

Cassava processing

In normal years, most cassava isprocessed by direct sun-drying for1 to 4 weeks, depending on sunshine.The roots are first peeled and leftwhole if small or split if large. Theresulting dried root pieces are knownas makopa, and are either sold forcash or pounded into a flour used formaking ugali, a stiff porridge.Legumes, small fish, cassava leaves(kisamvu), or other green vegetablesconstitute the relishes regularly eatentogether with cassava or maize ugali.

During the food shortage, thenormal processing method wasreplaced by two shortened methods.Chinyanya was faster and more widelyused, according to the focus groups.Peeled roots were pounded into pieces,sun-dried for some hours, thenrepeatedly pounded and dried until aflour was obtained within half to

170


1 day. The second method, “smallmakopa,” involved cutting fresh rootsinto finger-sized pieces and dryingthem on hot rocks until they could bepounded into flour. This method took1 or 2 days, depending on thesunshine. The relish used tosupplement the ugali made from suchshort-processed roots, was limited tokisamvu during the food shortage.

All households had consumedcassava during the 24 h before theinterviews (Table 1). Although mosthouseholds consisted of farmingfamilies (91%), during the foodshortage, some relied on cassava eitherbartered or given free of charge. Of the29% of households that had consumedmaize, almost a quarter had mixed itwith cassava flour to make ugali. Allhouseholds admitted they had madesome shortcuts in processing cassavaby producing “small makopa” and 65%of households stated that they hadused the chinyanya method.

Nevertheless, 9% of householdsfermented cassava by soaking the roots

in water, and another 18% fermentedpeeled roots in covered heaps.

Acute poisoning and dietarycyanide exposure

Key informants stated that the acutepoisoning following cassava-basedmeals frequently occurred in theMasasi villages between March andSeptember 1988. All those interviewedagreed that they had seen or heardof villagers who were poisoned aftereating cassava-based meals. Of the35 households interviewed in thesecond round, 80% confirmed thatmost family members had sufferedacute poisoning on one or moreoccasions. The pattern of symptoms,time of onset after meals, and durationof poisoning, as determined by theextensive interviews of households inMumbaka village (Table 2) areconsistent with information obtainedfrom other interviews.

Clinical records for July 1988showed that the Ndanda MissionHospital treated several outpatients

Table 2. Results of interviews of six households regarding acute poisoning in Mumbaka village, Masasidistrict, southern Tanzania, 1989.

Poisoning parameters Household code numbera

1 2 3 4 5 6

Number of persons affected 6 5 6 6 4 3Number of times poisoned in 1988 3 3 10 10 1 1Interval between meal and onset of symptoms (h) 1 4 1 6 2 2Estimated duration of poisoning (h) 4 8 24 8 10 8Cassava was processed into:

Chinyanya y y y y y y“Small makopa” y

Symptoms of poisoning:Vomiting y y y y y yDizziness y y y y y yNausea y y y y y yPalpitations y y y y yWeakness y y y y yDiarrhoea y y y yHeadaches y y y yDifficulties in seeing y y y

a. y = yes.

171


and admitted three patients forcassava poisoning. The first was a7-year-old girl who sufferedabdominal pain and vomiting aftereating bitter cassava. She recoveredthe following morning without specifictreatment.

Two days later, a 4-year-old girlwas admitted semiconscious anddehydrated, but without fever. Shehad suffered a sudden onset ofintense vomiting caused by eatingpieces of cassava that were beingdried. A routine neurologicalexamination was normal. Shereceived antibiotic treatment againstsuspected aspiration pneumonia andparenteral fluid, and recovered within24 h.

Several days later, a 10-year-oldboy was admitted unconscious after asudden onset of poisoning symptomsfrom eating bitter cassava. Becauseantidotes were unavailable, he wastreated with dextrose saline infusionand cortisone but died 3 h later.

In July 1988, the Masasi DistrictHospital also admitted three caseswith similar symptoms and history.All recovered within 24 h.

Table 3 shows that, during thefood shortage, the plasma SCN valuewas more than 10 times and theurinary SCN more than 100 timeshigher in Masasi subjects than inthose from the Kilimanjaro village.When 32 subjects in Masasi were

reexamined in the same month of1989, a normal year, their meanurinary SCN was only 6% of the meanfound the year before. All householdsstill consumed cassava daily but thisyear the ugali was made from properlydried, normal-sized makopa (Essers etal., 1992).

Cassava processing experiment

The two shortened methods, chinyanyaand “small makopa,” were used toprocess the 19 pairs of batches of split,peeled roots from 20 cassava plants ofthe same bitter variety from the samefield, as described on p. 169.

Each of the 19 batches processedto chinyanya flour was pounded andsun-dried four successive times andsieved after each pounding to obtainflour. Each of the 19 batchesprocessed into “small makopa” wassplit into small finger-sized pieces thatwere sun-dried on hot rocks and thenpounded into flour at the end of theday.

Table 4 compares the cyanogencontent of flours obtained in theexperiment by the two shortenedmethods with that of flour samplescollected from 12 households in 1989,a normal year. Glucosides were veryhigh in the “small makopa” flour andcyanohydrins were high in thechinyanya flour. The samples of“normal” flour had relatively high levelsof glucosides but very low cyanohydrincontent.

Table 3. Thiocyanate (SCN) levelsa in subjects from Masasi, southern Tanzania, who eat cassava, andsubjects from Kilimanjaro, northern Tanzania, who eat banana.

SCN sample Cassava diet, Cassava diet, Banana diet,drought year normal year normal year

(n = 95) (n = 32) (n = 201)

Plasma 335 ± 12 28 ± 4Urine 1,120 ± 75 68 ± 9 7 ± 1

a. Values are given in µmol/L and as mean ± the standard error of mean.

172


Table 4. Cyanogen levels in cassava floura processed by three different methods, southern Tanzania.

Contents Processing experiment Household flour madefrom normal-sized

Chinyanya “Small mapoka” mapoka, normal year(n = 19) (n = 19) (n = 12)

Glucosides 90 ± 17 768 ± 107 120 ± 70(12-296) (121-1,837) (93-879)

Cyanohydrins 48 ± 5 15 ± 4 7 ± 2(16-120) (0-61) (0-17)

Hydrogen cyanide 6 ± 1 7 ± 0.5 6 ± 0.5(2-12) (5-10) (4-9)

Total cyanogens 144 ± 18 971 ± 107 133 ± 71(56-336) (131-1,855) (8-901)

pH 6.6 ± 0.1 6.9 ± 0.1 6.3 ± 0.2 (6.2-6.9) (6.6-7.2) (5.2-7.0)

Moisture (%) 13.5 ± 1 10.8 ± 0.2 10.6 ± 0.4(6.0-23.6) (9.3-12.7) (9.5-14.4)

a. Values are given as mean ± standard error of mean. Values in parentheses are ranges. Cyanogen values aremeasured as mg of CN equivalent/kg of dry weight.

this complicated analysis during ourstudy, and relied on thiocyanate.Levels of this easily determinedcyanide metabolite in the affectedpopulation were among the highestever reported in cassava eaters,suggesting strongly that cyanidecaused the poisonings. The fact thatthe thiocyanate levels fell almost tonormalcy the next year also supportsthe hypothesis of cyanide poisoning.

Cyanide

In contrast to earlier assumptions(Cheok, 1978), we concluded thatcassava poisoning is not the result ofingesting HCN. The reason is thatHCN evaporates at 28 °C, thus rapidlyescaping during drying, as verified bythe low levels found in all flouranalyzed. Further losses occur whenboiling ugali. Cyanide is also rapidlyabsorbed from the stomach, whereasthe poisonings reported here occurredone to several hours after ingestion.The length of this interval suggeststhat the cyanide exposure resultedfrom ingested cyanide precursors,such as glucosides or cyanohydrins,

Discussion

Results confirm that most householdsof the Masasi District suffered severefood shortage during March toDecember 1988, and depended almostentirely on cassava—the only crop tosurvive the drought. All interviewedhouseholds daily consumed ugali,most of it prepared from cassava flourmade from short-processed roots. Allacute poisonings resulted from eatingugali prepared from such flour,especially chinyanya flour.Undoubtedly, thousands of people inthe District were poisoned to somedegree during the food shortage.

General symptoms of the casepatients suggested cassava poisoning,which is characterized by an intervalof 1 to 4 h from meal to onset andsymptoms usually clearing within24 h (Cheok, 1978; Essers et al.,1992). Cyanide, originating fromcyanogenic glucosides occurringnaturally in the roots, is presumed tocause acute cassava poisoning. Butblood cyanide levels were neverdocumented. We could not perform

173


that probably yielded cyanide in thesmall intestine.

Glucosides

Ingested glucosides are an unlikelysource of cyanide exposure in theMasasi population. The reasons are,first, experiments with animalsindicate that ingested linamarin canbe absorbed unchanged and excretedintact in the urine (Barrett et al.,1977). Ingesting linamarin will onlyresult in cyanide exposure if suitablemicrobial glucosidases are present inthe gut, a mechanism still notconfirmed as occurring in humans.Second, poisonings were associatedmainly with consumption ofchinyanya flour, which had relativelylow glucoside levels. Third, the 1989subjects had low urinary thiocyanatelevels, despite the high levels ofglucosides found in the flourcollected from their households. Anadult’s estimated daily consumptionof 0.5 kg of normal, household,cassava flour corresponds toingesting 2,000 µmol of potentialcyanide. The 68 µmol of thiocyanateper liter of urine constitutes only 3%of this amount, indicating thatingested glucosides do not result incyanide exposure.

The rapid drying of “smallmakopa” in strong sunshine mayresult in high levels of glucosides,probably because the heat destroysthe linamarase enzyme or reducesmoisture content to a level thatinactivates this enzyme before tissuedisruption enables it to act on theglucosides. A daily consumption of0.5 kg of “small makopa” correspondsto a potential cyanide yield of morethan 300 mg, which is above thelethal dose (Hall and Rumack, 1986).Thus, the low mortality observedsuggests that glucoside ingestion is ofminor importance for cyanideexposure from insufficientlyprocessed cassava.

Cyanohydrins

We believe that cyanide exposureresults mainly from consumption ofcyanohydrins in ugali prepared fromchinyanya flour. Total cyanogenswere higher in “small makopa” flour,but cyanohydrins were higher inchinyanya flour.

Very little is known about the fateof different forms of cyanogens duringdigestion in humans. In the alkalineenvironment of the small intestine,cyanohydrins should rapidlydecompose to yield cyanide that isthen absorbed. Our results supportthis hypothesis. The cyanohydrinlevel found in chinyanya canpotentially yield about 1 mmol ofcyanide at the estimated dailyconsumption of about 0.5 kg. Thisquantity of cyanide matches the highurinary thiocyanate level of 1 mmol/Lfound during the poisonings. Thenext, normal, year, when thechinyanya method was not used, thecyanohydrin levels in household flourwere low, as were the urinarythiocyanate levels. Those whoconsumed cassava in the normal yearor bananas had thiocyanate valuesbelow 100 µmol/L, corresponding tothe levels of thiocyanate found innonsmokers1 in other countries(Lundquist et al., 1979).

Effects of cyanide exposure

Despite numerous cases of acutepoisoning in many Masasi villages,very few deaths were reported andwe could only document one.Hospital-based reports of cassavapoisoning give an impression of highmortality (Cheok, 1978; Tylleskar etal., 1991), but anecdotal informationsuggests low mortality in several otheroutbreaks of cassava poisoning forwhich we could not find scientific

1. Cigarette smoke contains small quantities ofcyanide.

174


documentation. Although manysubjects reached blood cyanide levelsat which symptoms of poisoningoccurred, few reached levels that weretwo to four times higher than lethal(Hall and Rumack, 1986). The smallnumber suggests that exposure tosudden high peaks of blood cyanidewas rare. The rarity can be explainedif the levels causing symptoms arepartly reached by a cumulative effectfrom several meals, and partly causedby a decreased cyanide-to-thiocyanateconversion as the intake of sulfuramino acids (which provides thesubstrate for conversion) drops. Agradual cyanide release followingmeals would contribute to flatexposure peaks that usually reachlevels causing only symptoms and,rarely, the higher levels causingdeath.

A different mechanism may alsooperate to reduce the mortality rate.The metabolism of linamarin and itseffects has not been studied inhumans. Some symptoms of acutecassava poisoning may be caused bynon-lethal effects of absorbed, intactlinamarin. If this hypotheticalmechanism operates in parallel tocyanide exposure, it may explain whythe frequency of diarrhoea seemshigher than that reported for cyanidepoisoning from other sources (Halland Rumack, 1986).

The epidemic paralytical disease,konzo, characterized by abrupt onsetof spastic paraparesis, has beenattributed to high dietary cyanideexposure from insufficiently processedcassava. We screened the populationin one village in Masasi for konzoand, as reported elsewhere (Tylleskaret al., 1991), found that theincidence of konzo was about 1 per1,000 people during the period of highcyanide intake. The minimal dietaryvariation, provided by relief food,probably protected the populationagainst a more extensive konzo

epidemic. The highest incidences ofthis disease have been found inpopulations who exclusivelyconsumed insufficiently processedcassava for several weeks to months(Howlett et al., 1992; Tylleskar et al.,1992).

In conclusion, the 1988 acutepoisonings in Masasi resulted fromconsuming insufficiently processedcassava roots, and which thereforecontained high residual amounts ofcyanohydrins. Roots from bittercassava varieties can be safely eatenafter effective processing. Even in theshortened processing methods,cyanohydrin levels may easily bereduced by improving thedisintegration and drying techniquesused.

Toxic effects from cassava mayoccur in several areas of easternAfrica where bitter varieties havebeen introduced and where ineffectiveprocessing methods are used,especially during food shortages(Essers et al., 1992). However,cassava’s agricultural potential maybe used to improve food security witha positive effect on nutrition andhealth if attention is paid to itspotential nutritional drawbacks.

An ongoing program aims toprovide practical measures to preventfurther outbreaks of acute poisoningsduring food shortages. This involvesmaking the authorities and thepopulation aware of the faults in theineffective processing methodspracticed in that area. The programwill then develop an extensionpackage for cassava processing thatwill use more efficient techniques,such as those for gari processingfrom West Africa. The extensionpackage includes selecting the bestand simplest message for thecommunity, strategies to introducenew processing methods, andapproaches to be used.

175


O’Brien, G. M.; Taylor, A. J.; and Poulter,N. H. 1991. An improved enzymicassay method for cyanogens in freshand processed cassava. J. Sci. FoodAgric. 56:277-289.

Rosling, H. 1987. Cassava toxicity and foodsecurity: a review of health effects ofcyanide exposure from cassava andof ways to prevent these effects.Report to the African HouseholdFood Security Program, UnitedNations Children’s Fund (UNICEF).International Child Health Unit,Uppsala University, Uppsala,Sweden. 40 p.

Scrimshaw, S. C. M. and Hurtado, E. 1987.Rapid assessment procedures fornutrition and primary health care:anthropological approaches toimproving programmed effectiveness.Reference series, vol. 11. LatinAmerican Center, University ofCalifornia (UCLA), Los Angeles, CA,USA.

Seenappa, M. and Mlingi, N. 1988.Household food security and the roleof cassava: a case study fromTanzania. In: Nutrition and foodsecurity. Vol. 2. Proceedings of theThird Africa Food and NutritionCongress, Harare, Zimbabwe,5-8 September, 1988. p. 734-763.

Sunderesan, S.; Nambisan, B.; and Easwari,A. 1987. Bitterness in cassava inrelation to cyanoglucoside content.Indian J. Agric. Sci. 57:37-40.

Tylleskar, T.; Banea, M.; Bikangi, N.; Fresco,L.; Persson, L-A.; and Rosling, H.1991. Epidemiological evidence fromZaire for a dietary aetiology of konzo,an upper motor neutron disease.Bull. W. H. O. 69:581-90.

__________; __________; __________; Poulter,N.; Cooke, R.; and Rosling, H. 1992.Cassava cyanogens and konzo, anupper motor neuron disease foundin Africa. Lancet 339:208-11.

References

Banea, M.; Poulter, N.; and Rosling, H. 1992.Shortcuts in cassava processing andrisk of dietary cyanide exposure inZaire. Food Nutr. 14:137-143.

Barrett, M. D.; Hill, D. C.; Alexander, J. C.;and Zitnak, A. 1977. Fate of orallydosed linamarin in the rat. Can. J.Physiol. Pharmacol. 55:134-36.

Cheok, S. S. 1978. Acute cassava poisoningin children in Sarawak. Trop. Doc.8:99-101.

Cooke, R. D. 1978. An enzymatic assay for thetotal cyanide content of cassava(Manihot esculenta Crantz). J. Sci.Food Agric. 29:345-352.

Essers, A. J. A.; Alsen, P.; and Rosling, H.1992. Insufficient processing ofcassava induced acute intoxicationsand the paralytic disease konzo in arural area of Mozambique. Ecol. FoodNutr. 27:17-27.

Gondwe, A. T. D. 1974. Studies onhydrocyanic acid content of somelocal varieties of cassava and sometraditional cassava products. East Afr.Agric. For. J. 40:161-7.

Hall, A. H. and Rumack, B. H. 1986. Clinicaltoxicology of cyanide. Ann. Emerg.Med. 15:1067-74.

Howlett, W. P.; Brubaker, G. R.; Mlingi, N.;and Rosling, H. A. 1992.Geographical cluster of konzo inTanzania. J. Trop. Geogr. Neurol.2:102-108.

Lundquist, P.; Martensson, J.; Sorbo, B.;and Ohman, S. 1979. Method fordetermining thiocyanate in serum andurine. Clin. Chem. 25:678-81.

__________; __________; __________; and__________. 1983. Adsorption ofthiocyanate by anion-exchange resinsand its analytical application. Clin.Chem. 29:403.

176


CHAPTER 21

GARI, A TRADITIONAL CASSAVA SEMOLINA

IN WEST AFRICA:ITS STABILITY AND SHELF LIFE AND THE

ROLE OF WATER

N. Zakhia*, G. Chuzel**, and Dany Griffon*

consumed with milk (for breakfast) oradded to hot sauces.

As demand for gari from urbanmarkets is increasing, a better qualityproduct with an extended shelf-life isneeded. The shelf-life of packed andstored gari largely depends on storagetemperature and relative humidity,and on the product itself (moisturecontent and water activity). Theseparameters determine the rate ofgari’s microbial and physicochemicaldeterioration. The sorption isotherm,that is, the equilibrium between wateractivity and moisture content of afoodstuff, is a good indicator of theproduct’s stability at differentambient temperatures and relativehumidities (Bandyopadhyay et al.,1980).

Our study used adsorptionisotherms to determine gari’s optimalshelf-life. Thus, we could propose aset of packaging recommendations toensure better storage of gari in thetropics.


Samples of traditional Togolese gari(about 8% w.b. water content) werecollected from a small factory nearLomé and packaged in sealedpolyethylene bags. They were sent toMontpellier, France, and stored at

Abstract

Gari is a cassava semolinatraditionally processed and widelyconsumed in West Africa. Itsshelf-life is largely dependent on bothits water content and storagetemperature and relative humidity.This paper deals with gari’sadsorption properties in relation towater availability at 15, 25, and30 °C. From the results, optimalpackaging conditions can beestimated, thus providing anextended shelf-life for gari.

Introduction

Gari is a cassava semolinatraditionally prepared and widelyconsumed in West Africa. Gariprocessing consists of pressing thejuice out of peeled and grated cassavaroots for 2 to 4 days, allowing anatural lactic fermentation to takeplace. The fermented mash is thencooked in an open clay vessel untilthe starch gelatinizes sufficiently(Chuzel, 1989; Zakhia, 1985). Gari isa “ready-to-use” food, generally

* CIRAD/SAR, Montpellier, France.** CIRAD/SAR, stationed at the Faculdade de


177

Gari, A Traditional Cassava Semolina in West Africa:...

Table 1. Experimental data for adsorption equilibrium of gari at 15, 25, and 35 °C (± SD) with saltsources and references.

15 °C 25 °C 35 °C

Equilibriumrelative

humidity

0.119b

0.234c

0.333b

0.431c

0.607c

0.741b

0.755c

0.856b

0.911b

Moisturecontent

(% d.b. ± SD)

5.2 ± 0.4 7.4 ± 0.4 8.1 ± 0.4 9.1 ± 0.211.4 ± 0.715.2 ± 0.815.4 ± 0.520.7 ± 0.924.5 ± 0.8

Equilibriumrelative

humiditya

0.1100.2240.3300.4280.5770.7080.7530.8430.902

Moisturecontent

(% d.b. ± SD)

4.8 ± 0.1 6.7 ± 0.3 7.6 ± 0.2 8.6 ± 0.311.0 ± 0.214.3 ± 0.216.1 ± 0.220.3 ± 0.626.7 ± 0.6

Equilibriumrelative

humidity

0.112b

0.230d

0.320b

0.410e

0.545c

0.680d

0.751f

0.829c

0.894g

Moisturecontent

(% d.b. ± SD)

4.4 ± 0.3 5.9 ± 0.2 7.3 ± 0.3 8.6 ± 0.210.3 ± 0.613.7 ± 0.816.6 ± 0.720.8 ± 1.226.8 ± 1.2

a. Stokes and Robinson, 1949.b. Acheson, 1965.c. Greenspan, 1977.d. Rockland, 1960.e. Griffin, 1944.f. Clarke and Glew, 1985.g. Baxter and Cooper, 1924.

Salt

LiClCH

3COOK

MgCl2

K2CO3

NaBrSrCl2

NaClKClBaCl2

2 °C until evaluation. Adsorptionisotherms were determined at 15, 25,and 35 °C, using the standardmethod recommended by the WaterActivity Group (WAG) of the EuropeanUnion (EU)1 (Wolf et al., 1985).Equipment used comprised nine,sealed, glass jars, containingsaturated salt solutions ranging fromaw 0.1 to 0.9 (Table 1). This aw rangeis required practice for predicting theshelf-life of dried packaged products(Chirife et al., 1979).

In France, before measurementswere taken, 3 g of gari samples werepredried over phosphorus pentoxide(P2O5) for 10 days at about 20 °C(ambient temperature) to lower theirwater content to a minimum. They

were then sprayed with a solution ofsodium azide (0.5%) to inhibit thegrowth of microorganisms at highwater activities (aw > 0.8). They wereplaced in dishes resting on trivetsstanding in jars of salt solutions. Thesealed jars were then submerged inan insulated water bath, thetemperature of which was controlledto within ± 0.2 °C.

Equilibrium time was 21 days, asrecommended by WAG’s standard(i.e., 7 days for aw < 0.6 and 14 daysfor aw > 0.6), and in accord with theequilibration time required forcassava mash (Gevaudan et al.,1989). Equilibrated samples wereoven-dried (103 °C, 24 h) in triplicateto determine moisture content.

The WAG uses the GAB(Guggenheim-Anderson-De Boer)model, which is believed to be thebest for providing equations fordescribing food isotherms up to aw 0.9(Bizot, 1983; Van den Berg, 1985).The GAB equation is as follows(Labuza et al., 1985):

1. The Water Activity Group conducts theproject “COST 90 bis” as part of the programfor European Cooperation in the Field ofScience and Technical Research, sponsoredby the former European EconomicCommunity, now the European Union.

178


waX

= αaw + βaw + γ

where:

X = water content (% d.b.)

Xm = percentage of water contentcorresponding to theoccupation of all primaryadsorption sites by one watermolecule. “Xm” is also calledthe “monolayer.”

C = the Guggenheim constant.C = CN exp [(H1 - Hm)/RT]

K = a correction factor for themultilayer molecules. K = KNexp [(H1 - Hq)/RT]

H1 = the heat of condensation ofpure water vapor

Hq = the total heat of sorption of themultilayer water molecules

Hm = the total heat of sorption of themonolayer

The GAB model’s coefficients α,β, and γ were determined for eachtemperature by using a nonlinear,least-square regression asrecommended by Schär and Rüegg(1985). The values of GABconstants Xm, C, and K were alsocalculated. The confidence of fitwas judged by the relative rootmean square error (% RMS).


Adsorption isotherms

Table 1 and Figure 1 show theaverage values and standarddeviations of the equilibrium moisturecontents for the water activitiesstudied at 15, 25, and 35 °C, andtheir fitted GAB isotherms.

At low water activities (aw < 0.6),the experimental curves agree withthe sorption theory, that is, atconstant aw, an increase oftemperature causes a small decreaseof moisture content. At high wateractivities (0.7-0.9), moisture contentincreases with rise in temperature.The three calculated isothermsintersect at the following points:

isotherms 15 °C and 25 °Cwhere X = 13.0% and aw = 0.66



β = −11

2X Cm

( )

α = −KX Cm

( )1

1

γ = 1X K Cm . .

(1)

Figure 1. Adsorption isotherms of gari at threetemperatures. Experimental pointsare means of triplicates. = 15 °C; = 25 °C; = 35 °C.

30

20

10

0

Moi

stu

re c

onte

nt

(g H

2O

/100 g

dry

matt

er)

0 0.2 0.4 0.6 0.8 1.0

Water activity (aw)

²

179


The crossing of isotherms at highwater activities with increasingtemperature has already beenobserved in some foods and mayresult from the product’s chemicalcomposition and its treatments (e.g.,heating, drying, andpregelatinization). Other foodsshowing this phenomenon aresucrose and fructose (Loncin et al.,1968), potato slices (Mazza, 1982),carrots (Mazza, 1983), Jerusalemartichoke (Mazza, 1984), and sultanas(Saravacos et al., 1986). Theexplanation is that some sugarsincrease their solubility withtemperature, thus binding more waterat higher temperatures andincreasing the equilibrium moisturecontent.

The technological treatmentsinvolved in gari production (grating,fermentation, and squeezing) inducedamage to about 3% to 6% of thecassava starch (Zakhia, 1985).During roasting, the starch is heatedin the presence of water, but theinitial moisture content of about1 g g-1 (d.b.) of cassava mash does notallow the starch to completelygelatinize (Chuzel, 1989; Gevaudan etal., 1989). But crystallinity is lostand extensive swelling of the starchgranules occurs. A complexmetastable network forms, consistingof amorphous regions (containingplasticizing water) and hydratedmicrocrystalline regions that had notdissolved during the partialgelatinization and which serve asjunction zones (Levine and Slade,1988). All these factors stronglyaffect the polymer-water interactions(Radosta et al., 1989).

The sorption mechanism forstarch is almost entirely governed byactive sites, that is, the glucoseresidues of the starch polymer (Hilland Rizvi, 1982). We suggest,therefore, that increases in bothtemperature (to 35 °C) and water

content (for aw between 0.5 and 0.7)initiate a “collapse” that makes thesoluble starch (amorphous fractionsand branched segments) leach out.This increases the number ofavailable adsorption sites (glucoseresidues) and explains why garibecomes more hygroscopic at highertemperatures and water activity.Moreover, the degree of starchdamage in gelatinized starchyproducts is measured by thesolubility and swelling indices at30 °C, which depend on the ability ofstarch to absorb water (Anderson etal., 1969).

Slade and Levine (1988) andOrford et al. (1989) have alsodiscussed the physicochemical effectof water, acting as a plasticizer of theamorphous regions in the starchnative granule, on the temperature ofthe vitreous transition that occursduring native starch gelatinization.Only the water included in the starchgranule (about 10% w.b.) is involvedin this process; the vitreous transitiontemperature decreases sharply withincreasing water content. Thisplasticizing effect could also explainthe observed adsorption behavior ofgari. The amorphous matrix of gari,which is partially plasticized at roomtemperature by excess water (> 12%d.b.), would become more plastic ifthe temperature increased to 35 °C.The mobility of chains is thenenhanced and the free volumes insidethe polymer increase. As these freevolumes may absorb more water, thesorption sites are then more available.

Gari storage

In tropical countries, high relativehumidity and temperature makelong-term storage very difficult. Sopredicting the shelf-life of packagedgari in terms of its storage conditionsand quality becomes important. TheHeiss and Eichner (1971a; 1971b)model allows calculation of the

180


Ws = Weight of the product (kg ofdry matter in the package).

S = Slope of the product isotherm(assumed linear over the rangebetween “Xe” and “Xc”).

Using the Heiss and Eichnermodel, we estimated the shelf-life ofgari packaged in 1-kg polyethylenebags (A = 0.124 m2, Kx = 2.28 10-6 kgH2O.m-2.Pa-1.day-1) at threetemperatures (15, 25, and 35 °C) andfour initial moisture contents (6%,8%, 10%, and 12% d.b.) (Table 2). Asambient storage conditions, weconsidered a relative humidity of 0.9(which is the safe storage borderlinehumidity) and an aw of 0.7 (which isthe safe storage borderline aw

generally used for most products)(Pixton, 1982). Adeniji (1976)observed a significant growth ofmould in gari stored at 27 °C, in arelative humidity of 0.7 and having anequilibrium content of 14.5% (d.b.).For aw 0.7, our sorption curves at15, 25, and 35 °C give 14.0%, 14.2%,and 15.3% (d.b.) as equilibriummoisture contents (Table 2). Thesevalues agree with those of Adeniji(1976). Moisture contents around14% (d.b.) allow gari to maintain acrispness that consumers greatlyappreciate (Chuzel, 1989; Ekundayo,1984).

Moisture contents beforepackaging (Xi) are those generallyfound in local gari sold in tropicalmarkets. lkediobi and Onyike (1982)mention moisture contents rangingfrom 4% to 19% (d.b.). Thecrossing-over of adsorption isothermsdescribed above leads to the followingparadox: gari stored at 35 °C seems tobe better than that stored at 25 °C fora moisture content of 12% (d.b.).This moisture content is usuallyobtained with traditional gariprocessing (Chuzel, 1989), althoughstorage time was less than 1 monthunder the given conditions.

potential storage time based on acritical aw for a particular systemunder given storage conditions. Thismodel equation is:

Θce i e c

x s o

X X X XK A W S

= − −ln (( ) /( ))( / ) (P / )

where:

Θc = Potential shelf-life of theproduct (time in days for thepackaged product to suffermicrobial and biochemicaldeterioration with loss ofsensory quality).

Xe = Equilibrium moisture content(g g-1, d.b.) of the product (if itis left in contact with theatmosphere outside thepackage). “Xe” depends ontemperature, relative humidityand on the product adsorptionisotherm.

Xc = Safe storage moisturecontent of the product (g g-1,d.b.), that is, the moisturecontent corresponding to thesafe storage borderline aw.“Xc” is calculated for theborderline aw by successiveiterations from the GABregression equation until thedifference between twocalculated “Xc” is lower than0.01.

Xi = Initial moisture content of theproduct when it is packaged(g g-1, d.b.).

Kx = Permeability of the package tomoisture vapor(kg H2O.m-2.Pa-1.day-1).

Po = Vapor pressure at storagetemperature (in Pa).

A = Surface area of the package(m2).

(2)

181


Table 2. Estimated shelf-life (days) for safe storage of gari at aw 0.7 at four initial moisture contents (d.b)and three storage temperatures.

Initial moisture content (d.b.) Equilibration temperature (°C)a

15 (14.0) 25 (14.2) 35 (15.3)

6 181 90 578 166 85 52

10 140 73 4512 89 24 30

a. Values in parentheses are equilibrium moisture contents (d.b.) for aw 0.7 (from experimental adsorptionisotherms).

We also focused on thepermeability of packaging materials.Polyethylene, especially thehigh-density type, tends to inhibitwater vapor transfer, but is permeableto oxygen and carbon dioxide, whichmay oxidize gari or cause loss of itsaromas. Polypropylene is lesspermeable to water vapor and oxygen,but, because it is more expensivethan polyethylene, it is less suitablefor storing gari as a daily foodstuff forlow-income consumers.

Conclusions

Gari processing modifies thestructure of native cassava starch sothat, at high water activity, itbecomes more hygroscopic astemperature increases. Our studypointed out a “collapse” (not yetobserved in other starchy products),caused by the partial gelatinizationof gari starch. The GAB regressionequation was adequate for fittingsorption isotherms of gari.

The shelf-life of gari wastheoretically estimated for threestorage temperatures (15, 25, and35 °C) at a relative humidity of 0.9(which is usual in the tropics). For alow-cost storage of at least3 months at about 30 °C, werecommend packaging gari at aninitial moisture content of about 8%(d.b.) in polyethylene bags. Thebags should be sealed and tightly

packed in cardboard boxes for bothphysical convenience and to preventoxidation and reaction to light.

Further research should becarried out to determine the safestorage water activity in the tropics,taking into account the initialmicrobial flora of cassava roots andthe local quality requirements for gariquality, that is, color, crispness,flavor, and food customs.

References

Acheson, D. T. 1965. Vapor pressure ofsaturated aqueous salt solutions.In: Wexler, A. (ed.). Humidity andmoisture, vol. 3. Reinhold, NY.p. 521-530.

Adeniji, M. O. 1976. Fungi associated withthe deterioration of gari. Niger. J.Plant Prot. 2:74-77.

Anderson, R. A.; Conway, H. F.; Pfeiffer,V. F.; and Griffin, E. L. 1969.Gelatinization of corn grits by roll andextrusion cooking. Cereal Sci. Today14(1):4-7.

Bandyopadhyay, S.; Weisser, H.; and Loncin,M. 1980. Water adsorptionisotherms of foods at hightemperatures. Lebensm. Wiss. &Technol. 13:182-185.

Baxter, G. P. and Cooper, W. C., Jr. 1924.The aqueous pressure of hydratedcrystals. II. Oxalic acid, sodiumsulfate, sodium acetate, sodiumcarbonate, disodium phosphate,barium chloride. J. Am. Chem. Soc.46:923-933.

182


Bizot, H. 1983. Using the “GAB” model toconstruct sorption isotherms. In:Jowitt, R.; Escher, F.; Hallstrom, B.;Meffert, H. F. T.; Spiess, W. E. L.; andVos, G. (eds.). Physical properties offoods. Applied Science Publications,London, UK. p. 43-54.

Chirife, J.; Boquet, R.; and Iglesias, H. 1979.The mathematical description ofwater sorption isotherm of foods inthe high range of water activity.Lebensm. Wiss. & Technol.12:150-152.

Chuzel, G. 1989. Etude des traitementstechnologiques intervenant lors de latransformation du manioc en gari.Ph.D. dissertation. Ecole nationalesupérieure agronomique deMontpellier (ENSAM), Montpellier,France. 195 p.

Clarke, W. E. and Glew, D. N. 1985.Evaluation of the thermodynamicfunction for aqueous sodium chloridefrom equilibrium and calorimetricmeasurements below 154 °C. J. Phys.Chem. Ref. Data 14(2):429-610.

Ekundayo, C. A. 1984. Microbial spoilage ofpackaged gari in storage. MicrobiosLett. 26:145-150.

Gevaudan, A.; Chuzel, G.; Didier, S.; andAndrieu, J. 1989. Thermophysicalproperties of cassava mash. Int. J.Food Sci. Technol. 24:637-645.

Greenspan, L. 1977. Humidity fixed points ofbinary saturated aqueous solutions.J. Res. Natl. Bur. Stand. A. Phys.Chem. 81 A(1):89-96.

Griffin, R. C. 1944. Technical Association ofthe Pulp and Paper Industry(TAPPI). TAPPI Data Sheet 109-109a.NY.

Heiss, R. and Eichner, K. 1971a. Moisturecontent and shelf-life. I. Food Manuf.46(5):53-56.

__________ and __________. 1971b. Moisturecontent and shelf-life. II. Food Manuf.46(6):37-38, 41-42.

Hill, P. E. and Rizvi, S. S. H. 1982.Thermodynamic parameters andstorage stability of drum dried peanutflakes. Lebensm. Wiss. & Technol.15(4):185-190.

Ikediobi, C. O. and Onyike, E. 1982. The useof linamarase in gari production.Process Biochem. 17(4):2-5.

Labuza, T. P.; Kaanane, A.; and Chen, J. Y.1985. Effect of temperature on themoisture sorption isotherms and thewater activity shift of two dehydratedfoods. J. Food Sci. 50:385-391.

Levine, H. and Slade, L. 1988. Water as aplasticizer; physicochemical aspectsof low moisture polymeric systems.Water Sci. Rev. 3:79-185.

Loncin, M.; Bimbenet, J. J.; and Lenges, J.1968. Influence of the activity ofwater on the spoilage of foodstuffs.J. Food Technol. 3:131-142.

Mazza, G. 1982. Moisture sorption isothermsof potato slices. J. Food Technol.17:47-54.

__________. 1983. Dehydration of carrots:effects of pre-drying treatments onmoisture transport and productquality. J. Food Technol. 18:113-123.

__________. 1984. Sorption isotherms anddrying rates of Jerusalem artichoke(Helianthus tuberosus). J. Food Sci.49:384-388.

Orford, P. D.; Parker, R.; Ring, S. G.; andSmith, A. C. 1989. Effect of water asa diluent on the glass transitionbehavior of malto-oligosaccharides,amylose and amylopectin. Int. J. Biol.Macromol. 11:91-96.

Pixton, S. W. 1982. The importance ofmoisture and equilibrium relativehumidity in stored products. Trop.Stored Prod. Inf. 43:16-29.

Radosta, S.; Schierbaum, F.; Reuther, F.;and Anger, H. 1989. Polymer-waterinteraction of maltodextrins, part 1.Water vapour sorption and desorptionof maltodextrin powders. Starch/Stärke 41(10):395-401.

Rockland, L. 1960. Saturated salt solutionsfor static control of relative humiditybetween 5 °C and 40 °C. Anal. Chem.32(10):1375-1376.

Saravacos, G. D.; Tsiourvas, D. A.; andTsami, E. 1986. Effect of temperatureon the water adsorption isotherms ofsultana raisins. J. Food Sci.51(2):381-383.

183


Schär, W. and Rüegg, M. 1985. Theevaluation of G.A.B. constants fromwater vapor sorption data. Lebensm.Wiss. & Technol. 18:225-229.

Slade, L. and Levine, H. 1988. Non-equilibrium melting of nativegranular starch. Part I. Temperaturelocation of the glass transitionassociated with gelatinization ofA-type cereal starch. Carbohydr.Polym. 8:183-208.

Stokes, R. H. and Robinson, R. A. 1949.Standard solutions for humiditycontrol at 25 °C. Ind. Eng. Chem.41:2013.

Van den Berg, C. 1985. Development ofB.E.T.-like models for sorption ofwater on foods theory and relevance.In: Simatos, D. and Multon, J. L.(eds.). Properties of water in foods.Nato Asi series, no. 90. MartinusNijhoff, Dordrecht, the Netherlands.p. 119-131.

Wolf, W.; Speiss, W. E. L.; and Jung, G.1985. Standardization of isothermmeasurements. In: Simatos, D. andMulton, J. L. (eds.). Properties ofwater in foods. Nato Asi series,no. 90. Martinus Nijhoff, Dordrecht,the Netherlands. p. 661-679.

Zakhia, N. 1985. Etude de I’operation decuisson-séchage du gari; mémoireingénieur. Ecole nationale supérieuredes industries agricoles etalimentaires, Section RégionsChaudes (ENSIA-SIARC), Montpellier,France. 97 p.

SESSION 4:

BIOCONVERSION AND

BYPRODUCT USE

187

Fermentation in Cassava Bioconversion

Introduction

Cassava fermentation is traditionallypracticed in the tropics. But bothtechnology and productcharacteristics differ according toregion and sociocultural conditions:gari in East and West Africa,chikwangue or fufu in Central Africa,and sour starch in Latin America. Butthey have in common the aim toeliminate the poisonous cyanidecomponents and conserve cassava bylactic acidification.

The essential role of lactic acidbacteria in the three products wasdemonstrated by studies carried outby the Institut français de recherchescientifique pour le développement encoopération (ORSTOM) through the

CHAPTER 22

FERMENTATION IN CASSAVA

BIOCONVERSION1

M. Raimbault*, C. Ramírez Toro**, E. Giraud***,C. Soccol.†, and G. Saucedo††

STD2 Program of the European Union(EU), otherwise known as “Improvingthe Quality of Traditional FoodsProcessed from Fermented Cassava”(Raimbault, 1992; Saucedo et al.,1990).

When producing gari, lacticacidification of cassava is rapid anddetoxification is sometimesincomplete. Controlling throughinoculation would improve quality.For fufu or chikwangue, retting isessential for texturing and detoxifyingthe cassava. Lactic acid fermentationis heterolactic, operating inassociation with secondary alcoholicand anaerobic fermentation toproduce alcohol and organic acidssuch as butyrate, acetate, andpropionate that develop specialaromatic and organolepticcharacteristics. As for gari,fermentation for sour starch(especially in Colombia and Brazil) ishomolactic, but takes 3 or 4 weeks.Amylolytic lactic acid bacteria havebeen isolated from chikwangue byORSTOM scientists and from sourstarch by CIRAD scientists.

A. Brauman isolated a new strain,Lactobacillus plantarum A6, which wasdescribed by Giraud et al. (1991). Itsphysiological and enzymologicalcharacteristics for cultivation oncassava starch media, amylaseproduction, and biochemical

* Institut français de recherche scientifiquepour le développement en coopération(ORSTOM), stationed in Cali, Colombia.

** Laboratorio de Bioconversión, Departamentode Procesos Químicos y Biológicos, Facultadde Ingeniería, Universidad del Valle, Cali,Colombia.

*** ORSTOM, Montpellier, France.† Laboratório de Procesos Biotecnologia,

Departamento de Tecnologia Química,Faculdade de Engenharia, UniversidadeFederal de Paraná, Brazil.

†† Departamento de Biotecnología, UniversidadAutónoma Metropolitana (UAM), Iztapalpa,Mexico.


188


properties have now been described(Giraud et al., 1992; 1993a; 1993b).

ORSTOM scientists have beenresearching solid fermentationcultivation of fungi on cassava andamylaceous components for more than10 years. Soccol et al. (1994) showedthat protein enrichment is possible bycultivating various strains of Rhizopus,even on crude, nongelatinized cassavaflours. Saucedo et al. (1992a; 1992b;1992c) studied, at the ORSTOMLaboratory, Montpellier, the growthand alcohol fermentation of cassavastarch in solid-state fermentation,using a highly promising amylolyticyeast.

Swedish and African researchershave described the beneficial effects oflactic acid fermentation on theprophylactic and keepingcharacteristics of those traditionalfoodstuffs made from fermentedcassava, maize, and mixed cereals,and of baby foods. These foods tendto increase children’s resistance todiarrhoea.

All these studies are beingcontinued in new projects comprisingthe EU-STD3 Program. Other EUstudies are being conducted oncassava quality, environment, physicalprocessing, and transformation at alow industrial scale to take advantageof the economic and commercialopportunities in Latin America.

Solid-State Fermentation ofCassava and Starchy

Products

For more than 15 years, an ORSTOMgroup has worked on a solid-statefermentation process for improvingthe protein content of cassava,potatoes, bananas, and other starchycommodities used for animal feed.Fungi, especially from the Aspergillusgroup, are used to transform starchand mineral salts into fungal proteins(Oriol et al., 1988a; 1988b; Raimbaultand Alazard, 1980; Raimbault andViniegra, 1991; Raimbault et al.,1985). Table 1 shows the overallchanges in composition between theinitial substrate and final products.Through such techniques acassava-fermented product with an18%-20% protein content (dry matterbasis) was obtained.

More recently, Soccol et al.(1993a; 1993b), also at the ORSTOMLaboratory, obtained good resultswith the Rhizopus fungi, of specialinterest in traditionally fermentedfoods. In particular, they studied theeffect of cooking before fermentationon the availability of starch, proteincontent, and the rate of starch’sbioconversion into protein (Table 2).They found that a selected strain ofRhizopus oryzae could transformuncooked cassava, which containsonly 1.68% protein, into a fermentedcassava containing 10.89% protein.

Table 1. Effects of Aspergillus niger on protein and sugar contents of different starches (percentage of drymatter) after 30 h of fermentation in solid-state culture.

Substrate Initial composition Final composition

Proteins Sugar Proteins Sugar

Cassava 2.5 90 18 30Banana 6.4 80 20 25Banana waste 6.5 72 17 33Potato 5.1 90 20 35Potato waste 5.1 65 18 28

189


Table 2. Growth of Rhizopus oryzae in solid-state cultivation on cassava granules after various cookingtreatments.

Treatmenta Dry matterb Total sugarc Proteinsc

Initial Final Initial Final Initial Final

I 60.90 46.48 80.01 46.78 1.20 11.69

II 59.18 45.35 84.11 60.72 1.61 12.40

III 57.95 42.12 82.44 52.57 1.56 13.93

IV 55.63 43.88 82.49 56.62 1.47 11.89

V 45.57 37.88 82.04 56.62 1.68 10.89

a. Treatment: I = Cassava autoclaved for 30 min at 120 °C, frozen, dried, and ground

II = Cassava flour (40% water) autoclaved for 30 min at 120 °C

III = Cassava flour (30% water) autoclaved for 30 min at 120 °C

IV = Cassava flour (30% water) vapor cooked for 30 min at 100 °C

V = Untreated crude cassava flour

b. g/100 g total weight.

c. g/100 g dry matter.

SOURCE: Soccol et al., 1994.

Table 3 shows results of amylasebiosynthesis in solid or liquidculture, using raw or cookedcassava. The amount ofglucoamylase was 10 to 15 timeshigher in solid than in liquid culture,and higher in raw starch mediumthan in cooked cassava.

This work is being continued inthe EU-STD3 Program at theBioconversion Laboratory of theUniversidad del Valle, Cali,Colombia. It focuses on simplifyingcassava processing by learning moreabout the specificity of Rhizopusstrains in degrading the raw starchgranule. But clean flours of rawcassava are needed. The commonflours of cassava contain too muchnatural microflora to allow microbialstudies with fungi; they must first besterilized and (unfortunately)gelatinized. Ramírez et al. (1994)developed raw cassava flour with avery low content of bacteria andfungi, and little gelatinization.

To measure gelatinization, thesimple method of Wotton et al.

(1971) was adopted and a goodcorrelation coefficient for thecalibration curve was obtained.Table 4 shows the effect of thermictreatment and microwaves on starchgelatinization in cassava flour (watercontent typically lower than 10%).Where water content was very low,gelatinization was also low.

The same thermic treatment ofdry cassava flour eliminated thenatural microflora contained in rawflour, from 109 bacteria/g of dry flourto fewer than 103 bacteria/g afterheating the flour for 30 min at90 °C. With gelatinization limited toless than 5% under such conditions,obtaining clean, raw cassava flour ispossible in the laboratory.

Figures 1 and 2 show the effectsof various physical and thermictreatments on the bacteria content ofcassava flour. Cassava flour will beused as a solid substrate forcultivating Rhizopus strains, and tocompare the capacity of selectedstrains to grow on raw or gelatinizedcassava starch.

190

Ca

ssava

Flou

r an

d S

tarch

: Progress in

Resea

rch a

nd

Develop

men

t

Table 3. Effect of cooking and type of culture on the growth and amylases of various strains of Rhizopus oryzae cultivated on cassava granules.

Strain of Liquid-state culturea Solid-state culturea

RhizopusRaw cassava Cooked cassava Raw cassava Cooked cassava

α − Gluco- Protein α − Gluco- Protein α − Gluco- Protein α − Gluco- Proteinamylase amylase (g/100 g amylase amylase (g/100 g amylase amylase (g/100 g amylase amylase (g/100 g(U/g DM) (U/g DM) DM) (U/g DM) (U/g DM) DM) (U/g DM) (U/g DM) DM) (U/g DM) (U/g DM) DM)

28168 42.20 9.60 3.90 157.20 3.10 10.00 39.30 55.30 10.60 178.40 46.22 12.30

34612 40.40 7.30 4.60 168.50 5.70 9.30 55.00 70.00 12.60 170.00 47.00 14.10

28627 76.00 7.80 4.00 145.40 3.30 9.60 98.00 108.00 11.40 167.00 37.00 13.80

a. DM = dry matter; U = enzyme units.

SOURCE: Soccol et al., 1994.

191


Table 4. Effect of temperature and microwaves on starch gelatinization of cassava flour.

Temperature Time Gelatinization rate (%)a

(min)Exp. 1 Exp. 2 Exp. 3 Mean

Test 1 75.439 84.063 88.911 82.80(80% gel.)

Test 2 25.411 26.184 29.702 27.10(20% gel.)

80 °C 60 3.529 3.444 2.714 3.2385 °C 30 3.529 3.357 3.487 3.4685 °C 3.444 3.486 3.444 3.4690 °C 30 3.572 3.444 3.572 3.5390 °C 60 9.454 9.064 9.107 9.2195 °C 30 6.961 5.546 5.803 6.10

100 °C 30 4.965 4.602 4.001 4.52105 °C 30 6.961 5.503 5.301 5.92120 °C 30 4.816 4.730 4.473 4.67140 °C 30 4.773 3.100 3.100 3.66160 °C 30 3.529 3.487 4.301 3.77

Autoclaving 15 3.572 3.100 4.301 3.66(121 °C)

Microwaves 5 2.886 2.410 2.842 2.71(Pot. 70)

Microwaves 5 2.971 2.242 2.242 2.49(Pot. 100)

Microwaves 15 3.879 3.057 3.915 3.62(Pot. 30)

a. Exp. = Experiment. Mean is across the experiments.

Duration of treatment (minutes)

Figure 1. Total microflora (plate count analysis)in cassava flour, according totreatment. ( = ultra-violet radiation; = microwaves; = 80 °C; = 85 °C; = 90 °C.)

123123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123123

123123123123123123123

123123123123123

123123123123

123123123123123

123123123123123123

123123123123123123123

Via

ble

bact

eria

(n

/g

of flo

ur)

0 80 85 90 95 100 105120140 160180

Temperature (°C)

Figure 2. Effect of temperature on bacterialpopulation in cassava flour.

1010

109

108

107

106

105

104

103

102

109

108

107

106

105

104

103

102

101

Via

ble

bact

eria

(n

/g

of flo

ur)

0 30 60 90 120

192


Lactic Acid Fermentationof Cassava

Lactic acid fermentation is importantfor many traditional fermented foods,silage, and animal feed, and forrecycling agroindustrial byproducts.Because of its acid, bacteriostatic,and bactericidal properties,fermentation preventsmicroorganisms, whether parasitic,saprophytic, or pathogenic, frombreaking down vegetable material.

In tropical countries, lacticfermentation not only plays animportant role in the traditionaltransformation of starchy foods,such as cassava, but also in thetransformation and conservation ofother foods, and fish and itsbyproducts. Two types of lacticfermentation exist:

(1) Homolactic, when more than80% of total acidity andmetabolites formed consists oflactic acid, and

(2) Heterolactic, when thepercentage of acetic acid,propionic acid, and ethanol ismore significant, and lactic acidrepresents 50%-80% of totalacidity.

Lactic bacteria produce two typesof lactic acid: L(+) and D(-). Only theL(+) form is assimilated by humans.

Previous studies, realized duringthe EU-STD2 Program in 1988-1991(Raimbault, 1992), consisted ofimproving traditional fermented foodmade from cassava in Africa andLatin America. Three kinds oftraditional foods were considered:gari, chikwangue, and sour starch.We demonstrated the essential roleof lactic acid bacteria in alltraditional processes.

Amylolytic lactic bacteria wereisolated from fermented cassava.

The first strain of Lactobacillusplantarum to be described as havingvery high amylolytic capacity wasobtained from fermented cassava byA. Brauman in the Congo. Detailedphysiological and biochemicalcharacterization of this new strainis expected to be published soon byE. Giraud.

Mbugua and Njenga (1991) andSvanberg (1991a; 1991b), workingin Tanzania and at the UppsalaUniversity, respectively, havereported on the effect of lactic acidfermentation on the pathogenmicroflora content of traditionalAfrican foods.

Some of their results arereported in Table 5 and Figure 3,which show how lactic acidbacteria reduce the number offood-poisoning pathogens such asspecies of Staphylococcus,Salmonella, and Shigella, andEscherichia coli. High levels of suchpathogens are sometimes found intraditional foods after processingunder unhygienic conditions,especially those for malting maizeduring the rainy season in parts oftropical Africa.

Lactic fermentation oftraditional foods reduces pathogenicbacteria from 108 to 103. The sameauthors also found a significantcorrelation between the resistanceof young children to diarrhoea andeating acidified gruels.

We are bioconverting, throughprobiotics and bactericides, cassavaflour and starch containingamylolytic lactic acid bacteria toisolate new strains from traditionalfoods. At the same time, we arebroadening knowledge on thecultivation of lactic acid bacteria instarchy substrates. We hope suchinformation will help elaborate newfood and feed products.

193


Table 5. Effect of lactic acid fermentation on the content of pathogenic bacteria in traditional fermentedfoods in Africa.

Time (h) Log number of bacteria/g food

Control Nonfermented, Fermented foodacidified

food Flour Gruel(nonviable) (viable)

Shigella flexneri0 6.8 6.7 6.4 6.03 6.6 5.8 5.1 4.07 7.0 4.2 5.5 3.3

24 7.0 4.1 3.7 2.7

Salmonella typhimurium0 8.5 8.1 8.3 7.73 8.0 6.7 6.0 7.17 7.9 5.3 4.4 6.3

24 8.9 4.0 2.0 2.0

SOURCE: Lorri and Svanberg, 1988.

production. This may be because,first, cassava cultivation yieldsrelatively few, commercially significantbyproducts, compared with, forexample, sugarcane which yieldsenormous quantities of bagasse, avaluable source of energy fordistillation. Second, cassava starchneeds to be hydrolyzed into sugar forbioconversion into ethanol by thecommon Saccharomyces cerevisiae.This implies an additional, costly step.

For cassava to be an economicallyviable energy source, its processingcosts must be reduced. Solid-statefermentation is one, simple, and newmethod of reducing costs: the use of anamylolytic yeast that eliminateshydrolysis.

At the ORSTOM Laboratory,Saucedo et al. (1992a; 1992c)developed a new process for the solidculture of an amylolytic yeast,Schwanniomyces castelii (Figure 4).The main advantage of this techniqueis its continuous recuperation ofethanol in a cold trap condenser. Thegas produced in the reactor is pumpedthroughout the system, thus ensuringits continual removal from the medium

Log

via

ble

bact

eria

(n

/g

gru

el)

Figure 3. Evolution of pathogenic bacteria duringthe lactic fermentation of uji, afermented cassava gruel (after Mbuguaand Njenga, 1991). ( = Staphylococcusaureus; = Salmonella typhimurium; = Escherichia coli; = Shigelladysenteriae.)

Time (h)

10

9

8

7

6

5

4

3

Alcoholic Fermentation ofCassava and Starch Products

Cassava is a potential producer ofethanol, considering its potentiallyhigh yields and low costs. Yet fewreports concern the industrialapplication of cassava for ethanol

0 10 20 30 40 50 60

194


Table 6. Comparison of various processes for ethanol production from cassava in liquid or solid substrate.

Process Hydrolysis Sugar Ethanol Recovered Theoretical(g/L) (g/L) (g/L) (%)

Liquid substrate, using S. cerevisiae a, b + 145 72.50 72.50 83.2

Solid substrate, using S. cerevisiae b, c + 165 41.73 41.73 65.0

Solid substrate, using Rhizopus koji d - 200 110.00 110.00 83.0

Solid substrate, using Schw. castelii e, f - 300 68.40 212.60 64.0

a. Saraswati, 1988.b. S. = Saccharomyces.c. Jaleel et al., 1988.d. Jujio et al., 1984.e. Schw. = Schwanniomyces.f. Saucedo et al., 1992a.

Figure 4. Producing ethanol through solid-substrate fermentation of cassava starch. The reactor containsa solid support impregnated with a starchy suspension and inoculated with the fermentationagent, an amylolytic yeast known as Schwanniomyces castelii. The resulting gas is pumped to acondenser where ethanol is extracted. The residual gas is sent to a humidifier.

potential of cassava as a substratefor ethanol production. The solid-state technique has to be carefullyconsidered. Results obtained withthe fungus Rhizopus koji areparticularly significant. Thepotential of Schwanniomyces is alsointeresting because amylolytic yeastwould be easier to control at thesmall-scale industrial level.

Column tohumidify gas

Pump

Continuousextraction ofethanol

Cold trapcondenser

Reactor

Ethanol

and limiting its toxic effects on theyeast’s metabolism. The resultsobtained by Saucedo et al. (1992a;1992b; 1992c) were promising, butthe technology and feasibility of theprocess for commercial operationneed further research.

Table 6 shows the resultsobtained by various authors on the

195


__________; Gosselin, L.; and Raimbault, M.1992. Degradation of cassavalinamarin by lactic acid bacteria.Biotechnol. Lett. 14(7):593-598.

__________; __________; Marín, B.; Parada,J. L.; and Raimbault, M. 1993a.Purification and characterization ofan extracellular amylase fromlactobacillus plantarum strain A6.J. Appl. Bacteriol. 75:276-282.

__________; __________; and Raimbault, M.1993b. Production of aLactobacillus plantarum starter withlinamarase and amylase activities forcassava fermentation. J. Sci. FoodAgric. 62:77-82.

Jaleel, S. A.; Srikanta, S.; Ghildyal, N. P.;and Lonsane, B. K. 1988.Simultaneous solid phasefermentation and saccharification ofcassava fibrous residue forproduction of ethanol. Starch/Stärke40(2):55-58.

Lorri, W. S. M. and Svanberg, U. 1988.Improved protein digestibility incereal based weaning foods by lacticacid fermentation. Harare, Zimbabwe.

Mbugua, S. K. and Njenga, J. 1991.Antimicrobial properties of fermentedUJI as a weaning food. In: Westby, A.and Reilly, P. J. A. (eds.). TraditionalAfrican foods: quality and nutrition.International Foundation of Science(IFS), Sweden. p. 63-67.

Oriol, E.; Raimbault, M.; Roussos, S.; andViniegra-González, G. 1988a. Waterand water activity in the solid statefermentation of cassava starch byAspergillus niger. Appl. Microbiol.Biotech. 27:498-450.

__________; Schetino, B.; Viniegra-González,G.; and Raimbault, M. 1988b. Solidstate culture of Aspergillus niger onsupport. J. Ferment. Technol.66:1-6.

Raimbault, M. 1992. Etudes physiologiqueset génétiques des bactérieslactiques dans les fermentationstraditionnelles du manioc. Finalreport CEE/STD2, no. TS2A-00226.Institut français de recherchescientifique pour le développement encoopération (ORSTOM), Montpellier,France. p. 1-53. (Internaldocument.)

Conclusions onBioconverting Cassava and

Potential Products

To bioconvert cassava starch andflour to elaborate new products,ORSTOM, CIRAD, and collaboratinginstitutes are emphasizing twoapproaches: solid-state fermentation,and lactic acid fermentation.

The first is of great interestbecause of its potential to simplifyprocesses and reduce costs, and itslarge reactor volume. Both Rhizopusand Schwanniomyces (or otheramylolytic) yeasts can be used in asolid-state cultivation process. Thisimplies a three-phase reactor with asolid fiber support, a liquid phasecontaining the substrate insuspension and salts, and a gaseousphase for exchanging volatilecomponents, that is, oxygen, water,and ethanol.

In lactic acid fermentation, weare investigating the culture controlof amylolytic lactic acid bacteria inmixed and composite starters able toremain competitive in a natural,nonaxenic environment. Theprophylactic role of lactic acidbacteria is also of great interest.

Finally, we are studyingmicroorganisms able to degradenative cassava starches withoutneed of gelatinization, as in innatural biotransformation andbiodegradation. We will alsostudy the amylolytic capacity ofRhizopus spp., yeasts, and lactic acidbacteria.

References

Giraud, E.; Brauman, A.; Kéléke, S.; Lelong,B.; and Raimbault, M. 1991.Isolation and physiological study ofan amylolytic strain of Lactobacillusplantarum. Appl. Microbiol.Biotechnol. 36:379-383.

196


__________ and Alazard, D. 1980. Culturemethod to study fungal growth insolid fermentation. Eur. J. Appl.Microbiol. Biotechnol. 9:199-209.

__________ and Viniegra, G. 1991. In:Chahal, D. S. (ed.). Modern andtraditional aspects of solid statefermentation in food, feed and fuelfrom biomass. p. 153-163.

__________; Revah, S.; Pina, F.; and Villalobos,P. 1985. Protein enrichment of cassavaby solid substrate fermentation usingmolds isolated from traditional foods.J. Ferment. Technol. 63(4):395-399.

Ramírez, C.; de Stouvenel, A.; and Raimbault,M. 1994. Effect of physical treatmentson microflora content in cassava flour.Poster presented at the InternationalMeeting on Cassava Flour and Starch,held in January 1994 at Cali,Colombia.

Saraswati. 1988. The experience of pilot plantof ethanol from cassava in Indonesia.Regional Workshop on Upgrading ofCassava/Cassava Wastes byAppropriate Biotechnologies, Bangkok,Thailand, 1987. Thailand Institute ofScientific and Technological Research,Bangkok, Thailand. p. 41-49.

Saucedo, G.; González, P.; Revah, S.; Viniegra,G.; and Raimbault, M. 1990. Effect ofLactobacilli inoculation on cassava(Manihot esculenta) silage:fermentation pattern and kineticanalysis. J. Sci. Food Agric.50:467-477.

__________; Lonsane, B. K.; Navarro, J. M.;Roussos, S.; and Raimbault, M.1992a. Potential of using a singlefermenter for biomass build-up, starchhydrolysis and ethanol production:solid state fermentation systeminvolving Schwanniomyces castelii.Appl. Biochem. Biotechnol. 36:47-61.

__________; __________; __________; __________;and __________. 1992b. Importance ofmedium pH in solid state fermentationsystem for growth of Schwanniomycescastelii. Lett. Appl. Microbiol.15:164-167.

__________; __________; and Raimbault, M.1992c. Maintenance of heat andwater balance as scale-up criterionfor production of ethanol bySchwanniomyces castelii in solidstate fermentation system. ProcessBiochem. 27:97-107.

Soccol, C.; Iloki, I.; Marín, B.; andRaimbault, M. 1994. Comparativeproduction of alpha-amylase,glucoamylase and proteinenrichment of raw and cookedcassava by Rhizopus strains insubmerged and solid statefermentations. J. Food Sci. Technol.31:320-332.

__________; Marín, B.; Roussos, S.; andRaimbault, M. 1993a. Scanningelectron microscopy of thedevelopment of Rhizopus arrhizuson raw cassava by solid statefermentation. Micol. Neotrop. Apl.6:27-39.

__________; Rodríguez, J.; Marín, B.;Roussos, S.; and Raimbault, M.1993b. Growth kinetics of Rhizopusarrhizus in solid state fermentationof treated cassava. Biotechnol.Tech. 7(8):563-568.

Svanberg, U. 1991a. Lactic fermentation ofcereal-based weaning gruels andimproved nutritional quality. In:Westby, A. and Reilly, P. J. A. (eds.).Traditional African foods: qualityand nutrition. InternationalFoundation of Science (IFS),Sweden. p. 53-60.

__________. 1991b. The potential role offermented cereal gruels inreduction of diarrhoea amongyoung children. In: Westby, A. andReilly, P. J. A. (eds.). TraditionalAfrican foods: quality and nutrition.International Foundation of Science(IFS), Sweden. p. 33-38.

Wotton, M.; Weedon, D.; and Munck, N.1971. A rapid method for estimationof starch gelatinization in processedfoods. Food Technol. Aust.23:612-614.

197

Cassava Lactic Fermentation in Central Africa:...

CHAPTER 23

CASSAVA LACTIC FERMENTATION INCENTRAL AFRICA:MICROBIOLOGICAL AND BIOCHEMICAL

ASPECTS

A. Brauman*, S. Kéléke**, M. Malonga***,O. Mavoungou***, F. Ampe†, and E. Miambi***

cyanogenic compounds (e.g.,concentration decreased from400 ppm in fresh cassava to 20 ppmin fermented mash); (2) a significantlysis of cassava cell walls due to thesimultaneous action of endogenouspectin methylesterase and bacterialpectin lyase; and (3) the production oforganic acids (C2 to C4), mainly lactateand butyrate, that contribute to thetypical flavors of chikwangue andfufu.

In the study, most microflorainvolved in retting were facultative,anaerobic, fermentative bacteria,among which lactic bacteria werepredominant. From the second day offermentation, endogenousLactobacillus species were totallysupplanted by Leuconostocmesenteroides and Lactococcus lactis.Anaerobic bacteria such asClostridium butyricum were also foundand seemed responsible for initiatingbutyrate production. Yeasts playedno significant role, but theirincreasing number at the end of theprocess (Candida species) probablyinfluenced the conservation of endproducts.

Despite the significant number ofamylolytic bacteria (105-106 b/ml), theamylase activity found in the rettingjuice came from the roots anddisappeared after 48 h offermentation. The main enzymes of

Summary

Retting is a lactic fermentation duringwhich cassava roots are soaked forlong periods in water. Despite theimportance of this fermentation, nokinetic study of it has beenundertaken. Our study thereforeexamined the biological and physicalchanges of cassava roots duringretting to provide a basis for itspossible mechanization.

The study was carried out to(1) enumerate and characterize themain microorganisms of the process;(2) determine the evolution ofphysicochemical parameters duringretting; and (3) measure theproduction of organic products andsome principal enzyme activities.

Retting can be characterized bythree essential transformations of theroots: (1) a degradation of endogenous

* Institut français de recherche scientifiquepour le développement en coopération(ORSTOM), Paris, France.

** Laboratoire de microbiologie, Directiongénérale de la recherche scientifique ettechnique (DGRST), Brazzaville, Congo.

*** Laboratoire de biologie cellulaire, Faculté dessciences, Université Marien-N’Gouabi,Brazzaville, Congo.

† Laboratoire de microbiologie et debiotechnologie, ORSTOM, stationed inBrazzaville, Congo.

198


this process were cassava pectinmethylesterase, bacterial pectinase,and endogenous linamarase.

The pH became stable at about4.5 after 48 h and the partial oxygenpressure dropped to 0.2 mg/L after10 h.

These results suggested thatretting is a typical heterolacticfermentation with a significantproduction of butyrate.

Introduction

Processed cassava (Manihotesculenta Crantz) is eaten in Westand Central Africa in such forms asgari, lafun, fufu, chikwangue, andtapioca. In the Congo, the world’ssecond largest cassava consumerafter Zaïre (Trèche, n.d.), cassavaroots account for 47% of the calorieintake (Trèche and Massamba,n.d.b).

The two main products associatedwith fermented cassava are fufu andchikwangue. The former is a flourobtained from sun-dried cassavamash that is pulverized. This flourmay be mixed with boiling water andserved in bowls with sauce and fish ormeat. Chikwangue, a cassava bread,is obtained after multiplepostfermentation steps, includingdefibering and pugging (Trèche andMassamba, n.d.a).

Both products require afermentation in which the roots soakfor 3 to 6 days in tap water. Duringthis process, cyanogenic compoundsare eliminated, flavor compoundsare elaborated, and the roots soften(Okafor et al., 1984; OladeleOgunsa, 1980). Softening isindispensable for further rootprocessing but the mechanismsinvolved are not yet fullyunderstood.

Significant differences exist inretting processes throughout CentralAfrica and even in the Congo. Peeledor unpeeled roots are retted in rivers,standing water, large barrels of water,or even buried in soil. Thefermentation temperature varies withseason and location. Suchdifferences, combined with the lowreproducibility of the local processors,lead to a variability in quality andtaste of cassava foods (Trèche andMassamba, n.d.a).

To increase the quality of thesetraditional products and provide abasis for the possible mechanizationof the process, the European Union(EU) Program-STD2, known as“Improving the Quality of TraditionalFoods Processed from FermentedCassava” was set up in 1990 inCentral Africa and South America.Our laboratory was to describe themechanisms of root transformationduring retting with a view tooptimizing product quality andfermentation speed.

In this paper, we present the mainresults obtained during this EUprogram, describe the microbiologicaland biochemical evolution throughoutthe process, and define the origin(vegetal or microbial) of the mainenzymes.

Material and Methods

Origin of plant material

Cassava roots (Manihot esculenta var.MM 86, or ‘Ngansa’) were harvestednear Brazzaville, Congo, 18 monthsafter planting.

Retting procedures

About 100 kg of washed and peeledroots were placed in a barrel and thevolume made up to 50 L with rainwater. A second barrel, filled only

199


with rain water, was used as controlfor physicochemical measurements(T °C, pH, pO2). Samples were takenevery 12 h for the first 2 days andthen every 24 h until retting wascompleted.

Sample preparation for bacterialenumeration

Sampling was carried out byrandomly selecting six root sections,which were then cut into 0.5-cmcubes and mixed under sterileconditions. Of this mixture, 60 gwere extracted and diluted in 540 mlof sterile, peptonized water (dilution10-1). The solution was then mixed ina Blendor (Turnmix ME 88,SOFRACA, France) and seriallydiluted in sterile, peptonized water foraerobic counts and in anaerobicHungate tubes containing sterile,reduced water, flushed with 20% CO2

and 80% N2 for anaerobic counts.

Methods of bacterial quantification

Two types of enumeration wereperformed: “most probable number”(MPN) enumeration and plate countson solid medium. The MPN methodwas used to either ascertain thegrowth of fermentative andpectinolytic bacteria or count themetabolites produced during growthon appropriate media for anaerobic,lactate-using bacteria. For each MPNdetermination, four successivedilutions of root samples wereinoculated in three or four tubes perdilution. Results were calculatedaccording to the McCready tables(McCready, 1918).

For plate counts, 0.1 ml samplesof appropriate dilutions wereinoculated in triplicate on agarmedium in plates. All the plateswere incubated at 30 °C and thenumber of colony-forming unitsdetermined after 48 or 72 h ofincubation.

Bacterial enumeration

Lactic acid bacteria (l.a.b.).The l.a.b. were enumerated on MRSagar medium (de Man et al., 1960),supplemented with 0.1% of anilineblue. In each petri dish, 0.1 ml ofappropriate root sample dilution wascovered with medium and kept at45 °C. Enumeration was carried outafter a 48-h incubation at 30 °C.Subcultures were further purified byrepeated plating.

Strains were differentiated intovarious bacterial groups by thefollowing tests: microscopyexamination, gram reaction, catalasetest, and oxygen metabolism(fermentative or oxidative) test in softMRS agar. Strains which were grampositive, catalase and oxidasenegative, nonmotile rods or cocci, andcolored by aniline blue wereconsidered as lactic bacteria.

Glucose- and lactate-fermentingbacteria. These bacteria (g.f.b. andl.f.b., respectively) were enumeratedon a basal medium that contained theequivalent of 2 g/L glucose or 5 g/Lof lactate (used as a carbohydratesource); 0.5 g/L of trypticase andyeast extract; 0.5 g/L of cysteine HCl(used as a reductive agent); 0.1 g/Lof sodium acetate; 0.005 g/L ofresazurine; 20 ml of Widdel mineralsolution (Widdel and Pfennig, 1984);and 1 ml of Widdel trace elementsolution (Widdel and Pfennig, 1984).

The Hungate technique (Hungate,1969), modified for using syringes(Macy et al., 1972), was usedthroughout the study. After boiling,the medium was cooled under acontinuous flow of oxygen-free N2,adjusted to a pH of 7.2 with NaOHsolution, and distributedanaerobically into Hungate tubes.The medium was sterilized for 35 minat 110 °C. Before inoculation, 1% ofNa2S-9H2O (5%) was added as a

200


reductive agent to each tube.Inoculations were performed withsyringes filled with oxygen-free N2,using a gas manifold.

Yeast. A potato-dextrose agarmedium (PDA, DIFCO Laboratory) wasprepared, containing 0.05 g/L ofchloramphenicol and with a final pHof 3.5, adjusted with tartaric acid(10%). The agar’s surface was thendried. From an appropriate microbialdilution, 0.1 ml was spread, intriplicate, on plates containing themedium. The plates were thenincubated for 72 h at 30 °C.Subcultures were further purified byrepeated plating on PDA. Isolateswere characterized to the genus level,and Api tests (API 5030 stripsBiomerieux, France) were used todetermine fermentation carbohydratesources.

Physicochemical parameters

Penetrometry index.Penetrometry was used to indicateroot softening during retting. Aprevious study showed that apenetrometry index of 15 mm/5 scorresponded to the end of retting asit is traditionally evaluated (Braumanet al., n.d.). A penetrometer (PNR10-SUR, Berlin) was used to measurethe consistency of the roots. Every10 h, and for each experiment, sixroot sections were randomly chosen.Penetrometry depth was estimatedwith six repetitions for each rootsection.

The pH and partial oxygenpressure of the retting juice. Every10 h, 50 ml of retting juice wasextracted to test the pH (measuredwith CG 838 pH-meter from SCHOTTGeräte, Germany) and estimatepartial oxygen pressure (measuredwith OXI 91 from WTW, Germany).

The pH and partial oxygenpressure of the roots. A 20-g

sample was added to a Waringblender and mixed with 120 mldistilled water at low speed for 15 sand at high speed for 1 min. Themixture was then filtered through aGF/A filter and the volume made upto 200 ml with distilled water.Extracts were taken in duplicate at0 h, 48 h, and at the end of retting.Acidity was titrated with 0.01 MNaOH.

Biochemical analysis

Enzyme assays. A sample of40 g of cassava mash was added to aWaring blender, together with 80 mlof 0.1 M citrate buffer (pH = 6.5) andthe mixture homogenized. Themixture was held overnight at 4 °Cand centrifuged at 12,000 g for30 min. The supernatant waslyophilized and resuspended in 1/10volume of citrate buffer.

βββββ-glucosidase activity. Thiswas measured with a chromogen,p-nitrophenol-β-d-glucopyranoside,at 20 mM in 0.1 M of Na-phosphatebuffer (pH = 6.8) for 1 h at 25 °C.The reaction was stopped by addingan equal volume of 0.2 M sodiumborate (pH = 9.8), and p-nitrophenolwas determined with aspectrophotometer at 400 nm (Hoseland Bartz, 1975).

Linamarase. This was assayedwith linamarin as substrate and bymeasuring the appearance of CN-

(Giraud et al., 1992). To 400 µl ofextract, 100 µl of 50 mM linamarinin 0.1 M citrate buffer (pH = 6.0)were added. At regular intervals,50 µl aliquots were added to 50 µl of0.1 M NaOH to stop the reaction,and stored at 4 °C. Cyanide wasliberated by adding 50 µl of 0.1 MH2SO4 and 850 µl distilled water toeach aliquot, and was measuredwith a spectroquant kit (Merck,Darmstadt, Germany). One unit oflinamarase was defined as the

201


amount of enzyme that released1 µmol of CN- per minute.

Activity of pectinesterase (PE;pectin pectylhydrolase,EC 3.1.1.11). This was assayed bytitrating 1 ml of extract in 1% pectinat 30 °C (Grindsted RS400-DM 74%),and in 0.1 M NaCl and 1 mM NaN3.pH was increased to 7.0 with 0.01 MNaOH. One unit corresponds to theneutralization of 1 µmol of COO-/min.

Polygalacturonate lyase (PGL)activity. PGL activity was assayedby the Starr et al. (1977) procedure.This assay does not differentiatebetween endo-PGL (poly (1,4-α-d-galacturonide) lyase, EC 4.2.2.2) andexo-PGL (poly (1,4-α-d-galacturonide)exolyase, EC 4.2.2.9). One unit ofPGL corresponds to the formation of1 µmol of one unsaturated bond ingalacturonide between C4 and C5.

Polygalacturonase (PG; poly(1,4-ααααα-d-galacturonide)glycanohydrolase, EC 3.2.1.15).This was assayed by viscometry. To40 ml of 1% pectin in 100 mM ofacetate buffer (pH = 4.7), 0.5 ml ofextract was added. The rate ofreduction in viscosity was measuredat 25 °C in a viscometer (Haakemodel; VT 500, rotation: 150.93 s-1

and system MV-MV1). One unitcorresponds to the release of 1 µmolof hexose/min. Total activities areexpressed as units per 100 g ofcassava.

Action of pectic enzymes invivo. Sterilized slices of cassava wereinoculated with 50 µl of enzymeextract or 5 µl of purified pectolyticenzymes (endopolygalacturonaseP-5146 from Aspergillus niger;pectolyase P-3026 from A. japonicum;and pectinesterase P-0764 fromorange peel) (Sigma, Saint-QuentinFallavier, France). The inoculatedslices were placed in sterile beakerscontaining 10 ml of 0.01 M of citrate

buffer (pH = 5.0). Penetrometerreadings were estimated after 24 hand 48 h at 30 °C.

Cellulase, amylase, andxylanase activities. These activitieswere also assayed at 37 °C and pH of5.8, using the Somogyi procedure(Somogyi, 1945). The substrates weremicrocrystalline cellulose (100 mg)and xylan (18 mg/ml).

Other analytical methods

Total and free cyanides were assayedby the Cooke et al. method (1978).Protein was determined with amodified Lowry procedure (Bensadounand Weinstein, 1976).

Organic compounds

Sugars, volatile fatty acids (VFA), andlactate and ethanol concentrations inthe roots were determined byhigh-performance liquidchromatography (HPLC) of thesupernatant, as described by Giraudet al. (1991). The resulting columns(BioRad Laboratories, Richmond,California) were:

(1) Fast carbohydrate column formonosugars analysis (100 x 7,8 min) with 0.6 ml flow of milliQwater (pH = 6.0) at 70 °C;

(2) Aminex HP 42 A (300 x 7.8 minBiorad) for polyosides analysiswith 0.3 ml flow of milliQ water(pH = 6.0) at 70 °C;

(3) Aminex HP x 87H column with0.8 ml/min flow of H2SO4 6 mM at60 °C.


Kinetic studies of retting

We now present the results of ourglobal study of lactic fermentation.Kinetic parameters such as total andfermentative microflora,physicochemical parameters, and

202


substrates and metabolites producedhave been measured throughout theprocess. These results are the meanof seven rettings performed in barrelsunder the same conditions.

Evolution of physicochemicalparameters

The main physicochemicalparameters were assayed throughoutthe process (Figure 1). The partialoxygen pressure dropped to wellbelow 1 mg/L after 10 h and the pHbecame stable (at 4.5) within 48 h.Conversely, root softening, indicatedby the penetrometry index, appearedafter 2 days of fermentation andevolved exponentially. This processseems to require anaerobic and acidicconditions to proceed. Microscopicexamination shows that the cassavacell walls were extensively disruptedat the end of the process,demonstrating the attack ofdepolymerizing enzymes.

The concentration of endogenouscyanogenic compounds decreasedfrom 300 mg/kg as HCN (dry matter

ppm

300

200

100

00 1 2 3 4 5

Figure 2. Total cyanide evolution.( = linamarin; = cyanhydrines +free cyanides; = free cyanides.)

Time (days)

basis) in fresh cassava to 20 in thefermented mash (Figure 2). In allassays, total cyanogens werealmost eliminated (90%). Theseresults demonstrated that, underthe standard conditions of localtransformations in Central Africa,detoxification occurred normallywithout need of an additionalprocess.

Figure 1. The evolution of physicochemical parameters during retting. ( = pH; = pO2;

= penetrometry index.)

12

10

8

6

4

2

0

Pen

etro

met

ry in

dex

7

6

5

4

5

4

3

2

1

0

Time (days)

0 1 2 3 4

pH

pO

2 (m

g/L)

203


Evolution of substrates andmetabolites

The main substrates degraded(Figure 3) were oligosaccharides(fructose, glucose, and saccharose).The low level of polyosides generatedby starch degradation (e.g.,maltotriose and maltose) underlinethe weak degradation of the starchymass during retting. Saccharoseseems to be the main substratedegraded by the fermentativemicroflora.

The main organic acid producedwas lactate. However, significantlevels of ethanol, acetate, andbutyrate were also found (Figure 4).They seem to be generated mostly bythe heterolactic fermentation of theoligosaccharides present in thecassava roots, except for butyrate,which could have come from ananaerobic fermentation mediated byClostridium species. Butyrateconcentration could vary from 0.4 to2.5 g/100 of dry matter in differentfermentations carried out under the

Figure 4. Organic acids and alcohol evolution during retting. ( 123123123 = butyrate;

123123 = ethanol;

123123123 = acetate; = lactate.)

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0

Time (hours)

123123123123

12341234123412341234

123123123123

123123123123

123412341234

123412341234

123123123

123412341234

1231231231231234123412312341231231234

123123123123

Con

c. g

per

100 g

of

dry

matt

er

Figure 3. Oligo- and monosaccharide evolution during retting. ( = maltotriose; = maltose;123123 = saccharose;

1212 = glucose;

1212 = fructose.)

123123123123123123123

12121212121212121212

123123123123

123123123123123

123123123123123123123123123

123123123

123123123123

121212121212

123123123

12121212121212

123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123

123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123

123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123123123123123

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

00 12 24 36 48 62

Time (hours)

Con

c. g

per

100 g

of

dry

matt

er

0 12 24 36 48 62

204


same conditions. Because of theirorganoleptic qualities, butyrate andlactate seem to be the most typicalproducts of this process.

Microflora evolution

Fermentative and lacticmicroflora. In the enumerations,only fermentative bacteria werecounted because retting was seen aslargely anaerobic (Figure 1). Thefermentative microflora evolvedduring the first 2 days of fermentationand remained stable to the end. Thetotal fermentative microflorarepresented by the glucose-fermentingbacteria was dense, reaching 1012 b/gafter 48 h of fermentation. The nextmost predominant flora were thel.a.b. (Figure 5), reaching 104 to108 b/g of DM on fresh roots. Thevariation of endogenous l.a.b.,composed mainly of Lactococcus andheterolactic Lactobacillus species, didnot influence the evolution of l.a.b.during fermentation.

Lactate-fermenting bacteria.One metabolite formed duringfermentation is butyrate (Figure 4).This compound is a typical product ofcarbohydrate fermentation by

anaerobic spore formers (Clostridiumspecies). To evaluate this population,enumeration was done anaerobicallyon lactate because (1) lactate is themajor substrate found in retting; and(2) it is not used as a substrate by thel.a.b. Surprisingly, the results of thisenumeration showed that thepopulation of lactate-fermentingbacteria remained constant and atlow levels (103 b/g of DM) throughoutthe retting (Figure 5). The presence ofbutyrate and acetate in the positivetubes, and the isolation of strictlyanaerobic, sporulating, gram-positiverods with the same fermentationpattern as Clostridium butyricum,confirmed that Clostridium species arepresent in retting. However, their rolein the process remains to be studiedbecause of their reduced numbers inthe enumeration and lactate does notseem to be their natural substrate inretting.

Yeasts. The only flora thatappeared after 48 h of fermentationand still developed until the end ofretting were yeasts. Theirmetabolisms allow them to grow atthe low pH imposed by the l.a.b.Their numbers remained low duringthe fermentation (about 103 b/g of

Figure 5. Evolution of fermentative microflora during retting. ( 123123123 = glucose-fermenting bacteria;

121212 = lactic acid bacteria;

123123123 = lactate-fermenting bacteria; = yeast.)

1234123412341234123412341234123412341234123412341234

123123123123123123123123123123

123412341234123412341234123412341234

123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234

1234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123123

123412341234123412341234123412341234

12341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123123123123123123123123

1234123412341234123412341234

1234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234

12341234123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123

14

12

10

8

6

4

2

0

Log

. b/g

of d

ry m

att

er

Time (hours)

0 24 48 60 72

205


DM), suggesting that they do notplay a significant role in retting.When the retting finished, the yeastscovered the entire water surface andbecame the main flora of thepostretting stage. Their increasingnumbers at the end of the process(mostly Candida species) maytherefore influence the conservationof end products.

Origin of enzymes involved inretting. The main enzymes foundin this process were pectinase andlinamarase, and to a lesser extent,amylase (data not shown). Nocellulase or xylanase activities werefound in retting. To elucidate theorigin of cyanogen elimination andthe mechanism of root softening,two fermentations were carried outsimultaneously: one “natural,” usedas a control (CF), and one sterile(SF). pH and oxygen pressure of SFwere set on those of CF. Pectinaseand linamarase activities wereassayed throughout the experiment.For SF, cassava roots were sterilizedwith HgCl2 and soaked in sterilewater.

Origin of softening. Nosoftening was obtained in sterilefermentation (Figure 6). Highendogenous pectin methyl esteraseactivities were found in cassavaextracts from both fermentations(Figure 7). Depolymerizing enzymes,endopolygalacturonase (active atlow pH), and pectate lyase werefound only in the “natural”fermentation (Figures 8 and 9). Noother depolymerizing enzymes, suchas cellulase or xylanase, nor otherhydrolases were found. Moreover,softening could be performed byinoculating commercialpectinesterase and depolymerizingpectolytic enzymes on fresh andsterile cassava roots.

We suggest, therefore, that rootsoftening is a result of the combinedaction of both endogenous pectinmethyl esterase and exogenousbacterial depolymerizing enzymes.But further studies are needed toshow the precise contribution ofeach pectic enzyme to rootsoftening.

Pen

etro

met

ry in

dex

(m

m/s)

15

10

5

00 20 40 60 80

Time (hours)

µmol

of

CO

O- p

er m

in/m

g of

pro

tein

s

1.0

0.8

0.6

0.4

0.2

0

123123123123123123123123123123123123123123123123123123

12341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123

12341234123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123

123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123

12341234123412341234123412341234123412341234123412341234

0 9.5 20 27 44

Time (hours)

Figure 6. Comparative evolution of softeningbetween a sterile ( ) and a naturalretting ( ).

Figure 7. Pectinesterase activity during retting.( 123123 = sterile fermentation;

1212 = control

fermentation.)

206


50

40

30

20

10

0µmol

of

CO

O- p

er m

in/m

g of

pro

tein

s

123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

12345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678

123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

0 9.5 20 27 44

Time (hours)

Figure 8. Pectate lyase activity during “natural”fermentation.

Origin of cyanogen elimination.Of total cyanogenic compounds, 50%were eliminated in SF and 97% inCF (Figure 10). Enzyme assaysfurther confirmed endogenouslinamarase activity (Table 1).Linamarase activity (measured asβ-glucosidase activity) in CF wassignificant in fresh roots (specificactivity 9.4 units/mg protein). Thistotal activity then decreased after afew hours. In SF, total activityremained constant, but at a low level.The difference in β-glucosidaseactivity in the fresh roots between SFand CF may be attributed to theinhibitory effect of the HgCl2 used tosterilize the roots. However, as nearly25% (Table 1) of the totalβ-glucosidase activity present in thesterile roots can degrade more than50% of the total cyanide content ofthe fresh roots, we can assume thatthe level of linamarase activitypresent in the intact roots wassufficient to detoxify the roots.

Origin of the amylolytic activity.The amylase activity remainedconstant in SF, but disappeared after36 h of fermentation in CF (Figure 11).Our data suggest that the amylaseactivity detected in retting does nothave a bacterial origin as suggestedby different authors (Collard and Levi,1959; Oyewole and Odunfa, 1992;Regez et al., 1987).

123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

1234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

12345671234567123456712345671234567

0.3

0.2

0.1

0

Time (hours)

12345123451234512345123451234512345123451234512345123451234512345

1234512345

123456123456123456123456123456123456123456123456123456123456123456

12345123451234512345123451234512345123451234512345

123451234512345123451234512345123451234512345123451234512345

1234512345123451234512345123451234512345123451234512345123451234512345123451234512345123451234512345123451234512345123451234512345

1234512345123451234512345123451234512345123451234512345123451234512345123451234512345123451234512345123451234512345

12345123451234512345123451234512345123451234512345123451234512345123451234512345

123451234512345123451234512345

1234512345123451234512345

Per

cen

tage

of to

tal cy

an

ide 120

100

80

60

40

20

00 9.5 20 27 44

Time (hours)

Figure 10. Total cyanide evolution in control ( 121212 ) and sterile (

123123123 ) fermentations.

µm

ol o

f ga

lact

uro

nic

aci

d p

erm

in/m

g of

pro

tein

s

0 9.5 20 27 44

Figure 9. Endopolygalacturonase activity during“natural” fermentation.

207


20,000

10,000

0

communication) suggest thatClostridium species (such asClostridium butyricum) could beinvolved with Bacillus species (suchas Bacillus polymyxa) in rootsoftening as pectinase producers. Wedid not see any involvement ofGeotrichum spp. or Corynebacteriumspp., as have other authors (Collardand Levi, 1959; Okafor et al., 1984;Regez et al.,1987). Yeasts (mostlyCandida species) were more involvedin postretting.

Our biochemical analyses showedthat retting is a fermentation in whichboth endogenous and microbialenzymes coact to soften the roots anddegrade cyanogenic, endogenouscompounds. Our results suggestedthat cell-wall degradation is initiatedby endogenous pectinesterase, locatedin intercellular spaces and releasedby pH decrease. This is followed bymicrobial polygalacturonase and lyasedepolymerizing pectic chains. Thepresence of pectic enzymes in cassavaretting has previously been reported(Okafor et al., 1984; Oyewole andOdunfa, 1992). But this work givesthe first evidence of the vegetal originof pectinesterase and of the in vivoactivity of depolymerizing enzymes.

The amylase activity measured inretting seems to be of vegetal origin.But its low level of activity anddisappearance within the first 30 h ofretting suggest that it is notimportant to the retting process.

Results of cyanide measurementsindicate that endogenous linamarase(measured as β-glucosidase activity) isthe main enzyme responsible fordetoxification. We can assume, asMaduagwu (1983) suggested, that thelevel of linamarase activity present inintact roots is sufficient to detoxifythem of their cyanogen contentwithout help from any microbiallinamarase. Nevertheless, if bacteriado not directly detoxify cassava roots,

Conclusions

These results suggest that retting is acomplex heterolactic fermentation,with an interaction between lacticbacteria, Clostridium species, andpossibly Bacillus species. Heterolacticbacteria (such as Leuconostocmesenteroides) are the mostimportant and numerous microflorain the process; they are responsiblefor the physicochemical properties ofretting (e.g., pO2 and pH) and theproduction of the main organic acids(acetate and lactate). Clostridiumspecies seem to be involved inbutyrate formation, which is essentialfor the organoleptic properties of thefinal products. Moreover, recentresults (S. Kéléke, 1994, personal

Table 1. β-glucosidase activities in control andsterile fermentations. (Activities areexpressed in mmol per min/100 g ofdry matter).

Time (h) Fermentation

Control Sterile

0 9.12 2.159.5 5.58 2.55

20.0 6.10 1.7527.0 7.68 2.3044.0 7.24 1.38

Figure 11. Amylase activity in control ( ) andsterile ( ) fermentations.

0 20 40 60 80

Tot

al act

ivit

y (U

/L)

Time (hours)

208


they could help degrade linamarin bydestroying cell walls.

Findings from our study havehelped other researchers:

(1) Isolate and characterize the firstamylolitic Lactobacillus plantarum(strain A6) (Giraud et al., 1991);

(2) Improve fufu processing bysignificantly reducing retting time,and increase the organolepticqualities of the final product(Ampe et al., 1994);

(3) Adapt the process for areas withlow water availability (Miambi etal., n.d.).

References

Ampe, F.; Brauman, A.; Trèche, S.; andAgossou, A. 1994. The fermentationof cassava: optimization by theexperimental research methodology.J. Sci. Food Agric. 65:355-361.

Bensadoun, A. and Weinstein, D. 1976.Assay of protein in the presence ofinterfering materials. Anal. Biochem.70:241-250.

Brauman, A.; Kéléke, S.; Mavoungou, O.;Ampe, F.; and Miambi, E. n.d. Etudesyntétique du rouissage traditionneldes racines de manioc en Afriquecentrale (Congo). In: Agbor, E.;Brauman, A.; Griffon, D.; and Trèche,S. (eds.). Cassava food processing.Institut français de recherchescientifique pour le développement encoopération (ORSTOM) Editorials,Paris, France. (In press.)

Collard, P. and Levi, S. 1959. A two-stagefermentation of cassava. Nature(Lond.) 183:620-621.

Cooke, R. D.; Blake, G. G.; and Battershill,J. M. 1978. Purification ofcassava linamarase. Phytochemistry(Oxf.) 17:381-383.

de Man, J. C.; Rogosa, M.; and Sharpe, M. E.1960. A medium for the cultivation ofLactobacilli. J. Appl. Bacteriol.23:130.

Giraud, E.; Brauman, A.; Kéléke, S.; Lelong,B.; and Raimbault, M. 1991. Isolationand physiological study of anamylolitic strain of Lactobacillusplantarum. Appl. Microbiol.Biotechnol. 36:379-383.

__________; Gosselin, L.; and Raimbault, M.1992. Degradation of the cassavalinamarin by lactic acid bacteria.Biotech. Lett. 14(7):593-598.

Hosel, W. and Bartz, W. 1975. DFglucosidases from Cicer arientum L.Eur. J. Biochem. 57:607-616.

Hungate, R. E. 1969. A roll tube method forthe cultivation of strict anaerobes. In:Norris, J. R. and Ribbons, D. W.(eds.). Methods in microbiology,vol. 3B. Academic Press, NY.

McCready, M. H. 1918. Tables for rapidinterpretation of fermentation tuberesults. Can. J. Public Health 9:201.

Macy, J. M.; Snellen, J. E.; and Hungate,R. E. 1972. Use of syringe methodsfor anaerobiosis. Am. J. Clin. Nutr.25:1318-1323.

Maduagwu, E. N. 1983. Differential effects onthe cyanogenic glycoside content offermenting cassava root pulp byβ-glucosidase and microbialactivities. Toxicol. Lett. (Amst.)15:335-339.

Miambi, E.; Machicout, M.; Trèche, S.; andBrauman, A. n.d. Le rouissage sanseau, une nouveau procédé detransformation des racines demanioc. In: Agbor, E.; Brauman, A.;Griffon, D.; and Trèche, S. (eds.).Cassava food processing. Institutfrançais de recherche scientifiquepour le développement en coopération(ORSTOM) Editorials, Paris, France.(In press.)

Okafor, N.; Ijioma, B.; and Oyolu, C. 1984.Studies on the microbiology ofcassava retting for fufu production.J. Appl. Bacteriol. 56:1-13.

Oladele Ogunsa, A. 1980. Changes in somechemical constituents during thefermentation of cassava roots(Manihot esculenta Crantz). FoodChem. 5:249.

209


Oyewole, O. B. and Odunfa, S. A. 1992.Extracellular enzyme activitiesduring cassava fermentation for“fufu” production. World J. Microbiol.& Biotechnol. 8:71-72.

Regez, P. F.; Ifebe, A.; and Mutinsumu, M. N.1987. Microflora of traditionalcassava foods during processing andstorage: the cassava bread(chikwangue) of Zaire. Microb.Aliment. Nutr. 5:303-311.

Somogyi, M. 1945. Determination of bloodsugar. J. Biol. Chem. 160:61-68.

Starr, M. P.; Chatterjee, A. K.; Starr, P. B.;and Buchanan, G. E. 1977.Enzymatic degradation ofpolygalacturonic acid by Yersinia andKlebsiella species in relation toclinical laboratory procedures. J.Clin. Microbiol. 6:379-386.

Trèche, S. n.d. Importance du manioc enalimentation humaine dansdifferentes régions du monde. In:Agbor, E.; Brauman, A.; Griffon, D.;and Trèche, S. (eds.). Cassava foodprocessing. Institut français derecherche scientifique pour ledéveloppement en coopération(ORSTOM) Editorials, Paris, France.(In press.)

__________ and Massamba, J. n.d.a. Laconsommation du manioc au Congo.In: Agbor, E.; Brauman, A.; Griffon,D.; and Trèche, S. (eds.). Cassavafood processing. Institut français derecherche scientifique pour ledéveloppement en coopération(ORSTOM) Editorials, Paris, France.(In press.)

__________ and __________. n.d.b. Les modesde transformation traditionnels dumanioc au Congo. In: Agbor, E.;Brauman, A.; Griffon, D.; and Trèche,S. (eds.). Cassava food processing.Institut français de recherchescientifique pour le développement encoopération (ORSTOM) Editorials,Paris, France. (In press.)

Widdel, F. and Pfennig, N. 1984.Dissimilatory sulfate- orsulfur-reducing bacteria. In: Krieg,N. R. and Holt, J. G. (eds.). Bergey’smanual of systematic bacteriology,vol. 1. Williams and Wilkins, MD,USA. p. 663-679.

210


CHAPTER 24

A LACTIC ACID BACTERIUM WITH

POTENTIAL APPLICATION IN CASSAVA

FERMENTATION

E. Giraud*, A. Brauman**, S. Kéléke***,L. Gosselin*, and M. Raimbault†

* Laboratoire de biotechnologie, Institutfrançais de recherche scientifique pour ledéveloppement en coopération (ORSTOM),Montpellier, France.

** ORSTOM, Paris, France.*** Laboratoire de microbiologie, Direction

générale de la recherche scientifique ettechnique (DGRST), Brazzaville, Congo.

† ORSTOM, stationed in Cali, Colombia.

Introduction

Lactic microflora play an importantrole in the preparation of traditionalfoods based on fermented cassava,such as gari, chikwangue, fufu, andsour starch. But this microflora’sfunction in preserving foods,eliminating cyanogenic compounds,and improving organoleptic qualities isnot yet clear. Traditional technologiesare still used to manufacture thesefoods. As fermentation occursnaturally with lactic microflora, thequality of the food products is notuniform.

The mass inoculation of cassavaroots with one or several selectedstrains would permit a better controlover natural fermentation, thusresulting in a product of improvedquality. Because cassava containsmainly starch (more than 80% of drymatter), the selection of a lactic acidbacterium capable of metabolizingstarch (i.e., amylolytic) is essential.

But few lactic acid bacteria canconvert starch into lactic acid.Examples of amylolytic lactic acidbacteria are Streptococcus bovis, S.equinus, Lactobacillus amylophilus, L.amylovorus, L. acidophilus, L.cellobiosus, and others isolated fromanimal digestive tracts and plantwastes (Champ et al., 1983; Cotta,1988; Nakaruma, 1981; Nakarumaand Crowell, 1979; Sen and

Abstract

An amylolytic lactic acid bacterium,identified as Lactobacillus plantarum,was isolated from cassava roots(Manihot esculenta var. Ngansa) duringretting. Cultured on starch, the straindisplayed a growth rate of 0.43 perhour, a biomass yield of 0.19 g/g, anda lactate yield of 0.81 g/g. The growthkinetics were similar on starch andglucose. Enough enzyme wassynthesized, and starch hydrolysiswas not a limiting factor for growth.The synthesized amylolytic enzymewas purified by fractionatedprecipitation with ammonium sulfateand by anion exchangechromatography. It was identified asan α-amylase with an optimal pH of5.5 and an optimal temperature of65 °C. The use of such a strain as acassava fermentation starter for gariproduction had the following effects: achange from a heterofermentativepattern observed in naturalfermentation to a homofermentationone, a lower final pH, a faster pHdecline rate, and a greater productionof lactic acid (50 g/kg of dry matter).

211

A Lactic Acid Bacterium...

Chakrabarty, 1986; Sneath, 1986).Almost no information exists on thephysiology of these microorganisms.

Below we describe how we isolatedand identified a new amylolytic lacticacid bacterium from fermentingcassava roots. We also investigated thephysiology of this bacterium and theproperties of the amylase produced.

Methods

Isolating and identifying strains

Peeled roots were immersed in rainwater. Sampling was carried out4 days after fermentation by randomlyselecting six roots cut into 0.5-cmcubes and mixed under sterileconditions. A sample of 60 g wasdiluted in 540 ml of sterile peptonesolution. Then 0.1 ml of decimaldilutions were spread on JP2 medium(see below) in petri dishes. Afterincubation for 48 h at 30 °C, thedishes were exposed to iodine vapor todetect the starch hydrolysis areas.Isolated strains were then purified bythree successive transfers on JP2medium, and cultures routinelychecked for purity by microscopicobservation.

Microorganisms were identifiedby:

(1) the configuration of the lacticacid produced after treatment(Ivorec-Szylit and Szylit, 1965)with the enzymes dehydrogenasel and d (Boerhinger Mannheim);

(2) the microorganisms’ homolactic orheterolactic character, asdetermined by acetic acid or

(3) presence or absence of catalase;(4) microscopic and macroscopic

examination of morphology,mobility, and spores;

(5) Gram stain;(6) arginine dissemination;(7) growth at 15 and 45 °C; and

(8) fermentation of different carbonsources (API 50CH #5030 strips,Biomérieux, France).

“Bergey’s Manual” (Sneath, 1986)was used to evaluate results andidentify the different strains.

Strains and culture media

Three strains were used as reference:Lactobacillus plantarum (Lacto Labo,France), Streptococcus equinusCNCM 103233, and Lactobacillusamylophilus CNCM 102988T.

JP2 medium (g/L). Thisconsisted of:

M66 universal peptone 2.5Soya peptone obtainedby papain digestion 5Casein peptone obtainedby pancreatic digestion 2.5

Yeast extract 5Meat extract 2.5MgSO4,7H2O 0.1NaCl 3(NH4)2SO4 2K2HPO4 0.2Prolabo soluble starch 3Tween 80 (in ml) 0.4

The pH was adjusted to 6.75before sterilization.

Physiological studies wereperformed, using a deMan-Rogosa-Sharpe (MRS) basalmedium (de Man et al., 1960) andchanging the carbon sources to 5%glucose and 5% starch.

Culture conditions. Strains werecultured in a 2-L bioreactor(Biolafitte, France) at 30 °C andagitated at 200 rpm. The pH wasadjusted to 6.0 by adding NaOH (5 N).Inoculation at 10% v/v was performedwith a 20-h pre-culture in the samemedium used for fermentation.

212


Analytical methods

The biomass concentration wasdetermined by measuring the opticaldensity (OD) at 540 nm related to thedry weight measured after twowashing and centrifugation cycles anddrying at 105 °C for 24 h. For starchcultures, hydrolysis of residual starchwas performed with a mixture ofamylases (thermamyl + dextrosyme,supplied by Novo). The dry weightand OD were then determined asabove. Lactic acid, glucose, aceticacid, and ethanol concentrations inthe supernatant were assayedby high-performance liquidchromatography (HPLC). Compoundswere separated by using an AminexHPX 87H column (Bio RadLaboratory) with a 0.8 ml/min flow(pump LDC 3200) of H2SO4 (0.012 N)solution at 65 °C. Analyses werecarried out with a refractive indexdetector (Philips PU 4026). Totalsugars in media containing starchwere also determined with anthrone,using the Dubois et al. (1956)method.

Amylase assay. The α-amylaseactivity was measured by incubating0.1 ml of appropriately dilutedenzyme solution with 0.8 ml of asolution containing 1.2% of Prolabosoluble starch in 0.1 mol/Lcitrate-phosphate buffer (pH = 5.5) at55 °C. The reaction was stopped byadding 0.1 ml of 1 mol/L H2SO4. Afterincubation, residual starch contentswere determined colorimetrically afterdifferent periods at 620 nm by adding0.1 ml of the reaction mixture to2.4 ml of an iodine solutioncontaining 30 g/L of KI and 3 g/L of I2

and diluted to 4% with distilled water.

An enzyme unit is defined as theamount of enzyme that permits thehydrolysis of 10 mg of starch in30 min under the conditionsdescribed above. Proteinconcentration was estimated with the

Bradford (1976) method, using aBiorad Kit (Cat No. 500-0001,Ivry-sur-Seine, France) and bovineserum albumin as standard.

Purification of amylase.Fermentation was stopped afterculture for 9 h. Cells were removed bycentrifugation (at 15,000 g for 15 minat 4 °C), and the supernatant fluid(750 ml) filtered through a cellulosefilter (0.45 µm pore size, HAWP type,Millipore, Saint Quentin les Yvelines,France) to remove cell debris.

Powdered ammonium sulfate wasthen slowly added to the supernatantfluid under constant stirring at 4 °C.Most of the amylase activity wasprecipitated at between 50% and 70%saturation.

After the ammonium sulfatefractionation, the precipitatedprotein collected by centrifugation(at 15,000 g for 30 min at 4 °C) wasresuspended in 50 mmol/L KH2PO4/Na2HPO4 standard buffer (pH = 6.8).The enzyme solution was washed andconcentrated with a PM-10 Amiconultrafiltration membrane. It was thenloaded onto a diethylaminoethyl(DEAE) cellulose column (DE-52;Whatman Laboratory Sales, Hillsboro,Oregon, USA). The column(25 x 250 mm, flow rate 2.5 ml/min,25 °C) was previously equilibratedwith the standard buffer. The enzymewas eluted, using a concave, sodiumchloride gradient (0-1.0 mol/L).Fractions (5 ml) were collected. Thefractions that were enzymatically themost active were pooled, dialyzedovernight at 4 °C against the standardbuffer, and used for further studies.They were kept at -30 °C. No activitywas lost for at least 3 months undersuch conditions.

Polyacrylamide gelelectrophoresis. This was carried outaccording to Laemmli’s method (1970),with a 10% running gel and 4%

213


stacking gel. Electrophoresis undernondenaturating conditions wasperformed in the absence ofsodium dodecyl sulfate (SDS) andβ-mercaptoethanol in any buffer. Gelswere run at a constant 150 V for 1 h at25 °C. Proteins were stained by thesilver method (Oakley et al., 1980).

Amylase stain. Afterelectrophoresis, gel was incubatedfor 1 h at 30 °C in 0.1 mol/Lcitrate-phosphate buffer (pH = 5.5),containing 1% of soluble starch. Aftertwo washes with distilled water, lightlanes (representing starch hydrolysisareas of amylase activity) were detectedby immersing the gel in Lugol’ssolution.

Molecular mass determination.SDS-PAGE electrophoresis was used todetermine the approximate molecularmass of amylase. Marker proteins(Biorad, Cat. No. 161-0315) used weremyosin (200,000), β−galactosidase(116,250), phosphorylase-b (97,400),bovine serum albumin (67,000), andovalbumin (45,000).

Assays on gari. Fresh importedcassava roots from Cameroon wereobtained from Anarex (Paris, France).Gari was prepared from peeled, washedcassava roots, which were chopped andminced in a food mixer (SEB). Thepulp obtained was packed tightly intoplastic, sterile, screw-cappedcontainers (60 ml; OSI, A12.160.56)and placed at 30 °C.

Three batches were prepared:(1) natural fermentation, using theendogenous microflora present;(2) fermentation after inoculation withL. plantarum A6 (108 cfu/g of driedcassava), which had been cultured inbioreactors on cellobiose MRS medium;(3) fermentation after inoculation withL. plantarum Lactolabo (108 cfu/g ofdried cassava), which had beencultured in bioreactors on MRScellobiose. Cells were washed in

physiological solution before cassavainoculation.

A container from each batch wasmonitored every day to test the followingparameters:

(1) The pH was measured on a 10-gsample and homogenized indistilled water (20 ml). Moisturewas measured by drying a 10-gsample at 105 °C for 24 h.

(2) The number of lactic acid bacteria(l.a.b.) was estimated on a 10-gsample homogenized in90 ml of physiological sterilesolution. Colonies werecounted on MRS agar, using aspread-plate technique on petridishes and after incubation at30 °C and 48 h.


Isolation and identification ofLactobacillus plantarum A6

Seven amylolytic microorganisms wereisolated on JP2 medium from rettedcassava roots. Two were revealedby HPLC to have a capacity toproduce lactic acid from starch.Table 1 lists their morphological,physiological, and biochemicalcharacteristics. The ability of thesecultures to use 49 differentcarbohydrates was studied withAPI 50CH #5030 strips. The resultswere compared, by computer, with thepercentage of positive reactions ofdifferent Lactobacillus species as perAPI. A 99.9% rate of similarity withL. plantarum was observed and henceidentifying these cultures as strains ofL. plantarum. The two strains, A6 andA43, displayed precisely the same sugardegradation profiles, which suggeststhat they are probably the same.

The amylolytic activities on JP2medium of L. plantarum A6, S. equinus,and L. amylophilus indicated that the

214


Table 1. Characteristics of Lactobacillus plantarum strains A6, A43, and Lacto Labo (check).

Strain A6 A43 Check

Ratio of d:l lactic acid 69:31 66:34 73:27Homolactic + + +Catalase - - -Bacterium shape Short rod Short rod Short rodGram stain + + +Spore - - -Mobility - - -Dissemination of arginine - - -Growth at 15 °C + + +Growth at 45 °C - - -

starch hydrolysis zone was largest forL. plantarum A6. It was thereforeselected for further studies.

Lactobacillus plantarum A6 growthkinetics

The growth of L. plantarum A6 onglucose MRS medium (Figure 1) is fullycomparable with that of L. plantarum(Lacto Labo). The growth rate (0.43/h)and biomass productivity (0.75 g/L perhour) were slightly lower than those ofthe standard (Lacto Labo) strain, butthe biomass and lactate yields were

almost identical. The strain thereforedoes not seem to require nutrientsother than those of the commonstrain, suggesting that massproduction is possible.

On starch MRS medium, thestrain exhibits the same kineticprofiles (Figure 2) and the same yieldsas the standard strain. The rate ofstarch hydrolysis was greater than theuptake rate, leading to a 3 g/Lmaltose peak during the seventh hourof fermentation (results not shown).Thus, hydrolysis of starch is not alimiting factor.

Characterizing the amylolyticenzyme

To characterize the amylolytic activityexhibited by L. plantarum A6, acomparison was made of the HPLCprofiles after starch hydrolysis by thecell-free extract and commercialamylolytic enzymes (Aspergillus oryzaeα-amylase, Sigma A0273; potatoβ-amylase, Sigma A7005, andAspergillus niger amyloglucosidase,Sigma A3514). Under theseconditions, the main products ofstarch hydrolysis analyzed by HPLCwere glucose from amyloglucosidase,maltose from β-amylase, and amixture of glucose, maltose, andoligosaccharide (retention time of5.2 min) from α-amylase. The

Time (hours)

Glu

cose

, la

ctate

(g/

L)

Bio

mass

(g/L)

25

20

15

10

5

0

50

40

30

20

10

00 2 4 6 8 10 12 14

Figure 1. Fermentation of Lactobacillusplantarum A6 on MRS glucose( = glucose; = lactic acid; = biomass ). Temperature = 30 °C;

pH = 6.0.

215


purified fraction possessed anamylase activity. These procedureswere therefore considered sufficientfor purifying the extracellularamylase activity of L. plantarum A6.The SDS-PAGE analysis of thepurified fraction resulted in adistribution between a clearlydefined band (50 kDa) and a diffuseband with a molecular weight ofclose to 150 kDa.

Hypotheses. Several hypothesescan explain these many amylaseforms. We find the most satisfactoryis that which suggests that thepurified extract consists of apopulation of aggregates of a50-kDa amylase. This interpretationis based on the fact that most of thebacterial amylases described have amolecular weight of this order(Fogarty, 1983). This type ofaggregation of purified enzyme wasobserved in Bacillus subtilis amylase(Robyt and Ackerman, 1973), withzinc being the factor inducingclumping. The clumping factorremains to be defined in our case.

Further study is needed tosupport this hypothesis. The

50

40

30

20

10

0

25

20

15

10

5

0

5

4

3

2

1

0

Sta

rch

, la

ctate

(g/

L)

Bio

mass

(g/L)

Am

ylase

act

ivit

y (U

/m

l)

0 2 4 6 8 10 12 14

Figure 2. Fermentation of Lactobacillus plantarum A6 on starch MRS medium ( = starch ; = lactic acid; = biomass; = amylase activity). Temperature = 30 °C; pH = 6.0.

Time (hours)

breakdown profile of starch by theenzyme from L. plantarum A6 issimilar to that of α-amylase, therebyindicating that the enzymesynthesized by L. plantarum A6 isextracellular α-amylase.

Purification of amylase

The results of purifying the amylaseproduced by the strain L. plantarumA6 are summarized in Table 2. Thefirst step in purification wasconventional (NH4)2SO4 fractionation.The 50%-70% fraction revealedmaximum enzyme activity and wasselected for further purification byDEAE-cellulose. The elution profiledisplayed only one amylase activitypeak. The purification proceduredescribed above makes it possible, inonly two stages, to obtain a proteinfraction containing most of theamylase activity of L. plantarum A6enriched by a factor of nearly 20.

Testing the homogeneity of thefraction by electrophoresis undernative conditions revealed a majorprotein and three others that werequantitatively unimportant. However,all the proteins detected in the

216


Table 2. Purification of α-amylase of Lactobacillus plantarum strain A6 cultivated in a modified MRSmedium containing 2% (w/v) soluble starch and 0.5 g/L CaCl2 at 30 °C.

Materials Volume Protein Activity Specific activity Yield Purification(ml) (mg) (U) (U/mg) (%) (fold)

Culture filtrate 750.0 82.5 35100 425 100.0 1.0

(NH4)2SO

4

(50%-70% fraction) 39.0 18.1 25935 1433 73.9 3.4

Ultrafiltrate 8.8 10.4 16016 1540 45.6 3.6

DEAE-cellulose(117-130 fractions) 61.8 1.5 12484 8270 35.6 19.5

amount of enzyme isolated was notlarge enough for further investigation.Immunological characterization wouldprobably determine the type of relationbetween the different amylase formsobserved and thus confirm thehypothesis.

Effects of pH and temperatureon amylase activity. The effect of pHon enzyme activity was studied in a3.0 to 7.5 pH range with 0.1 mol/Lcitrate-phosphate buffer at 55 °C. Theenzymatic activity profile according totemperature was determined within a10 to 80 °C temperature range understandard conditions (see above). Theoptimal pH was 5.5 and the optimaltemperature was 65 °C (Figures 3and 4).

Compared with the characteristicsof the lactic acid bacterial amylasesdescribed in the literature, theproperties of the enzyme synthesizedby L. plantarum A6 are different. Theenzyme from a Leuconostoc spp.studied by Lindgren and Refai (1984)had a pH optimum of 6.0 and atemperature optimum of 40 °C. Twoactive enzyme fractions were clearlyseparable by isoelectric focusing. Theenzyme isolated from L. cellobiosus(Sen and Chakrabarty, 1986) had amolecular weight of 22.5 to 24 kDa, apH optimum from 6.3 to 7.9, and atemperature optimum of 40 to 50 °C.But the characteristics of the

pH

Rel

ati

ve a

ctiv

ity

(%)

Figure 3. Effects of pH on amylase activity at55 °C.

amylase from L. plantarum A6 are verysimilar to those of Bacillus subtilis(Fischer and Stein, 1960; Fogarty,1983; Robyt and Ackerman, 1973;and Welker and Campbell, 1967):extracellular enzyme, identicaloptimal pH (5.5), identical optimaltemperature (65 °C), presence oftyrosyl phenolic groups at the activesite, and presence of multiple forms(aggregates).

We speculated that theexceptional capacity of L. plantarumA6 to break down starch might have

100

80

60

40

20

03 4 5 6 7 8

217


Figure 4. Effects of temperature on amylase activity at pH = 5.5. (A) Relative activity versus temperature;(B) Arrhenius plot.

been a result of transfer of geneticmaterial between Bacillus subtilis andL. plantarum, which could be possible,because both are microorganismsfound in the natural microflora offermented cassava (Nwanko et al.,1989), and whose amylase activitiesare very similar. Further investigationwould answer this question.

Inoculation effect of Lactobacillusplantarum A6 on cassavafermentation

Three different assays were carriedout: (1) natural cassava fermentation,(2) cassava inoculated with L.plantarum A6, and (3) cassavainoculated with a control strain, L.plantarum Lacto Labo.

Evolution of pH, organic acids,and lactic acid bacteria. In allthree assays, a rapid pH decrease wasobserved from the start offermentation (Figure 5). The naturally

0 20 40 60 80 100 120

Temperature (°C)

100

90

80

70

60

50

40

30

20

10

0

Rel

ati

ve a

ctiv

ity

(%)

A

5

4

3

2

B

1,000/T (K-1)

Inact

ivit

y (%

)

2.8 3.0 3.2 3.4 3.6

fermented cassava showed a steep fallfrom 6.2 to 4.3 (assay 1), and bothinoculation assays (2 and 3) from6.2 to 3.9. This pH shift wascorrelated with lactic acid production,which was the principal metaboliteproduced (Figure 6). These dataconfirm that the lactic acid bacteriaare the predominant fermentativemicroflora. In all three assays, thisflora reached 5.109 cfu/g after 24 h offermentation (Figure 5).

In the natural cassavafermentation, within the first 24 h, asimultaneous production of lactic andacetic acids and traces of propionicand butyric acids and ethanol wereobserved. But, although the acetatecontent reached its maximum level(10 g/kg DM) and remained constantafter the first day of fermentation, thelactate concentration beganincreasing from the second day of theprocess. This suggests thatfermentation is primarily related to

218


an heterolactic flora growth, which issupplanted by a more acid-toleranthomolactic flora.

This hypothesis is supported byOyewole and Odunfa (1990), whostudied the characteristics anddistribution of lactic acid microfloraduring the preparation of fufu. Theyreported a predominant developmentof Leuconostoc mesenteroides, whichwas subsequently replaced byL. plantarum. They suggested thatthis sequence resulted becauseL. mesenteroides was unable totolerate increasing acidity.

In the inoculated fermentations,the lactic acid content was higher.The production kinetics of this acidwere identical in both L. plantarumstrains during the first 24 h. But, onthe second day, this concentrationreached its maximum (40 g/kg DM)and remained constant in the controlstrain. In contrast, in the amylolyticstrain (L. plantarum A6), lactateproduction continued to rise,increasing by 25%.

Traces of ethanol, propionate, andbutyrate were also found in theinoculated fermentation assays.Furthermore, the lower acetateproduction showed that a massiveinoculation with an L. plantarumstrain inhibited the development of thenatural heterolactic microflora.

Conclusions

The presence of amylase in lactic acidbacteria has already been reported.But, as far as we know, no author hasdescribed any amylolytic strain of L.plantarum. When investigating thebacterial microflora of fermentedcassava roots, Regez et al. (1988)isolated numerous L. plantarumstrains, but did not report anyamylolytic strains. Scheirlinck et al.(1989) studied the integration of the

pH

10.0

9.5

9.0

8.5

8.0

7.5

7.00 20 40 60 80 100

7

6

5

4

3

Log

10 lact

ic a

cid b

act

eria

/g

Con

cen

trati

on (g/

100 g

of dry

matt

er)

0 20 40 60 80 100

6

5

4

3

2

1

0

Time (hours)

Figure 6. Evolution of lactate and acetateconcentration during cassavafermentation. ( = lactate and

= acetate in natural fermentation;' .'= lactate and = acetate infermentation inoculated withLactobacillus plantarum A6; = lactateand = acetate in fermentationinoculated with L. plantarum Lacto Labo.)

Time (hours)

Figure 5. Changes in pH and numbers oflactic acid bacteria (l.a.b.) duringcassava fermentation. ( = pH and

= l.a.b. in natural fermentation; = pH and = l.a.b. in fermentation

inoculated with Lactobacillusplantarum A6; = pH and = l.a.b.in fermentation inoculated with L.plantarum Lacto Labo.)

219


α-amylase gene of Bacillusstearothermophilus in the genome of anL. plantarum strain, but did not verifythe expression, stability, andcompetitiveness of the transformedstrain in a natural medium.

In our research, we had isolateda natural amylolytic strain of L.plantarum from cassava roots. Ourdata, as reported here, suggest thatthis new lactic acid bacterium is ofparticular interest, not only for itstaxonomy, but also for its capacity todevelop rapidly and massively instarch-based media.

Finally, preliminary trials ofinoculating cassava with L. plantarumA6 for gari production demonstratethat this strain may play a significantrole in developing organolepticqualities, and in standardizing andpreserving the final product because ofthe large amounts of lactic acidproduced and the resulting faster andsignificant drop in pH values.

References

Bradford, M. M. 1976. A rapid and sensitivemethod for the quantitation ofmicrogram quantities of protein,utilizing the principle of protein dyebinding. Anal. Biochem. 72:248-254.

Champ, M.; Szylit, O.; Raibaud, P.; andAit-Abdelkader, N. 1983. Amylaseproduction by three Lactobacillusstrains isolated from chicken crop.J. Appl. Bacteriol. 55:487-493.

Cotta, M. A. 1988. Amylolytic activity ofselected species of ruminalbacteria. Appl. Environ. Microbiol.54:772-776.

de Man, J. C.; Rogosa, M.; and Sharpe, M. E.1960. A medium for the cultivation oflactobacilli. J. Appl. Bacteriol.23:130-135.

Dubois, M.; Gilles, K. A.; Hamilton, J. K.;Rebers, P. A.; and Smith, F. 1956.Colorimetric method for determinationof sugars and related substances. Anal.Chem. 28:350-356.

Fischer, E. H. and Stein, E. A. 1960.α-Amylases. In: Boyer, P. D.; Lardy,H.; and Myrbäck, K. (eds.). Theenzymes, vol. 4. Academic Press, NY.p. 313-343.

Fogarty, W. M. 1983. Microbial enzymes andbiotechnology. Applied SciencePublishers, Barking, Essex, UK.

Ivorec-Szylit, O. and Szylit, M. 1965.Contribution à l’étude de ladégradation des glucides dans lejabot du coq: mise en évidence etdosage des stéréo-isomères d et llactates. Ann. Biol. Anim. Biochim.Biophys. 5:353-360.

Laemmli, U. K. 1970. Cleavage of structuralproteins during the assembly ofthe head of bacteriophage T4. Nature(Lond.) 227:680-685.

Lindgren, S. and Refai, O. 1984. Amylolyticlactic acid bacteria in fish silage.J. Appl. Bacteriol. 57:221-228.

Nakaruma, L. K. 1981. Lactobacillusamylovorus, a new starch-hydrolyzingspecies from cattle waste-cornfermentations. Int. J. Syst. Bacteriol.31:56-63.

__________ and Crowell, C. D. 1979.Lactobacillus amylophilus, a newstarch-hydrolyzing species fromswine waste-corn fermentation. Dev.Ind. Microbiol. 20:531-540.

Nwanko, D.; Anadu, E.; and Usoro, R. 1989.Cassava-fermenting organisms.MIRCEN J. Appl. Microbiol.Biotechnol. 5:169-179.

Oakley, B. R.; Kirsh, D. R.; and Morris, N. R.1980. A simplified ultrasensitivesilver stain for detecting proteins inpolyacrylamide gels. Anal. Biochem.105:361-363.

Oyewole, O. B. and Odunfa, S. A. 1990.Characterization and distribution oflactic acid bacteria in cassavafermentation during fufu production.J. Appl. Bacteriol. 68:145-152.

Regez, P. F.; Zorzi, N.; Ngoy, K.; andBalimandawa, M. 1988. Evaluationde l’importance de quelques souchesde Lactobacillus sp. pourl’acidification de differents aliments àbase de manioc. Lebensm.21:288-293.

220


Sneath, P. H. A. (ed.). 1986. Bergey’s manualof systematic bacteriology, vol. 2.Williams and Wilkins, Baltimore, MD,USA.

Welker, N. E. and Campbell, L. L. 1967.Comparison of the α-amylase ofBacillus subtilis and Bacillusamyloliquefaciens. J. Bacteriol.94:1131-1135.

Robyt, J. F. and Ackerman, R. J. 1973.Structure and function of amylase.II. Multiple forms of Bacillus subtilisα-amylase. Arch. Biochem. Biophys.155:445-451.

Scheirlinck, T.; Mahillon, J.; Joos, H.; Dhaese,P.; and Michiels, F. 1989. Integrationand expression of α-amylase andendoglucanase genes in theLactobacillus plantarum chromosome.Appl. Environ. Microbiol.55:2130-2137.

Sen, S. and Chakrabarty, S. L. 1986. Amylasefrom Lactobacillus cellobiosus D-39isolated from vegetable wastes:purification and characterization.J. Appl. Bacteriol. 60:419-423.

221

Cassava Wastes:...

CHAPTER 25

CASSAVA WASTES:THEIR CHARACTERIZATION, AND

USES AND TREATMENT IN BRAZIL1

M. P. Cereda* and M. Takahashi**

Introduction

Cassava is widely grown in Brazil. Itis used fresh, that is, directly, incooking; processed into a typicalflour, known as farinha; and forstarch extraction. All the resultingfood products have no or nontoxiclevels of cyanide (Table 1). Mostcyanide is carried away by the wastes,whether liquid or solid.

The crop is grown in diverseproduction systems, ranging fromsmall farms to plantations.Depending on their quantity andcomposition, cassava residues candamage the environment and evenconstitute profit losses. Culinary use,for example, does not producesignificant amounts of residues. Incontrast, industrial use may causeenvironmental problems. Even tinyfactories such as the casas de farinhacan produce significant quantities ofresidues, because of their tendency to

* Faculdade de Ciências Agronômicas (FCA),Universidade Estadual Paulista (UNESP), SãoPaulo, Brazil.

** Instituto Agronômico do Paraná (IAPAR),Paraná State, Brazil.


cluster in certain areas or cities. Forexample, sour or fermented-starchfactories are concentrated by thehundreds in two districts of MinasGerais State: Divinópolis and PousoAlegre. Paranaví, a district of ParanáState, has a concentration of about150 flour factories of different sizes.

Cassava Structure andComposition

The literature on cassava’s structureand chemical composition is variable.Nevertheless, the data overall suggestthat the cassava root is caloric, andgenerates about 1,500 cal/kg fromabout 350 g/kg of carbohydrates.The average values of othercomponents are about 50 g/kg.Phosphorus and calcium contents arehigher. Iron may occur, but in lowquantities. Hegarty and Wadsworth(1968) state that raw cassava usuallyhas an iron content of 1 to2 mg/100 g of dry matter, but warnthat if the analytical equipment usedis made of iron, then such contentmay reach as high as 3.2 mg/100 g.

Table 2 shows the differences incomposition when cassava leaves areconsidered. Oke (1968) detailscassava root composition, mainly asmineral contents, as follows:

222


Humidity: 71.50%

Dry matter (g/100 g):crude protein = 2.60crude fiber = 0.43ash = 2.40lipids = 0.46carbohydrates = 94.10

Ash minerals (g/kg):nitrogen = 0.84potassium = 1.38phosphorus = 0.15calcium = 0.13

Other minerals (mg/kg):sodium = 56.00iron = 18.00boron = 3.30molybdenum = 0.90magnesium = 12.00copper = 8.40zinc = 24.00aluminum = 19.00

Other components:oxalate = 0.32%phytic acid = 76.00%HCN = 38.00 mg/100 g

The potassium content is greaterthan that of calcium, phosphorus,and iron. The idea that cultivatingcassava weakens the soil is probablybased on this fact.

Group B vitamins occur incassava varieties with yellow pulp.These varieties are normally used byfactories only in the northern states ofBrazil. Cassava roots have highvitamin C content, but it can bedestroyed in factory processing orcooking.

The literature differs on thenitrogen fraction of cassava roots. Thetraditional methodology evaluatesproteins by multiplying crude nitrogenby a factor. Cereal or other, vegetable,factors are calculated in this way. Oke(1968) does not consider this accurateenough because a factor for cassavaamino acids has yet to be established.Despite being low, cassava proteinsare overestimated because rootnitrogen fractions include both aproteinic fraction and nonproteiniccompounds. The nitrogen of thelinamarin radical (CN), for example,could be wrongly considered as part ofthe protein evaluation of raw cassavaor cassava fractions.

Sreeramamurthy (1945) reportsthat the traditional solvents of proteinmethodology fail to extract somenitrogen, part of which is of a proteinicnature. For example, copperhydroxide separates and precipitatesonly 10% of total protein. The cassavaproteinic fraction contains arginine,tryptophan, and cystine, andimportant amino acids. Cassava rootprotein is small in quantity ratherthan low in quality when comparedwith casein, egg albumin, and theprotein fractions of cabbage andsweetpotatoes. In contrast, Rogers(1965) suggests that cassava protein islow in histidine, proline, glycine, andamino acids containing sulfur (e.g.,

Table 1. Composition (in percentage) of some typical cassava products, Brazil. Numbers are rounded.

Component Product

Farinha Starch Sour Chipsflour starch

Humidity 1.2 1.1 1.6 0.9Dry matter:

Carbohydrates 93.0 97.3 95.6 94.0Proteins 1.3 0.6 1.5 1.7Lipids 0.1 0.3 0.3 0.3Crude fiber 3.3 0.6 0.7 1.1Ash 1.1 0.1 0.3 0.4Cyanide 0.0 0.0 0.0 1.6

SOURCE: Faculdade de Ciências Agronômicas (FCA), Universidade Estadual Paulista (UNESP), unpublished data.

223

Cassava Wastes:...

Table 2. Central American cassava cultivars: root and leaf composition. Numbers are rounded.

Component Root size Leaves

Long, thin Medium Short, thick

Humidity (%) 62.10 61.10 62.10 77.20Dry matter (g/100 g):

Fiber 1.60 1.25 1.14 2.54Lipids 0.65 0.20 0.24 1.31Nitrogen 0.32 0.17 0.11 1.10Carbohydrates 32.95 34.18 34.70 10.33Ash 1.20 1.20 0.86 1.77

Other components (mg/kg):Calcium 46.00 27.00 27.00 206.00Phosphorus 78.00 66.00 43.00 95.00Iron 1.60 0.50 0.50 3.50Carotene 0.01 0.01 0.01 4.53Thiamine 0.09 0.06 0.05 0.15Riboflavin 0.04 0.04 0.30 0.30Niacin 0.82 0.72 0.60 2.02Ascorbic acid 32.00 40.75 41.40 211.00

SOURCE: Calculated from Martelli, 1951.

methionine, cystine, threonine,isoleucine, and tryptophan).

Cassava juice is milky, smells ofcyanide, and consists of 91.00%water, 0.13% essential oils containingsulfur, 2.30% gum, 1.14% saponins,1.66% glycosides, and 3.80%nonspecified components.

Oke (1968) reported cassava lipidsfrom 0.1% to 1.0%, made up of 35%palmitic, 3% stearic, 39% oleic, 18%linoleic, and 5% linolenic acids.

The literature rarely mentionscassava fiber. Despite cassava rootsbeing fibrous, the processing methodthat uses acid and alkalinehydrolysis yields only about 2.0%fiber, whereas other methods (such asneutral detergent analysis orenzymatic analysis) yield almost 20%.

Carbohydrate is the highestfraction of cassava root composition,with starch constituting the largestpart. Oke (1968) puts the nonstarchyfraction of the carbohydrates at 3.5%,

of which 1.79% is composed ofreducing sugars (0.93% glucose,0.43% fructose, and 0.43% maltose)and 1.71% nonreducing sugars(1.70% saccharose and 0.01%raffinose). Oke considers starchcontent as being 35% of the freshmatter, and possibly higher if totalcarbohydrates are calculated bydifference. Amylase hydrolyzescassava starch to 48% in its granularor raw form and to 78% whenpreviously boiled.

Sugar originating from starch mayincrease if fermentation takes place.According to Amido (Um novocaminho..., 1973), fermentationduring starch extraction andpurification causes loss of starchbecause it turns into soluble sugars.

Sreeramamurthy (1945) concludedthat cassava roots are mainly starchy.They contain less than 1% protein,have a very low lipid content, and arepoor in minerals and group Bvitamins, although fresh roots haveconsiderable vitamin C content.

224


oxidative phosphorylation pathway,combining with cytochrome-oxidaseto inhibit electronic transportationand thus the formation of adenosinetriphosphate (ATP).

For animals, calculating thequantity sufficient to cause death(lethal dose) is done by experimentand expressed in mg per kilo of liveweight. Oke (1969) mentions that1 mg/kg of live weight is consideredthe limit for humans, and is used toclassify cassava roots into poisonousor nonpoisonous, according to theamount of cyanogenic potential in theroot. The literature mentions valuesranging from 15 to 400 mg ofhydrocyanic acid per kg of freshcassava roots, although averagevalues are 30 to 150 ppm (Carvalhoand Carvalho, 1979).

Oke (1969) suggests that, inprocessed foods made from cassava,the hydrolytic enzyme of the plantlinamarase remains active andcatalyzes a reaction that releasesmolecules of glucose, acetone, andhydrocyanic acid in proportions of1:1:1. Linamarase has an optimal pHof 5.5 to 6.0. Glucose can act as anantidote because it changes thereaction’s direction and cooperateswith glucoside synthesis.

Microorganisms consume freeglucose in preference to glucoside.Coop and Blakey (1948), cited by Oke(1969), confirmed this hypothesis.When an extract of cassava in asolution containing 1 to 3 ppm of HClwith a pH of 6.5, was placed in thepresence of 2% glucose, the releasedcyanide content did not change. Norwas the extract toxic when incubatedwith rumen liquid. Determining thepH is important, because reactionrate depends on it. Animals, ingeneral, have a detoxificationmechanism that can prevent deathwhen reaction is slow. It operates inswine, whose stomachs are

Toxic Cassava Glycosides

Cooke (1979) describes bothlotaustralin and linamarin, the toxicglycosides found in the cassava plant(Table 3), as being able to generatehydrocyanic acid. Although freecyanide is well known to be toxic, thetoxicity of glycocyanide is stillunknown.

Oke (1969) reported thatlinamarin is a β-glucoside of acetonecyanohydrin, and lotaustralin ofethyl-methyl-ketone-cyanohydrin.The more representative glucoside islinamarin, which constitutes 80% oftotal glucosides. He also suggestedthat glucosides in linked form are nottoxic to the plants themselves.

Oke hypothesized that theglucosides are intermediatecompounds in protein synthesis, suchas from amino acids that areconstituted from the nitrate absorbedby roots. Thus, the cyanogenicglucosides are stable intermediatesthat do not accumulate if conditionsfor protein synthesis are favorable.Glucoside synthesis probably startswith glycine.

The toxic action of cyanide(released when cell walls aredamaged) on animals is explained bythe cyanide’s affinity to iron,combining with hemoglobin to formcyanohemoglobin. In higher plantsand microorganisms (Cereda et al.,1981), cyanide interferes with the

Table 3. Cyanogen glycoside concentration(mg/kg of tissue) in cassava tissues ofsweet and bitter cultivars (Manihotesculenta Crantz).

Cultivars Seeds 10-day-old Mature Rootsplantlets leaves

Sweet 0 285.0 468.0 125.0Bitter 7.5 245.0 310.0 185.0

SOURCE: Nartey, 1981.

225

Cassava Wastes:...

monogastric, with a pH of 3.0, butdoes not effectively prevent death inbovines, which have polygastricstomachs, with a pH of 7.0.

Microorganisms can develop onsubstrates that contain cyanide ifthey have an anaerobic metabolism—an alternative mechanism to therespiratory chain (Cereda et al.,1981)—or if they can detoxify cyanideby splitting the radical into carbonand nitrogen (Jensen andAbdel-Ghaffar, 1969). This fact mayexplain the related fertilizing effect ofwaste-water spillage from cassavaprocessing.

Cassava Wastes

Cassava wastes are plant residuesgenerated by processing. Wastequality and quantity vary greatlybecause of such factors as plant age,time after harvesting, kind ofindustrial equipment, and itsadjustment.

In Brazil, cassava roots are mostlyprocessed into flour (which generatesmore solid residues) and starch (moreliquid residues). Some solid wastesare brown peel, inner peel, unusableroots, crude bran, bran, bagasse, andflour refuse. Among the liquid wastesis manipueira, which is formed during

Figure 2. Free and total cyanogens (ppm of HCN) in products of a cassava flour factory (EquipamentoZaccharias), using cassava cultivar IAC 12 829 at 24 months old, São Paulo, Brazil.

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

12345671234567123456712345671234567123456712345671234567

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

250

200

150

100

50

0

Tot

al cy

an

ide

123456123456123456123456123456123456123456123456

1234567123456712345671234567123456712345671234567123456712345671234567

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

200

160

120

80

40

0

Fre

e cy

an

ide

Waste Ground Rootswater mass

Waste Ground Rootswater mass

flour making by pressing bulkquantities of cassava roots. It is alsoformed during starch extraction, butthe water used in the process dilutesthe manipueira, reducing its organicload and cyanide content, but vastlyincreasing its output. Water fromwashing roots is also considered asliquid waste. Figures 1 to 5 show therelationships between cassava

Figure 1. Cyanide content (ppm of HCN) of plantparts, products, and wastes ofprocessing cassava cultivarIAC 12 829.

891

Water waste

Sieved mass

Pressed mass

Ground mass

Roots

Chips

Leaves

193

65

75

127

15251

226


Ground mass714 kg

58% m.c.

Waste water(manipueira)

289 kg, 90% m.c.Total cyanide = 120 ppm

Crueira(type of solid waste)

40 kg, 48% m.c.

Pressed mass425 kg

36% m.c.

Oven-drying

Figure 3. Mass balance of a fermented-starch factory, Colombia. (m.c. = moisture content.)(After Arguedas and Cooke, 1982.)

Flour(farinha)253 kg

0.74% m.c.

Roots1,000 kg

52% m.c.

Peelings68 kg

68% m.c.

Grinder + waterSieve + water

Waste water(diluted manipueira)

10,620 L95% m.c.

Total cyanide = 60 ppm5,000 biological oxygen

demand

Wet starch149 kg

52% m.c.

Figure 4. Mass balance of a fermented-starch factory, which used cassava variety Branca de SantaCatarina at 24 months old, Minas Gerais State, Brazil. (m.c. = moisture content.)

Evaporatedwater

133 kg

Discards3 kg

55% m.c.

Bran1,120 kg85% m.c.

Peelings196 kg

47% m.c.

Discards90 kg

57% m.c.

Roots1,000 kg

52% m.c.

227

Cassava Wastes:...

123123123123123

123412341234123412341234123412341234123412341234123412341234

123123123123123123123

123123123123123123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123123123123

1234123412341234123412341234123412341234123412341234123412341234123412341234

1212121212121212121212

12341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123123123

123123123

123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123123123

123123123123123123123123123123123

123412341234123412341234123412341234

123123123123123123123123123123123

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0Mibo Grounted Foster Hand cut Average

1234123412341234123412341234123412341234123412341234123412341234123412341234

1234123412341234123412341234123412341234

1234123412341234123412341234123412341234123412341234123412341234123412341234

123412341234123412341234

123451234512345123451234512345123451234512345

123412341234123412341234

12341234123412341234123412341234123412341234

12345123451234512345123451234512345123451234512345

1234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123

123412341234123412341234123412341234

1234512345123451234512345123451234512345123451234512345123451234512345123451234512345

5.0

4.0

3.0

2.0

1.0

0

Drying floors

Cya

nid

e (p

pm

)

Covered trays

Mibo Grounted Foster Average

Figure 5. Cyanide reduction in cassava chips processed at three factories and by hand cutting, Brazil.( 123123 = 5 kg/m2

123123123

= 10 kg/m2123412341234

= 15 kg/m2 = Average.)

processing, cultivars, and wastes inthe material balance of cassava flourand sour starch production.

Solid wastes

Peelings. The brown peel,sometimes called bark, of cassavaroots corresponds, in technical terms,to the periderm and varies between2% and 5% of the root total. It is thinand cellulosic, and although usuallydark brown, can be white orcream-colored. A small quantity ofinner peel, or cortical parenchyma,may come off with the bark, causinglosses in starch factories. In farinha

factories, if the inner peel is highlyfibrous, it is best taken off. Inindustrial terms, peelings areresidues and refer to the mixture ofboth inner peel and bark. Table 4shows the average composition ofseveral samples of peelings. Peelingscan be used as fertilizer or animalfeed.

Discards. These are producedduring selection, so as not tooverwork the rasper. Theircomposition is similar to that ofcassava roots but is more fibrousbecause they contain the peduncle.Moisture content is 55%-60%. The

228


Table 4. Chemical composition of cassava peelings. Average values of several samples are given. Dashesindicate that no data were available.

Component Peelings

Outer (bark) Inner Mixture

Humidity (%) 48.3 65.6 72.3Dry matter:

Volatile solids (%) - - 26.2Ash (%) 4.0 3.0 1.4Soluble carbohydrates (%) - - 7.9Starch (%) 0 58.0 32.0Lipids (%) 3.0 2.0 0.6Nitrogen (%) 0.6 1.3 2.1Fiber (%) 41.0 6.0 -Lignin (% SV) - - 6.5Free cyanide (ppm) - - 23.9Total cyanide (ppm) 0 320.0 120.0Phosphorus (ppm) 60.0 - 60.0Potassium (ppm) 430.0 - 430.0Calcium (ppm) 280.0 - 280.0Magnesium (ppm) 80.0 - 80.0Iron (ppm) 5,538.0 - 26.0Copper (ppm) 9.0 - 9.0Zinc (ppm) 21.0 - 21.0Manganese (ppm) 104.0 - 103.0Sulfur (ppm) 110.0 - 320.0Boron (ppm) 18.0 - 18.0

Volatile acidity (mg acetic acid/L) - - 5,548.0Alkalinity (mg bicarbonate/L) - - 2,191.0C/N ratio - - 6.4C/P ratio - - 0.3

SOURCE: Motta, 1985.

quality of discards depends on thecultivar and on root age. Togetherwith bran, discards may be used rawas animal feed, thus bringing extraincome for the industry. The valuesshown in Figures 3 and 4 may beoverestimated, because the process isstill being investigated.

Bran or bagasse. This solidwaste is made up of fibrous rootmaterial, and contains starch thatphysically could not be extracted. Itis produced as starch is separated. Ithas a large absorption capacity andmay contain about 75% moisture.Table 5 shows the chemicalcomposition of bran after partialdrying, with differences according tothe technology used. Table 6 shows

the composition of sun-dried branfrom fermented-starch factories inMinas Gerais, Brazil.

Crude bran. Another type ofsolid waste is crude bran (farinhão orcrueira), which is made up of pieces ofroot and inner peel. In cassava-flourprocessing (at EquipamentoZaccharias, São Paulo State), theseare separated out by sieving beforebeing oven-dried. Table 7 (p. 231)shows the composition of such waste.At other factories (e.g., Mádia, ParanáState), these residues are replaced byfine threads (fiapos) made up ofcassava fibers. Another solid residueis cassava-flour refuse, the gratedmass that daily falls and collects onthe floor.

229

Cassava Wastes:...

Table 5. Differences in chemical composition of bran according to technology adopted. Dashes indicatethat no data were available.

Componenta Type of technology

Royalb Minasc Fiaposb

Humidity (%) 9.42 14.82 9.52

Dry matter:

Ash (%) 0.83 3.77 0.66

Soluble carbohydrates (%) 0.01 - -

Starch (%) 69.76 74.99 63.85

Lipids (%) 0.65 0.28 0.83

Nitrogen (%) 0.24 1.86 0.32

Fiber (%) 11.08 7.81 14.88

Total cyanide (ppm) 0 0 -

Phosphorus (ppm) - 30.00 -

Potassium (ppm) - 280.00 -

Calcium (ppm) - 90.00 -

pH 4.00 - -

a. No data were available for the following components: volatile solids, lignin, free cyanide, magnesium, iron,copper, zinc, manganese, sulfur, boron, volatile acidity, alkalinity, C/N and C/P ratios, chemical oxygendemand, or titratable acid.

b. Large factory.c. Small, traditional factory.

Liquid wastes

Lagoon mud. Table 7 also showsthe composition of sedimented lagoonmud and liquid wastes. Sometimes

these are sun-dried and used asfertilizer. The use of thin starch isalso uncommon.

Manipueira. Dilutedmanipueira is a liquid waste fromcassava starch extraction andsour-starch manufacture. It may bewaste water from root washing,after the washer/husker hasremoved soil and peelings and thewater is decanted or filtered. Theaverage factory volume is 2.62 m3.Waste water may also be extractedfrom pressed and grated cake inflour manufacturing and from theroots themselves. It is also abyproduct of starch extraction(average factory volume is 3.68 m3).

The average composition ofmanipueira sampled from differentstarch factories in São Paulo State isvariable, as shown in the followinglist (numbers are rounded):

Table 6. Average composition of sun-dried branfrom 20 fermented-starch factories(traditional) from Pouso Alegre andDivinópolis, Minais Gerais, Brazil.

Component Average values (%)a

Pouso Alegre Divinópolis

Starch 63.6 2.78Soluble carbohydratesb 0.2 0.10Protein 2.3 0.34Phosphorus 0 0.01Calcium 0.1 0.03Potassium 0.3 0.06Lipids 0.6 0.35Fiber 8.3 2.06

a. Numbers are rounded.b. Expressed in percentage of glucose.

SOURCE: Escola Superior de Agricultura de Lavras(ESAL), unpublished data.

230


existing in the disintegrated rootmass.

The water used in starch factoriescarries high concentrations of theseglycosides (linamarin andlotaustralin) (Sobrinho, 1975). Theyare hydrolyzed by linamarase enzymeand acid, making the cyanide a freeradicle (CN) (Williams, 1979).

According to Sobrinho (1975),liquid waste thrown onto soil or intowaterways causes pollution. If thepollution rate of starch factories isexpressed as biological oxygendemand (BOD) over 5 days, at 20 °C,and calculated as 24 g per habitantper day, it would be equivalent to thatcaused by 150-250 habitants perday—very high. In Santa CatarinaState, the pollution caused by thesewastes corresponds to 460 habitantsper day (Anrain, 1983).

Conclusions

Cassava wastes can be used indifferent ways. The solid residues canbe used as animal feed; the literatureshows that cassava waste can replacea part or all of the feed components.Manipueira can be used in agricultureas a herbicide, nematicide,insecticide, or fertilizer. Anaerobicdigestion is well studied in Brazil andis more advantageous than aerobicdigestion. Manipueira comes fromflour industries, and the bestprocessing method uses the separatedphases reactor. We now need tostudy how to optimize the acidicphase.

Cassava waste can also be usedfor biomass production. The yeastTrichosporon sp. can be isolated bynatural fermentation with acyanide-resistant respirationpathway, and potentially can produceboth a proteinic and a fat biomass.

Component Value

Humidity (%) 93.7Dry matter (g %):

Total solids 6.3Volatile solids 5.2Starch 0Soluble carbohydrates 0.5Lipids 0.5Ash 1.1Crude nitrogen 0.5Fiber 0.3Lignin 6.0Free cyanide 43.7Total cyanide 444.0

Dry matter (ppm):Phosphorus 160.8Potassium 1,863.5Calcium 227.5Magnesium 405.0Iron 15.3Copper 1.1Zinc 4.2Manganese 3.7Sulfur 19.5Boron 5.0

Chemical oxygen demand 6,365.5

Volatile acidity(mg acetic acid/L) 2,703.7

Alkalinity(mg bicarbonate/L) 1,628.0

C/N ratio 7.6

C/P ratio 34.4

Titratable acidity(ml NaOH N%) 3.3

pH 4.1

Cyanogen content tends to behigh, but varies according to cultivar.The organic load is also high, andvaries with the type of processingused (Table 8). All residual starch isremoved from manipueira beforetreatment. It has most soluble andsome insoluble substances insuspension and this residue carriesalmost all the cyanogenic glycosides

231

Cassava Wastes:...

The microorganisms can also beused to produce such biomasses asorganic acids (citric or lactic),biological insecticides, andenzymes.

Solid wastes also have apotential use for foodstuffs. Theproduction of high-fiber biscuitsfrom bagasse is being studied at theFaculdade de Ciências Agronômicas(FCA) of the Universidade EstadualPaulista (UNESP).

Table 7. Chemical composition of different types of cassava wastes, averaged over several analyses.Numbers are rounded. Dashes indicate no data were available.

Componenta Type of waste

Farinhãob Varredurac Lagoon mud

Humidity (%) 11.7 - 4.9Dry matter (g %):

Soluble carbohydrates 1.1 - 61.4Lipids 68.5 - 1.8Nitrogen 1.7 - 0.1Fiber 0.5 0.5 1.8Lignin (% SV) - - 9.7

Dry matter (ppm):Free cyanide - - 0Total cyanide - - 0Phosphorus 70.0 70.0 540.0Potassium 700.0 640.0 240.0Calcium 130.0 90.0 140.0Magnesium 60.0 50.0 60.0Iron 41.0 32.0 23,800.0Copper 2.0 3.0 63.0Zinc 8.0 8.0 75.0Manganese 20.0 18.0 105.0Sulfur 30.0 30.0 46.0Boron 20.0 7.0 14.0

pH 5.4 - 5.4Titratable acidity (ml NaOH N %) 3.7 - 3.9

a. No data were available for volatile solids, ash, and starch.b. Farinhão = solid waste made of pieces of cassava roots and inner peels.c. Editor’s note: No explanation of this term was provided by the authors.

Table 8. Composition of extraction water (mg/L)from the Fleischmann-Royal factory,Conchal, São Paulo State, Brazil.

Component Range measured

Chemical oxygen demand 6,280 -51,200Biological oxygen demand 1,400 -34,300Total solids 5,800 -56,460Soluble solids 4,900 -20,460Suspended solids 950 -16,000Fixed solids 1,800 -20,460Organic matter 1,500 -30,000Reducing sugars 2,800 -8,200Total phosphate 155 -598Total nitrogen 140 -1,150Ash 350 -800Sedimentable solids (1 h) 11 -33Cyanide content 22.0 -27.1pH 3.8 -5.2

SOURCE: Lamo and Menezes, 1979.

232


References

Anrain, E. 1983. Tratamento de efluentes defecularia em reator anaeróbico defluxo ascendente e manta de lodo. In:Anais do XII Congreso Brasileiro deEngenharia Sanitaria Ambiental,Balneário de Camburiu. Fundação deAmparo à Tecnologia e ao MeioAmbiente, Balneário de Camburiu,SP, Brazil. p. 1-21.

Arguedas, P. and Cooke, R. D. 1982.Concentraciones de cianuro residualdurante la extracción de almidón deyuca. Yuca Bol. Inf. (Cent. Int. Agric.Trop.) 10:7-9.

Carvalho, V. D. and Carvalho, J. G. 1979.Princípios tóxicos de mandioca.Inf. Agropecu. 5:82-88.

Cereda, M. P.; Brasil, O. G.; and Fioretto,A. M. C. 1981. Actividade respiratóriaem microorganismos isolados delíquido residual de fecularias. Paperpresented at the 11º CongressoBrasileiro de Microbiología,Florianópolis, Bazil.

Cooke, R. D. 1979. Enzymatic assay fordetermining the cyanide content ofcassava and cassava products.Cassava Information Center, CIAT,Cali, Colombia. 14 p.

Hegarty, J. V. and Wadsworth, G. R. 1968.The amount of iron in processedcassava (Manihot esculenta). J. Trop.Med. Hyg. 71:51-52.

Jensen, H. L. and Abdel-Ghaffar, A. S. 1969.Cyanuric acid as nitrogen sources formicroorganisms. Arch. Mikrobiol.67:1-5.

Lamo, P. R. and Menezes, T. J. B. 1979.Bioconversão das águas residuais doprocessamento de mandioca paraprodução de biomassa. Col. Ital.10:1-14.

Martelli, H. L. 1951. Mandioca, planta devalor. A. Fazenda, NY. 46:40.

Motta, L. C. 1985. Utilização de residuos deindustrias de farinha de mandiocaem digestão anaerobia. Thesis forMaster of Agriculture. “Julio deMesquita Filho,” UniversidadeEstadual Paulista, Botucatu, SP,Brazil. 130 p.

Nartey, F. 1981. Cyanogenesis in tropicalfeeds and feedstuffs. In:Vennesland, B.; Conn, E. E.;Knowles, C. J.; Westley, J.; andWissing, F. (eds.). Cyanide in biology.Academic Press, London, UK.p. 115-132.

Oke, O. L. 1968. Cassava as food in Nigeria.World Rev. Nutr. Diet. 96:227-250.

__________. 1969. The role of hydrocyanicacid in nutrition. World Rev. Nutr.Diet. 11:170-98.

Rogers, D. J. 1965. Some botanical andethnological considerations ofManihot esculenta. Econ. Bot.19(4):369-377.

Sobrinho, P. A. 1975. Auto-depuração doscorpos d’agua. In: CursoPoluição das Aguas, São Paulo.Companhia de Tecnologia deSaneamento Ambiental (CETESB),Associação Brasileira de EngenhariaSanitária (ABES), and BancoNacional de Habitação (BNH), SãoPaulo, SP, Brazil.

Sreeramamurthy, V. V. 1945. Investigationson the nutritive value of tapioca(Manihot utilissima). Indian J. Med.Res. 33:229-238.

Um novo caminho para a mandioca: Químicay derivados. 1973. Amido (São Paulo)32:26-28.

Williams, H. J. 1979. Estimation of hydrogencyanide released from cassava byorganic solvents. Exp. Agric.15(4):393-400.

233

Cassava Starch Extraction:...

CHAPTER 26

CASSAVA STARCH EXTRACTION:A TYPICAL RURAL AGROINDUSTRY WITH AHIGH CONTAMINATION POTENTIAL

Olga Rojas Ch.*, Patricia Torres L.*, Didier Alazard**, Jean-Luc Farinet***, and María del Carmen Z. de Cardoso*

* Facultad de Ingeniería, Universidad del Valle(UNIVALLE), Cali, Colombia.

** Institut français de recherche scientifiquepour le développement en coopération(ORSTOM), stationed in Cali, Colombia.

*** Département des cultures annuelles (CA),CIRAD, Montpellier, France.

effluents. Its operating principle isbased on immobilizing microorganismson a lignocellulose support. Thehydrodynamic characteristics of threetypes of supports—sugarcane bagasse,bamboo, and paja de monte(underbrush straw)—were determinedin the laboratory. Paja de monteshowed the best characteristics.

These studies will becomplemented with the monitoring ofexisting microflora as changes occur inoperational parameters.

Introduction

Agroindustrial processes generatelarge volumes of waste waters andsolid residues whose quality variesaccording to the process used.Generally, farm activities useabundant water to wash and treatproducts, at which point the water isloaded with harmful elements andcompounds. These are directlydischarged into rivers and streams,representing a risk for theenvironment, and the reduced qualitymakes the water less suitable for otheruses.

Colombia is a mainly agriculturalcountry, with small and medium-sizedagroindustries widely scattered. Thismakes conventional water treatment

Abstract

Every year, about 5,500 t of starch areproduced in Colombia from about27,000 t of cassava roots. Starchproduction usually involves simpletechnology, consuming an average of23 m3 of water per ton of cassava.This generates a contaminating loadof about 180 kg of chemical oxygendemand (COD) per ton of roots. Anaverage of 13.5 t of COD is dischargedinto Colombian rivers each day.

Processing generates two liquidresidues: the first results from thewashing and peeling of cassava roots,and generally contains a large amountof inert material with low COD; thesecond results from draining thestarch sedimentation tank, and has ahigh contaminating load of COD andbiochemical oxygen demand (BOD).

A pilot project was proposed totreat waste waters, using an anaerobicfilter and a transfilter. The transfiltertechnology has been tested in France,yielding good results with household

234


systems onerous to use, given thesmall volumes of products processed.But versatile water treatment systemsare now available at low cost and areattractive alternatives within the reachof small industries.

The feasibility of applyingtransfilter systems (anaerobicprocessing) to purify these dischargesis being studied by the Universidaddel Valle (UV), Cali, Colombia, incollaboration with the Institut françaisde recherche scientifique pour ledéveloppement en coopération(ORSTOM), France, and is financiallysupported by the European Union(EU). A pilot reactor will be located ina starch factory in the CaucaDepartment. The CorporaciónAutónoma Regional del Valle delCauca (CVC) will build a pilotanaerobic filter at the same site, soboth can be evaluated from technicaland economic viewpoints.

Waste Waters from CassavaStarch Extraction

Production and identification ofwaste waters

About 200 cassava-starch productionfactories are located in the CaucaDepartment, most in the north. Theirannual consumption of cassava rootsis about 27,000 t, from which theyextract about 5,500 t of starch. Plantprocessing capacity ranges from 500to 2,500 kg fresh roots per day(Janssen and De Jong, 1981).

Cassava-starch extraction involvesseveral stages: root washing andpeeling, rasping, screening, starchsedimentation, and, for sour starch,fermentation (Figure 1). Roots arewashed in a tank or drum. They canalso be peeled in a drum, but thisoperation removes only 60% to 70% ofthe peel, the remainder being peeled

Figure 1. Characterization of waste water from cassava starch extraction.

Starch(0.27 m3)

Cassava(1 t)

Rasped cassava(820 kg)

Peeled cassava(820 kg)

Starchy slurry(20 m3)

Dried starch(230 kg)

Sedimentation

Wash water

Water(20 m3)

Waste water(15 m3)

Bran(60 kg)

Waste water

Peelings180 kg

Dried scum(26 kg)

235


by hand. This first stage of washingand peeling generates the first wastewaters.

The roots are rasped, and the pulpsieved through a nylon mesh thatcovers the inside of the screeningdrum. The starchy slurry is left tosettle for 20 to 24 h in sedimentationtanks, until the starch layer is 25 to30 cm thick. The liquid is drained anddiscarded, and the extracted starchpasses to fermentation tanks, whichare completely filled and then coveredwith a thin layer of water.Fermentation takes about 4 weeks(Pinto, 1978).

Use of solid residues

When water is separated from starchin the sedimentation tank, a layer ofgreyish material is formed over thestarch, called mancha by starchmanufacturers. This film, orproteinaceous fraction, can be easilyremoved and dried. It is frequentlyused as animal feed and is widelyaccepted in the market.

Characteristics of waste waters

Waste water results from twoprocessing stages: the washing ofpeeled cassava and the draining of thesedimentation tanks (Figure 1). Theformer contains a large amount of inertmaterial and has a low COD, and thelatter, high organic loads of BOD andCOD.

Analyses of waste water samplestaken from sedimentation tanks atdifferent starch factories were carriedout by the UV and the CVC. Averagevalues were obtained to indicate theapproximate composition of suchwater. According to this information,the volume of waste water dischargedper processing plant per day rangesfrom 18 to 48 m3. The average overallcontaminant load is about 13.5 t ofCOD per day, or 3.45 t of BOD per day.

The use of anaerobic digestion topurify waste waters from cassavastarch extraction

The results of characterizing thedrainage water from sedimentationtanks permitted an analysis of thepossibility of applying anaerobictreatment to this type of residue. Theaverage COD value (900 mg/L) ishigh compared with that of BOD(300 mg/L), suggesting a highCOD-to-BOD ratio and the presence ofa high COD content, resistant tobiological degradation. But the UV andCVC’s preliminary research indicatesthat this factor is less important interms of anaerobic biodegradation.When specifically tested to determinethe percentage of organic matterbiologically degradable underanaerobic conditions, the percentage ofbiodegradability was found to be 83%.

The results of waste water analysesshow that a sufficient amount ofnitrogen, a major element in biomassgrowth, is present in the residues. Butthe amount of phosphorus, anotheressential macronutrient, is deficient(Table 1).

The pH of drainage water from thesedimentation tank ranges from 3.9 to4.7, which means the residue must beneutralized before being fed to thereactor.

The low cyanide concentration inthe waste water (average 2.12 mg/L)suggests that the microbial biomasscan adapt to this inhibitor.Methanogenic bacteria first react byreducing methane (CH4) production,but, within a few days, they adapt tothe cyanide and finally decompose it.

Based on this finding, the UV andthe CVC conducted studies to see ifanaerobic processes are applicable fortreating this type of discharge(Table 2). Additional studies are nowbeing conducted on a pilot scale.

236


Table 1. Characteristics of waste waters resulting from cassava starch extraction (average values).

Parameter Average value Rangea

CODb (mg O2/L) 9,100 4,000 - 12,800

BODc (mg O2/L) 3,100 1,500 - 8,600

COD/BOD (ratio) 2.9 -Cyanides (mg CN-/L) 2.12 1.2 - 4.04Total solids (mg/L) 5,740 2,680 - 10,020Volatile solids (mg/L) 4,870 2,020 - 9,320Total organic carbon 2,420 870 - 5,300pH (units) - 3.9 - 4.7Temperature (°C) - 19 - 22Sedimentation (ml/L) 29 15 - 47Nitrogen (mg.N/L) 105 29 - 233Phosphorus (mg.P/L) 2.34 0.3 - 6.0

a. This range is broad due to the amount of material processed.b. COD = Chemical oxygen demand.c. BOD = Biochemical oxygen demand.

SOURCE: Raddatz, 1986.

Methanogenic activity in mud wasmeasured in the 12-liter reactor,increasing significantly from an initialrate of 0.063 kg COD-CH4/kg VSS(volatile suspended solids) to a rate of0.188 kg COD-CH4/kg VSS. Althoughcyanide concentration was measuredonly a few times, it was calculated asdecreasing by about 69%.

Transfiltering

This is a type of anaerobic treatment,using a lignocellulose base to filter

particulate material present in thewaste water, and to fix thosemicroorganisms responsible forbiodecomposing waste organic matter.The support bed decomposes withtime so it has to be changed regularly;this also prevents silting.

Three operations occursimultaneously within the transfilter:(1) waste waters are purified byfiltration and the organic materialpresent in the waste is digested;(2) biogas (an energy resource) isproduced; and (3) the lignocellulose

Table 2. Results of laboratory and pilot studies on the feasibility of anaerobic treatment of waste waterfrom cassava processing, carried out by the Universidad del Valle and the Corporación AutónomaRegional del Valle del Cauca, Colombia.

Descriptiona 12-liter reactor 23-liter reactorb

n Average ∂n-1 n Average ∂n-1

Effluent flow (ml/min)c 52 16.8 6.40 14 10.8 4.6COD

Af (mg/L)c 52 3,294 2,732 14 1,640 1,263

CODEf (mg/L)c 52 659 56 12 105 26OVL (kg COD/m3.day)c 52 5.02 2.46 12 4.31 3.2Removal COD (%)c 36 95 2 12 85.4 12.1Biogas production (L/day) 53 16.4 8.4 - - -

a. CODAf = chemical oxygen demand in affluent flow; CODEf = chemical oxygen demand in effluent flow;OVL = organic volume load.

b. The effluent was recycled by about 30%.c. These units refer to average values of the COD.

SOURCES: Escandón, 1988; Hernández, 1987.

237


material (a solid waste appropriate forcompost) is digested. Figure 2 is adiagram of a transfilter reactor(Farinet, 1993).

The UV is conducting laboratoryresearch on the transfilter processbased on waste waters from cassavastarch extraction. A pilot reactor willlater be built at a starch factory in theCauca Department.

So far, with sugarcane bagasse,paja de monte, and bamboo as supportbeds, the following hydrodynamiccharacteristics were determined:volume of waste water eliminated fromthe supports, and load loss from cleanwater flowing through the filteringmedium, as affected by water velocityand density of medium (compression).

Results showed very low loadlosses for higher velocities (35 m/h),and for stronger compressions(120 kg/m3 for bagasse and100 kg/m3 for paja de monte).Maximum load loss was 7 cm forbagasse 1 m long and 6.3 cm for paja

de monte. No load loss was observedin bamboo (Gotin, 1993). Technically,any of these materials can therefore beused, if due attention is paid to theoperating criteria.

To complement the research ontypes of support, further studies onthe filtration capacity of paja de montewill be made at a pilot starch factory.The aim is to determine the maximumcompression of the support at whichoptimal filtration efficiency is obtainedfor a given volume and period inrelation to time of silting the filter.The supernatant effluent from thestarch sedimentation tanks will be thewaste water treated.

To define constraints to designingthe pilot reactor, feasibility studies onmethanizing the filter effluent will beconducted in the laboratory, based onprevious results.

Bacterial microflora will also bestudied for their composition,distribution, and nature of thedifferent groups of bacteria involved

Figure 2. A transfilter reactor, used for the anaerobic treatment of waste water from cassava starchextraction. (From Farinet, 1993.)

Purifiedwater

Biogas Biogas

SupportFermented (digested)

support material

Wastewater

Piston

238


and their interactions. Considerableknowledge is already available on themetabolic pathways of themicrobiological process of anaerobicfermentation (hydrolysis,acidogenesis, acetogenesis, andmethanogenesis).

We will quantify and characterizethe active bacterial microfloraresponsible for biodegrading theeffluent at each stage of the process.The evolution, and the density, ofeach group of bacteria will also beassessed throughout the operation ofthe digestor.

References

Escandón, F. 1988. Tratamiento de aguasresiduales del proceso deelaboración de almidón de yuca, enun reactor de flujo ascendente ymanto de lodos. ConvenioUniversidad del Valle-CorporaciónAutónoma Regional del Valle delCauca (CVC), internal publication.Universidad del Valle, Cali,Colombia.

Farinet, J. L. 1993. Traitement des eauxusées par le procédé transfiltre.Rapport d’essais sur prototype.Département des cultures annuelles(CA), CIRAD, Montpellier, France.

Gotin, G. 1993. Caractérisationhydrodynamique de supports naturelsen vue de les employer en biofiltration.Thesis. Ecole nationale supérieure debiologie appliquée à la nutrition etl’alimentation (ENSBANA), Dijon,France. 43 p.

Hernández, L. I. 1987. Tratamiento anaerobiode las aguas residuales del proceso deproducción de almidón de yuca ydesechos del café. ConvenioUniversidad del Valle-CorporaciónAutónoma Regional del Valle delCauca (CVC), internal publication.Universidad del Valle, Cali, Colombia.

Janssen, W. and De Jong, G. 1981. Cassavaand cassava starch: the production,processing, and marketing of cassavaand sour cassava starch in Mondomo,Colombia. CIAT, Cali, Colombia.177 p.

Pinto, R. 1978. Extracción de almidón de yucaen rallanderías. ICA (Inst. Colomb.Agropecu.) Informa 12(9):3-6.

Raddatz, W. 1986. The possibility of anaerobictreatment of wastes andwastewater from small and mediumagroindustries: sisal and cassavastarch production. Convenio deCooperación, Corporación AutónomaRegional del Valle del Cauca (CVC)-Deutsche Gesellschaft für TechnischeZusammenarbeit (GTZ). CVC, Cali,Colombia.

SESSION 5:

TECHNOLOGY DEVELOPMENT

241

Improving Cassava Sour Starch Quality in Colombia

CHAPTER 27

IMPROVING CASSAVA SOUR STARCH

QUALITY IN COLOMBIA1

C. Brabet*, G. Chuzel**, D. Dufour*,M. Raimbault†, and J. Giraud††

Bread-making potential (BMP) isthe main criterion of quality for sourstarch and is defined as the ability ofthe starch to swell during baking(Laurent, 1992).

Although quality and rapidity aretwo major issues in cassava starchproduction, sour starch is stillproduced according to traditionalmethods. Hence, sour starch ishighly variable in product quality,limiting its use in food industries.

Fermentation and sun-dryingcritically influence the BMP of sourstarch (Brabet and Dufour, 1996;Larsonneur, 1993). Developingadequate control of and suitablepractices for these two processingsteps would help stabilize andimprove sour starch’s economic valueand strengthen the status of thisagroindustry.

Cassava processors sometimesimprove sour starch quality byinoculating batches with surfacewater from fermentation tanks inwhich good quality products havebeen produced. But this practice stillresults in irregular quality of sourstarch.

We therefore studied the naturalfermentation of cassava starch indetail, in an attempt to relate thenature of microflora and their effect

Introduction

Fermented or “sour” starch extractedfrom cassava is used in Colombia toprepare traditional, gluten-free,cheese breads such as pandeyucaand pandebono. Starch extractionconsists of peeling, washing, andgrating fresh cassava roots. The pulpis then screened under running waterto obtain starch milk or lechada. Thestarch is then sedimented out andplaced into wooden or tiled tanks(about 1 m3), where it fermentsnaturally over 20 to 30 days underanaerobic conditions and at anaverage temperature of 21 °C. Theresulting sour starch is thensun-dried to obtain a stable productwith 10%-15% moisture, and ismarketed (Brabet and Dufour, 1996;Jory, 1989).

* CIRAD/SAR, stationed at the CassavaUtilization Section, CIAT, Cali, Colombia.

** CIRAD/SAR, stationed at the Faculdade deCiências Agronômicas (FCA), UniversidadeEstadual Paulista (UNESP), São Paulo,Brazil.

† Institut français de recherche scientifiquepour le développement en coopération(ORSTOM), stationed in Cali, Colombia.

†† Laboratoire de microbiologie et biochimieindustrielles (GBSA), Université deMontpellier II, Montpellier, France.


242


on final product quality. We thencarried out a cassava starchinoculation trial, using amylolyticlactic acid bacteria (ALAB), isolatedand selected from previousfermentation kinetic studies. Ourpurpose was to standardize productquality and reduce fermentation time.

We also carried out studies toconfirm the role of sun-drying in theacquisition of BMP, and to determinethe key factors responsible. The trialsinvolved sun-drying kinetic studies,oven-drying at 40 °C and at 55 °C,drying under cover, oven-drying at40 °C under ultraviolet (UV) light, anddrying with water added.

Cassava Starch Fermentation

Natural fermentation

Natural fermentation of cassavastarch is characterized by thepresence of a predominantly lacticacid flora (108-109 cfu/g dry matter ofstarch), confirmed by the rapid anddrastic decrease of pH (7 to 3.5 in5 days), while total acidity increasesbecause of a mainly lactic acidproduction (Brabet and Dufour,1996). Lactic acid flora has an activecatabolism but its level is constantduring fermentation.

At the start of fermentation,starch is the main source offermentable sugar. Gómez (1993)isolated 75 lactic acid bacterialstrains, exhibiting good amylolyticactivity, from natural cassava starchfermentation. This ALAB strain bankis currently being molecularly andbiochemically characterized.

Previous works have shownmodifications of cassava starchphysicochemical and rheologicalcharacteristics during fermentation(Brabet and Dufour, 1996; Brabetand Mestres, 1991; Camargo et al.,

1988; Cárdenas and de Buckle, 1980;Larsonneur, 1993; Nakamura andPark, 1975).

Effect of a starter culture oncassava starch fermentation andquality

A fermentation with starchinoculation was carried out at the“SDT Agroindustrial,” astarch-processing plant in LaAgustina, Cauca Department,Colombia, using cassava varietyCM 523-7. An amylolytic lactic acidbacterial strain, ALAB 20, used forthe starch inoculation trial, wasisolated from a previous naturalcassava starch fermentation andidentified as Lactobacillus crispatus,using the Gómez (1993) API 50CHsystem. Flores (1993) studied thephysiological parameters of this ALAB20 strain during a lactic acidfermentation on an MRS-starchmedium in a bioreactor. (The glucosein the medium was replaced bysoluble starch.)

The fermentation tank waspartitioned into two: one part fornatural fermentation and the otherfor inoculated-starch fermentation.Inoculated and noninoculatedlechadas (aqueous starchsuspensions) were first sampled.Then, samples of inoculated andnoninoculated starch were taken at 1,2, 3, 4, 5, 7, 10, 14, and 20 days offermentation.

Changes in amylolytic andtotal lactic acid flora. Total lacticacid flora on MRS medium showed nosignificant difference betweeninoculated and noninoculated starch(Figures 1 and 2). This flora reached108-109 cfu/g of dry matter of starchon the second day and remainedconstant until the end offermentation. In contrast, as aproportion of total flora, amylolyticflora on MRS-starch medium (MRS

243


120

100

80

60

40

20

0Log

(cf

u/g

dry

matt

er o

f st

arc

h) 10

9

8

7

6

50 5 10 15 20 25

Days of fermentation

Figure 1. Changes in anaerobic microflora onMRS and MRS-starch media innatural fermentation. ( = total lacticacid flora on MRS agar; = total floraon MRS-starch agar; = amylolyticflora on MRS-starch agar.)

0 5 10 15 20 25


Figure 3. Evolution of amylolytic flora asproportion of total flora onMRS-starch medium during cassavastarch fermentation. ( = naturalfermentation; = inoculated-starchfermentation.)

Pro

por

tion

of

am

ylol

ytic

flo

ra (%

)

Log

(cf

u/g

dry

matt

er o

f st

arc

h) 10

9

8

7

6

50 5 10 15 20 25


Figure 2. Changes in anaerobic microflora onMRS and MRS-starch media ininoculated-starch fermentation.( = total lactic acid flora on MRSagar; = total flora on MRS-starchagar; = amylolytic flora onMRS-starch agar.)

6.0

5.5

5.0

4.5

4.0

3.5

3.00 5 10 15 20 25


pH

Figure 4. Evolution of pH during cassava starchfermentation. ( = naturalfermentation; = inoculated-starchfermentation.)

medium where glucose was replacedby 20 g/L of soluble starch) were ina higher proportion in inoculatedstarch than in noninoculated(Figures 1, 2, and 3). Furthermore,flora were heterogenous during thenatural fermentation, whereas theinoculated fermentation resulted ina predominance of the ALAB 20strain.

pH and lactic acid production.In the inoculated-starch fermentation,acidification was more notable duringthe first 5 days of fermentation, butpH value finally stabilized at 3.5 (pKa

of lactic acid) in both fermentations(Figure 4). The inoculatedfermentation produced slightly morelactic acid (Figure 5) during the firstdays of fermentation.

244


12

10

8

6

4

2

00 5 10 15 20 25


Figure 5. Evolution of lactic acid content duringcassava starch fermentation.( = natural fermentation; = inoculated-starch fermentation.)

Lact

ic a

cid (m

mol

/100 g

dry

matt

er o

f st

arc

h)

0 5 10 15 20 25

6

5

4

3

2

Spec

ific

vol

um

e (c

m3/g)

Figure 6. Evolution of cassava starchbread-making capacity duringfermentation. ( = naturalfermentation; = inoculated-starchfermentation.)

Sour starch specific BMP increasedfrom 2 cm3/g (wet starch) to at least5 cm3/g in 4 h of sun-drying(Larsonneur, 1993). The same starchsample, when oven-dried at 40 °C(slow drying) or 55 °C (rapid drying),or dried under cover for 8 h, did notexpand (specific BMP of 2-2.5 cm3/g).

These results demonstrate theneed for sun-drying if sour starch isto acquire BMP, and the importanceof solar radiation. The results alsoexplain why Brazilian plantsproducing cassava sour starch do notartificially dry starch during the rainyseason but ferment it instead forvarious months until the dry seasonarrives.

Oven-drying trials of sour cassavastarch at 40 °C and under a UV lamp(Cole-Palmer, G-09817-20, 4 W,254 nm and 366 nm) were conductedfor 8 and 16 days. Under UVradiation, the sour starch’s capacityfor bread making increased to a valueclose to that of the sun-dried starchcontrol, whereas oven-dried samplesexpanded little:

Treatment Bread-makingpotential (cm3/g) at:

8 h 8 days 16 days

Sun-drying 6.82 6.82

Oven-drying at 40 °C 2.46 3.18

Oven-drying at 40 °Cand under UV:

254 nm 3.94 4.95

366 nm 3.78 4.75

These results show that UVradiation is one of the different typesof sun radiation able to develop theBMP of cassava sour starch.Compared with 8 h of sun-drying, thelengthy period (8 and 16 days) neededto increase the bread-making capacity

Bread-making potential ofstarch. Inoculation of cassava starchwith ALAB 20 allowed the final BMPto be reached 10 days earlier,compared with natural fermentation.But the final BMP of the starch wasnot improved (Figure 6).

Sun-Drying Cassava SourStarch

Importance of ultraviolet radiation

Kinetic studies of drying cassava sourstarch in the sun (8 h) were realized.


245


of sour starch may be explained bythe low power (4 W) of the UV lampused.

The role of water in sun-dryingsour starch

Water content of cassava sour starchduring sun-drying plays an importantrole in improving the starch’sbread-making capacity. For example,cassava sour starch oven-dried at40 °C for 8 h, then rehumidified to50% and sun-dried for another 8 h,had a higher bread-making capacity(5.10 cm3/g) than the same starchsample dried under the sameconditions but without the additionalwater (3.75 cm3/g).

Better results are obtained(7.4 cm3/g) if sour starch is sun-driedat 40 °C for 8 h, then sun-dried foranother 8 h, but with water addedevery hour for 3 h. In contrast, theexpansion of the starch in thesun-dried control (8 h) was5.03 cm3/g.

Conclusions

To improve the quality of cassavasour starch, the followingrecommendations should be made tothe rallanderos (cassava sour starchproducers):

(1) To ferment. Starch should befermented for at least 20 days.The pH should be controlled at3.5. The fermentation tank shouldbe covered with about 5 cm ofwater to ensure anaerobicconditions and lactic acidfermentation.

(2) To dry. Sour starch should bedried under sunny conditions.Starch samples should be turnedover to ensure exposure of allstarch granules.

The preliminary results of starchinoculation trial demonstrated thatthe use of ALAB 20 as a starterculture helped reduce fermentationtime. Replicated starch inoculationtrials, using the same strain, will beundertaken to confirm these results.Other lactic inocula will also beinvestigated for reducing fermentationtime and improving cassava sourstarch quality.

From the results cited above, theconcept of an artificial dryingapparatus, using UV radiation andcontrolling starch moisture, can bevisualized. This would makestandardizing sour starch drying andquality possible, which would nolonger be at the mercy of the weather.

Studies are being conducted toevaluate the influence of cassavavariety and root storage on sourstarch quality. Climatic conditionsand the water used during productionmay also have effects.

Acknowledgments

We wish to thank J. Mayer andA. L. Chaves (BiotechnologyLaboratory, CIAT, Cali, Colombia) fortheir cooperation in analyzing organicacids, using high-performance liquidchromatography (HPLC). We alsothank F. Alarcón (CIAT) andA. Beitz (Universidad del Valle, Cali,Colombia) for their activeparticipation in the starch inoculationtrials. We specially thank A. L. Jaime(Universidad del Valle) for her help.

References

Brabet, C. and Dufour, D. 1996. El almidónagrio de yuca: producción y estudiosde los propiedades fisicoquímicas. In:Proceedings of the Simposio enCarbohídratos, del 4 al 6 octubre1993, Quito, Ecuador. p. 197-203.

246


__________ and Mestres, C. 1991. Evaluaciónde las modificaciones estructuralesdel almidón de yuca durante lafermentación: medida de laviscosidad intrínseca y técnica decromatografía de permeación en gel.In: Proceedings of the taller “Avancessobre Almidón y Yuca”; abstracts,17-20 June, Cali, Colombia. CIAT,Cali, Colombia. p. 1-6.

Camargo, C.; Colonna, P.; Buleon, A.; andRichard-Molard, D. 1988.Functional properties of sour cassava(Manihot utilissima) starch/polvilhoazedo. J. Sci. Food Agric. 45:273-289.

Cárdenas, O. S. and de Buckle, T. S. 1980.Sour cassava starch production: apreliminary study. J. Food Sci.45:1509-1512, 1528.

Flores, C. 1993. Estudio preliminar delcomportamiento fisiológico yenzimático en bioreactor de cuatrobacterias amilolíticas aisladas delalmidón agrio de yuca (Manihotesculenta Crantz) en Colombia.Informe de trabajo. CIAT, Cali,Colombia. 22 p.

Gómez, Y. 1993. Bacterias lácticasamilolíticas presentes en lafermentación del almidón agrio deyuca. Thesis. Facultad de Ciencias,Departamento de Biología,Universidad del Valle, Cali, Colombia.69 p.

Jory, M. 1989. Contribution à l’étude de deuxprocessus de transformation dumanioc comportant une phase defermentation: le gari au Togo,l’amidon aigre en Colombie. Mémoirede Mastère en technologie alimentairerégions chaudes. Ecole nationalesupérieure des industries agricoles etalimentaires (ENSIA) and CIRAD,Montpellier, France. 45 p.

Larsonneur, S. 1993. Influence du séchagesolaire sur la qualité de l’amidonaigre de manioc. Projet de fin d’études.Génie biologique, produits biologique etalimentaires, Université de technologiede Compiégne, France. 87 p.

Laurent, L. 1992. Qualité de l’amidon aigre demanioc: validation d’une méthoded’évaluation du pouvoir de panificationet mise en place d’une épreuvedescriptive d’analyse sensorielle.Projet de fin d’études. Génie biologique,produits biologiques et alimentaires,Université de technologie deCompiégne, France. 68 p.

Nakamura, I. M. and Park, Y. K. 1975. Somephysico-chemical properties offermented cassava starch (“polvilhoazedo”). Starch/Stärke 27(9):295-297.

247

Investigating Sour Starch Production in Brazil

CHAPTER 28

INVESTIGATING SOUR STARCH

PRODUCTION IN BRAZIL

R. C. Marder*, R. de Araujo Cruz**, M. A. Moreno***,A. Curran*, and D. S. Trim*

Abstract

In sour starch producing countriessuch as Brazil and Colombia, mostproduction is from small andmedium-sized plants. If the sector isto develop, it must adapt to changingcircumstances, environmental factors,and market demand for improvedproduct quality. Data on currentprocessing operations are essential foridentifying and prioritizingdevelopment and modernizationneeds.

This paper presents the resultsof a detailed, in-depth investigationof sour starch production in southernBrazil. We first describe themajor processing operations: rootpreparation, disintegration, screeningfor fiber removal, sedimentation, anddrying. Then we discuss the inputsand outputs for each operation,the composition of products andintermediates within the process,and, in particular, the volume andcomposition of waste waters—a factorof increasing environmental concern.

The data are then related to starchproduction technology used in other

countries. Areas identified for futuredevelopment and improvementinclude quality definition andstandardization, marketing andpromotion, and pollution abatementmeasures.

Introduction

Production of sour starch

In Brazil, sour starch (polvilho azedo)is manufactured principally in theState of Minas Gerais (MG), withadditional production in São Pauloand Paraná States. Plants aretypically small to medium-sized,processing about 10-20 t/day ofroots, although larger plants canprocess as many as 50 t/day.

An estimated 80 plants operate inthe municipalities of Cachoeira daMinas and Conceição dos Ouros, inthe region of Pouso Alegre, southernMG. Typical plants produce about3 t/day of starch, although somelarger plants produce 10-15 t/day.The Empresa de Assistência Técnica eExtensão Rural (EMATER) (personalcommunication, 1993) suggests thatthe region produced about 18,000 tin 1986, and is now producing12-13 thousand tons per year.

Sour starch production is alsoconcentrated around Divinópolis

* Natural Resources Institute (NRI), Kent, UK.** APTA Consultancy Group, São Paulo, Brazil.*** Universidad del Valle (UNIVALLE), Cali,

Colombia.

248


where about 10 small plants operate,each employing three to four peopleand producing more than 100 t ofstarch per year (EMATER, personalcommunication, 1993). These existalongside more than 100 very smallstarch production units, based onfamily farms, which sell to the localmarket, supermarkets, bakeries, andhouseholds. The total 1985production in the Divinópolis regionwas estimated to be about 10,000 t.Current production is probably belowthis figure if trends mirror those of thePouso Alegre region. No apparentchange has occurred in the numberof the region’s plants during the last3 years.

Sour starch is used in certainsnacks, mainly pão de quiejo andbiscoitos, on sale in public eatingplaces such as cafés and bus stations.The market for these products isstagnating, because of increasingcompetition from other snackproducts and the effects of the currenteconomic climate on consumerspending. However, pão de quiejo hasrecently begun to be marketedthrough a fast-food chain whichspecializes in the product, andsupermarkets have begun stocking itas a frozen product.

Although this expanded markethas resulted in a more buoyantdemand for sour starch, it has not hadthe kind of effect on the industry thatmight have been expected. In ParanáState, especially, producers of pão dequiejo are increasingly replacing sourstarch with industrially produced,unfermented, sweet starch (fecula), orsun-dried sweet starch (polvilho doce).

If the local sour starch industry isto survive, it must adapt to changingcircumstances—e.g., increaseddemand for an improved qualityproduct, increasing costs of inputs,and concern for environmentalconservation—by improving its

processing methods. Before changescan be made, accurate and detaileddata are needed on current processingoperations. Cereda and Takahashi(Ch. 25, this volume) have gatheredinformation on processing operationsin farinha and native starch industriesin Brazil. But comprehensive data areunavailable on operations andproblems experienced in small andmedium-sized sour starch plants. We,therefore, conducted a detailedanalysis of plant operations in twoplants in Minas Gerais.

Production Technology

The small- to medium-scaleproduction of sour starch in Brazil isschematically similar to that of sourstarch in Colombia as described bySalazar de Buckle et al. (1971)(Figure 1).

Lorry loads of fresh roots aredelivered to the plant and fed into arotary washer fitted with overheadwater sprays for part of its length. Aswell as washing off dirt and debris,the tumbling action of the rootsas they pass along the washer alsoremoves most of the bark. Washedroots are transferred to thehopper-fed, root disintegrator, viaan inspection conveyor, at which anoperator cuts up excessively largeroots, and removes remaining barkand stems.

All plants employ similarlyconstructed disintegrators known asthe “Jahn rasper” (Grace, 1977;Radley 1976). This machine consistsof a hollow, cylindrical drum, withtooth-edged steel blades sandwichedbetween local hardwood slats fixedlongitudinally to its surface. Thedrum is mounted between twocircular steel endplates on a centralshaft and housed inside a steelcasing, the base of which includes ascreening plate.

249


Water

Roots

Disintegration

Screening

Drying

Packaging

Milling

Washing Waste water

Figure 1. Sour starch manufacture in a typical plant in Minas Gerais State, Brazil.

Pulp

Water

Water vapor

Water

Milling

Fermentation

Sedimentation

Recycled liquor from the starchseparators is continuously fed into thedisintegrator. The resultant slurry ofcrushed roots passes through thescreening plate into a sump tank fromwhich it is pumped to the separators.

All plants employ a two-stageseparation process to remove theliberated starch from the fibrous pulp(massa). The majority of plantsemploy two centrifugal separators,which have replaced the traditional,rotating brush-and-screen washers.

The centrifugal separator consistsof a rotating conical screen, housedinside a shaped, mild-steel casing,

tapering from front to back. Theconical screen is a metal framecovered with a nylon mesh. Thenarrow end of the cone is closed witha fixed metal plate connected to thedrive shaft. Slurry is pumped into thecenter of the separator (toward thefixed plate) and forced through thescreen to an outlet at the bottom ofthe casing into a sump tank. Water issprayed into the slurry from jetspositioned around the screen.

In the sump tank, the slurryreceives extra water to facilitatepumping it over a flatbed reciprocatingscreen to remove any remaining fiber(larger plants employ an additional

Starchmilk

Recycledliquor

WaterWater

Waste water

Separation(First stage)

Separation(Second stage)

250


centrifugal separator). The slurry thenenters a second separator for furtherstarch extraction. Liquor dischargedfrom the second separator is returnedto the disintegrator, and thesuspension of pulp, or “starch milk,”is discharged to storage tanks.

The milk then flows intosedimentation channels or “tables”(Bruinsma et al., 1981). Dimensionsfor the channels vary considerablyfrom plant to plant: in length, from150 to 200 m; in width, from 0.6 to1.0 m; and in depth, from 0.4 to0.6 m. The channels are usually linedwith ceramic tiles because both starchand starch milk attack concrete. Thechannels are roofed to protect thestarch from rain or sunlight.

The milk is directed into one endof the channels and the supernatantliquor flows over a weir at the otherend to be discharged as waste waterinto nearby watercourses, seepagepits, or infiltration channels.

After overnight settling,supernatant remaining in thesedimentation channels is dischargedby removing the weir. The surface ofthe settled starch is sometimeswashed to remove those uppermostlayers containing high concentrationsof dirt, protein, and fiber impurities.Over several days, the channels areallowed to fill with successive layersof starch until space is available inthe fermentation tanks. The starchis then dug out of the channel,transferred to the tanks, covered withwater, and left to ferment for aminimum of 30 days. The tanks arealso lined with ceramic tiles and areusually constructed in series adjacentto the sedimentation channels. They,too, are roofed to prevent exposure tosunlight and rain.

After fermentation, the starch isremoved from the tanks, broken upwith a spike mill, and dried on hessian

sacks laid on raised, drying tables,usually made of bamboo. The dryingstarch is agitated manually at regularintervals. When dry, the starch iscollected, milled to a powder, andpackaged into 50-kg bags or, in someplants, into packs for direct sale toretailers.

Monitoring Plant Operations

Measuring process parameters andsampling procedures

With the agreement and cooperationof plant management and staff,processing operations at two plantswere monitored for 3 weeks.Monitoring activities were:

(1) Measuring water flows withinprocessing operations to determinewater consumption at each stage;

(2) Periodic sampling of fresh roots,disintegrator slurry, starch milk,waste fibrous pulp, fermented anddried starch;

(3) Sampling of water supplies andgenerated effluents throughout theprocess to characterize pollutionloads.

During monitoring, no attemptwas made to influence plantmanagement and staff in their work.

Sample analysis

Moisture content of root, starch cake,and pulp samples were determined byoven drying at about 45 °C to constantweight. Dried samples were storedand later analyzed for starch content,using enzyme hydrolysis (AOAC, 1965)and crude fiber (Harris, 1970).

Immediately after collection, thepH of all water and effluent sampleswas measured with a hand-held meter(CIBA Corning Diagnostics Ltd.,Halsted, UK). The samples were then

251


taken to a local laboratory andanalyzed for chemical oxygen demand(COD), and suspended and dissolvedsolids.


Root washing

Table 1 shows the proportions of dirt,bark, peel, and parenchyma in theroots received at the two plants andTable 2 shows the composition of thewashed roots.

The root washers employed in thetwo plants had a similar design: asemicircular, slatted trough, 7 m longwith a 0.95-m diameter. It had fixed,4-bladed paddles, mounted 0.3 mapart on an overhead, central rotating

shaft, which was driven at 150 rpm bya 2.2-kW (3 HP) motor. In plant A, thetrough was fitted with an overheadwater spray for the latter two-thirds ofits length, whereas in plant B, thespray covered only the last third.

The flow of roots through thewashers was, in effect, the same forboth plants at 0.55 kg/s of freshroots. But plant A used a much largervolume of water for washing: about1.95 L/s (or 3.55 m3/t of roots),compared with 0.70 L/s (or 1.27 m3/t)for plant B (Table 3). The washer’sperformance at plant A, as measuredby the percentage removal of bark,was considerably more effective (about95%), compared with that of plant B(78% to 80%). At plant B higher levelsof dirt and bark fragments were visiblyobservable in the sedimented starch.

Large sweet-starch plants in Braziland India employ similar washers. Insmaller Brazilian and Colombianplants, root washing is performed inbatches, using rotating, slatted drumswith a continuous supply of water. But in medium-sized plants in India,roots are passed through a flatbedconveyor washer, removing only thedirt and leaving the bark (Trim et al.,1993). For sago production, both thebark and peel are removed manually.

Table 2. Composition of roots, starch, and pulp at two plants producing sour starch, Minas Gerais, Brazil.

Sample Component (% DM)a

Total Starch Crude Fat Protein Ashsolids (%) fiber

Plant A:Washed roots 34.45 89.35 1.92 0.44 2.55 1.32Dried starch 88.10 96.59 0.35 - - -Pulp 7.70 85.59 8.45 0.16 1.36 0.96

Plant B:Washed roots 36.85 90.11 2.16 0.19 2.68 1.10Dried starch 88.73 96.43 0.41 - - -Pulp 7.23 82.21 12.14 0.24 1.79 1.26

a. (% DM) = percentage of dry matter.

Table 1. Composition (%) of residues fromwashing fresh cassava roots at twoplants producing sour starch, MinasGerais, Brazil.

Component Plant A Plant B

Dirt 0.50 0.49Bark 2.35 1.84Peel 15.01 17.47Parenchyma 82.14 80.20

252


Table 3. Total water consumption at two plantsproducing sour cassava starch, MinasGerais, Brazil.

Operation Water consumption

m3/t of m3/t ofroots product

Plant A:Root washing 3.55 15.24Starch extraction 3.78 16.22

Total 7.33 31.46

Plant B:Root washing 1.27 4.70Starch extraction 4.50 16.67

Total 5.77 21.37

Root disintegration

The drums in the disintegrators atthe two plants were of similarconstruction, 0.32 m in diameter and0.28 m in width. The blades werelongitudinally spaced, at about12 mm, around the circumference ofthe drum. Each drum had 80 to85 blades. The disintegrators wereboth directly driven by an electricmotor and rotated at 2,500 rpm.However, plant A employed a smallermotor (11.2 kW or 15 HP) than didplant B (18.6 kW or 25 HP). Bothplants employed 1.5-kW (2 HP)centrifugal pumps to transfer theslurry from the disintegrator sump tothe separators.

The total solids content of thedisintegrated root slurry at plant Awas 8.2% and at plant B, 7.6%.Although these values are similar, thedisintegrator at plant B produced amuch finer slurry, indicating a higherdegree of root maceration. This wasreflected in the starch extractionefficiency (i.e., the fraction of starchreleased in disintegration) at plant A(81%), compared with plant B (84%).These figures are considerably higherthan those quoted by Bruinsma et al.(1981) of 61% to 68% for small- tomedium-scale production, but close

to that reported by Trim et al. (1993)of 83% for an Indian sago plant.However, in the Indian plant, twoperforated drum disintegrators wereused in series to improve starchextraction.

The operation of the disintegratorat plant B was much smoother, withless notable strain on the motor,because of variation in feeding theroots. Furthermore, the plant operatorthought that the throughput ofroots in the disintegrator could beincreased, thus increasing maximumoutput.

Starch separation

Plant A employed two identicalcentrifugal separators, bothbelt-driven from a common 3.7-kW(5 HP) motor and rotating at 650 rpm. The rotating conical screen in eachwas 0.70 m in length, 0.25 m indiameter at the narrow end, and0.76 m at the wider end. Itwas covered with nylon mesh(PA-120-125/ASTM1). The steelcasing was 1 m squared in front,tapering to 0.5 m squared at the back.

Water was fed to the firstseparator at 0.90 L/s and to thesecond at 0.72 L/s. Fresh water wasalso added to the sump tank betweenthe separators at a rate of 0.44 L/s. The total water added was therefore2.06 L/s (3.78 m3/t of roots). Starchmilk was discharged from thefirst separator directly into thesedimentation channel at a rate of1.99 L/s with a concentration of solidsat 7.1%.

Plant B used a single centrifugalseparator, identical to those at plantA, for the primary stage, and arotating brush-and-screen washer for

1. ASTM = American Society for Testing andMaterials.

253


the second stage. The operatingspeed of the centrifugal separator at760 rpm was higher than at plant A,although using a similarly poweredmotor.

The brush-and-screen washerconsisted of a semicircular, screenedtrough (5.65 m long and 0.42 m indiameter), above which a shaft,rotating at 530 rpm, was centrallymounted. Plastic brushes werespaced at 90 mm intervals along theshaft, which was rotated by a 2.2-kW(2 HP) centrifugal pump.

Water was sprayed into theseparator at a rate of 0.74 L/s andinto the washer at 0.55 L/s; 1.20 L/sof water was fed into the sumptank between the two. Total waterconsumption was therefore 2.48 L/s(4.51 m3/t of roots). Starch milk wasdischarged from the washer at a rateof 2.52 L/s with a concentration ofsolids at 6.80%.

The solids content of the wastepulp was 7.70% in plant A and 7.23%in plant B (Table 2). The starchcontent of the pulp at plant A was85.59% and at plant B, 82.21%. Thehigher concentration from plant Aagain indicates less efficientdisintegration of the roots. Trim et al.(1993) measured a starchconcentration of 72% in the pulpdischarged from a system ofreciprocating screens.

The concentration of free starch inthe waste pulp was very low at bothplants (0.32% at plant A and 0.18% atplant B), indicating a high efficiency ofextraction of free starch from the pulp.

Sedimentation and fermentation

Plant A used six channels,constructed side by side andconnected in series to minimize spaceand ease unloading. Each channelwas 32 m long, 0.74 m wide, and

0.4 m deep. Plant B had four similarlyconstructed channels, each 45 m long,0.82 m wide, and 0.5 m deep. Bothsets of channels had a weir, 0.15 mhigh, at one end. The residence timefor starch milk flowing into emptychannels was 3.0 h for plant A and2.4 h for plant B. The solids contentof batches of sedimented starchremoved from the channels averaged59.9% for plant A and 59.1% forplant B.

In India, tanks are used instead ofchannels for sedimentation, largelybecause of historical reasons (Trim etal., 1993). After overnight settling andremoval of the supernatant liquor, thestarch cake had a concentration ofsolids at 50%, but after washing, theconcentration was 55%.

Starch and crude fiberconcentrations of the settled cake inthe two plants were similar, averaging96.7% for starch and 0.3% for crudefiber (dry matter basis).

The changes occurring in thestarch as a result of fermentation arethe subject of much recent research(e.g., Brabet et al., Ch. 27, thisvolume) but were not studied in thisinvestigation. Although a minimumfermentation time of 30 days isnecessary, starch often remained inthe tanks at the two plants for longerperiods because of the lack ofavailable drying space. Suchprolonged fermentation had nodetrimental effect on starch quality.The temperature of the fermentingstarch at the two plants rangedbetween 12 and 13 °C.

Drying

Both plants employed traditionaldrying tables, raised about 1 m fromthe ground. They were essentiallybamboo mats (esteiras), 4.0 m longand 1.2 m wide, tied to bamboo beamsmounted on wooden stakes. Plant A

254


had 600 esteiras, with a total dryingarea of 2,800 m2, and plant B had760 esteiras (3,650 m2). The fermentedstarch was spread on cotton sacksstretched across the tables with a wetstarch loading of about 1.8 to2.0 kg/m2.

In summer, drying may take 6 to7 h, but in winter it may take 2working days or 13 h. Starch that isstill damp by the end of the day isgathered up in cotton sacks and placedin storage sheds overnight.

Figure 2 shows the drying curvesfor batches of starch dried at the twoplants (moisture contents are given ona wet basis). As calculated fromTable 2, the moisture content (drybasis) of the dried starch produced atplant A was 11.9% and that at plant B,11.3%. The dried starch contents were96.6% at plant A and 96.4% at plant B(dry matter basis) (Table 2).

Many plants in the area areinvesting in drying tables made of wiremesh within wooden frames. The meshprovides improved ventilation aroundthe starch and so reduces drying time.

Material balance

The material balance for theprocessing operations at the twoplants was calculated from themeasured data of process flows andfrom results of laboratory analyses(Figures 3 and 4). Figure 3 shows thatthe total mass flow of dried starch inplant A was 23% and Figure 4, 27% inplant B. A more accurate comparisonis that of starch recovery efficiency—the fraction of starch in the rootsrecovered in the product. The overallstarch recovery was about 67% forplant A and 72% for plant B.

Product quality

Table 2 gives the composition of thestarch products. The results indicateno significant difference in starchpurity in products from the twoplants, especially in root washing,despite their different processingprocedures.

Processors commonly definequality in terms of the degree ofwhiteness and the acid taste of thesour starch, but no data exist toconfirm that these criteria are linkedto commercial value. Producersbelieve quality improves the more theprocessing environment is clean, themore water used, and the purer theprocessing water. Spring water isusually preferred to well or river water.The lower temperature of springwater is also believed to improvefermentation. Intense sunlight andagitated air movement around thestarch on a second or third day’sdrying may deteriorate quality byencouraging growth of mold.

Water consumption andcharacteristics of waste water

Plant A used more water for rootwashing (3.55 m3/t of roots) than didplant B (1.27 m3/t) (Table 3).However, plant A used appreciably

Figure 2. Sour starch drying curves at two plantsin Minas Gerais, Brazil. ( = plant A;

= plant B.)

Drying time (hours)

Moi

stu

re c

onte

nt

(% w

b)

50

45

40

35

30

25

20

15

10

5

00 1 2 3 4 5 6 7 8 9 10 11 12 13

255


Figure 3. Material balance, based on 1,000 kg of roots, for plant A in Minas Gerais, Brazil. (T = total massflow; W = water flow; S = starch flow; N = flow of nonstarch components.)

less water for root disintegration andstarch separation (3.78 m3/t) thandid plant B (4.50 m3/t). Total waterconsumption was 7.33 m3/t for plantA, and 5.77 m3/t for plant B. Theflow of water recycled from theseparators to the disintegrator wasmarginally different in the twoplants: plant A recycled 3.50 m3/tand plant B, 4.22 m3/t. In sagoproduction in India (Trim et al.,1993), water consumption was6.40 m3/t, of which 4.10 m3/t wasused for disintegration andseparation.

Table 4 shows the results ofanalyses of the two principal

effluents from the plants. These dataconfirm the highly polluting nature ofthese waste waters. The COD of thewaste waters from plant A was4,800 mg/L and from plant B,3,500 mg/L. The CODs of thesupernatant liquor discharged fromthe sedimentation channels were11,500 mg/L for plant A and14,800 mg/L for plant B, that is,much higher than the 6,700 mg/Lnoted in liquor discharged fromsedimentation tanks in India (Trim etal., 1993).

Analyses also indicated that thesupernatant liquors containedsignificant levels of cyanide

Washing

RootsT-1000, W-637, S-299, N-64

Disintegration

Sedimentation

Drying

Dried starchT-233, W-28, S-199, N-6

WaterW-3550

WaterW-3780

Waste waterT-3640, W-3591, S-18, N-31

Recycled liquorT-3500, W-3442, S-46, N-12

PulpT-1150, W-1070, S-67, N-13

Water vaporW-109

Waste waterT-3200, W-3171, S-15, N-14

Separation

Fermentation

Starch cake

Root slurryT-4410, W-4038, S-327, N-45

Washed rootsT-910, W-596, S-281, N-33

Starch cakeT-340, W-135, S-199, N-6

Starch milkT-3540, W-3306, S-214, N-20

256


Disintegration

WaterW-1270

Waste waterT-1360, W-1313, S-21, N-26

WaterW-4500

SeparationPulpT-950, W-882, S-56, N-12

Recycled liquorT-4220, W-4166, S-40, N-14

Sedimentation Waste waterT-4060, W-4032, S-14, N-14

Water vaporW-130

Fermentation

Drying

Dried starchT-270, W-30, S-233, N-7

RootsT-1000, W-617, S-324, N-59

Figure 4. Material balance, based on 1,000 kg of roots for plant B in Minas Gerais, Brazil. (T = total massflow; W = water flow; S = starch flow; N = flow of nonstarch components.)

Washing

Table 4. Characteristics of waste waters at two plants processing cassava sour starch in Minas Gerais,Brazil.

Sample Characteristica

COD DS SS pH HCN(mg/L) (mg/L) (mg/L) (mg/kg)

Plant A:

Waste waters 4,778 401 1,297 5.93Supernatant liquor 11,538 1,516 7,351 5.11 43

Plant B:

Waste waters 3,475 618 1,797 6.21Supernatant liquor 14,778 3,370 4,979 5.38 62

a. COD = chemical oxygen demand; DS = dissolved solids; SS = suspended solids.

Root slurryT-5130, W-4740, S-343, N-47

Starch cakeT-400, W-160, S-233, N-7

Starch cake

Washed rootsT-910, W-574, S-303, N-33

Starch milkT-4460, W-4192, S-247, N-21

257


organized and the equipment usuallywell maintained. Some significantchanges in processing have beenadopted over recent years, mostnotably the introduction of centrifugalseparators for the recovery of starchfrom the macerated roots. Aneffective means of technology transferexists through the localized nature ofthe industry, the local equipmentsupply and maintenance workshops,and plant workers setting up theirown processing plants.

Plant operators see the mostimportant issues as being:

Availability and price of cassavaroots;Access to “soft” loans to financeworking capital;Labor costs;Packaging costs;Marketing and promotion;Perception of improved quality byconsumers;Efficient and cost-effective effluenttreatment systems.

Processors also considerfermentation and drying to beprocessing bottlenecks. The longfermentation periods tie up scarceworking capital, and sun-drying issometimes unreliable and requiresconsiderable space. But, if improvedtechnology for rapid fermentation andartificial drying become reality, thenlarge-scale industrialists in otherareas may be able to undercutsmall-scale producers in productprice by having access to cheaper rootsupplies and reaching economies ofscale. Such undercutting wouldmean the collapse of a largeproportion of the industry in MinasGerais.

The market for sour starchproducts is growing slowly, butcompetition in supply is intensifyingand quality is becoming moreimportant. Producers in Minas

compounds, measured at43-62 mg/kg. These values are muchhigher than those measured in India(20-35 mg/kg). The roots used inIndia for sago production are peeledbefore disintegration, thus carryingaway larger quantities of cyanogens.

Effluent treatment

The effluent problem is a majorenvironmental issue in both PousoAlegre and Divinópolis. Many plantsdischarge their effluent directly intosmall streams feeding the local river.Fish and animals have been killed bypolluted watercourses, and the StateWater Authority is concerned aboutthe dangers of polluting drinkingwater supplies.

Federal legislation requires thatplants install effluent treatmentsystems capable of removing at least85% of the pollution load. Some localauthorities have threatened legalaction against plants that do notinstall treatment systems, despite thefact that no effective treatmentsystems are available that are alsoeconomically feasible. In reality,however, plant closures are unlikelybecause of local socioeconomicfactors, and pollution will continueuntil cost-effective solutions arefound.

The most commonly used disposalsystems include seepage pits (usuallythree pits used in series) or infiltrationchannels, which allow water to seepthrough the soil. The solid material isremoved periodically and usedas fertilizer. Some plants use theeffluent for irrigating their cassavacrop. The long-term effects of thesemethods are unknown.

Conclusions

The two plants studied, and most ofthe others visited, were efficiently

258


Gerais may well encounter futureproblems as a result of increasingcompetition from new producers(especially in São Paulo State), whohave greater financial resources,access to higher levels of technology,and are located near cheap andabundant supplies of roots.

Future priorities for researchshould be concentrated in threeareas:

(1) Product quality. Quality factorsneed to be clearly defined andstandards established.Relationships between processinputs, operations, and qualityfactors need to be identified andevaluated.

(2) Markets. Promotional efforts arerequired to expand consumerawareness of sour starch and itsspecialized properties and uses.

(3) Water pollution. Affordabletechnology for water conservation,waste reduction, and treatmentoperations needs to be developedto minimize pollution.

References

AOAC (Association of Official AnalyticalChemists). 1965. Official methodsof analysis. 10th ed. Arlington, VA,USA.

Bruinsma, D. H.; Witsenburg, W. W.; andWurdemann, W. 1981. Cassava. In:Selection of technology for foodprocessing in developing countries.Centre for Agricultural Publishing andDocumentation (PUDOC), Wageningen,the Netherlands. p. 113-158.

Grace, M. R. 1977. Cassava processing. FAOplant production and protection series.Food and Agriculture Organization ofthe United Nations (FAO), Rome. 155 p.

Harris, L. E. 1970. Determination of cell wall(neutral detergent fiber) and cellcontents. In: Nutrition researchtechniques for domestic and wildanimals, vol. 1. Utah State University,Logan, UT, USA. p. 2801-2802.

Radley, J. A. 1976. Starch productiontechnology. Applied SciencePublications, London, UK. 587 p.

Salazar de Buckle, T.; Zapata M., L. E.;Cárdenas, O. S.; and Cabra, E. 1971.Small-scale production of sweet andsour starch in Colombia. In: Weber,E. J.; Cock, J. H.; and Chouinard, A.(eds.). Cassava harvesting andprocessing; proceedings of a workshopheld at CIAT, Cali, Colombia.International Development ResearchCentre (IDRC), Ottawa, Canada.p. 26-32.

Trim, D. S.; Nanda, S. K.; Curran, A.;Anantharaman, M.; and Nair, J. 1993.Investigation of cassava starch andsago production in India. Paperpresented at the InternationalSymposium on Tropical Root Crops,6-9 Nov., Thiruvananthapuram, India.

259

Implementing Technological Innovations...

CHAPTER 29

IMPLEMENTING TECHNOLOGICAL

INNOVATIONS IN CASSAVA FLOUR AND

STARCH PROCESSING:A CASE STUDY IN ECUADOR1

Vicente Ruiz*

de Asociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY). This team copiedand adapted some prototypeequipment and tools on-site in theAssociation’s processing plants. Theproducts—cassava starch and flour—were efficiently produced andentered national and internationalmarkets.

Flour Processing

Technology introduced fromColombia

In late 1985, trials showed thatcassava meal could be technically andeconomically produced, using atechnology introduced from CIAT,Colombia. The technology consistedof chipping, drying, and grindingdried cassava. Chips, produced by aThai-type, mechanical, disc chipper,are dried on outdoor concrete floorsand then ground in hammer mills.

Technology currently used byUATAPPY

In addition to cassava meal, threeother types of flour are produced:white industrial flour, table flour, andsieved whole-grain flour. Thetechnology used to produce theseflours differs from that for cassavameal (Table 1). To produce white

Background

Before 1985, the only cassavaprocessing technology known inEcuador was mechanical rasping andhand-sieving to extract starch fromthe roots. Since then, newtechnologies have been introduced,and existing ones improved, toincrease processing efficiency andopen new markets for both cassavastarch and flour.

These new technologies includechipping, drying, and grindingcassava roots to produce meal andflour from peeled cassava roots, andsieving coarse-grained flours toproduce fine ones. Improvedequipment for starch processinginclude raspers with saws,continuous flow washer-peelers,vibrating screens, and sedimentationchannels.

In Manabí, Ecuador, aparticipatory approach has been usedto facilitate the adoption of improvedtechnologies. The first step was totrain the technical team of the Unión

* Unión de Asociaciones de TrabajadoresAgrícolas, Productores y Procesadores deYuca (UATAPPY), Manabí, Ecuador.


260


Table 1. Comparison of steps in the processing of different cassava flours, once roots are received, usingcurrent technology, Manabí, Ecuador.

Process Flour(technique)

Cassava White Table Sievedmeala industrial flour flour whole-grain flour

Peeling(manual) X X X

Washing(manual, mechanical) X X

Chipping(Thai-type disc chipper) X X X X

Drying(concrete floors) X X

Drying(trays) X X

Milling(hammer mill) X X X X

Sieving(vibrating or centrifuge screen) X X

Packaging(polypropylene) X X X X

a. Original technology, introduced from Colombia. The other products are produced with more recenttechnology.

industrial flour, the roots are peeledby hand before being fed to thechipper. The rest of the process is thesame as for cassava meal. Toproduce table flour, the roots arepeeled and washed before chipping,and then dried naturally on trays, orartificially. Once the dried chips areground, the resulting flour is sievedthrough a vibrating or centrifugescreen. Sieved whole-grain flour isproduced by passing the mealthrough a vibrating screen as fortable flour.

Starch Processing

Traditional technology used inEcuador

Manual starch extraction in Ecuadordates back about 50 years.Traditionally, to extract cassavastarch, roots are peeled and washedby hand, grated by hand or

mechanically with an engine-drivenwooden drum covered with aperforated zinc plate, then sieved byhand. Sedimentation is carried out inwooden or concrete tanks, and thestarch dried on concrete floors or onpaper (Figure 1).

Technology currently used byUATAPPY

The UATAPPY is currently extractingcassava starch with mechanizedtechnology developed with thetechnical assistance of CIAT and theFundación Adelanto Comunitario(FACE), and with the financialsupport of the Fundación para elDesarrollo Agropecuario(FUNDAGRO). Cassava roots undergothe following procedures: washingand peeling, in either batch orcontinuous flow, with Brazilian-typewashers; mechanical rasping withBrazilian-type saw blades; sieving,both by hand and vibrating screens;

261

Implementing Technological Innovations...

Roots received

Peeling Manual

Washing Manual

Grating

Sieving Manual

Sedimentation Concrete tanks

Sieving

Rasping Rasper with blades

Sedimentation Channels

Drying

Packaging Paper

Drying

Packaging

ManualMechanized

Mixed (manual andmechanical)Vibrating screen

PlasticConcrete floors

PaperPolypropylene

Concrete floorsPaperZinc

Process Technique Process Technique

Traditional technology batch system

Current technology gravity system

Products:

Starch for human consumptionIndustrial starch (first grade)Industrial starch (second grade)

and sedimentation in concretechannels lined with ceramic tiles.Drying is carried out naturally onplastic sheets placed on bambooplatforms. When a finer qualityproduct is required, milling isdone in hammer mills (Figure 1).

Experiments

Sour starch production trials werefirst carried out in December 1992and renewed in November 1993 attwo starch factories, both UATAPPYmembers. Artificial drying trials with

Figure 1. Differences in traditional and current technologies for cassava starch extraction, Manabí,Ecuador. Current technologies include innovations introduced from Colombia and Brazil.

Washing Continuous-flow washerBatch washer

262


a flash dryer will be conducted,together with mechanized sieving,using vibrating or centrifuge screens.

Training and InstitutionalSupport

To introduce and adapt new cassavaprocessing technologies, especially forstarch, UATAPPY received technicaland financial support fromFUNDAGRO and CIAT. Its technicalteam has received training nationallyand in Colombia and Brazil on

elements of processing andtechnology.

Results

(1) Product quality (flours and starch)has improved, allowing newmarkets to be opened at nationaland international levels.

(2) Higher yields have been obtainedand efficiency has improved.

(3) Production capacity, especially ofstarch, has increased.

263

The Influence of Variety and Processing...

This chapter outlines the results sofar.


The cassava cultivars used in thisresearch were selected after thecassava core collection, held at CIAT,Cali, Colombia, was evaluated(Wheatley et al., 1993). Cultivars wereselected to represent a broadvariability in root contents ofcyanogens, dry matter, and amylose.

Experiments were designed todetermine how flour preparationinfluences the resultant quality of theend product. In October 1992,1,500 kg of fresh roots of the cassavacultivar CM 3306-4 were harvested atCIAT. The roots were processed intoflour as outlined:

Fresh roots > Washing and peeling

Milling < Drying < Chipping <

> Sieving > Flour

In August and September 1993,two more cultivars of cassava(CM 3306-4, M Ven 25) were harvested10 months after planting and similarlyprocessed.

Before drying, the chips from eachcultivars were divided into three lots of

CHAPTER 30

THE INFLUENCE OF VARIETY AND

PROCESSING ON THE PHYSICOCHEMICAL

AND FUNCTIONAL PROPERTIES OF

CASSAVA STARCH AND FLOUR

A. Fernández *, J. Wenham **, D. Dufour ***,and C. C. Wheatley†

Abstract

The influence of certain processingconditions on the quality, functionalproperties, and product potential offlour made from three cassava cultivarsare being evaluated as part of aproject (DGXII) funded by theEuropean Union (EU). Thecollaborators in this project are theUniversidad del Valle (UNIVALLE),Colombia; CIRAD-SAR, France; theNatural Resources Institute (NRI), UK;and CIAT, Colombia.

The influence of dryingtemperature (40, 60, and 80 °C),milling procedure (hammer, roller, pin,and paddle), and particle size(< 250 µm and < 160 µm) on thequality, functional properties, andproduct potential of flour from threecassava cultivars are being evaluated.The influence of genetic variability onstarch quality is also being evaluated,using starches made from cultivarschosen from the cassava core collectionestablished at CIAT.

* Universidad del Valle (UNIVALLE), Cali,Colombia.

** Natural Resources Institute (NRI), Kent, UK.*** CIRAD/SAR, stationed at the Cassava

Utilization Section, CIAT, Cali, Colombia.† Centro Internacional de la Papa (CIP),

stationed at Bogor, Indonesia.

264


475 kg. They were then dried in alayer, about 15 cm thick, on the floor(3 m2 area) of an airflow bin dryer.Drying temperatures used were 40,60, and 80 °C.

Four different types of mill wereused to grind the resultant dry chips:(1) hammer mill (set at 5,800 rpm andwith a 1

8 -inch screen), (2) roller mill(first pass with rollers set at 300 µmapart and second pass with rollers setat 30 µm), (3) pin mill, and (4) apaddle auger in a cylindrical sifterwith a 5-mm and 250-µm screen. Theflours produced were divided into twosieve fractions to give two particlesizes: smaller than 250 µm, andsmaller than 106 µm.

The flours produced by thedifferent treatments were analyzedwith a Brabender amylograph (BA). Gelatinization profiles weredetermined from 6%-starch solutions,using a heating and cooling rate of1.5 °C/min. The temperature wasincreased to 95 °C, held for20 min, and cooled at a rate of1.5 °C/min to 50 °C. The viscographsobtained were used to calculate thefollowing parameters: the initialtemperature of gelatinization, peakviscosity, ease of cooking, gelinstability, and gelatinization index.

In October 1992, 29 cassavacultivars were harvested at CIAT 10to 12 months after planting. Starchsamples were extracted as outlined:

Fresh roots > Washing and peeling

Starch removed < Rasping <by adding waterand filtering

> Starch > Dryingsedimentation

Starch <

Starch samples were also preparedfrom another 33 cassava cultivars,

including the same 29 cultivars,harvested at CIAT in July 19939 months after planting.

The starch samples extracted fromthe cassava cultivars in 1992 wereanalyzed with a BA. Gelatinizationprofiles were determined from6%-starch solutions as describedabove. The amylose contents weredetermined, using an iodo-colometrictest and a calibration curve preparedfrom potato amylose and amylopectin.The crystallinity of the starch granuleswas determined with an X-raydiffraction system. The diffractiondata were collected over an angularrange from 4° to 32° 20’.

Starch samples from both harvestswill be examined for granular sizedistribution,.amylose-to-amylopectinratio, chain length, degree ofpolymerization, X-ray diffractionpatterns and absolute crystallinity,differential scanning calorimetry (DSC)analysis, pasting and rheologicalcharacteristics, swelling power andsolubility, and water-binding capacity.

Results

Flour

The various processing proceduresused in these experiments allinfluenced the gelatinization profilesof the resultant flours (Figures 1 to 4;Table 1). Whether the differencesobtained in gelatinization propertiesare enough to significantly influencethe potential uses of the flours is yet tobe determined. At the time of writing,the flours prepared from the cultivarsharvested and processed in Augustand September 1993 were not yetanalyzed.

Starch

Figure 5 shows a sample of the X-raydiffractograms obtained from starches

265


Figure 1. Viscoamylograph of cassava flour in relation to drying temperatures ( = 40 °C; = 60 °C; = 80 °C). Brabender curves were obtained from flour suspension at 6% of dry matter. Theflour had a particle size smaller than 250 µm, after milling with rollers.

Figure 3. Viscoamylograph of cassava flour in relation to particle size ( = 250 µm; = 106 µm).Brabender curves were obtained from flour suspension at 6% of dry matter. Chips were dried at60 °C and milled in a hammer mill.

350

300

250

200

150

100

50

0

Time (minutes)

0 27 32 37 42 47 52 57 64 74 84 94

Bra

ben

der

vis

cosi

ty u

nit

s

Time (minutes)

0 27 32 37 42 47 52 57 64 74 84 94

Bra

ben

der

vis

cosi

ty u

nit

s

400

300

200

100

0

500

400

300

200

100

0

Time (minutes)

0 27 32 37 42 47 52 57 64 74 84 94

Bra

ben

der

vis

cosi

ty u

nit

s

..

..Figure 2. Viscoamylograph of cassava flour in relation to drying temperatures ( = 40 °C; = 60 °C; = 80 °C). Brabender curves were obtained from flour suspension at 6% of dry matter. Theflour had a particle size smaller than 250 µm, after milling with a hammer mill.

266


Time (minutes)

0 27 32 37 42 47 52 57 64 74 84 94

Bra

ben

der

vis

cosi

ty u

nit

s

Angle (2-theta)

10 20 30

Figure 5. A sample of wide-angle, X-ray diffractograms of native starches from cassava cultivars harvestedat CIAT, October 1992.

M Col 1132

M Col 72

M Bra 897

M Bra 881

M Bra 162

M Col 22

........

500

400

300

200

100

0

Figure 4. Viscoamylograph of cassava flour in relation to milling method ( = hammer mill; = pin mill; = roller mill; = paddle mill). The flour was made from chips dried at 60 °C, and flourparticle size was smaller than 250 µm. Brabender curves were obtained from flour suspension at6% of dry matter.

267


Table 1. Cassava flour functionality characteristics in relation to its particle size, the drying airtemperature, and the milling procedure of the chips.

Flour characteristic Milling equipment and drying temperatures (°C)

Hammer Roller Pin Paddle

40 60 80 40 60 80 40 60 80 40 60 80

Flour composition:

Flour Aa:Starch (%, db) 83 82 79 85 83 81 84 82 81 88 86 82Fiber (%, db) 0.9 0.8 1.0 1.4 1.0 1.6 1.2 0.7 1.3 1.0 1.2 1.3Ash (%, db) 1.5 1.7 1.7 1.8 1.8 1.8 1.5 1.6 1.7 1.6 1.5 1.4

Flour Bb:Starch (%, db) 87 85 83 86 86 86 86 86 85 92 91 87Fiber (%, db) 0.6 0.8 0.4 0.4 0.5 0.6 1.1 0.8 1.1 0.6 0.8 1.1Ash (%, db) 1.4 1.5 1.5 1.7 1.3 1.6 1.5 1.5 1.7 1.5 1.3 1.5

Gelatinization temperature (°C):

Flour A 65.5 65.5 65.5 65.5 65.5 65.5 64.0 65.5 65.5 65.5 65.5 65.5Flour B 65.5 65.5 65.5 65.5 65.5 65.5 65.5 64.0 65.5 65.5 65.5 65.5

Maximum viscosity:

Flour A 371 380 380 255 323 295 380 380 340 408 410 410Flour B 385 420 380 285 350 327 387 405 360 400 430 425

Viscosity at 95 °C:

Flour A 365 365 366 251 321 289 377 363 340 390 380 385Flour B 375 387 367 284 338 320 377 380 358 380 390 385

Viscosity after 20 min at 95 °C:

Flour A 172 175 185 102 152 131 202 183 239 168 180 190Flour B 177 190 185 119 160 152 177 190 228 170 185 200

Viscosity at 50 °C after cooling:

Flour A 285 300 322 180 252 210 292 313 319 295 318 350Flour B 308 327 320 180 285 260 297 340 335 305 340 380

Ease of cookingc:

Flour A 19 17 16 20 18 20 19 17 16 17 16 15Flour B 18 16 16 20 17 16 18 17 15 17 16 14

Gel instabilityd:

Flour A 199 205 185 153 171 164 178 197 101 240 230 220Flour B 208 230 195 166 190 175 210 215 132 230 245 225

Gelatinization indexe:

Flour A 113 125 137 78 100 79 90 130 80 127 138 160Flour B 131 137 135 61 125 108 120 150 107 135 155 180

a. Flour A = flour with particles smaller than 250 µm.b. Flour B = flour with particles smaller than 106 µm.c. Ease of cooking = time to maximum viscosity - time to gelatinization.d. Gel instability = maximum viscosity - viscosity after 20 min at 95 °C.e. Gelatinization index = viscosity at 50 °C after cooling - viscosity after 20 min at 95 °C.

268


crystallinity and amylose content,together with the analysis reported byCIAT of root dry matter and cyanogencontents. Table 3 gives thegelatinization profiles of starch

made from the roots of the October1992 harvest. All spectra of the 29cultivars analyzed showed an A-type,X-ray diffraction pattern. Table 2 givesvalues calculated for starch

Table 2. Dry matter of fresh roots, cyanogen content of fresh parenchyma, amylose content, andcrystallinity of starch obtained from 29 cassava cultivars harvested at 10-12 months at CIAT,Palmira, Colombia, October 1992.

Cultivar Dry matter Total cyanogens Amylose Crystallinity (%) (as HCN, mg/kg, db) (% in starch) (%)a

M Bra 162 32 1,012 17 39

M Bra 881 31 832 20 41

M Bra 897 36 98 21 38

M Col 22 35 85 23 37

M Col 72 33 248 22 41

M Col 1132 21 69 26 39

M Col 1486 37 120 22 43

M Col 1684 37 752 23 38

M Col 2066 30 58 24 43

M Col 2215 43 243 25 44

M CR 35 45 17 24 41

M Mal 1 38 411 25 39

M Mal 2 27 413 24 40

M Mex 59 34 311 21 39

M Nga 2 22 632 23 47

M Per 196 33 393 21 42

M Tai 1 33 629 22 38

M Ven 25 27 1,628 22 43

M Ven 77 32 223 23 40

CG 1-37 35 182 22 44

CG 165-7 23 402 22 44

CG 402-11 18 169 20 45

CG 915-1 37 149 24 41

CG 1118-121 27 27 25 39

CG 1141-1 40 337 24 44

CM 489-1 23 86 41

CM 2766-5 32 82 43

CM 2772-3 27 114 42

CM 3306-4 39 82 43

a. Based on the separation and integration of the areas under the crystalline and amorphous X-ray diffractionpeaks.

269


Table 3. Values of total cyanogen content in parenchyma, amylose content, starch crystallinity, andstarch functionality characteristics for six cassava cultivars harvested in October 1992 at CIAT,Palmira, Colombia.

Characteristic Cultivar

CM 3306 CG 1-37 M Ven 77 CG 165-7 M Tai 1 M Ven 25

Total cyanogen in 82 182 223 402 629 1,628parenchyma(as HCN, mg/kg, db)

Amylose (%) 26 22 23 22 22 22

Crystallinity (%) 43 44 40 44 43 43

Gelatinization 64.0 64.0 65.5 62.5 62.5 62.5temperature (°C)

Maximum viscosity 975 775 610 800 780 730

Viscosity at 95 °C 415 320 330 350 340 310

Viscosity after 20 min 260 225 195 220 230 190at 95 °C

Viscosity at 50 °C 520 460 380 435 410 330after cooling

Ease of cookinga 4 4 7 5 5 5

Gel instabilityb 715 550 415 580 550 540

Gelinization indexc 260 235 185 215 180 140

a. Ease of cooking = time to maximum viscosity - time to gelatinization.b. Gel instability = maximum viscosity - viscosity after 20 min at 95 °C.c. Gelatinization index = viscosity at 50 °C after cooling - viscosity after 20 min at 95 °C.

samples analyzed. The resultsobtained show a similar trend to thatreported by Wheatley et al. (1993).Differences in starch viscositycharacteristics were observed betweencultivars with high and low cyanogeniccontent.

Research is continuing with the33 starch samples obtained fromcultivars harvested in July-August1993.

Reference

Wheatley, C. C.; Orrego, J. I.; Sánchez, T.;and Granados, E. 1993. Qualityevaluation of the cassava corecollection at CIAT. In: Roca, W. M.and Thro, A. M. (eds.). Proceedings ofthe First International ScientificMeeting of the Cassava BiotechnologyNetwork, Cartagena de Indias,Colombia, 25-28 August 1992.Working document no. 123. CIAT,Cali, Colombia. p. 255-264.

270


CHAPTER 31

ESTABLISHING AND OPERATING ACASSAVA FLOUR PLANT ON THE

ATLANTIC COAST OF COLOMBIA1

Francisco Figueroa*

price structures of cassava and wheatin Colombia, producing cassava flourat a price competitive with that ofwheat flour was economically feasible(Tables 1, 2, and 3). Hence, the nextphase, that of the pilot-project, wasinitiated.

In the research phase, bakedproducts had been considered as themain market, where cassava flourwould substitute 15% of wheat flour.But, because bakers saw a high riskof decreased product quality whenusing cassava flour, phase II wasfocused on other food categorieswhere cassava flour would notpresent high risks.

With phase II, the production,processing, and marketingcomponents of the cassava floursystem were integrated under the realconditions of a cassava-growingregion in Colombia. These resultscan be used by both public andprivate enterprises to promote thereplication of rural, cassavaflour-producing plants and theproduct’s use in the national foodindustry.

The institutions participating inthe project are CIAT, Cali;Universidad del Valle, Cali; the Fondode Desarrollo Rural Integrado (DRI) ofthe Colombian Ministry ofAgriculture; and the Fundación para

Background

CIAT has developed a strategy todesign and implement cassavaprojects, integrating aspects of thecrop’s production, processing, andcommercialization in northernColombia. Within this framework,three phases of development can bedistinguished:

(1) Research: developing technologyfor cassava processing, andstudying in detail the technology’smarket opportunities, both on anational and regional basis.

(2) Pilot project or market test:producing and marketing on asmall scale under real marketconditions.

(3) Commercialization or expansion:consolidating the market for newproducts and replicating theprocessing units.

A project to develop, under thisstrategy, a rural cassava flourindustry was begun, and its progressso far is reported here.

Results of phase I (research)indicated that, under the cost and

* Fundación para la Investigación y elDesarrollo de Tecnologías Apropiadas al Agro(FUNDIAGRO), Colombia.


271

Establishing and Operating a Cassava Flour Plant...

Table 1. Variable costs (US$) of cassava flour in January 1994, Chinú, Colombia.

Item Unit/t Unit cost Cost/t

Raw material 3.5 t 43.00 150.50Labor 60 man-hours 0.40 24.00Package 25 units 0.30 7.50Electricity 140 kW/h 0.10 14.00Mineral coal 550 kg 0.04 22.00Water 7 m3 0.40 2.80

Variable costs 220.80

Table 2. Fixed costs (US$) of cassava flour,Chinú, Colombia, January 1994.

Item Cost/month Cost/t

Managera 123.00 6.15

Production chiefb 12.00 0.60

Watchman 121.00 6.05

Maintenance 125.00 6.25

Other expenses 15.00 0.75

Fixed costs 396.00 19.80

a. The cost is shared by the chip and flour plants. b. Bonus for production.

carried out by small-scale, ruralproducers and inhabitants. It isimplemented in three phases, andpromotes cassava’s transformation inagroindustry by integrating functionsof production, processing, andcommercialization. It is supported bygovernmental and nongovernmentalorganizations.

Phase I of the Flour Project:Research (1985-1987)

Colombia’s economic situation, theprospects for cassava, and thenational potential for cassava-basedproducts were studied to select themost promising product and choosean appropriate site. The Atlanticcoastal region (northern Colombia)was also studied as having thegreatest potential for developing theproject. Aspects such as cassavaproduction, farmer organizations,and markets were taken intoaccount to choose the best site forthe pilot plant.

Aim

The objective of this phase was todetermine the economic and technicalconditions required for the project.

Activities

Studies were made of the cassavaproduction and marketing systems onColombia’s Atlantic coast. On-farm

la Investigación y el Desarrollo deTecnologías Apropiadas al Agro(FUNDIAGRO). The donor agency isthe International DevelopmentResearch Centre (IDRC), Canada.

Methodology Used in theIntegrated Cassava Project

The integrated cassava project is arural development strategy. It is

Table 3. Production costs (US$) of flour inChinú, Colombia, January 1994.

Item Cost/t

Variable costs 220.80

Fixed costs 19.80

Total production costs 240.60

272


trials were conducted withimproved cassava productiontechnology. Economic studieswere made of the wheat millingand baking industries. Theexperimental cassava flour plantwas designed and developed.Trials were made of equipmentand processing. Laboratory trialswere made on flour quality andconsumer acceptance.

Results

The results demonstrated thetechnical and economic feasibilityof producing cassava flour to

compete with wheat flour inColombia.

Phase II: Pilot Project(1988-1992)

A pilot plant was set up in Chinú,Department of Córdoba (Figure 1),with technical conditions forsemicommercial operation under realmarket conditions.

Aims

The major objective was to validatethe technology under real field

Figure 1. Site for the cassava-flour production pilot plant in northern Colombia. The pilot plant is partof phase II of a project to develop new, market-oriented, cassava-based products and theirmarkets.

PacificOcean

Ecuador

Brazil

Peru

473 km

BarranquillaAtlanticOcean

Cartagena

Sincelejo

Cali

120 km220 km

Chinú

Montería

80 km30 km

367 km

Medellín

258 km

Santafé de Bogotá

PanamaVenezuela

273


conditions, integrating production,processing, and marketing. Otherobjectives were to (1) gather reliabledata on production costs and on theinvestment needed to establish thistype of plant; (2) produce enoughcassava flour to promote its useamong consumers; and (3) use theplant as a display model to expandthis technology to other regions ofColombia.

Activities

The main activity was to establishthe pilot plant. Criteria for siteselection included aspects ofcassava production, landavailability, potential for increasingcassava yields, processing, rawmaterial availability (production,seasonality, access to fresh market),service infrastructure (water,electricity, roads), proximity toterminal markets, institutionalpresence and support, potentialproject impact, and socioeconomicimportance of cassava.

Alternative sites were surveyed,four potential zones selected, then asite, with farmer organizations closeby, was chosen. The pilot plant wasredesigned, in which combinednatural and artificial drying waseliminated. A designer and builderwere contracted and the redesignedplant built.

Results

The pilot plant began operatingwith adjustments in production,processing, and marketing. Aviable and functional model wasobtained.

Phase III: CommercialExpansion (1993 Onward)

A market study for the new productwas designed and developed, clients

were contacted, and test trialsconducted with them.Commercializing cassava flour in themeat processing and adhesiveindustries began.

At the time of writing, projectexpansion to other areas of Colombiahad not yet started, marketexpansion was still to come, togetherwith a further consolidation of thenew rural agroindustry.

Aim

The objective was to market cassavaflour and consolidate a ruralagroindustry that would benefitfarmers, not only in northernColombia, but also in other regions.

Activities

A marketing plan was designed andexecuted, and market segmentsselected. A bibliographical reviewwas made of cassava flour uses.Commercial contacts wereestablished and sales volume andconditions determined.

Results

Commercialization of cassava flourhas begun. The model has beenevaluated and adjusted and newsites selected. The project isexpanding to other zones.

A Cooperative Carries Outthe Project

The Cooperativa de Productores delos Algarrobos (COOPROALGA),based in Chinú, is a first-orderorganization with 43 members, allsmall-scale farmers dedicated togrowing cassava intercropped withmaize or yam. Most members payrent for land and the remaining 20%own it.

274


COOPROALGA manages twoplants, one producing cassava chipsfor animal feed, and the other thepilot cassava flour plant (Figure 2).

Characteristics of the Plant,Process, and Product

The flour plant

The cassava flour plant is awarehouse with an office, bathrooms,a tool room, a coal storage room, andareas where cassava roots arereceived, washed, chipped, and dried.The ground area of the plant is2,058 m2.

The plant has two water storagetanks, one underground with acapacity of 39 m3 and the otherelevated, holding 6 m3. All the plant’sresidual waters flow in twoindependent lines. The plant’s wallsare of concrete blocks, and the roofhas a metal framework and is tiledwith asbestos.

Construction of the plant had costUS$29,484.00 in March 1990. The

Universidad del Valle and CIATdesigned the main processingequipment, which was built in Cali.

Processing

A batch process was implementedand includes the followingoperations: harvest, transport,reception, weighing, selecting,preparing, washing, chipping,drying, premilling, and milling. Theresulting cassava flour is thenpackaged and stored (Table 4 andFigure 3).

Each batch is processed in2 days. On the first day, the rootsare harvested, transported, selected,and prepared. On the second day,they are washed, chipped, dried, andmilled, and the resulting flourstored.

The product

Before harvesting, the farmer prunesthe cassava plant, removing aerialparts, and on the next day heharvests and packs the roots, andtakes them to the plant.

Figure 2. The organization of the pilot cassava flour plant set up in Chinú, northern Colombia.

Second-orderorganization

Accountant Head of Marketing

COOPROALGAOther cooperatives

Treasurer

Manager

Production chief

Workers (4),Watchman (1)

Coordinator forpurchases

275


Table 4. Processing 1 t of cassava flour in a pilot plant, Chinú, northern Colombia.

Day Hours Activity Man-hours (no.) Workers (no.)

1 5:00 - 11:00 Harvest (3.5 t) - - 9:00 - 14:00 Root transportation - -9:00 - 14:00 Reception and weighing 2 1

14:00 - 18:00 Selection and preparation 20 5

2 7:00 - 11:00 Washing and chipping 8 2 7:00 - 11:00 Loading the drying chamber 4 16:00 - 7:00 Cleaning the burners 1 17:00 - 8:00 Drying starts 1 18:00 - 20:00 Drying (chip turning) 20 3

3 6:00 - 7:00 Cleaning and maintenance 2 26:00 - 7:00 Unloading the dryer 1 27:00 - 8:00 Milling and packaging 1 2

Total 60 6

Cassava roots are received in 50to 60 kg sacks, and should have beenharvested on the day of receipt. Theyshould also be free of diseases,deterioration, or severe mechanicaldamage, and should be from varietiescontaining high dry matter content.

After washing, chipping, drying,milling, and sieving through 150microscreens, cassava flour is finallyobtained.

Figure 3. Procedures in cassava flour processing at the pilot plant, Chinú, northern Colombia.(Dotted lines refer to secondary processes.)

276


Abstract

The potential of cassava flour todiversify markets for cassavaproducers is investigated. The effectsof different root processing regimes oncyanogen contents and microbiologicalcounts—major factors governingquality in cassava flour—wereinvestigated at CIAT. Chipping,rasping, and different dryingtechnologies were evaluated in termsof product quality. Three types ofchippers, five raspers, and drying bysun, oven, or bin were used. Raspingand drying reduced the cyanogenicglucoside contents of the roots by 90%to 100%, but microbiological countswere high for all drying technologies.The chipping trials indicate that sundrying on trays produced chips ofsimilar microbiological quality toartificial drying.

Introduction

Cassava is grown in many parts ofthe developing world, mainly bysmall-scale farmers, for both food andincome. Often such farmers have

CHAPTER 32

IMPROVING PROCESSING TECHNOLOGIES

FOR HIGH-QUALITY CASSAVA FLOUR

D. M. Jones*, D. S. Trim*, andC. C. Wheatley**

limited scope for other crops, becauseof harsh climate, poor soils, or both.Markets for fresh roots for directconsumption are stagnant ordiminishing in many places becauseof increasing urbanization andchanges in eating habits. Demand forroots for starch and chips for animalfeed, although existing where suchindustries operate, is limited. Cassavaflour is a product that could helpdiversify and strengthen cassavamarkets for these small producers.

The main industrial marketopportunities for cassava flour are inthe substitution of other rawmaterials, primarily wheat flour orstarches, for further processing intofinal products. In some areas, smallerregional markets exist for local,cassava-based food specialties. Topenetrate these markets, cassava flourmust be of at least comparable qualityto the product it is potentiallyreplacing. Possible clients are unlikelyto risk changing feed stocks if it is atall possible that the quality of theirend product will be adversely affected.

Factors of Flour Quality

Microbiological quality

Wheat flour tends to be of highmicrobiological quality, because theeconomic product (the grain) develops

* Natural Resources Institute (NRI), Kent, UK.** Centro Internacional de la Papa (CIP),


277

Improving Processing Technologies for High-Quality Cassava Flour

above the ground; is cultivated withmodern, large-scale, farming practices;and is harvested at relatively lowmoisture content. In contrast, cassavaroots are usually cultivated with basicfarming practices, picking up amicrobial load from the soil, and havea much higher moisture content thangrains. Hence, cassava flour is likelyto have higher levels of microbiologicalgrowth.

Table 1 gives selected flourstandards. The Colombian standardpermits the same maximum bacterialloads for both wheat and cassavaflours. Cassava flour has a lowermaximum permitted moisture thanwheat flour, because of its perceivedgreater susceptibility to contamination.

Cyanogens

Cassava flour also contains residuallevels of cyanogenic compounds(cyanogens), mostly cyanogenic

glucosides (CG), cyanohydrins, andhydrogen cyanide (HCN). Theglucosides initially present in thefresh roots are broken down, duringprocessing, to the other cyanogensgiven above (Bokanga, 1992).Cyanogen concentrations areexpressed as mg HCN equivalent perkg of dry matter, unless otherwisestated. Nonglucosidic cyanogen (NGC)concentration describes the combinedconcentrations of cyanohydrins andHCN. The cyanogen levels remainingvary with the raw materialconcentration and the processingtechnologies employed (Fish and Trim,1993). These levels are not a majorconcern for nonfood use.

Hydrogen cyanide is toxic, butis usually present only in smallquantities because of its volatility.Evidence suggests that cyanidepoisoning and intoxication resultingfrom consumption of cassava flourmay be caused by high residual

Table 1. Quality standards for selected flours.

Quality criterion Cassava flour Wheat flour

Colombiana Africanb Colombianc Tanzaniand

Chemical composition(maximum permitted levels):

Moisture (%) 120 130 1400

Starch (% minimum) 620

Ash (%) 20 30 0.700

Crude fiber (%) 2.50 20 200

Sand (%) 30 100

Crude cellulose (%) 50

Total HCN (mg/kg) 500

Microbial content (cfu/g):

Aflatoxins 00

Aerobic plate count at 35 °C 2 x 105 2 x 105 1 x 105

Coliform bacteria 1 x 102 1 x 102

Escherichia coli 00 00 00

Salmonella 00 00 00

Molds and yeasts 1 x 103 1 x 103 1 x 103

a. ICONTEC, 1990.b. FAO and WHO, 1992.c. ICONTEC, 1967.d. Tanzania Bureau of Standards, 1989.

278


cyanohydrin levels, which thendecompose after ingestion (Banea,1993; Mlingi et al., 1992). The effectof consuming CG on health is lessclear and has not yet beenthoroughly investigated.

Few official standards existspecifically for cassava chips andflour for human consumption. TheColombian standard for driedcassava sets a maximum totalcyanogen content of 50 mg/kg (freshbasis), measured as HCN (ICONTEC,1990). The regional standard beingdeveloped for Africa (FAO and WHO,1992) sets a maximum totalcyanogen content of 10 mg/kg (freshbasis). The standards are expectedto evolve with the product, andfurther guidance may be found inthe proceedings of the CassavaSafety Workshop held in 1994.

Research on ProcessingTechnologies

The quality of the cassava flourproduced at the CIAT pilot plant wasrigorously evaluated in terms ofresidual cyanogens andmicrobiological quality (results notshown). Research was then carriedout at CIAT, with the followingobjectives:

(1) To investigate the modification ofchip size as a means ofincreasing the elimination oftotal cyanogenic potential (CNP)during flour production.The degree of cyanogenelimination achieved by the pilotplant effectively sets themaximum initial cyanogenconcentration in the feed rootsacceptable by a plant of thistype. Increasing the eliminationof cyanogens withoutfundamentally changing theprocess would ensure that thecassava flour produced meets

the standard, and would increasethe range of varieties that theplant’s processing operations cansatisfactorily detoxify. Manuallypeeling the roots was notinvestigated at this stage.

(2) To investigate means of processinghigh cyanide varieties of cassavainto flour with safe residualcyanogen levels.High cyanogen varieties are moresuited to some agroecologicalzones, and are preferred to lowcyanogen varieties in someregions. The operations ofchipping and drying do noteliminate enough cyanogens toprocess high cyanogen varietiessatisfactorily.

Effect of Chip Size onResidual Cyanogens in

Bin-dried Chips

Methods

Trials were carried out with threedifferent chipping disks: thestandard disk (CIAT-designed); amodified version with reduced chipaperture to give thinner chips; and agrating disk designed by the Ecolenationale supérieure des industriesagricoles et alimentaires (ENSIA),France (Monroy-Rivera, 1990). Roots11 months old were harvested the daybefore the trial and stored outdoorsovernight (normal factory plantpractice). The roots were washed in adrum washer, which also effectivelydehusks the roots, and chipped. Thewet chips were bin-dried at 60 °C, atloading densities of 75 and 85 kg/m2

(Figure 1). Six samples were takenfrom both fresh and dried chips, andanalyzed with the modified Cookemethod (O’Brien et al., 1991).

Results

Table 2 gives the cyanogen contentsmeasured during these trials.

279


Fresh roots

Washer30 L/minfor 5 min

Figure 1. Procedures used in cassava-chipping trials.

Modified chippingdisc

Standardchipping disc

Gratingdisc

Bin dryer85 kg/m2

60 °CAir flow:

0.5 kg/s.m2

Bin dryer75 kg/m2

60 °CAir flow:

0.5 kg/s.m2

Standard disk. The total CNPin the fresh chips were reduced byabout 34% by chipping with thestandard disk, followed by dryingwithin 5 hours. This figure isconsistent with the results obtainedby using the same disk at the pilotplant, where proportionally greaterreductions in CNP were achievedwith longer drying times. This wasdespite processing a different varietyunder different climatic conditions.

Modified disk. The modifieddisk achieved a similar level ofreduction in CG, but with a loweroverall CNP reduction of 28%.

The modified chips had anaverage thickness of 4.3 mm,compared with 6.1 mm for thenormal chips, providing a greatercut-surface area. A greater initialelimination of the CG was thereforeexpected in the modified fresh chipsbecause of the higher percentage ofdamaged root tissue. Undersuitable conditions, a faster dryingrate was also expected, leading tosurface drying of the chips in a

shorter period, and earliertermination of cyanogenic reactions.

The low degree of eliminationobtained with the modified chipsindicates that the effect of fastdrying is masking any effect ofchipping, which would be moreobvious at the slower drying ratesobtained at higher loading densities.

Grating disk. The grating diskshowed a higher reduction (56%)than the standard chipper (34%) ofCNP, with chipping and drying.This is consistent with the greaterextent of tissue damage achieved.Reduction of CG with chipping anddrying was consistent at 59%-61%.The reduction in CG is dictated bythe quantity of glucosides broughtinto contact with linamaraseenzyme, which, in turn, depends onthe extent of tissue damage.Cyanogenic glucosides inundamaged tissue remain intact.The chips produced by the gratingdisk were more fragile than the pilotplant ones and less suitable for bindrying.

280

Ca

ssava

Flou

r an

d S

tarch

: Progress in

Resea

rch a

nd

Develop

men

t

Table 2. Cyanogen concentrationsa measured during cassava chipping trialsb.

Cyanogenic contents Standard disk at loading density Grating disk at loading density(kg/m2): (kg/m2):

85 75 85 75

CNP NGC CG CNP NGC CG CNP NGC CG CNP NGC CG CNP NGC CG

Fresh chips 1,269 253 1,016 1,096 172 924 832 152 680 858 320 539 1,118 198 920(mg HCN equiv./kg dry matter)

Dried chips 812 35 776 746 38 709 598 7 591 396 40 356 480 41 439(mg HCN equiv./kg dry matter)

Reduction with chipping (%) 20 16 18 37 18

Reduction with chipping and drying (%) 36 39 32 35 28 29 54 59 57 61

a. CNP = total cyanogen potential; NGC = nonglucosidic cyanogen content; CG = cyanogenic glucoside content.b. Each value is an average of six samples; percentage of reduction in both CNP and CG is based on fresh chips CNP; all trials used roots of M Ven 25, a high cyanogen variety.

Modified disk atloading density of

85 kg/m2

281


Summary

Elimination of CNP from chips madeby the standard disk increased withdrying time, regardless of cassavavariety or location.

The grating disk eliminated 22%more CNP than the pilot plant disk atthe same loading density. Gratedchips, however, are more fragile thanstandard chips and less suitable forbin drying.

Effect of Different Rasperson Residual CNP inTray-Dried Pulps

The effect of different raspers on thedegree of cyanogen eliminationachieved with rasping and drying wasinvestigated. Rasping almostcompletely destroys the root tissuestructure, much more so thanchipping. The trials used roots ofM Ven 25, a very high cyanogenvariety, to establish the upper limitsof cyanogen elimination.

Methods

Five different raspers were used:

(1) A conventional, wooden Jahnrasper, in which serrated bladesare mounted laterally on a woodendrum.

(2) A punched-drum rasper,consisting of a metal sheet withoutward facing jagged holes(punched through with a nail),fixed around a wooden drumframe.

(3) A pin rasper (experimental), ametal drum scored diagonally inboth directions across its lengthwith metal pins protruding about5 mm from the drum’s surface.

(4) An abrasion rasper (experimental),with a layer of carborundum,about 10 mm deep, fixed around adrum.

(5) A plastic, Jahn rasper(experimental), in which metalserrated blades are mountedlaterally on a plastic drum.

Four of the rasper drums testedwere interchangeable within thesame rasper frame, designed toinvestigate their relative starchextraction efficiency. The drumswere 400 mm in length and 270 mmin diameter. The plastic Jahn rasperdrum was a smaller, self-containedunit, 275 mm in length and 200 mmin diameter. An ordinary 5-HP motorwas used for all the raspers. Thewooden Jahn rasper and thepunched-drum rasper are in commonuse in the cassava starch industry.

Roots were harvested at9 months and stored as for thechipping trials. The roots werewashed in clean but untreated waterand dehusked manually. Fifteenkilograms of the washed roots wererasped without adding water. Theresulting pulp was mixed anddried at 8 kg/m2 on two traysin a despatch tray dryer at60 °C (Figure 2). Four samples eachof the fresh and dried pulps weretaken for evaluation of cyanogenconcentrations. This procedure wasfollowed for each rasper.

Results

Table 3 gives the cyanogenconcentrations measured during thistrial.

Cyanogen contents of raspedpulps. The reduction in CG withrasping only was variable, with boththe Jahn raspers reducing the CG by65%, and the punched drum by 43%.When the pulps were both raspedand dried, the CNPs were reducedby 94%-96% for all raspers,regardless of the degree of reductioneffected by rasping alone. Theresidual CNPs in the pulps ranged

282

Ca

ssava

Flou

r an

d S

tarch

: Progress in

Resea

rch a

nd

Develop

men

t

Table 3. Cyanogen concentrationsa during cassava-rasping trials, measured in mg CN equiv./kg dry matterb.

Rasper drum Fresh roots Fresh pulp Dried pulp Reduction with rasping and drying (%)

Type Feed CNP NGC CG CNP NGC CG CNP NGC CG CNP CG(kg/min)

Wooden Jahn 28.6 2,318 271 2,047 2,195 1,409 786 104 29 74 66 96 97Punched drum 32.6 2,417 243 2,175 2,267 968 1,299 152 88 64 46 94 97Abrasion 2.1 2,024 235 1,789 1,932 1,372 559 132 29 103 72 94 95Metal pin 15.0 2,608 315 2,293 2,236 881 1,355 163 27 137 48 94 95Plastic Jahn N/A 2,234 293 1,941 2,045 1,358 687 111 25 87 69 95 96

a. CNP = total cyanogenic potential; NGC = nonglucosidic cyanogen content; CG = cyanogenic glucoside content.b. Each value is an average of six samples of fresh roots and four samples of pulp; percentage of reduction in both CNP and CG is based on fresh root CNP; all trials used roots of

M Ven 25, a high cyanogen variety.

Reduction ofCG with rasping

(%)

283


Fresh roots

Manual washing

Metal pin rasper

Plastic Jahnrasper

Abrasionrasper

Punched-drumrasper

Wooden Jahnrasper

reduction in CNP achieved with theabrasion rasper was not significantlydifferent to that achieved with theother raspers. The pulp was also moreliquid and difficult to handle than theothers.

Summary

Except for the abrasion rasper, theraspers evaluated were suitable forprocessing roots with high cyanogencontents to flour with low cyanogencontent. The wooden Jahn and thepunched-drum raspers arecommercially available.

Effect of Different DryingTechniques on Residual

Cyanogens and theMicrobiological Quality of

Dried Pulps and Chips

The effects of different dryingtechniques (sun and artificial) on themicrobiological quality and on thecyanogen concentrations of chips andrasped pulps were evaluated.

Rasped pulp is not suitable for bindrying, and the effect of rasping on themicrobiological quality of the driedproduct is unknown. Because smalleroperations may not be able to justify

from 104 to 163 mg/kg. In previousmilling trials, the residual CNPconcentration in first-grade flour wasabout 36% of the level in freshly driedchips. Assuming this to hold true forpulps, the flours would have CNPsbetween 38 and 58 mg/kg, thusmostly meeting the Colombianstandard of 50 mg/kg.

This level of total cyanogenelimination probably approaches themaximum possible in practice, giventhat variations occur because offluctuating conditions. No significantdifferences in the overall elimination ofCNPs was found between the raspers.

Root throughput. The abrasionrasper’s root feed was 10% below thatof the Jahn or punched-drum raspers,thus making it unsuitable forcommercial flour production. Theshredding action employed by both theJahn and punched-drum raspersremoves a deeper layer of root tissuewith each contact than does theerosive action of the abrasion rasper,resulting in a larger root feed.

The pulp produced by theabrasion rasper was finer and morehomogenous than the other pulps,indicating a greater degree of tissuecomminution. However, the final

Figure 2. Procedures used in cassava-rasping trials.

Tray dryer8 kg/m2

60 °C

284


the investment and cost of artificialdrying, the effect of sun drying on themicrobiological quality of products wasalso evaluated.

Methods

Three trials were carried out on10-month-old roots of cassava varietyM Ven 25. The first two trials used thewooden Jahn and punched-drumraspers for root comminution. Theroots were washed and dry-rasped asin the rasping trials described above.The pulps were dried at loadingdensities of 5 and 10 kg/m2 on raisedtrays and on a concrete floor in the

sun, and in an oven at 60 °C. Thefinal trial was carried out with themodified chipper, with the chips driedat 5 kg/m2 in the same way (Figure 3).Chips were also bin-dried at 70 kg/m2

and 60 °C (Figure 4). The chips andpulps were mixed manually every2 h during drying. Three compositesamples of each dried product weretaken for microbiological analysis.The samples were analyzed for aerobicplate counts (APC) (35 °C), sporecounts (35 °C), and yeasts and moldsthe following day (ICMSF, 1978). Foursamples each of the fresh and driedpulps were also taken for cyanogenevaluation.

Figure 4. Procedures used in cassava chipping and drying trials.


Modifiedchipper

SunConcrete floor

5 kg/m2

SunRaised tray

5 kg/m2

Bin60 °C

70 kg/m2

Oven60 °C

5 kg/m2

Fresh roots


SunConcrete floor

10 kg/m2

Oven60 °C

5 kg/m2

SunRaised tray

5 kg/m2

Oven60 °C

10 kg/m2

SunRaised tray10 kg/m2

SunConcrete floor

5 kg/m2

Punched-drumrasper

Wooden Jahnrasper

Fresh roots

Figure 3. Procedures used in cassava rasping and drying trials.

285


drop with storage and may also bereduced by the heat generated bymilling to flour.

Microbiological quality of drypulps and chips. All of the dried,rasped, pulp samples had high APCs(108 cfu/g), as did the chips whichwere sun dried on a concrete floor.The oven-dried, and raised-tray,sun-dried chips were of acceptablequality (105 cfu/g), and the bin-driedchips had only slightly highercounts. The rasped pulps provide abetter substrate for microbial growththan the chips, as the cell contents(e.g., sugars and proteins) have allbeen released by rasping.

However, the APCs of fresh chipshave been measured at around

Results

Table 4 gives the cyanogenconcentrations measured during thetrials and Table 5, the microbiologicalcounts.

Cyanogenic contents of driedchips and pulps. Compared withthe rasping-only trials, the puncheddrum reduced CG (88%) more thanthe wooden Jahn rasper (53%).Rasping and drying reduced the CGby 90%-100%. Sun-dried pulpstended to have higher residual NGCthan did oven-dried pulps, possiblybecause of the higher rate of removalof HCN during forced-circulation ovendrying, which would increase the rateof breakdown of cyanohydrin to HCN.Residual cyanohydrin levels tend to

Table 4. Cyanogen concentrationsa during drying trials (rasped pulp only), measured in mg CN equiv./kgdry matterb.

Pulp sample from:

Wooden Jahn rasper Punched-drum rasper

CNP NGC CG Reduction with CNP NGC CG Reduction withrasping and rasping anddrying (%) drying (%)

CNP CG CNP CG

Fresh pulp 1,302 696 606 54 1,562 1,383 179 87

Dried pulp:

Oven, 5 kg/m2, 154 30 124 88 91 37 17 20 98 9960 °C

Oven, 10 kg/m2, 105 31 74 92 94 33 28 5 98 >9960 °C

Sun, 5 kg/m2, 99 47 52 92 96 53 45 7 97 >99raised tray

Sun, 10 kg/m2,raised tray 67 51 15 95 99 60 50 9 96 99

Sun, 5 kg/m2,concrete floor 84 77 8 94 99 68 61 7 96 >99

Sun, 10 kg/m2,concrete floor 85 82 3 95 >99 81 73 8 95 99

a. CNP = total cyanogenic potential; NGC = nonglucosidic cyanogen content; CG = cyanogenic glucoside content.b. Each value is an average of six samples of fresh roots and four samples of pulp; percentage of reduction in both

CNP and CG is based on fresh pulp CNP; all trials used roots of M Ven 25, a high cyanogen variety.

Cyanideconcentration

286


Table 5. Microbiological quality of dried pulp and chips generated in cassava-drying trials.

Microbiological counta

Aerobic Spore count Yeast and moldplate count at 35 °C count

at 35 °C

Wooden Jahn rasper:

Oven, 60 °C 5 3.43 x 108 3.18 x 105 4.99 x 104

Oven, 60 °C 10 3.58 x 108 2.61 x 105 5.40 x 104

Sun, raised tray 5 4.57 x 108 6.63 x 104 8.18 x 104


Sun, floor 5 5.64 x 108 9.20 x 104 1.03 x 105

Sun, floor 10 3.51 x 108 3.28 x 104 3.08 x 104

Punched-drum rasper:

Oven, 60 °C 5 1.01 x 108 8.02 x 104 6.33 x 103

Oven, 60 °C 10 2.12 x 108 3.74 x 104 2.65 x 104



Sun, floor 5 1.13 x 108 1.38 x 104 1.15 x 105

Sun, floor 10 5.93 x 108 1.87 x 104 7.13 x 104

Modified chipper:

Oven, 60 °C 5 5.45 x 105 4.50 x 102 2.67 x 102

Bin, 60 °C 70 2.04 x 106 8.67 x 102 7.33 x 102


Sun, floor 5 4.01 x 108 1.71 x 105 2.15 x 105

a. Counts expressed as colony forming units per gram (cfu/g), wet weight basis; average of three compositesamples.

Loading density(kg/m2)

Rasping and dryingmethod

105 cfu/g (Table 6). Previouspilot-plant experience has shownthat, with long drying times (22 h),the APCs of the chips are at108 cfu/g, but reducing the dryingtime to 10 h reduces the APCs to105 cfu/g. Faster drying of the pulpmay therefore offer a means ofreducing the counts. The shortestpulp drying time of 6 h wasinsufficient to affect the counts.

Raised-tray sun drying of chipsgave a product of goodmicrobiological quality with APCssimilar to those of oven-dried chips.Thus, this method may havepotential for reducing costs undersuitable climatic conditions (sitespecific).

Summary

Rasping and drying of cassava rootsis an effective means of reducing theCNP present in high cyanogen cassavavarieties. However, the greater degreeof root disintegration leads toincreased microbiological growth.

Conclusions

Processing with the grating diskreduced CNP by 22% more than thestandard disk. However, drying gratedchips at high loading densities may bedifficult.

Fast drying stopped theelimination of cyanogens early in the

287


Table 6. Microbiological quality of processed samples from pilot plant and CIAT trials, November 1991.

Sample Microbiological countsa

Aerobic plate Spore count Coliforms Fecal coli-formscount at 35 °C at 35 °C (MPNb) (MPNb)

CIAT:

Soil 7.7 x 107 8.1 x 105 >1.1 x 103 15

Root peelc 3.0 x 107 6.2 x 104 >1.1 x 103 <3

Parenchyma 1.2 x 103 1.5 x 102 <3 <3

Pilot plant:

Soil 6.8 x 107 5.7 x 107 >1.1 x 103 40

Root peelc 1.4 x 107 3.0 x 105 >1.1 x 103 7

Well water 4.6 x 103 8.3 x 101 <3 <3

Tank waterd 4.7 x 103 2.6 x 102 <3 <3

Fresh chips 4.9 x 105 2.8 x 103 1.1 x 103 500

a. Counts expressed as cfu/g (wet weight basis) for processed samples and as cfu/ml for water samples.b. MPN = most probable number.c. Root peel includes bark and peel.d. Tank water treated with 10-20 mg/L free chlorine.

SOURCE: D. S. Trim and P. Wareing, 1991, personal communication.

drying period of the modified-diskchips, masking any effect chip sizemight have had. Greater reductionin CNP is likely at higher loadingdensities.

Rasping and drying is aneffective means of processing evenvery high cyanogen roots to a flourthat meets the Colombian standard.Further work is needed to improvethe product’s microbiologicalquality.

In suitable climatic conditions,raised-tray sun drying of chips givesa product of good microbiologicalquality.

References

Banea, M. 1993. Cassava processing,dietary cyanide exposure and konzoin Zaire. Thesis for Master ofMedical Sciences degree.International Child Health Unit(ICH), Uppsala, Sweden. 65 p.

Bokanga, M. 1992. Mechanisms of theelimination of cyanogens from cassavaduring traditional processing. In: Westby,A. and Reilly, P. J. A. (eds.). Proceedingsof a Regional Workshop on TraditionalAfrican Foods - Quality and Nutrition,25-29 Nov. 1991, Dar es Salaam.International Foundation for Science(IFS), Uppsala, Sweden. p. 157-162.

FAO and WHO (Food and AgricultureOrganization of the United Nations andWorld Health Organization), CodexAlimentarius Commission. 1992. Codexstandard for edible cassava flour—African regional standard—CODEX STAN 176-1991. Eighth sessionof the Codex Committee on Cereals,Pulses and Legumes, CX/CPL 92/9,June, 1992. FAO/WHO Food StandardsProgram, Rome, Italy. 17 p.

Fish, D. M. and Trim, D. S. 1993. A review ofresearch into the drying of cassavachips. Trop. Sci. 33:191-208.

ICMSF (International Commission onMicrobiological Specifications forFoods). 1978. Microorganisms in foods,1: their significance and methods ofenumeration. 2nd ed. Academic Press,London, UK.

288


ICONTEC (Instituto Colombiano de NormasTécnicas). 1967. Harina de trigo parapanificación. In: Industriasalimentarias, 2nd rev., vol. 10.NTC 267. Bogotá, Colombia. p. 55-67.

__________. 1990. Yuca seca para consumohumano. In: Frutas, legumbres yhortalizas. NTC 2716. Bogotá,Colombia.

Mlingi, N. L. V.; Assey, V. D.; Poulter, N. H.;and Rosling, H. 1992. Cyanohydrinsfrom insuffiçiently processed cassavainduces ‘KONZO’, a newly identifiedparalytic disease in man. In: Westby,A. and Reilly, P. J. A. (eds.).Proceedings of a Regional Workshopon Traditional African Foods - Qualityand Nutrition, 25-29 Nov. 1991, Dares Salaam. International Foundationfor Science (IFS), Uppsala, Sweden.p. 163-169.

Monroy-Rivera, J. A. 1990. Eliminación decompuestos cianogénicos durante elsecado de yuca. Informe de lostrabajos realizados en el CIAT. Ecolenationale supérieure des industriesagricoles et alimentaires (ENSIA),Massy, France. 51 p.


Tanzania Bureau of Standards. 1989.Tanzania wheat flour specification.TZS 439:1989. Dar es Salaam,Tanzania.

289

Cassava Flour in Malawi: Processing, Quality, and Uses

CHAPTER 33

CASSAVA FLOUR IN MALAWI:PROCESSING, QUALITY, AND USES

J. D. Kalenga Saka*

Introduction

Cassava (Manihot esculenta Crantz) isa major root crop in the tropics, andits starchy roots are a significantsource of calories for more than500 million people worldwide (Cock,1985). In Malawi, cassava is thesecond most important staple aftermaize (DEPD, 1987): about 30% ofthe population depends on cassavafor calories (Sauti, 1982). The cropgrows easily in all parts of thecountry, but especially along theshores of Lake Malawi where it is themost important staple food. Since the1991/92 drought, which devastatedMalawi, the Government hasintensified the country’s production ofcassava, a drought-resistant crop, toguarantee food security.

Cassava is eaten in variousforms; these determine the methodsof processing, which aim to(1) provide products that arestorable and easy to transport tomarket; (2) improve the taste of finalproducts; (3) reduce potentialcassava toxicity; and (4) provideproducts such as flour forsubsequent conversion to a varietyof end products (Hahn, 1989;Lancaster et al., 1982). In Malawi,two methods are employed to makecassava flour, resulting in two kindsof flour: kondowole and ntandaza(Saka, n.d.; Williamson, 1975).

Abstract

The quality of flour processed fromcassava (Manihot esculenta Crantz) bytwo methods commonly used inMalawi was determined. The first,simple sun-drying, gives a flourknown as ntandaza; the other—soaking in water, followed by sundrying—provides kondowole flour.Processing affects both the nutritionalquality and cyanogen content of thefinal products. The soaking stepsignificantly reduces mineral andprotein contents and raises thecarbohydrate level (P > 0.05) to91.1% ± 1.1% for ntandaza flour and95.3% ± 0.7% for kondowole.

The soaking step, followed by sundrying, reduces the cyanogen contentmore than sun drying alone. Insoaking + sun drying, less than 10 mgHCN/kg dry wt were detected in thefinal products, representing a98.0% ± 1.6% reduction of initialcyanogen content. Simple sun dryingreduced total cyanogen content by82.9% ± 5.2%.

The uses of cassava flour inbakery, brewing, and making cassavasima are described.

* Chemistry Department, Chancellor College,University of Malawi, Zomba, Malawi.

290


Kondowole flour is prepared bysoaking peeled cassava roots for 2 to7 days; sun drying the soft mass(called maphumu) and pounding thedried mass to make the flour(Figure 1). This product is popularamong lakeshore populations living inKaronga District to as far south asNkhotakota District (Figure 2). Thesoaking of unpeeled roots is alsopracticed, but the flour gives productswhich taste bitter and appear darker.The flour is used in its pure form or ismixed with cereal flours (maize,sorghum, millet, wheat, or rice).

Among other things, the flour isused to make sima. Its preparationinvolves adding the flour tosimmering water and stirring thepaste to consistency. Both pure andcomposite flours are used for brewingsweet and alcoholic beverages. Whenmixed with wheat flour, the compositeflour is widely used to bake breads,scones, cakes, and biscuits. The purekondowole flour is also used inbaking. When mixed with cerealflour, the nutritional value of the

cassava-based products improves(Sauti et al., 1989).

Ntandaza flour is made by sundrying peeled and/or partially peeledroots for 1 week or several months.The roots may be dried whole, as cutpieces, or after pounding; the lastdries fastest. The dried product iscalled makaka and the resultant flouris commonly known as ntandaza.The flour is also referred to asntandasha and mtandasha,depending on the locality.

This method of processingcassava is predominant in central andsouthern Malawi (Figures 2 and 3). Avariation of the methodology involvesfirst covering the cassava roots withbanana leaves to induce moldformation. The moldy product is thensun dried to provide a darker andmoldy makaka (Van Drongelen,1992).

Although the ntandaza flour isused in the same way as kondowoleflour, its most important use is in

Figure 1. Processing kondowole flour from cassava roots, and its uses, Malawi.

Soft roots

Sima

Cassavaroots roots

PeeledPeel

Soft mass (maphumu)

Soak Sun drySqueeze or

pound

(unpeeled) Dried mass

Pounded mass

Pound

Sweet or alcoholicbeers

CookBake

BakeCook

Sima

(a cookedpaste)

beers

Whole roots orcut pieces

Soak

(1) Peel

(2) Sun dry(1) Sun dry (as balls or small pieces)

(2) Pound

bread, scones)Bakery (cakes,

Mix with maize, Composite Ferment Sweet oralcoholic

Ferment Kondowoleflour floursorghum, rice,

or millet flour

291


(2) Sun dry

Figure 2. Cassava-growing areas in Malawi ( = major growing areas; = scattered crops; = lake). 1 = Karonga District; 2 = Nkhotakota District; 3 = Zomba District; = town of

same name as district. (After Nyirenda.)

Cassavaroots wash

Peel and

(1) Sprinklewater

(2) Sun dry

Sweet oralcoholic beers

Figure 3. Processing ntandaza flour from cassava roots, and its uses, Malawi.

(1) Coverwithbananaleaves

Sun dry (1) Pound

(2) Sun dry

Dried mass (makaka)

Peeled roots (whole roots or cut pieces)

(1) Add millet, sorghum,or maize flour

(2) Mill or pound

FermentBake

Bake

Composite flour

Bakeryproducts

Ferment

Cook

Sima

Mix with sorghum,maize, or millet flour

Ntandaza flour(ntandasha,mtandasha)

Sima (a cooked paste)

Pound

Cook

Bake

Ferment

Compositeflour

Sweet oralcoholic beers

Tanzania

Zambia

Likoma and ChizumuluIslands (Mozambique)

Mozambique

Sima

292


brewing. The resulting beer isreported to be of superior quality.

Information on the quality ofcassava flour produced in Malawi waslimited until 1986, when our workbegan (Saka, n.d.). The processing ofcassava into various forms affects thenutritional value of the final products(Longe, 1980). The levels of totalcyanoglucosides, linamarin andlotaustralin are also affected duringprocessing (Lancaster et al., 1982).

Hydrolysis of cyanoglucosides byan endogenous enzyme, linamarase,liberates the highly toxic substance,hydrocyanic acid (HCN) viaacetocyanohydrin (de Bruijn, 1971).The presence of nonglucosidiccyanogens (NGC; acetocyanohydrinand HCN) limits cassava use (Nartey,1978). Cyanide has a lethal dose of0.5 to 3.5 mg HCN/kg of body weight.Although the reports of acute cyanideintoxication and death amongcassava-eating populations areinfrequent, ample evidence exists thatgoiter and cretinism (due to iodinedeficiency) are exacerbated, and thatdiseases such as tropic ataxicneuropathy and epidemic spasticparaparesis (konzo) are caused bylong-term ingestion of cyanide fromcassava (Rosling, 1987).

We studied the nutritional valueand cyanogen content of the twoMalawian cassava flours to ascertaintheir quality.

Material and Methods

Cassava samples

Tuberous roots were obtained fromthe Makoka Agricultural ResearchStation, Zomba, and from the “D. C.Munthali” Research Farm, BiologyDepartment, Chancellor College,Zomba. The roots analyzed fornutritional value were from

12-month-old plants, whereas thoseprocessed into kondowole andntandaza flours for cyanogendetermination varied from 20 to22 months in age.

Processing the flours

Kondowole. Four roots fromeach of three plants, totalling 12,from each of three varieties werepeeled, washed, and soaked indeionized water (volume not recorded)in plastic wash basins for 7 days.The resulting soft mass was washedwith clean water, broken down (byhand) into small pieces whileremoving floury material in theprocess, and left to dry on trays in thesun for 7 days. The dried productwas then ground in a blender andsieved.

The data (Tables 1 and 2)obtained for kondowole flour weretaken from 20 to 22-month-old plantsand the soft, soaked roots were madeinto balls and sun dried for 67 h.Samples of kondowole flour wereprovided by the Cassava CommodityTeam, Makoka Agricultural ResearchStation.

Ntandaza. Twelve roots wereselected as above, peeled, and eitherpounded or cut longitudinally andtransversely to produce chips. Thechips were sun-dried on trays and thedried material (makaka) wasprocessed into flour, using a blenderand sieve. The pounded roots weresun dried for 2 to 3 days (Table 1).

Chemical analysis

Analar grade chemicals and solventswere used. Fresh roots and cassavaflours were analyzed for moisture,ash, crude fiber, fat, crude protein,and minerals (Ca, P, Mg, and K),using standard procedures (Osborneand Voogt, 1978). The carbohydratecontent was calculated by difference.

293


Table 1. Cyanogen content of cassava flours (mg HCN/kg dry wt) produced in Malawi. Values are meansof samples, with SE in parentheses.

Flour type Moisture Cyanogens Total cyanogen (%) reduction

Total Non- Free (% of initial content)glucosidic

Kondowole (n = 21) 11.8 2.91 0.75 0.69 98.0

Ntandaza (n = 8):

Poundeda, 11.4 116.8 25.5 1.6 79.7sun dried (1.2) (6.2) (0.7) (0.2) (1.3)

Poundedb, 4.88 54.4 4.80 0.39 80.0sun dried (0.12) (2.5) (0.20) (0.06) (2.5)

Chipsc, 14.6 51.6 12.4 3.05 88.9sun dried (0.5) (3.1) (0.6) (0.20) (1.0)

a. ‘Nyambi’, a bitter variety, was peeled, pounded, and sun dried at 30 ± 1 °C for 48 h.b. ‘Gomani’, a bitter variety, was peeled, pounded and sun dried at 30 ± 1 °C for 72 h.c. ‘TMS 1230158 (OP)’, a bitter variety, was peeled, cut into chips and sun dried at 30 ± 1 °C for 72 h.

Table 2. Composition of cassava roots and products from our work and some literature sources (on drywt basis). Each value is the mean of 12 roots with ± SE.

Component Roots Ntandaza flour Kondowole flour(Malawi study)

Malawi Longe, Malawi Williamson, Longe,study 1980 study 1975 1980

Moisture (%) 55.9 ± 4.9 13.44 ± 2.66 11.80 10.77 ± 2.72 12.00 12.00Ash (%) 2.21 ± 0.45 2.15 ± 0.18 2.05 0.91 ± 0.30 1.79Crude fat (%) 1.23 ± 0.44 0.87 ± 0.33 0.46 0.70 ± 0.30 0.24Crude fiber (%) 2.29 ± 0.39 2.30 ± 0.70 1.62 ± 0.30Crude protein (%) 3.17 ± 0.62 3.39 ± 0.73 2.04 1.46 ± 0.30 1.70 1.51Carbohydrate (%) 91.1 ± 1.2 91.0 ± 1.1 90.30 95.3 ± 0.66 95.50 94.40P (mg/100 g) 82 ± 35 93 ± 27 40 ± 20Ca (mg/100 g) 54 ± 27 26 ± 12 17 ± 8 63.00Mg (mg/100 g) 40 ± 17 58 ± 16 32 ± 13K (mg/100 g) 768 ± 354 877 ± 358 330 ± 138

Cyanogen extraction and analysis

To 30 g of flour (60 g fresh roots) in ablender was added 0.1 M of chilledorthophosphoric acid (H3PO4)(200 cm3), with subsequent extractionaccording to Cooke’s (1978) method.The milky liquid was poured intocentrifuge tubes. Their weights wereadjusted until equal and the tubes

were then centrifuged at 8 x 103 g for10 minutes. The supernatant wascollected in sample bottles anddeep-frozen until analysis. The freshcassava was extracted in fourreplicates and the processed cassavain duplicates. For the assay of totalcyanogen content, samples wereprepared according to the acidhydrolysis method of Bradbury et al.

294


(1991). For NGC (cyanohydrin plusfree HCN) and free cyanide, theprocedure of O’Brien et al. (1991) wasused. In all cyanogen assays, asodium isonicotinate-sodiumdimethylbarbiturate coloring reagentwas used (Saka, 1992).

The moisture contents of freshand processed cassava weredetermined gravimetrically after ovendrying three replicate, 10-g-samplealiquots at 110 ± 5 °C for 16 h.


Table 2 presents the mean chemicaldata for kondowole and ntandazaflours and the literature data forcassava flours similarly processed.The results show that, despite certainsimilarities, the chemicalcompositions of the two flours weresignificantly different at P = 0.05.

At 1% level, neither the fat valuesnor the Ca content were significantlydifferent. The chemical data in Table2 reveal that, compared with freshroots, the two cassava flours areequally important sources ofcarbohydrates, but with generallylower values in protein, fat, and fiber.Their mean nutritional valuescompare well with published data(Longe, 1980) but higher fat levelswere obtained by Saka (n.d.). Thepresent data fill several gaps and alsoconfirm the limited availableinformation on Malawi cassava flour(Williamson, 1975).

Comparison of the chemicalcomposition of fresh roots (Saka, n.d.)and the two cassava flours (Table 2)indicates that sun drying alone, andsoaking in water followed by sundrying, affect the nutritional value ofcassava. Simple sun drying producedntandaza flour, whose dry matter, fat,Ca, and Mg levels were significantlydifferent (at P = 0.05) from those of

fresh roots. Whereas the dry matterand Mg contents were increased, thefat and Ca levels were decreased.

Soaking and subsequent sundrying of cassava provided kondowoleflour, whose composition wassignificantly different (at both P =0.05 and 0.01) from that of freshroots. During this process, thecarbohydrate content becamesignificantly higher while the rest ofthe analyzed constituents decreased.These were lost as dissolved materialduring soaking. These findings areconsistent with those reported byLonge (1980).

Table 1 provides the levels oftotal, nonglucosidic, and freecyanogens of kondowole andntandaza flours and presents thepercentage reductions in totalcyanogen content. The results showthat the method used to preparekondowole flour (involving asubmerged fermentation stage) wasmore efficient in reducing totalcyanogen content than that employedfor ntandaza flour. The production ofkondowole resulted in 98.0% ± 1.6%loss in the total cyanogen contentwhile an 82.9% ± 5.2% reduction wasachieved during the processing ofntandaza flour.

Mahungu et al. (1987) also noteda 99% reduction in cyanogen contentwith methods that involve soakingroots in water. Saka (1992) recentlyeliminated 70% to 80% of totalcyanogen content by sun drying1-cm3 cassava chips for 48 h. Theresidual, total cyanogen content ofkondowole flour was 2.91 ± 1.44 mgHCN/kg dry wt and of ntandaza flour,51.6 ± 3.1 to 116.8 ± 6.2. Thus, thentandaza flour contained muchhigher residual cyanogen contentthan did the kondowole. The finalcyanogen content depends onwhether the variety contains low(“sweet”) or high (“bitter”) cyanogen.

295


The less bitter, or sweet, varietieshave lower residual cyanogencontent when sun dried (Saka, n.d.).

The composition of the threeforms of cyanogens indicates thatfree HCN is a major component ofthe NGC in kondowole flour. Incontrast, in ntandaza, cyanohydrinis the major component.Cyanoglucosides also predominatein ntandaza flour.

High levels of acetocyanohydrinin sun-dried chips have also beenobserved by others (Mlingi et al.,1992). Consumption of this type ofcassava appears to lead to highthiocyanate levels in human urine(Mlingi et al., 1992). Plans arecurrently under way to develop orupgrade methods for reducing totalresidual cyanogen and cyanohydrinto levels comparable with those inkondowole flour.

Conclusions

Cassava and its flours are majorsources of carbohydrates, but havelow values in protein, fat, andminerals. The protein content couldbe improved by fortifying with cerealand legume grains. The use ofcassava flour in Malawi remainsrestricted to cooking sima, baking,and brewing. Diversifying andpromoting cassava flour use isdesirable.

Soaking and subsequent sundrying of cassava roots greatlyreduce the high cyanogen levels tolow, safe values for humanconsumption. This methodincreased the carbohydrate contentof the cassava, but other nutrientswere reduced considerably. Simplesun drying was less effective inreducing total cyanogens, especiallywhen the initial cyanogen content

was high. The final products mayremain potentially toxic for humanconsumption. Pounding of freshcassava and its subsequent sundrying seem to offer better prospectsin achieving low cyanogen content.

Acknowledgments

I wish to thank the Research andPublications Committee, University ofMalawi; R. F. N. Sauti, Team Leader,Cassava Commodity Team; theMalawian Ministry of Agriculture; theInternational Foundation for Science,Sweden, for funding; and Mrs. L. C.Saka for typing the manuscript.

References

Bradbury, J. H.; Egan, S. V.; and Lynch,M. J. 1991. Analysis of cyanide incassava using acid hydrolysis ofcyanogenic glucosides. J. Sci. FoodAgric. 55:277-290.

Cock, J. H. 1985. Cassava: new potential fora neglected crop. InternationalAgricultural Development Service(IADS) development-orientedliterature series. Westview Press,Boulder, CO. 191 p.

Cooke, R. D. 1978. An enzymatic assay forthe total cyanide content ofcassava (Manihot esculenta Crantz).J. Sci. Food Agric. 29:345-352.

de Bruijn, G. H. 1971. Etude du caractèrecyanogenetique du manioc. Papers.Wageningen Agricultural University,Wageningen, the Netherlands.140 p.

DEPD (Department of Economic Planningand Development). 1987. Agricultureand animal husbandry. In: Republicof Malawi Statement of DevelopmentPolicies 1987-1996. GovernmentPrinter, Zomba, Malawi. 22 p.

Hahn, S. K. 1989. An overview of Africantraditional cassava processing andutilization. Outlook Agric.18(3):110-118.

296


Lancaster, P. A.; Ingram, J. S.; Lim, M. Y.;and Coursey, D. G. 1982.Traditional cassava-based foods:survey of processing techniques.Econ. Bot. 38:12-45.

Longe, O. G. 1980. Effect of processing onthe chemical composition and energyvalue of cassava. Nutr. Rep. Int.21(6):819-828.

Mahungu, N. M.; Yamaguchi, V.; Almazan,A. H.; and Hahn, S. K. 1987.Reduction of cyanide duringprocessing of cassava into sometraditional African foods. J. FoodAgric. 1:11-15.

Mlingi, N. L. V.; Assey, V. D.; Poulter, N. H.;and Rosling, H. 1992. Cyanohydrinsfrom insufficiently processed cassavainduces ‘konzo’, a newly identifiedparalytic disease in man. In: Westby,A. and Reilly, P. J. A. (eds.).Proceedings of a Regional Workshopon Traditional African Foods - Qualityand Nutrition, 25-29 Nov. 1991, Dares Salaam. International Foundationfor Science (IFS), Uppsala, Sweden.p. 163-169.

Nartey, F. 1978. Manihot esculenta (cassava):cyanogenesis, ultrastructure andseed germination. MunksgaardInternational Pubs., Copenhagen,Denmark. 262 p.


Osborne, D. R. and Voogt, P. 1978. Theanalysis of nutrients in foods.Academic Press, London, UK. 251 p.

Rosling, H. 1987. Cassava toxicity and foodsecurity. Tryok Kontakt Pubs.,Uppsala, Sweden. 40 p.

Saka, J. D. K. 1992. Determination ofcyanogen content of cassava (Manihotesculenta Crantz), using sodiumisonicotinate-sodiumdimethylbarbiturate. Paper presentedat the Fifth International ChemistryConference in Africa, 27-31 July,University of Botswana.

__________. n.d. Nutritional value andhydrocyanic acid content of Malawicassava (Manihot esculenta Crantz)and cassava flour. Malawi J. Sci.Technol. (In press.)

Sauti, R. F. N. 1982. Country report: Malawi.In: Root crops in East Africa:proceedings of a workshop held atKigali, Rwanda, 23-27 Nov. 1980.International Development ResearchCentre (IDRC), Ottawa, Canada.p. 104-106, 122-128.

__________; Saka, J. D. K.; and Kumsiya,E. G. 1989. The composition andnutritive value of cassava-maizecomposite flour. In: Alvarez, M. N.and Hahn, S. K. (eds.). Proceedings ofthe Third Eastern and SouthernAfrica Regional Workshop Root andTuber Crops, 7-11 Dec., 1988,Mzuzu. International Institute ofTropical Agriculture (IITA), Ibadan,Nigeria. p. 71-75.

Van Drongelen, A. 1992. Reasons for choicesin cassava processing, the case ofMulanje. Wageningen AgriculturalUniversity, Wageningen, theNetherlands. 67 p.

Williamson, J. 1975. Manihot esculentaCrantz: useful plants of Malawi.University of Malawi, Zomba, Malawi.p. 155-157.

SESSION 6:

NEW PRODUCTS

299

The Potential for New Cassava Products in Brazil

Introduction

Cassava is an important crop inBrazil, with an annual production of22-25 million tons. Productionsystems, processing methods, and thedegree of technology employed varybetween the four major cassavaregions (Amazônia, Northeast,Central South, South), accordingto agroecological location andsocioeconomic conditions. Farinha, atoasted flour, comprises the principalmarket, accounting for 70%-80% ofcassava production, but price anddemand fluctuate greatly.

Price fluctuations influence thearea of land cultivated, adoption ofnew technology for production, andincome of cassava producers, mainlysmall-scale farmers. Diversificationwould help stabilize prices of bothcassava flour and fresh roots.Establishing new markets for cassavaand its products would enhance thevalue of cassava cultivation andestablish important links betweensmall-scale agriculture and expandingmarkets.

Given the various intermediateproducts of cassava (e.g., chips,flours, and starch); the array ofcurrent applications in human andanimal nutrition and in industry;and the numerous traditionalcassava preparations, it is possibleto visualize a broad range ofcassava-based markets. The keylies with new technologies and thedevelopment of novel products to fitcurrent and potential markets.

To innovate products withinthe existing matrix of traditionaland new products and theirrespective markets, the followingfactors should be considered:

(1) The evolution of a successfulstarch sector such as that ofFrance during the last20 years;

(2) Current trends in Brazil towardproduct diversification; and

(3) Proposed strategies for theshort and medium terms.

The Evolution of theStarch Sector in France

The modern French starchindustry, based on maize or potatostarch, provides a relevant exampleof an industry evolving in search ofnew products and markets in bothfood and nonfood sectors.

CHAPTER 34

THE POTENTIAL FOR NEW CASSAVA

PRODUCTS IN BRAZIL1

G. Chuzel*, N. Zakhia**, and M. P. Cereda***

* CIRAD/SAR, stationed at the Faculdade deCiências Agronômicas (FCA), UniversidadeEstadual Paulista (UNESP), São Paulo,Brazil.

** CIRAD/SAR, Montpellier, France.*** UNESP/FCA.


300


(1) An almost five-fold increase instarch utilization. The annualgrowth rate in the last decade ofstarch utilization in the EuropeanCommunity (now the EuropeanUnion) remained above 3.8%;

(2) A steadily increasing quantity ofstarch (from 52% to 58%) isdestined for nonfood uses;

(3) A profound change in usermarkets: increased use by paperindustries, and pure chemistryand pharmaceutical sectors, witha reduced use in the textileindustry.

To confront these developments,the starch industry has had to adaptcompletely its product range, creatingnew products and seeking newapplications. The industry haslearned how best to add value to,adapt, or modify the functional andphysicochemical properties of starches(e.g., viscosity, capacities for binding,thickening, adhesion, flocculation,and dispersion). A matrix of productversus market, and new versustraditional can be observed:

Before the 19th century

Only wheat starch was produced inFrance, principally for starchingfabrics, powdering wigs, and gluingpapyrus or paper, that is, exclusivelynonfood uses. The convergence ofglucose production, and that of beetsugar, with the industrial revolution ofthe 19th century, transformed thissmall-scale activity into a largeindustry, providing a wide range ofraw materials suited to a considerablybroadened range of applications. Thediscovery of dextrins in the 1830s,then of linters in the 1890s, and, mostsignificantly, modified starches in the1940s gave rise to the industry oftoday.

1960s to 1980s

The approach adopted by thedeveloping starch sector was “new andtraditional products for new andtraditional markets.” For example, themarket for starch products in Francefrom 1954 to 1987 (Figure 1) wascharacterized by:

1234123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123

123123123123 1234

123123

123123123123123 123

123412341234

250,000

200,000

150,000

100,000

50,000

0

Sta

rch

(t)

Foo

d in

du

stry

Pap

er

Cor

ruga

ted p

aper

Tex

tile

s

Adh

esiv

es

Ph

arm

aceu

tica

lsan

d c

hem

icals

Fou

ndry

Oth

ers

Markets

Figure 1. Starch markets in France, 1954-1987 ( = 1954;

121212 = 1987). Nonfood uses in 1954 comprised

52% of starch production, which totalled 145,000 t; in 1987, 58% of 710,000 t.Note: “Others” include such markets as drilling muds, flocculation agents, building materials,and mining.

301


Traditional products New products

Traditional markets Native and modified starches: Cationic starch:Food PaperPaperTextiles Borated dextrins:

Adhesives

New markets Native starch: Isoglucose:Corrugated paper, ceiling Beveragestiles, wall panels

CM starch:Pregelatinized starch: Pharmaceuticals

Flocculation agentsLipophilic starches:

Crosslinked, stabilized starches: Beverage emulsions,Frozen and microwave foods encapsulation

Organic acids, AA, enzymes

Approach adopted by the French starch industry during the 1960s to 1980s.

From the above matrix, thefollowing points are worth noting:

(1) Traditional markets for newproducts: The development ofcationic starches with increasedretention capacity hasconsiderably strengthenedexisting markets in the paperindustry.

(2) Traditional products in newmarkets: The food marketingsector has largely evolved duringthe last few years, opening newmarkets for such modifiedstarches as:

(a) Pregelatinized starches,cold-soluble starches;reticulated starches (morestable under cookingconditions, in thepreparation ofready-to-use foods);

(b) Oxidized starches, resistantto retrogradation, for frozenproducts;

(c) Reticulated and stabilizedstarches which preventundesirable effects associatedwith certain modes of cooking(e.g., heating by microwavecauses phase separations,varying degrees of swelling,breakage of the crust, andnonuniformity of flavors andaromas).

(3) New markets for new products:The development of isoglucose hasopened up large markets, especiallyin the U.S. drinks industry.Likewise, fermentation techniquesusing starch as a substratehave opened up chemical,pharmaceutical, and other markets,providing a wealth of derivedproducts.

For the next decade

The starch industry is now following asimilar approach to strengthen anddiversify its markets for the next decade:

302



Traditional markets Native and modified starches: Fat and sweetner substitutes:Food Paper (retention rate)Paper Low calorie foodsTextiles

New markets Starch, pregelatinized starch: PHB/V, Polylactic acid:Biodegradable plastics Biopolymers

Carboxylic starches, surfactants: BioconversionsThermoplastic starchesDetergents Cyclodextrins

Approach adopted by the French starch industry for the next decade.

Responding to future demands

This approach enables the starchsector to respond to emerging demandsfrom existing or potential users such asnutritional considerations, qualityrequirements, or environmentalconcerns.

The Cassava Industry in Brazil

Traditionally, cassava is consumed asfresh roots, or processed into farinha,polvilho azedo, or starch for food,paper, and textile industries. Althoughfarinha remains the principal marketfor cassava, the Brazilian cassavaindustry has taken a series ofinitiatives to diversify markets:

(1) New markets for traditionalproducts: A new market forpolvilho azedo (sour starch), anaturally fermented starch withbread-making properties, isdeveloping urban fast-food outlets;and for farinha in mining.

(2) New products for traditionalmarkets: In particular, the foodindustry is increasing its use ofnative or modified cassavastarches, such as cationic starchand maltodextrins. Frozen cassavachips is another new product.

These new products andapplications depend on previouslywell-identified target markets: theBrazilian cassava industries stillhave not moved toward newproducts for new markets. This ishighly risky in terms of researchand development, whethergenerating new technologies oridentifying new markets andmarketing strategies, particularlyas these industries lack thenecessary human and financialresources.

The French starch sector, todevelop as described above, devotesmore than 2% of its turnover toresearch and development—impossible to imagine in thecurrent Brazilian context.

In an attempt to overcome thislack of resources, the Centro RaizesTropicais (CERAT), together withthe Universidade Estadual Paulista(UNESP), brought together about50 researchers from severalBrazilian research institutions anddevelopment-support institutions,to help the industrial sector followthe “new products for new markets”approach, along the lines of thefollowing matrix:

303



Traditional markets Quality and new cassava varieties: High-fiber biscuitsFresh consumption

Fat substitutes:Farinha Meat products, ice-creams

Polvilho azedo Cyclodextrins

Starch:Food, paper, textiles

New markets Farinha, native starch: Glucose syrups:Grits substitutes in Food industrybeer brewing

Polvilho azedo: Maltose syrups (MFT):Premixes for food industry Brewing, acid-fermented

drinks, polysaccharides,packaging, uses ofbyproducts

Brazilian cassava industry: present and potential future products and markets.

The research initiated underproject EU-STD3 (value-addedproducts, byproducts, and wasteproducts of small and medium-scalecassava primary processing industriesin Latin America) falls within the scopeof this initiative, particularly in termsof new products for new markets.

Some of these research effortswill transfer to the industrial world ofsecondary processing (e.g., use offarinha or cassava starch in beerbrewing as a substitute for maize grits,formulation of sour starch-basedpremixes for production of, forexample, pão de queijo). The businesscommunity has already expressedinterest—an indication of the relevanceof the approach adopted.

Bibliography

Ansart, M. 1990. Le poids et la diversité desdébouchés industriels de l’amidon.Industrie Agro-alimentaire, juin1990. p. 541-545.

Leygue, J. P. 1992. Les utilisationsnon-alimentaires des céréales: quatredébouchés porteurs. Perspect. Agric.167:40-54.

Light, J. 1990. Modified food starches: why,what, where and how. Cereal FoodsWorld 35(11):1081-1092.

Swinnels, J. J. M. 1990. Industrial starchchemistry: properties, modificationsand applications of starches. AVEBEno. 05.00.02.006EF.

304


CHAPTER 35

EXTRUSION PROCESSING OF CASSAVA:FORMULATION OF SNACKS

N. Badrie and W. A. Mellowes*

Introduction

Cassava (Manihot esculenta Crantz) isgrown mainly in tropical developingcountries where it is a primary sourceof carbohydrates for millions of people(Coursey, 1978; Nestel, 1973). Theroots do not store well after harvestand usually begin to deterioratewithin 2 to 4 days (Odigboh, 1983).Processing helps solve the storageproblem (Sammy, 1971) andincreases the usefulness ofcassava.

Snack foods now comprise animportant part of the daily nutrientand calorie intake of manyconsumers. They can be sweet orsavory, light or substantial, and mayeven be endowed with attributes suchas “healthy” or “just for fun”(Tettweiler, 1991). Among WestIndians, spicy snacks are especiallypopular.

Extrusion processing is one of thefastest growing, and most important,food-processing operations of recentyears (Harper, 1981a; Paton andSpratt, 1984). The food industry hasinvested considerable research in theextrusion processing of a wide rangeof foodstuffs, developing manysuccessful products (Linko et al.,1981), including snacks, baby foods,cereals and starches, and/orvegetable proteins (Harper, 1981b).

Abstract

Acceptable snack-type extrudateswere produced, using flour fromcassava (Manihot esculenta Crantz) asthe main ingredient. Various formulasof cassava flour blended with otheringredients were tested. Extrusionprocessing was carried out, using alaboratory extruder (Wenger X-5,single-screw) under constantconditions, where feed moisture was11%, barrel temperature 120-125 °C,screw speed 520 rpm, and feed rate250 g/min. Sensory attributes ofcolor, flavor, and texture, and overallacceptability were rated by panelistson a 5-point scoring system. Analysisof variance indicated significantdifferences (P < 0.01) for sensoryattributes and for formulas. Flavorscored the highest, reflecting thepresence of popular spices in theblends. Formula F4 received thehighest scores for flavor and color andfor acceptability. All formulas wereacceptable, except for F7 and F8,which contained yeast. Color wasmost attractive when 0.1% turmericwas added.

* Food Technology Unit, Department ofChemical Engineering, Faculty ofEngineering, University of the West Indies,St. Augustine, Trinidad, West Indies.

305

Extrusion Processing of Cassava: Formulation of Snacks

Cooked, extruded snacks aregenerally prepared from cereals suchas de-germinated maize meal, andrice and wheat flour (Smith, 1976).

In Trinidad and Tobago, maizemeal is a major imported ingredientfor extruded foods such as ready-to-eat snacks. Limited work has beendone on the extrusion of cassava,resulting in the absence of cassavaextrudates on the local market. Theobjectives of this research were to(1) use cassava flour as the mainingredient for a snack product undersuitable processing conditions, and(2) determine, by sensory evaluation,the acceptability of extrudates ofvarious formulas.


Feed ingredients

Cassava roots of the local variety‘Maracas Blackstick’ were processedinto flour (Figure 1) within 48 hoursof harvesting. The flour was thenblended with small amounts ofadditional ingredients to yield avariety of formulas. These ingredientswere powdered spices, such as onion(0.2%, 0.5% w/w), garlic (0.2%,0.5%), chili (0.2%, 0.5%), turmeric(0.1%, 0.2%, 0.5%, and 1.0%), andpaprika (0.2%, 0.5%); sucrose (0.5%);uniodized salt (1.0%, 1.5%);monosodium glutamate (MSG, 1.0%),dried skimmed milk (0.5%); soybeanoil (4.0%); yeast (1.0%, 1.5%), anddefatted soybean flour (5.0%, 10.0%).

The feed sample of each formulawas left to equilibrate for 24 h andadjusted to the targeted feed moistureof 11% d.b. The samples were againleft to equilibrate at 4 °C for 24 h and,before extrusion, were allowed toreach ambient temperature. Aftercooking, the extrudates were packedin high-density polyethylene (HDPE)bags and stored at 4 °C in sealed

Cassava roots

Wash(remove surface dirt)

Hand-peel(immersing roots in water)

Slice(1-3 mm thick with a2 to 5-cm diameter)

Dry(at 60 °C for 3.5 to 4 h)

Grind

Sieve (particle size rangesfrom 0.25 to 0.84 mm)

Cassava flour8.5%-9.0% d.b.

Pack (in HDPE bags)

Store (in plasticcontainers with lids)

Blend ingredients(allow to equilibrate)

Adjust feed moisture(to 11% d.b.)

Extrusion

Pack extrudates(in HDPE bags)

Store (at 4 °C in plasticcontainers with lids)

Sensory evaluation

Figure 1. Procedures for extruding cassava flourblends. (HDPE = high-densitypolyethylene.)

plastic containers. Extrudates werepresented to panelists for sensoryevaluation within 2 days of extrusion.

A single-screw laboratoryextruder, with a 2.5-cm diameter, wasused (Wenger X-5, WengerManufacturing Company, Sabetha,Kansas). The screw was of decreasing

306


pitch with a compression ratio(channel depth in feed zone tochannel depth in metering zone) of2:1 and length to diameter (L/D) ratioof 15:1. The die diameter was5.0 mm and land length 9.0 mm. Thedie plate was attached to a breakerplate, 6.0 mm thick. The extruderconsisted of eight stainless steel,jacketed head sections. Each sectionin the barrel was uniformly suppliedwith steam generated from aSussman Electric Boiler (Hot-ShotModel, MB-6, Automatic SteamCorporation, NY). Barrel temperaturewas monitored by thermocouplesmounted inside the barrel, using atemperature recorder (Type BD41,Kipp and Zonen, Holland). A feedhopper with paddle agitator ensureduniform feed flow into the extruderbarrel.

Extrusion conditions

Badrie and Mellowes (1991b) hadalready established suitable extrusionconditions for cassava flour: blendsshould be extruded at constantconditions of feed moisture 11% d.b.,barrel temperature 120-125 °C, screwspeed 520 rpm, and feed rate of250 g/min.

Proximate analysis

The proximate composition of cassavaflour and the crude protein (%) ofdefatted soybean flour weredetermined by AOAC (1965)procedures, except for crude fiber(AACC, 1983). Total carbohydratewas determined by difference.Amylose was estimated with a rapidcolorimetric method (Williams et al.,1970).

Sensory evaluation

Extrudates were evaluated by 10panelists, who were students andfaculty staff of the University of theWest Indies. They were widely

experienced in the sensory evaluationof food products. They rated thesensory attributes of color, flavor andtexture, and overall acceptabilityaccording to a scale where1 = unacceptable; 2 = poor;3 = acceptable; 4 = good; and5 = excellent. In addition, commentswere required.

Extrudates of uniform size wereserved in sealed polyethylene bags,randomly coded by three digits. Twosamples were presented per sessionand water was provided for rinsingbetween samples. Scores assigned toeach quality attribute and to theformulas were subjected to analysis ofvariance to determine any significantdifferences. Sensory means wereseparated by Tukey’s test (Larmond,1977). Sensory evaluation wasconducted at the University’s FoodTechnology Laboratory, between10:00 a.m. and 11:00 a.m.


Proximate composition

The proximate composition of cassavaflour was crude protein, 1.5%-1.6%;crude fat, 0.6%-0.7%; crude fiber,1.7%-1.8%; ash, 1.5%-1.7%; totalcarbohydrate, 85.2%-86.2%; andstarch amylose, 16.4%. Crudeprotein of defatted soybean flourranged from 52.1% to 52.2%. When5% or 10% defatted soybean flour wasadded to cassava flour, the crudeprotein content rose from1.38% ± 0.02% to 5.20% ± 0.03 or to7.49% ± 0.05%, respectively.

Low protein (< 3%) staples suchas cassava do not provide adequateprotein for human requirements, evenwhen ingestion exceeds caloricrequirements. In contrast, diets withcereals (8%-10% protein) can meetadult protein requirements (Cheftel etal., 1985). Soya protein, rich in

307


amino acid lysine (Harper, 1981b),can be used to fortify cassava flour.

Soybean flour was added to thecassava flour blend to increase theprotein content, improve quality, andincrease the yellow color of theextrudate. Badrie and Mellowes(1992b) found that soybean flourmakes extrudates more attractive andyellower, resulting in a change ofMunsell color notation from 4.62Y6.38/1.75 to 5.04Y 6.46/2.19 at 5%soybean flour or to 5.30Y 6.46/3.10at 10%. Thermal processing of foodcan increase the potential forinteraction between lipids, proteins,carbohydrates, and their breakdownproducts (Bruechert et al., 1988).Maillard browning appeared the mostlikely reason for the color change.

Badrie and Mellowes (1992b) alsofound, however, that adding soybeanflour reduced extrudate expansionand increased bulk density.Extrudate expansion was negativelycorrelated with crude protein(P < 0.01, r = -0.88).

Establishing processing conditions

Sensory attributes of extrudatesdepend on extrusion conditions andfeed material. Badrie and Mellowes(1991b) established suitableprocessing conditions, evaluated onthe bases of extruder performanceand the physical and chemicalproperties of extrudates.

Optimal expansion (2.82)occurred at a feed moisture of11% d.b.—the minimum moisturenecessary to obtain a flow of theextrudate through the die (at120-125 °C, screw speed 520 rpm,and feed rate 250 g/min). Lower feedmoisture either blocked the rotationof the screw (there was no transitionfrom the original floury nature to the‘melted’ state typical of extrusion) orthe extrudates emerged from the die

in bursts. Only at 11% feed moisturewas a more uniform moisturedistribution and, thus, a more elasticdough achieved, resulting in a smoothsurface texture. Foods with a lowermoisture content also tend to be moreviscous, the greater pressuredifferential resulting in better puffing.

The optimal expansion of cassavaflour extrudate can be related to itsmicrostructure. Scanning electronmicroscopy on cassava flourextrudates (Badrie and Mellowes,1991a) at 11% feed moisture revealedwide porous air cells with thin cellwalls. Extrudate expansion waspositively correlated (P < 0.05,r = 0.80) to the water solubility index(WSI). At 11% moisture, the lowesttexture values were recorded.Low-moisture extrusion, according toHarper (1989), can cause moremechanical damage (shear stress) tothe feed, resulting in a softer texture.At 11% moisture, a more intense andattractive color (4.62Y 6.38/1.75) wasalso obtained.

Extrusion was stable between 100and 125 °C, producing uniform,puffed products. Temperatureincreases from 100-105 °C to120-125 °C brought correspondingincreases in extrudate expansion. Athigher temperatures (130-155 °C),extrudates became increasinglyirregular, degenerating to rapidlyejected fragments. Temperaturesabove 125 °C probably resulted in aweakened structure and led to arougher extrudate surface texture.

Establishing formulas

Because spicy snacks are a particularfavorite of West Indians, powderedflavorings of onion, garlic, chili,paprika, and turmeric were includedin the blends. Turmeric, a majoringredient of curry powder,also lent a more appealing yellowcolor. Other flavor enhancers were

308


sucrose, salt, and monosodiumglutamate. Dried skimmed milkprovided both protein and flavor.Soybean oil was added at 4% level—alevel at which Badrie and Mellowes(1992b) showed that lowest bulkdensity and highest extrudateexpansion resulted, linked toincreases in the WSI and totalreducing sugars. Hsieh et al. (1990),working with maize meal extrudateproduced by a twin-screw extruder,reported that adding salt and sugarenhanced radial and axial expansionbut reduced bulk density andbreaking strength.

For the first formula (F1), sensoryscores for all parameters were betterthan acceptable, that is, higher than3 (Table 1). But panelists’ commentsrevealed that the color was unevenlydistributed and too yellow.Extrudates were also too spicy andtoo salty, with a distinct taste ofturmeric. Expansion was acceptable,but texture was slightly hard and theextrudate overly dense.

For the second formula, F2,adjustments were made to F1: thelevel of turmeric was reduced from1% to 0.5%, and salt was reducedfrom 1.5% to 1.0%. Panelists againfound the extrudates hard, too spicy,too yellow, and tasting of turmeric,although all sensory scores wereacceptable (Table 1).

For formula F3, turmeric andpaprika were reduced from 0.5% to0.2%. A significantly better flavorand more acceptable color resulted(Table 1), but extrudates of F4 (0.1%turmeric and 1.5% spices) gained thehighest overall acceptability, scoringhighest in both color and flavor.

For F5, spice content was furtherreduced to 0.9%, but panelists foundthe extrudates too bland, and theflavor rating dropped from 4.25 (forF4) to 3.83.

To improve texture of theextrudates, the percentage of defattedsoyflour was increased from 5% to

Table 1. Sensory attribute scoring of cassava flour blend extrudates.†

Formula Color Flavor Texture Overall Overall meanacceptability‡ for formula

F1 3.18 bcdef 3.20 ef 3.42 abcd 3.24 bcdef 3.26 bcdef

F2 3.23 bcde 3.27 e 3.43 abc 3.30 bcde 3.37 bcde

F3 3.50 abc 3.92 ab 3.45 abc 3.59 ab 3.62 ab

F4 3.69 a 4.25 a 3.40 abcd 3.82 a 3.79 a

F5 3.58 ab 3.83 bcd 3.42 abcd 3.57 abc 3.60 abc

F6 3.48 abcd 3.81 bc 3.50 abc 3.55 abcd 3.59 abcd

F7 2.85 efg 2.91 efg 2.10 e 2.60 g 2.62 g

F8 2.95 efg 3.07 efg 2.67 e 2.88 efg 2.89 efg

Overall mean 3.31 a 3.53 ab 3.17 abc 3.32 bcof attribute§

† Columns: scores followed by the same letter are not significantly different among formulas at P < 0.05.Rows: values of overall mean attribute followed by the same letter are not significantly different amongattributes P < 0.05.LSD of formulas = 0.43; LSD of attribute = 0.25.

‡ On a scale where 1 = unacceptable, 3 = acceptable, and 5 = excellent.§ Mean of 10 replications.

309


10% (F6), thereby reducing the totalcarbohydrate level. The result was areduced extrudate expansion and anincreased bulk density, but nosignificant change in texture. Addingcassava starch to cassava flour (i.e.,increasing the total carbohydrate)tended to increase all texturalattributes, the extrudate becomingless elastic or springy (Badrie andMellowes, 1992b).

Badrie and Mellowes (1992b)showed that when soybean flour wasadded to cassava flour, thepercentage of noncarbohydratecomponents (in particular, crudeprotein) increased. Extrudateexpansion was negatively correlatedwith crude protein (r = -0.88,P < 0.01). Bulk density wasnegatively correlated (r = -0.96,P < 0.05) to extrudate expansion andpositively correlated to crude protein(r = 0.89).

To reduce the bulk density of theextrudates, instant dry yeast wasincorporated in the blend at 1.0% (F7)and 1.5% (F8). However, theseadditions resulted in unacceptablesensory scores. The effectiveness ofsodium bicarbonate (baking powder)or maize amylose on reducingextrudate bulk density will beassessed in a later study. Althoughamylose tends to provide surfaceregularity and lightness, cassavastarch or flour has too low an amylosecontent (16.4%) and produces denser,less radially expanded extrudates(Badrie and Mellowes, 1992a).

Sensory scores

Analysis of variance indicatedsignificant differences at both 5% and1% level for sensory attributes andformulas (Table 2).

Flavor received the highest overallmean score (3.53, i.e., better than

Table 2. Analysis of variance of sensory scoresfor cassava flour blend extrudates.

Source of df MS Fvariation

Attributes 3 0.18 5.63**

Formulas 7 0.65 20.31**

Error 21 0.032

Total 31

** Significant at P < 0.01.

acceptable), with texture (3.17)registering the lowest, although alsobetter than acceptable. Only F7 andF8 proved unacceptable (Table 1).Panelists commented on the uniqueflavor of cassava extrudates.However, they tended to rate texturelower because of their tendency tocompare cassava extrudates with thepopular maize extrudates. Cerealshave excellent expansive propertiesand are well suited to thermalextrusion.

As Stanley (1986) observed,texture is the major obstacle inremodelling ingredients intoacceptable foods. To produceproducts in the highly acceptable orexcellent categories, panelistsrecommended lightening texture andreducing bulk density and surfaceirregularity of extrudates.

The F4 overall rating of 3.79 wassignificantly (P < 0.05) higher thanother formulas.

Conclusions

Distinctive and acceptable extrudatescan be produced, using cassava flouras the main ingredient. The studypointed to texture, bulk density, andsurface regularity as areas requiringattention. Flavor achieved thehighest overall rating, attributable tothe 1.5% spice in most formulas.

310


Except for F7 and F8, to which dryyeast had been added, all formulaswere acceptable. Formula F4emerged as the best overall product,having scored the highest for flavorand color. Color was found to bemost attractive when 0.1% turmericwas added.

Other formulas, especially thoseincorporating local ingredients, canbe tried. Successful development ofextruded cassava products for thesnack food industry in the WestIndies could give rise to competitionwith popular, establishedmaize-based products, the meal forwhich must be imported. Extrusionis a rapidly growing food-processingoperation and extruded spicy snacksare popular in the West Indies, butmore trials and consumer-typesensory evaluations are necessarybefore cassava-based extruded snackscan enter the local market.

References

AACC (American Association of CerealChemists). 1983. Approved methods,vol. 1. 8th ed. St. Paul, MN, USA.

AOAC (Association of Official AgriculturalChemists). 1965. Official methodsof analysis. 10th ed. Washington, DC,USA.

Badrie, N. and Mellowes, W. A. 1991a.Texture and microstructure ofcassava (Manihot esculenta Crantz)flour extrudate. J. Food Sci.56(5):1319-1322, 1364.

__________ and __________. 1991b. Effect ofextrusion variables on cassavaextrudates. J. Food Sci.56(5):1334-1337.

__________ and __________. 1992a. Cassavastarch or amylose effects oncharacteristics of cassava (Manihotesculenta Crantz) flour extrudate.J. Food Sci. 57(1):103-107.

__________ and __________. 1992b. Soybeanflour/oil and wheat bran effects oncharacteristics of cassava (Manihotesculenta Crantz) flour extrudate.J. Food Sci. 57(1):108-111.

Bruechert, L. J.; Zhang, Y.; Huang, T. C.;Hartman, T. G.; Rosen, R. T.; andHo, C. T. 1988. Contribution of lipidsto volatiles generation in extrudedcorn-based model systems. J. FoodSci. 53(5):1444-1447.

Cheftel, J.; Cuq, J. L.; and Lorent, D. 1985.Amino acids, peptides and proteins.In: Fennema, O. R. (ed.). Introductionto food chemistry. Marcel Dekker,New York, USA. p. 245-369.

Coursey, D. G. 1978. Cassava: a major foodcrop of the tropics. Paperpresented at a workshop on cyanidemetabolism by the EuropeanOrganization at Canterbury, UK,14-18 August 1978.

Harper, J. M. 1981a. Extrusion of foods,vol. 1. CRC Press, Boca Raton, FL,USA.

__________. 1981b. Extrusion of foods, vol. 2.CRC Press, Boca Raton, FL, USA.

_________. 1989. Food extruders and theirapplications. In: Mercier, C.; Linko,P.; and Harper, J. M. (eds.). Extrusioncooking. American Association ofCereal Chemists, St. Paul, MN, USA.p. 1-16.

Hsieh, F.; Peng, I. C.; and Huff, H. E. 1990.Effects of salt, sugar and screw speedon processing and product variable ofcorn meal. J. Food Sci. 55(1):224-231.

Larmond, E. 1977. Laboratory methods forsensory evaluation of foods.Department of Agriculture, Ottawa,ON, Canada.

Linko, P.; Colonna, P.; and Mercier, C. 1981.High temperature - short timeextrusion cooking. In: Pomeranz, Y.(ed.). Advances in cereal science andtechnology, vol. 4. AmericanAssociation of Cereal Chemists, St.Paul, MN, USA. p. 145-235.

311


Nestel, B. 1973. Current utilization andfuture potential for cassava. In:Nestel, B. and MacIntyre, R. (eds.).Chronic cassava toxicity, vol. 1.Proceedings of an inter-disciplinaryworkshop held in London.International Development ResearchCentre (IDRC), Ottawa, ON, Canada.p. 11-26.

Odigboh, E. U. 1983. Cassava production,processing and utilization. In: Chan,Jr., H. T. (ed.). Handbook of tropicalfoods. Marcel Dekker, New York, USA.p. 145-200.

Paton, D. and Spratt, W. A. 1984.Component interactions in theextrusion cooking process: influenceof process conditions on thefunctional viscosity of the wheat floursystem. J. Food Sci. 49(5):1380-1385.

Sammy, G. M. 1971. Some problems in theestablishment of fruit and vegetableprocessing in Trinidad and Tobago.In: Sammy, G. M. (ed.). Postgraduateseminar on food technology, session2. Food Technology Series No. 5.Faculty of Engineering, University ofthe West Indies. p. 1-16.

Smith, O. B. 1976. Extrusion cooking. In:Altschul, A. M. (ed.). New proteinfoods, vol. 2(B). Academic Press, NewYork, USA. p. 86-120.

Stanley, P. W. 1986. Chemical and structuraldeterminants of texture offabricated foods. Food Tech.40(3):65-68, 76.

Tettweiler, P. 1991. Snack foods worldwide.Food Tech. 45(2):58-62.

Williams, P. C. Z.; Kuzina, F. D.; and Hlynka,I. 1970. A rapid colorimetricprocedure for estimating the amylosecontent of starches and flours. CerealChem. 47(4):411-420.

312


CHAPTER 36

THAI CASSAVA FLOUR AND STARCH

INDUSTRIES FOR FOOD USES:RESEARCH AND DEVELOPMENT1

Saipin Maneepun*

Introduction

Cassava has become a major cashcrop for Thailand: in 1993, productionwas about 21 million tons, andincreasing (TDRI, 1992). About 90%of total production is exported, mainlyto Europe. Although cassava is mostlyprocessed into pellets and chips foranimal feed, the volume of theseproducts has decreased slightly duringthe last decade in favor of cassavaflour and starch for domestic industryand export (Table 1).

In 1990/1991, although 96cassava starch processing factorieswere registered, only 55 wereoperating. Total production capacitywas 1.5-1.8 million tons, of which50% was for export, 25% forfood-processing, and 25% for nonfoodindustries.

In 1991, the 55 cassava starchfactories could be classified into 46starch factories and 9 modified-starchfactories. Monosodium glutamate(MSG) processing ranks highestamong those cassava starchprocessing industries manufacturingamino acids (Table 2; Rodsri, 1993).

Recently, lysine has become availableand is used as a nutrient in feedmills. At present, the ThaiGovernment is planning to supportresearch that would help diversify theuse of cassava starch in variousindustries to prevent falling prices.Cassava starch industries onlyproduce 50%-70% of their totalcapacity, and thus have potential forfurther development.

Production andDevelopment of Cassava

Flour

Two kinds of cassava flour-processingfactories operate in eastern Thailand:traditional and modern.

Traditional factories

Flour-processing factories were firstbuilt in the early history of cassavaproduction in Thailand. The model ofoperation—a family business thatemploys now-obsolete technology—isstill common in some parts ofThailand. Cassava roots are firstcrushed, then soaked in water. Theresulting starch is extracted, sundried, and pulverized into flour,which, however, is inferior in qualityand bulk. It is used for making suchproducts as noodles, desserts, andsago, most of which are sold on localmarkets.

* Institute of Food Research and ProductDevelopment (IFRPD), Kasetsart University,Bangkok, Thailand.


313

Thai Cassava Flour and Starch Industries for Food Uses:...

Table 1. Export of cassava sago, chips, pellets, and starch from Thailand, 1982-1991, in metric tons.

Year Sago Chips Pellets Starch Total prod.(x 1000)

1982 2,397 523,059 6,892,786 396,754 17,7881983 2,948 279,913 4,554,332 359,298 18,9891984 5,831 137,808 5,975,136 449,183 19,9851985 7,566 123,702 6,474,503 482,309 19,2631986 5,243 35,699 5,842,468 435,154 15,2551987 7,420 72,833 5,777,137 353,594 19,5541988 6,663 312,460 7,334,446 452,199 22,3071989 9,223 130,201 9,185,466 501,329 24,2641990 8,447 210,814 7,316,368 531,365 20,7011991 10,060 113,205 6,269,458 549,022 19,705

SOURCE: Rodsri, 1993.

(1) Cassava roots are weighed andmeasured for their starch content.

(2) The roots are precleaned by a soilseparator, then passed through acleaning machine and peeler.Peeling makes extracting starcheasier.

(3) The roots are then crushed andliquid starch is extracted, leaving acake, which is sun dried beforebeing used as a supplementaryfeed or for producing cassava chipsand pellets.

(4) The liquid starch is purified bypassing it through a sulfur vaporto rid the starch of sap.

(5) Water is then filtered out and thestarch dried mechanically. It isthen packed and shipped tomarkets.

On the average, 1 kg of fresh rootsyields 200 g of starch and between40 and 90 g of cake. Cassava flourquality depends on the manufacturingprocess. If the process is efficient andclean, the flour will be of high quality.Quality is judged by granule size; flourcolor, smell, and purity; fiber and ashcontents of flour; humidity, acidity,and viscosity of liquid starch; andcake. At present, a cassava breedingprogram has been established todevelop varieties that have high starchcontent: 22%-24%, depending on thegrowing season.

Table 2. Use of cassava starch in Thailand.

Uses Percentage oftotal use

Direct consumption 26Industry:

Monosodium glutamate 19Glucose syrup 15Paper 9Food 7Sago 6Textiles 2Timber 1

Other 15

Total 100

SOURCE: Rodsri, 1993.

Modern factories

Many flour-processing factories havebeen modernized since World War II toproduce high-quality cassava flour forexport and use in domestic industry.The role of cassava thus changedfrom being a local crop for domesticconsumption to a major commercialcrop for export. At present, modernprocessing equipment is beingdeveloped for export to neighboringcountries such as China, Vietnam,and Indonesia. The capacity ofsuch equipment ranges from 50 to100 t/day (Figure 1). In modernprocessing:

314


Fresh cassava rootsWasher

Root sieve

Inclined chain conveyor

Inclinedbelt

conveyor

Root chopper

Rasper

Separator

Screw pressPressed pulp

ExtractorExtractor

Flashdryer

Tapioca starchproduct

Centrifuge

Drying cyclone

Drying cyclone

Horizontalbelt conveyor

Cooling cyclone

Sifter andbagging

Screwfeeder

Standard Specifications forStarch for Export and

Local Markets

The Ministry of Commerce hasestablished specifications forinspecting and controlling cassavaflour and starch for export.

Certified products must beinspected according to Ministrystandards and/or importerspecifications. “Cassavaflour/starch” is defined as that“obtained from cassava root [as] awhite or cream-colored powder thatdoes not include modified starch.”Cassava flour and starch are gradedat three levels, each with its ownspecifications (Table 3; Ministry ofCommerce, 1993).

Standards for cassava flour andstarch used locally were establishedin 1974 by the Thai IndustrialStandard, Ministry of Industry:“Flour/starch obtained from cassavaroots (Manihot utilissima) has starchgranules of microscopic appearance,consisting of a cluster of two to eightgranules, each granule measuring5 to 35 µm and having an averagediameter of 15 µm. Most starchgranules are oval or truncated at oneend to form a kettledrum shape whilethe other end has a cutting edge withthe inner surface concave orirregularly flat. Starch granulesclearly show an eccentric hilum withsegment lines.” The quality of localflour and starch is classified intothree grades (Table 4; Ministry ofIndustry, 1978).

Figure 1. Processing tapioca starch, Thailand. (From Thai Tapioca Starch Industries Trade Association,1976, personal communication.

315


Table 3. Export standards for cassava flour and starch at three grades, Thailand, 1993.

Characteristics Supreme grade First grade Second grade

Starch (day basis) (%) 85 83 80Moisture content (%) 13 14 14Ash (%) 0.2 0.3 0.5Fiber (cm3/50 g flour) <0.2 <0.5 <1.0pH 4.5-7.0 3.5-7.0 3.0-7.0Particle size (mesh hole = 150 µm) (%) 99 97 95

SOURCE: Ministry of Commerce, 1993.

Table 4. Local standards for cassava flour and starch at three grades, Thailand, 1978.

Characteristics Grade 1 Grade 2 Grade 3

Starch (dry basis) (%) 97.5 96.0 94.0Moisture content (%) 13 14 14Ash (%) 0.15 0.30 0.50Ash (acid insoluble) (%) 0.05 0.10 0.15Protein (%) 0.3 0.3 0.3Fiber (cm3/50 g flour) 0.2 0.5 1.0pH 4.5-7.0 3.5-7.0 3.0-7.0Particle size (mesh hole = 150 µm) (%) 1 3 5

SOURCE: Ministry of Industry, 1978.

Both bulk and retail packers canstamp the appropriate grade mark onpackets for consumers’ selection.Retail packs are for home cooking.Information on using composite flour(that includes cassava flour) infoodstuffs is readily available at localbookshops.

Various food products from rice,bean, and wheat flours can improvetheir texture by substituting withcassava flour or starch.

Because cassava flour and starchare mostly processed with water, theycontain no hydrocyanic acid. TheThai Standard for Cassava/FlourStarch Committee does not accept thecodex standard for edible cassavaflour acceptable to African countries.The African Regional Standardpermits cassava flour to contain10 mg/kg of hydrocyanic acid(FAO and WHO, 1992).

Cassava Flour and Starch inLocal Food Products

Studies on incorporating cassavaflour into bakery goods for the localmarket require research on the eatinghabits of the population. Becauseconsumers prefer wheat-basedproducts, industries using cassavaflour have developed products madefrom mixtures of cassava and wheatflours. Such composite flours imparta unique taste and texture to theproducts.

Sponge cake made with cassavaflour

The effect of composite flour on thequality of sponge cake has beenstudied by Saencharoenrat(1990). He tested four kinds of wheatflour (chlorinated cake flour = CCF,unchlorinated cake flour, all-purposeflour, and bread flour) with different

316


levels of cassava flour: 0%, 20%, 40%,60%, 80%, and 100%. He found that,when the level of cassava floursubstitution was increased, proteincontent, ash content, and damagedstarch content decreased. Changes inmoisture content and pH, however,depended on the kind of wheat flour.The gelatinization temperature ofcomposite flours was in the samerange as that of the type of wheat flourused, whereas peak viscosity in gelformation increased. Waterabsorption, dough stability, resistanceto extension, and extensibility alsodecreased as the mixing toleranceindex increased. Sponge cakes madewith these composite flours were thenevaluated.

Results showed that the differentkinds of wheat flour and the levelsof substitution affected severalcharacteristics of cakes. Increasedlevels of substitution increased batterviscosity and specific volume of cake,but decreased specific gravity of batterand palatability of cakes. The kind ofwheat flour used affected ease ofcutting the cake (bread flour scoredthe highest), pH (lowest with CCF),and palatability (highest with CCF).Palatability scores agreed with totalcake scores. The ideal composite flourwas 40% cassava flour and 60% CCF,that is, no significant differences werefound (at P = 0.05) in palatabilitybetween cakes made from compositeflour and cakes from 100% wheatflour.

Cakes stored at room temperaturecould be kept for only 2 days, whereasrefrigerated cakes lasted at least7 days. Ease of cutting increased withstorage time, and moisture contentand palatability decreased.

Cookies made with cassava flour

Chananithithum (1986) found that themaximum substitution of cassavaflour for wheat flour in baking cookies

was 40%. Results showed that thespread was high, compared with thatof wheat flour. Most tasters found thecomposite-flour cookies to bepalatable.

Spread in composite-flour cookiescan be markedly reduced by usingan emulsifying agent. Patco-3(50% sodium stearic lactylate and 50%calcium stearic lactylate) and BV-15(commercial cookie improver), used at0.5% (flour basis), produced a cookiewith a spread factor not significantlydifferent from that of cookies madewith commercial cookie flour. Nor wereits organoleptic properties significantlydifferent from those of the commercialcookie.

The storage life of composite-flourcookies was not significantly differentfrom that of wheat-flour cookies.Composite-flour cookies could keep anacceptable texture for about 3 monthswhen stored in polythene bags,rigid plastic containers (polyvinylchloride [PVC]), or tin boxes. The40%-composite flour can reduce cookieprices by almost 2%, compared withwheat flour. The production ofcomposite-flour cookies has beenscaled up, using cassava flour as a rawmaterial.

Where cookies were made with15% full-fat soybean flour, cassavaflour could substitute wheat flour by asmuch as 50% (Boonyasirikool et al.,1987). The product still has more than10% protein. Palatability tests showedno significant difference between thesetwo types of cookies. Using vanilla andcocoa flavors can enhance productquality and make it highly acceptable.

Chemically Modified CassavaStarch for Use in the Food

Industry

Cassava starch is a typical root starchand is used in the production of

317


Various modified cassava starches havebeen developed and promoted for use inthe food industry.

Cassava starch phosphate

Industries manufacturing transparentnoodles, sauces, and custards havebeen encouraged to use cassava starchphosphate as a replacement for mungbean starch (Maneepun and Sirirojana,1990) and as a thickening agent insauces (Sirirojana, 1987) and custardspreads (Parvet, 1988).

Niyomvit et al. (1990) studied theuse of premixed cassava starchphosphate and native cassava starch

foodstuffs and adhesives. Direct use ofnative cassava starch is more frequentin home cooking than in industry.Root starch granules, when cooked,swell more and are more fragile (i.e.,they break down easily and thin outduring stirring) than are cereal starchgranules. The viscosity of these starchpastes can be determined by using aBrabender viscoamylograph. Afterstirring, tapioca starch shows thelowest viscosity (Figures 2 and 3).When certain chemicals areintroduced, they cross-link within thegranule, tighten up the molecularnetwork, restrict granule swelling, andso stabilize the viscosity of starchpastes against breakdown by agitation.

Vis

cosi

ty (B

U)

65 95 95 50 50

Temperature (°C)

1

4

3

2

Figure 2. Amylograph of cassava starch, and cassava starch crosslinked with sodium tripolyphosphateand sodium sulfate at different times. 1 = cassava starch; 2 = cassava starch crosslinked withphosphate for 2 h; 3 = for 4 h; 4 = for 6 h. (From Sirirojana, 1987.)

800

600

400

200

0

1,0000 10 20 30 40 50 60 70 80 90 100 110

Vis

cosi

ty (B

U)

Time (minutes)

35 50 65 80 95 95 95 95 80 65 50 35

Figure 3. Amylograms of starches. = acetylated cassava starch (6%); = wheat (9%); = maize = (8%); = potato (4%). (From Saencharoenrat, 1990.)

Temperature (°C)

..

30 60 90 120 150800

700

600

500

400

300

200

100

0

Time (minutes)

318


for the traditional Thai dessert, “kanomchen.” Their experiments showed thatthe cassava starches changed theirviscosity after cross-linking with thephosphate (Figure 2). Products wereprepared for tasting panels, whodetermined that the premixed cassavastarch phosphates, of 15% (2 h), 30%(4 h), or 15% (6 h), were highlyacceptable. The characteristicsrequired were transparency, easilyseparating layers, and stable texture.The Instron Food Tester showed thatthe texture was 60%-70% more stablethan that of unmodified, mixedcassava starch.

Acetylated cassava starch(A starch)

Most countries permit the use of Astarch as a direct food ingredient. Thedegree of substitution (DS) of starchacetate is determined by hydrolyzingwith excess sodium hydroxide. Thedigestion of sodium hydroxide isalmost equivalent to acetyl content(Institute of Food Research andProduct Development [IFRPD],unpublished data). Amylogramsshowed that the peak viscosity andstability of acetylated cassava starchare comparable with those of starchesfrom potato, maize, and wheat(Figure 3). The acetyl substitutionlowered the rate of retrogradation ofcooked pastes.

When cassava substituted30%-50% of starches from mung bean,potato, or sweetpotato in themanufacture of jelly bean sticks, theresulting product was smooth, and hadgood texture, gloss, and flexibility.

Acetylated and slightlycross-linked cassava starch(A/C starch)

This is called acetylated di-starchphosphate and is usuallycross-linked with phosphate. The DSof acetate can be determined by

hydrolyzing a hot-water-soluble,acetylated di-starch phosphatewith excess sodium hydroxide (IFRPD,1993, unpublished data). Thedigestion of sodium hydroxide is almostequivalent to acetyl content.

Amylograms showed that the peakviscosity and stability of viscosity werehigher in A/C starch than in potato,maize, or wheat starch (Figure 4).Swelling power is also greater in A/Cstarch than in wheat or maize starchbut lower than in potato starch(Figure 5). The A/C starch is stableunder conditions of freeze-thaw, highcold storage, and acidity. It has a shorttexture and high transparency. Theproduct used depends on the viscosityof final products and is recommendedfor use in sauces made with vinegar orfruit juice. Because A/C starch has agood affinity with raw meat and is hardto retrograde, it can be used for boiledfish paste, deep-fried fish paste, fishham, sausage ham, sausages, andmeat balls.

Both A and A/C starches can beused for making wheat noodles(oriental type), resulting in smoothnoodles whose texture when cooked isstable during cold storage.

Cassava starch ether (hydroxypropyl)

The cassava starch ether is used as afood ingredient in most countries.The degree of substitution ofhydroxypropyl in starch ether isascertained by hydrolyzing with hotsulfuric acid to propionaldehyde,which is then measured with aspectrophotometer after complexingwith ninhydrin (IFRPD, 1993,unpublished data). Amylogramsshowed that cassava starch ether hasa high peak viscosity and goodstability, which means it can reduceretrogradation of cooked paste(Figure 6). This starch ether is beingtried out in sauce-making (acidconditions).

319


Figure 6. Amylogram of cassava starch ether (hydroxypropyl). (From IFRPD Laboratory, 1993-1994,personal communication.)

0 10 20 30 40 50 60 70 80 90

Vis

cosi

ty (B

U)

Temperature (°C)

1,000

800

600

200

0

400

Time (minutes)

800

600

400

200

0

1,0000 10 20 30 40 50 60 70 80 90 100 110

Vis

cosi

ty (B

U)

Time (minutes)

35 50 65 80 95 95 95 95 80 65 50 35

Temperature (°C)

50 60 70 80 90

30

20

10

0

Sw

ellin

g pow

er (t

imes

)

Heating temperature (°C)

..Figure 5. Swelling power of starches. = acetylated and slightly crosslinked cassava starch;

= potato; = wheat; = maize. (From Saencharoenrat, 1990.)

One gram of starch and 49 g of water were mixed in a centrifuge tube. The tube was thenheated at several temperatures, each fixed for 30 min, while the mixture was stirred. Afterheating, the mixture was centrifuged at 3,000 rpm for 20 min. The precipitate was thenweighed. Swelling power was measured as the quantity of water 1 g of starch could absorb at agiven temperature.

Figure 4. Amylograms of starches. = acetylated and slightly crosslinked cassava starch (6%); = wheat (9%); = maize (8%); = potato (4%). (From IFRPD Laboratory,1993-1994, personal communication.)

..

45 63 90 90 50 50

320


Standards for modified starch forthe food industry

Standards were introduced toencourage cassava starch factoriesto produce modified cassava starchfor the local market (Ministry ofIndustry, 1992). First published inJanuary, 1992, these standards areupgrading modified-starch productsas manufacturers apply for gradecertification. The standards dealwith different types of modifiedstarch, physical and chemicaladditives used for modification,residues and limiting propertiesindicators, and methods of analysis.

Fifteen types of modifiedstarches exist: pregelatinized starch,dextrin, thin boiling starch, alkalinetreated starch, bleached starch,oxidized starch, di-starchphosphate, starch succinate,hydroxypropyl starch, starchacetate, monostarch phosphate,hydroxypropyl di-starch phosphate,acetylated di-starch adipate,acetylated di-starch phosphate, andcombination chemical processstarch.

Products of Chemical andMicrobial Processing of

Cassava Starch as IngredientsUsed by Food Industries

Some industrial processes use cassavastarch as raw material to manufacturefinal products that are themselvesused as ingredients in foodstuffs.Such products are typically eitherseasonings or sweeteners. and havebeen developed locally and overseas.Most of the technology has beenbrought from developed countries withexperts as consultants. Localinstitutions sometimes collaborate topromote further local development.

The processes are complex,involving chemical and microbialtechnologies that require sophisticatedmachinery. The final products arecostly. Table 5 shows estimates of thecassava starch consumed by theseindustries (TDRI, 1992).

The seasoning industry

Thailand has three seasoningfactories, one of which uses cassavastarch and the other two molasses as

Table 5. Estimates of cassava starch use in seasoning and sweetening industries, Thailand, 1992.

Products Cassava starch used Yield from 1 kg(t/year) cassava starch

(kg)

Seasoning industry:

Monosodium glutamate and lysine (1991) 97,977 0.42

Sweetening industry:

High fructose 15,000 1.00

Liquid glucose 30,000 0.90-0.95

Dextrose (monohydrous) 12,000 1.75

Dextrose (anhydrous) 100 0.50

Sorbitol 28,000 1.20

SOURCE: TDRI, 1992.

321


raw material for processing MSG.MSG and lysine products areexpected to be in high demand inthe future, especially fromfood-processing industries, for bothlocal and export markets. Lysineprocessing, a new industry, hasraised the consumption of cassavastarch in the last few years. Inmanufacturing MSG, cassava starchis hydrolyzed by using α-amylaseenzyme and α-amyloglucosidase tochange starch into glucose. Thefermentation is then continued withbacteria Micrococcus glutamicus orBrevibacterium spp., which are givenurea as nutrient supplement.Eventually, crystalline MSG isformed. The process was firstdeveloped in Japan, imported toThailand, and promoted within thelocal food industry.

The sweetening industry

Thailand first processed glucosesyrup in 1950, glucose powder in1976, and sorbitol in 1980. In 1989,the total sweetening industryconsisted of seven factories: fourproducing glucose syrup, oneproducing sorbitol, and theremaining two, various sweeteningproducts (Sathetkeingkai, 1989).

At present, the country’sproduction of glucose syrup is about76,000 t, which is sufficient for localneeds. Considering the potentialuses of glucose syrup, it could beused in the confectionery industry,which would add value to cassavastarch by as much as 55%. Theconsumption of confectioneryproducts is still low, and needs to bedeveloped and promoted both locallyand regionally.

Glucose syrup can be processedin various ways. At present, acontinuous process is beingdeveloped to replace the batchprocess by using several types of

reactors for starch digestion. Thisprocess, however, requires newequipment (such as a digestion tank,filter technique, and evaporator),which needs to be designed anddeveloped.

Sweetening products also needto be developed and their usepromoted in various food industriesthat manufacture, for example, softdrinks, beverages, ice creams, cannedfoods, and bakery products. Becauseof severe competition with thesugarcane industry, food-processingindustries are slow to developsweeteners from starches, includingcassava starch.

Conclusions

Because cassava production ispredicted to increase during the nextdecade, research and development areneeded on cassava use, for both localand overseas markets. The CassavaDevelopment Institute Foundationhas been established to study thenature of the crop’s production, use,and marketing. At present, cassava isstill faced with falling prices, whichaffect growers, industries, andtraders. Animal feed and nonfoodproducts can also be developed as themarket requires, thus adding value tothe crop.

References

Boonyasirikool, B.; Ratarpar, V.; andPhuphat, P. 1987. Qualityimprovement of Kaset-cookie. In:Proceedings of 25th Annual Meeting,Agroindustry Session, KasetsartUniversity, 3-5 February, Thailand.Faculty of Agroindustry, KasetsartUniversity, Thailand. p. 27-35.

Chananithithum, P. 1986. Partialsubstitution of wheat flour in cookieswith tapioca flour. M.S. thesis.Chulalongkorn University, Thailand.p. 84-92.

322


FAO and WHO (Food and AgricultureOrganization of the United Nations andWorld Health Organization), CodexAlimentarius Commission. 1992. Codexstandard for edible cassava flour—African regional standard—CODEXSTAN 176-1991. Eighth session of theCodex Committee on Cereals, Pulsesand Legumes, CX/CPL 92/9, June,1992. FAO/WHO Food StandardsProgram, Rome, Italy. 17 p.

Maneepun, S. and Sirirojana, V. 1990. Noveluses of modified cassava starch in theAsian food industry. In: Cassavabreeding, agronomy and utilizationresearch in Asia: proceedings of theThird Regional Workshop, Malang,Indonesia, October 22-27. CIAT, Cali,Colombia. p. 388-407.

Ministry of Commerce. 1993. Draft standardfor cassava flour/starch. Bangkok,Thailand. 12 p.

Ministry of Industry. 1978. Standard forcassava flour/starch. Bangkok,Thailand. 20 p.

__________. 1992. Standard for modifiedstarch for food industry. Bangkok,Thailand. 22 p.

.Niyomvit, N.; Kanchanapakornchai, A.; and

Rodporn, S. 1990. Production of“Kanom Cham” premix from tapiocaand modified tapioca. Food (Inst. Food

Res. Prod. Dev.) 20(2):105-114.Parvet, S. 1988. Product development of

coconut custard spread. M.S.thesis. Chulalongkorn University,Thailand. p. 70-90.

Rodsri, K. 1993. Trend of cassavadevelopment in agroindustry.J. Agroind. 4(3):16-22.

Saencharoenrat, C. 1990. Effects of tapiocaflour substitution in wheat flour onthe quality of sponge cake. M.S.thesis. Chulalongkorn University,Thailand. p. 90-118.

Sathetkeingkai, A. 1989. Trend ofinvestment of glucose syrup.Northeastern Economic Center,Industrial Economic Division,Permanent Secretary Office,Ministry of Industry, Bangkok,Thailand.

Sirirojana, V. 1987. Quality improvementof tapioca starch by chemicalmodification. M.S. thesis.Chulalongkorn University, Thailand.p. 70-87.

TDRI (Thailand Development ResearchInstitute). 1992. Cassava: in thenext decade. Bangkok, Thailand.p. 2-1 to 2-37.

323

Yuca Rava and Yuca Porridge:...

CHAPTER 37

YUCA RAVA AND YUCA PORRIDGE:THE FUNCTIONAL PROPERTIES AND

QUALITY OF TWO NOVEL CASSAVA FOOD

PRODUCTS1

G. Padmaja, C. Balagopalan,S. N. Moorthy, and V. P. Potty*

* Division of Postharvest Technology, CentralTuber Crops Research Institute (CTCRI),Trivandrum, India.


Introduction

Cassava (Manihot esculenta Crantz) isan important food staple for about500 million people of the tropicalworld (Cock, 1985). Cassava rootsare processed by several traditionalmethods, which vary widely fromregion to region. Usually, thesetechniques are intended to reduce thelevel of cyanogenic glucosides in theroots and improve palatability andshelf life of the resultant products(Cooke and Maduagwu, 1978). Whilefermented food products from cassavaare popular in many Africancountries, preparations fromdehydrated flour and those fromcooked fresh roots are preferred inAsian and many Latin Americancountries.

Yuca rava and porridge are twonovel food products made fromcassava roots at the Central TuberCrops Research Institute (CTCRI),Trivandrum, India. These productsare likely to capture the Indian foodmarket because of the ease and

relative economy of their preparation(Balagopalan et al., 1988). Beforebeing promoted in potential markets,these two products were evaluated fortheir quality, rheological and pastingbehavior, and residual cyanogencontents. The results are reported inthis paper.


Yuca rava and porridge wereprepared (Figure 1) from threecassava varieties: the low-cyanogencultivar H 1687; and twohigh-cyanogen cultivars, H 165 andH 226 (Table 1). Normally, varietiesrequiring less cooking are preferredfor preparing rava and porridge as thestarch in the roots does not gelatinizecompletely within the 10-min cookingtime. This technique of partialgelatinization, or parboiling, helps inthe partial swelling of starchgranules. The dried, parboiled chipsare powdered to obtain the finerfraction called “porridge” and thecoarse fraction called rava.

Conventionally, rava is preparedfrom round cassava chips which areput into boiling water. An attemptwas made to find out whethercyanogen retention in parboiled chipscould be reduced with a smaller chipsize. The round chips (1 cm thick)were either quartered to equal

324


Cassava roots

Washing and peeling

Parboiling bysteeping in

boiling water for10 min at 100 °C

Decanting

Parboiled chips

Sun drying for 36 h oroven drying for 24 h at 70 °C

Disintegrator

Sieving

Large fraction Medium-sized Small fractionfraction (porridge)(rava)

Figure 1. Manufacturing rava and porridge from cassava, India.

A parallel study was undertakento find out whether an initialpresoaking of cassava chips forvarying periods helps removecyanogens from the roots. Round andquartered chips, as well as strips,were soaked for 30 min, 1 h, 2 h, or3 h in standing water (1:4 w/v), thenthe water was drained off. Rava andporridge were prepared from thesechips as normal.

The total and intermediate(nonglucosidic, i.e., acetonecyanohydrin, plus free) cyanogens,and free cyanide were quantified inthe rava and porridge prepared from

Table 1. Initial content of cyanogens (mg/kgDM) in cassava cultivars.

Cultivar Cyanogensa

CG NGC FC

H 1687 88.97 37.27 31.21

H 165 271.15 45.67 31.25

H 226 214.12 26.81 20.98

a. CG = cyanogenic glucoside; NGC = nonglucosidiccyanogen (acetone cyanohydrin + free); FC = freecyanide.

portions or cut into strips of uniformwidth before being used to make ravaand porridge.

325


the three cultivars. The methods forextraction and initial stages ofcyanogen determination up to theformation of cyanogen chloride wereadopted from O‘Brien et al. (1991).The coupling of cyanogen chloridewas done with the barbituric acidpyridine reagents used in theNambisan and Sundaresan (1984)procedure.

The viscographic behavior of ravaand porridge samples from the threecultivars was studied with aBrabender viscoamylograph. Aconcentration of 10%-20% by weightof sample was studied at a heatingrate of 1.5 °C per minute. The peakviscosity at 97 °C and viscosity aftercooling were recorded. The swellingvolumes were determined by standardprocedure (Schoch, 1964). A sampleof 400 mg rava or porridge wassuspended in 40 ml water, heated to95 °C, maintained at thattemperature for 15 min, then cooledand centrifuged at 2,200 rpm for15 min. The volume of the gelatinousprecipitate obtained was taken as theswelling volume.

Starch and sugar contents of ravaand porridge were determined by astandard titrimetric procedure (S. N.Moorthy, personal communication).Using 80% alcohol, sugars wereextracted from samples by standingovernight. The remaining residue washydrolyzed with 2N HCl to convertstarch to sugars. These releasedsugars were quantified throughferricyanide titration (S. N. Moorthy,personal communication) and thestarch value computed, using a factorof 0.9.


Of the cassava-consuming areas ofthe world, India is perhaps unique inmaking pregelatinized, dried

preparations from cassava. Yuca ravaand yuca porridge are two novelproducts that can be easily prepared,have an acceptable shelf life, and aretasty.

Cyanogen changes

The residual cyanogenic glucosides(CG) in rava made from thelow-cyanogen cassava cultivarH 1687 ranged from 17.5-21 mg/kgDM and, in porridge, from14.5-24.5 mg/kg DM, according tothe type of chips (Table 2). Theseranges were much lower than thosefor the products made from the twohigh-cyanogen cultivars, H 165 andH 226. The initial CG values weremuch lower for H 1687 roots(88.97 mg/kg DM), compared with thehigh-cyanogen cultivars (Table 1).

For each variety, however, the CGcontent among the three types ofchips (round, quartered, and strips)did not vary significantly when theywere put directly into boiling water,parboiled for 10 min, dried, andpowdered to make rava and porridge(Table 2). The lack of variation maybe a result of the rapid loss oflinamarase activity at 100 °C as theroots were directly exposed to thistemperature.

The extent of cyanogen removalduring the boiling of cassava rootsdepends on boiling time, volume ofwater used, and size of root piece(Padmaja, 1993). Ezeala and Okoro(1986) reported that, after 35 min ofboiling, cyanogens were undetectablein the roots they used. The initialcyanogen level of 218 mg/kg haddropped to 97 mg/kg within the first5 min of boiling. However, theauthors had gradually raised thechips to boiling point, taking20-25 min. During that timelinamarase could act on theglucosides.

326


Cooke and Maduagwu (1978) alsoobserved that bound cyanogen wasremoved at a slower rate only duringboiling and 55% of it was retainedafter 25 min of boiling. Nambisanand Sundaresan (1985) reported thatduring 30 min of boiling, only45%-48% of total cyanogen waseliminated from the roots (varietiesH 165, H 2304, and H 1678).

We found that the CG contentdropped from about 89 to17-21 mg/kg DM in the low-cyanogencultivar H 1678 during 10 min ofboiling. For the high-cyanogencultivars, reductions were from 271 to86-108 mg/kg in H 165, and from214 to 87-111 mg/kg in H 226.

Fukuba et al. (1984) observedthat cultivar variations stronglyinfluenced cyanogen elimination in1-cm diced roots during boiling orsoaking treatments. They comparedthe effect of slow with rapid boiling.They found that, while 70% of totalcyanogen was eliminated from cubesbrought to boiling point, only30%-35% of total cyanogen was

eliminated if chips were put intoboiling water, even after 10 min ofcooking at 100 °C.

Soaking the different types ofcassava chips for varying periodsfrom 30 min to 3 h in standing waterdid not reduce the quantity ofcyanogen compared with unsoakedroots (Table 3). Nor did chip typeinfluence cyanogen eliminationduring soaking. Similar trends wereobtained for low- and high-cyanogencultivars. But cultivar variations arelikely to affect the cyanogenelimination during soaking. Fukubaet al. (1984) obtained 54%elimination of total cyanogen from1-cm diced roots of variety Bogor397 during 10 min of soaking, whileonly 5%-6% cyanogen waseliminated from other varieties.

The reduction in CG during thepreparation of rava and porridgeindicates that cyanogen hydrolysistakes place during parboiling. Thefree cyanide contents of the ravaand porridge obtained from eachvariety by various techniques vary

Table 2. Cyanogen content (mg/kg DM)a of rava and porridge made from different types of cassava chips(round, quartered, and strips).

Cultivar Rava PorridgeType of chips

CG NGC FC CG NGC FC

H 1687:Round 21.0 35.0 25.0 24.5 32.5 20.0Quartered 19.0 36.0 22.5 21.5 39.5 17.5Strips 17.5 26.5 17.5 14.5 25.5 19.0



a. CG = cyanogenic glucoside; NGC = nonglucosidic cyanogen (acetone cyanohydrin + free); FC = free cyanide.

327


Table 3. Effect of soaking chips on cyanogen content (mg/kg DM)a of rava and porridge made from threecassava varieties, India.

Variety Hours Rava PorridgeType of chips

CG NGC FC CG NGC FC

H 1687 (low-cyanogen)

Round: 0.5 24.5 32.5 23.0 25.3 34.8 24.3

3.0 28.5 29.0 22.5 34.3 28.3 19.0

Quartered: 0.5 25.5 29.3 19.8 25.0 29.8 19.0

3.0 32.5 25.0 19.3 32.0 30.8 19.8

Strips: 0.5 27.0 21.8 18.8 27.5 27.5 18.0

3.0 24.0 25.0 18.8 23.0 25.3 17.5

H 165 (high-cyanogen)

Round: 0.5 120.0 66.5 49.5 138.5 68.5 53.0

3.0 116.5 93.0 67.5 108.5 90.5 77.0

Quartered: 0.5 147.5 63.0 51.0 110.5 56.6 49.5

3.0 121.5 58.5 57.0 128.0 48.5 46.0

Strips: 0.5 160.5 67.5 56.0 133.0 60.5 43.0

3.0 105.5 58.5 51.0 144.0 55.0 50.0

H 226 (high-cyanogen)

Round: 0.5 78.0 37.0 17.0 128.0 30.0 19.5

3.0 73.5 21.5 14.5 86.5 30.5 14.5

Quartered: 0.5 78.0 36.5 25.5 122.0 38.5 26.0

3.0 70.0 29.0 19.0 98.5 51.5 40.0

Strips: 0.5 91.5 24.5 17.0 116.0 37.0 26.0

3.0 100.5 21.5 13.5 99.0 39.0 31.0

a. CG = cyanogenic glucoside; NGC = nonglucosidic cyanogen (acetone cyanohydrin + free); FC = free cyanide.

little (Tables 1 to 3). This indicatesthat the free cyanide formed is rapidlylost from boiling water, but that acertain amount of free cyanide isretained in dried, parboiled chips.

Rapid parboiling, by adding chipsto boiling water, is insufficient toeliminate cyanogen fromhigh-cyanogen cultivars. Rigorousprocessing is needed to minimizecyanogen retention in high-cyanogencultivars. But for low-cyanogencultivars, such as H 1687, rapidparboiling helps reduce operational

costs during manufacture of foodproducts.

Rheological and swellingproperties

Table 4 shows the rheology of ravaand porridge samples of the threevarieties. Significant differences inswelling volumes exist among theporridge samples of the threevarieties. Swelling volumes werehighest for porridge made fromH 226 and lowest for that fromH 1687. For rava samples, H 165

328


Table 4. Swelling volume and viscosity of rava and porridge made from cassava roots.

Cassava variety Swelling volume Viscosity propertiesType of chips (ml/g) (BU)a

Rava Porridge Maximum viscosity Viscosity breakdown

Rava Porridge Rava Porridge

H 1687: Round 5.00 3.15 360 650 60 280 Quartered 5.00 3.05 350 700 70 350 Strips 5.35 3.75 320 600 30 190H 165: Round 7.00 7.00 140 680 40 320 Quartered 6.20 6.25 160 820 20 400 Strips 6.25 5.75 150 800 30 360H 226: Round 3.25 7.00 160 960 20 280 Quartered 3.25 7.50 220 900 30 350 Strips 3.75 7.25 180 870 35 190

a. BU = Brabender units.

had the highest values and H 226,the lowest.

No relationship was foundbetween the shape of the chips usedto prepare rava and porridge andswelling volume. The values obtainedwere much lower than those normallyobserved for corresponding starchsamples which have 3-4 times theswelling volumes of rava. The lowerswelling volumes can be attributed tothe preliminary swelling of starchduring parboiling of the chips. Mostof the starch granules were alreadyswollen during the rava preparation.

The swelling volumes observedalso indicate an almost equal starchdistribution in both rava and porridgesamples. No significant differences inswelling volumes exist according tosoaking period of chips, which wasexpected. Raja and Mathew (1986)observed the sedimentation volume ofpowdered, parboiled chips to increaseslightly with longer boiling time. Butwe observed swelling volume todecrease with parboiling.

Viscosity data showed ravasamples to have a different viscositypattern to that of porridge samples.Porridge showed peak viscosity before60 °C, whereas rava behaved like astarch, with a peak viscosity around75-85 °C.

Recently, Raja and Ramakrishna(1990) found that parboiling affectsviscosity properties. Earlier, Raja etal. (1982) reported that the viscosityof powdered, parboiled chips waslower compared with powdered, driedchips. This corroborates our findings.

Starch and sugar changes

No significant differences wereobserved in the starch content of ravaand porridge samples according tocultivar. Starch content ranged from52%-66% in the rava samples and56%-70% in those of porridge.Neither was a relationship foundbetween chip type and starch contentof rava and porridge. Total sugarcontent was higher for the rava andporridge made from H 226 than from

329


Table 5. Starch and sugar changes (g/100 g DM) in rava and porridge made from cassava roots.

Cassava variety Rava Porridge

Type of chips

Starch Sugar Starch Sugar

H 1687: Round 66.18 14.60 64.38 15.74 Quartered 60.18 13.79 63.38 11.36 Strips 55.21 19.84 53.57 16.95H 165: Round 64.29 13.61 70.31 11.90 Quartered 60.00 12.20 59.60 14.71 Strips 64.29 11.76 62.50 9.09H 226: Round 67.67 14.81 56.60 20.20 Quartered 52.33 21.05 56.25 17.54 Strips 55.56 20.83 59.21 20.41

H 1687, and lowest for those fromH 165. These parameters did notseem to influence the rheological orswelling properties of these foodproducts (Table 5).

Conclusions

The rheological and swellingproperties of the rava and porridgefractions made from the threecassava cultivars suggest theirsuitability as a wheat substitute forbreakfast recipes and certain southIndian sweet dishes. However, theretention of cyanogens were foundto be slightly high in the case of thehigh-cyanogen cultivars. We expectto study detoxifying processes in anattempt to develop suitableprocessing technologies forhigh-cyanogen cultivars, therebyincreasing demand for foodproducts made from such cultivars.

References

Balagopalan, C.; Padmaja G.; Nanda,S. K.; and Moorthy, S. N. 1988.Cassava in food, feed and industry.CRC Press, Boca Raton, FL, USA.205 p.

Cock, J. H. 1985. Cassava: new potential fora neglected crop. Westview Press,Boulder, CO, USA. 191 p.

Cooke, R. D. and Maduagwu, E. N. 1978. Theeffect of simple processing on thecyanide content of cassava chips.J. Food Technol. 13:299-306.

Ezeala, D. O. and Okoro, N. 1986. Processingtechniques and hydrocyanic acidcontent of cassava-based humanfoods in Nigeria. J. Food Biochem.10:125-132.

Fukuba, H.; Igaraschi, O.; Biones, C. M.; andMendoza, E. T. 1984. Cyanogenicglucosides in cassava and cassavaproducts: determination anddetoxification. In: Uritani, I. andReyes, E. D. (eds.). Tropical rootcrops: postharvest physiology andprocessing. Japan Scientific SocietiesPress, Tokyo. p. 225-234.

Nambisan, B. and Sundaresan, S. 1984.Spectrophotometric determination ofcyanoglucosides in cassava. J. Assoc.Off. Anal. Chem. 67:641-643.

__________ and __________. 1985. Effect ofprocessing on the cyanoglucosidecontent of cassava. J. Sci. Food Agric.36:1197-1203.

O’Brien, G. M.; Taylor, A. J.; and Poulter,N. H. 1991. Improved enzymaticassay for cyanogens in fresh andprocessed cassava. J. Sci. Food Agric.56:277-289.

330


Padmaja, G. 1993. Cyanide detoxification incassava for food and feed uses. Crit.Rev. Food Sci. Nutr. 35(4)299-339.

Raja, K. C. M.; Abraham, E. T.;Sreemulanathan, H.; and Mathew,A. G. 1982. Studies on improvingtextural quality of cassava (tapioca)flour: proceedings of a symposium onpostharvest technology for cassava.Association of Food ScienceTechnology (AFST), Trivandrum,India. p. 108-116.

__________ and Mathew, A. G. 1986. Effect ofparboiling on hydration andsedimentation characteristics ofcassava (Manihot esculenta Crantz)chips. J. Food Sci. Technol. 23:39-41.

__________ and Ramakrishna, S. V. 1990.Compositional and pastingcharacteristics of plain-dried andparboiled cassava (Manihot esculentaCrantz). Food Chem. 38:79-88.

Schoch, T. J. 1964. Swelling power andsolubility of granular starches. In:Whistler, R. L. (ed.). Methods incarbohydrate chemistry, vol. IV.Academic Press, NY, USA.p. 106-108.

SESSION 7:

INTEGRATED PROJECTS

333

Integrated Cassava Research and Development Projects...

CHAPTER 38

INTEGRATED CASSAVA RESEARCH AND

DEVELOPMENT PROJECTS IN COLOMBIA,ECUADOR, AND BRAZIL:AN OVERVIEW OF CIAT’S EXPERIENCES

B. Ospina*, S. Poats**, and G. Henry***

cassava markets have diversified andoverall demand for cassava hasincreased. This has reduced pricevariability while increasing yields and,as a result, created incentives foradopting improved technologies. Poorfarmers’ incomes and employmentopportunities have also improvedthrough the promotion of small-scale,cassava-based, rural agroindustrieswith low opportunity costs, especiallyfor landless producers.

Introduction

In 1973, when CIAT’s CassavaProgram first took shape, few strongagricultural research programs inLatin America focused on cassava.Research was well behind, relative toother crops, and emphasized mainlyaspects of production (Pérez-Crespo,1991). The Cassava Programresearched germplasm developmentand agronomic practices from1973 to 1982. Results during thisperiod were encouraging, and clearlydemonstrated the technicalpossibilities of significantly increasingcassava production.

But farmers had no specialinterest in adopting new cassavaproduction technology to raiseefficiency or productivity. With anincreasing concentration of LatinAmerica’s population in urban

Abstract

This paper discusses CIAT’s 12-yearexperience in developing an integratedcassava research and developmentproject (ICRDP) approach. The origin,justification, methodology, results,and lessons learned from thisapproach are presented, using acomparative analysis of CIAT’sexperiences in Colombia, Ecuador,and Brazil. The ICRDPs have beeneffective vehicles for CIAT’s CassavaProgram to interact with variousnational research, rural extension,and development institutions.Existing production, processing, andmarketing technologies have beenvalidated and adapted to specificregional conditions with the ICRDPframework. New technologies havebeen generated through the synergy ofresearch and development thatICRDPs promote. Results havedemonstrated to research anddevelopment institutions, donors,governments, and policy makers thatcassava is a crop that can play animportant role in achievingdevelopment goals. Through theintegrated approach, traditional

* Cassava Program, CIAT, stationed at theCentro Nacional de Pesquisa de Mandioca eFruticultura (CNPMF), Brazil.

** Independent consultant.*** Cassava Program, CIAT, Cali, Colombia.

334


centers, preferences shifted fromcassava as a basic dietary staple tofoodstuffs easier to transport, store,and exchange. Thus, increasingcassava’s use in Latin America wasdependent on developing newproducts that would use cassava inits fresh state or transform it to astorable or higher value product, andin developing new markets for theseproducts (Lynam, 1978).

In 1979, CIAT took an innovativestep by adding the Utilization Sectionto the Cassava Program, thusextending its responsibilities for cropresearch beyond the developmentand transfer of germplasm andagronomic practices. This move wassimilar to that of many earlierprojects in numerous countries,especially in Southeast Asia, thataimed to exploit cassava’s industrialpotential by transforming itagroindustrially into meal, flour,starch, alcohol, or other derivedproducts. In Latin America, relativelyfew of these projects met with theanticipated success: some, trying toimprove production, ran intomarketing problems; others,investing in processing plants,encountered problems with the rawmaterial’s price or availability.

Analysis of these projectshighlighted the need for an integratedapproach to cassava production,processing, and market development.Cassava development could not beexpected unless all three areas weresimultaneously addressed in anintegrated fashion. Research anddevelopment activities needed tobegin by identifying potentialmarkets for cassava and its products.Once identified, then productdevelopment, processing, production,and commercialization should begin,to develop the market effectively.

The Utilization Section firstconcentrated on developing

technology to conserve fresh cassavaroots for human consumption and ondrying technology for the animal feedindustry. Research activities onsun-dried cassava chips at CIAT werestarted in an attempt to solve qualityproblems in dried cassava chips andpellets produced in Thailand andIndonesia and exported to theEuropean Union for incorporation intoanimal feed concentrates. Throughthis work, CIAT gained considerableexperience in the natural dryingtechniques used in Asian countries.

But this accumulated knowledgecould not be applied to LatinAmerican conditions immediately. Aseries of reviews had cast doubts onthe Program’s ability to reach farmerswith the technologies generated, andso attain increased productivity. Aftera series of internal planning exercisesfocusing on specific social objectives,a new research and developmentframework was formulated for theCassava Program. This included theneed to be directly involved incassava-based, rural developmentprograms, as a sine qua noncondition for the crop’s development(Cock and Lynam, 1990).

At the time the Cassava Programwas searching for Latin Americanpartners and sites to test thisapproach, the Colombian Ministry ofAgriculture, through the IntegratedRural Development (DRI) program,was pursuing CIAT’s collaboration tosolve problems related to increasingproduction and decreasing demandand prices for cassava in the NorthCoast, an extensive cassava-growingarea of Colombia. The two entitiesdeveloped a collaborative program.The experiences gained in this andsimilar projects in Asian countriesover the last 12 years has allowedCIAT to develop the integratedcassava research and developmentproject (ICRDP) methodologydiscussed in this paper.

335


The Importance of Cassavain Latin America

Latin America produces 21% of theworld’s cassava. According to theFood and Agriculture Organization(FAO, cited in CIAT, 1993), Brazil,Paraguay, and Colombia areresponsible for 92% of cassavaproduction in this region. Thecrop is generally produced inmarginal, rainfed areas and isgrown by small-scale farmers withlimited access to land, inputs, andimproved technology. In areaswhere cassava is grownextensively, farmers often have noalternative crops because ofclimate and soil limitations.

Marketing channels availableto cassava growers are usuallylimited to one or two traditionalmarkets per region for either freshroots or processed products suchas farinha da mandioca (toastedcassava flour) in Brazil. Associeties urbanize, demand forprocessed products may remainstable or even increase, creatingshortages and high prices. But theoverall demand for cassava tendsto decline, creating pricefluctuations and increasingcommercial risks. Lackingadditional market opportunities forfresh cassava, farmers have noincentives to adopt improvedproduction technologies.

Fortunately, cassava hasseveral characteristics that allow itto compete as a multiple source ofcarbohydrates, especiallycompared with other root crops, onthe basis of price, yield, nutritionalvalue, quality, and availability.Root dry matter content in cassavais higher than in other root crops(35%-40%), giving optimalconversion rates of 2.5:1 or better.Over 85% of root dry matter

consists of highly digestible starch.Cassava starch has agglutinantproperties that make it suitable forpelleting in animal feeds, such asfor shrimp or fish, replacingexpensive artificial agglutinants(Cock and Lynam, 1990).

The disadvantages of usingfresh cassava roots directly inproducts such as animalconcentrates are their bulk, rapidperishability, low protein content,and the presence of cyanogens inall root tissues. Thesedisadvantages can be overcome bysimple processing techniques, suchas chipping and natural drying.For example, sun drying eliminatesmost cyanogens from root tissues.Increasing cassava’s pricecompetitiveness with othercarbohydrate sources, anddifferentiating the uses of its highquality carbohydrate structure andcomposition, help overcome its lowprotein content.

Linking small-scale cassavafarmers to potential growthmarkets via new processingtechnology and new productdevelopment is an important optionthat can help meet several socialpolicy objectives such as incomegeneration among marginalfarmers and landless poor (Lynam,1987). But penetrating alternativemarkets needs competitivefarm-level prices, investment inprocessing capacity andmanagement, and a coordinatedexpansion in production,processing, and use.

The integrated projectmethodology, developed during thelast 12 years, aims at coordinatingchanges in farming systems withchanges in the marketing system,within the framework ofmulti-institutional integration.

336


Integrated Cassava Researchand Development Projects

(ICRDPs)

Definition

An ICRDP is defined as anintervention at institutional,technological, social, andorganizational levels to linksmall-scale cassava farmers to new orimproved growth markets. Thus,demand for production technology isstimulated with potential to improvesmall-scale farmer welfare.

Methodology

The ICRDP methodology has fourstages. These must be phasedsequentially to succeed (Figure 1).

Macroplanning. The overalleconomic situation of the country orregion initially targeted for an ICRDPis analyzed. The potential demand forcassava and derived products, thecrop’s ability to compete with otherproducts and markets, and thepotential for cassava production indifferent regions are considered.Information gathered in this phase

ensures that the correct target regionand the most promising markets areselected.

Microplanning. Information isgathered to define marketcharacteristics, production practicesand constraints, availability ofinstitutional support, existing farmerorganizations, cassava processingtechnologies, and regional governmentdevelopment priorities. Then, thetarget area is selected for the pilotproject.

Pilot phase. During this stage,available technologies can be entirelyreworked and adapted to localconditions. The project’s institutionaland organizational framework isdetermined and serves as theintersection point for cassavaproduction, processing, and productdevelopment research. Farmerorganizations are included at this stageand become the project’s permanentactors and decision makers. At theend of this stage, enough reliableinformation is available to test theassumptions made during planning.The commercial phase is then eitherjustified or rejected.

Figure 1. Planning integrated cassava research and development projects (ICRDPs).

Macroplanning

Commercial expansion

Productionresearch

Productdevelopmentresearch

Microplanning

Applied field Pilot evaluationtesting of for market andproduction producttechnology development

Pilot phase

337


Commercial expansion phase.The use of cassava processingtechnology and new or improvedproducts can now be replicated orexpanded, according to findings of thepilot phase. The new technology’scommercial costs and the resourcesneeded to promote its adoption on awider scale can now be calculated.This includes credit lines for cropproduction, establishing processingcapacity and operational capital, andinstitutional requirements to train andgive technical assistance to farmers.At the start of the commercial phase,a monitoring system should beestablished, based on the informationgathering mechanisms begun in thepilot stage. Finally, it must beremembered that the projectframework is not a permanentmechanism per se and the endresult of this stage should be aself-supporting, economicallysustainable, cassava-basedagroindustry.

Anticipated outcomes

The anticipated outcomes of theICRDPs were:

(1) That national research, extension,and development agencies wouldbecome involved in a concertedeffort to improve small-scalefarmer welfare through activitiesfocused on cassava;

(2) The development of cassavaprocessing and product marketsas income-generating activities;

(3) The creation of demand forimproved cassava productiontechnology.

Experiences and results

CIAT joined efforts with nationalcounterpart agencies to initiateICRDPs in nine Latin Americancountries (Table 1). These projectshave included different products,markets, and processing technologiesand have reached different stages ofdevelopment.

In Mexico and Peru, the projectsfailed. The Mexican failure wascaused by farmers not beingcommitted and involved from the startand because the production,processing, and commercializationactivities were insufficientlycoordinated. The Peruvian projectwas economically nonviable becausethe target area was far from marketsand another, more profitable,enterprise (cocaine processing) wascompeting.

In the Colombian, Ecuadorean,and Brazilian projects, the CIATCassava Program, through specialfunding, managed to have staffmembers directly involved in projectimplementation.

Table 1. Integrated cassava research and development projects in Latin America.

Country Dried chips Flour Starch Fresh roots Leavesfor animal for human forfeed Human Industrial Human Industrial consumption animal

consumption use consumption use feed

Argentina PilotBolivia PilotBrazil Commercial Pilot PilotColombia Commercial Pilot Pilot Pilot CommercialEcuador Commercial Commercial Commercial Commercial CommercialMexico FailedPanama CommercialParaguay Pilot Pilot PilotPeru Failed

338


The Colombian ICRDP

The North Coast is a major cassavaproduction zone in Colombia. In1990, it accounted for 52% of thenation’s cassava production,representing 13% of total land undercultivation and 20% of the region’stotal value of agricultural production(Henry et al., 1994). According toJanssen (1986), 40% of the totalsmall-scale farmer income fromagricultural production in this areais derived from cassava cultivation.On-farm consumption and freshcassava sold to urban markets havetraditionally been the two maincommercial outlets for the region’scassava crop, although some typical,processed, cassava-based productsfor human consumption also take asmall share of the cassava market.Industrial uses of cassava have beenvirtually nonexistent in the region.

In the late 1970s, the DRI wasalready promoting cassava as anagricultural policy option in theNorth Coast. It provided credit andtechnical assistance to increasecassava production.

This traditional,production-oriented approach wasrelatively successful. Cassavaproduction increased rapidly, mainlybecause of the effect that increasedcredit availability had onintensifying production by farmerbeneficiaries of the DRI program.The rapid growth in productioncaused saturation in local markets,and prices dropped to such levelsthat farmers were unable to findbuyers to recover their costs. Toresolve this problem, the DRIprogram set up a postharvestcommittee, who contacted CIAT forhelp in finding alternative marketsfor the region’s cassava production.

At the same time, CIAT’sCassava Program had already found

that a large and expanding marketfor animal feed existed in Colombia,and was analyzing the possible use ofdried cassava for this market. Thetwo entities teamed up to assess thepossibilities of entering thisalternative market.

Of the various possibilitiesanalyzed, the most promising seemedto be that of establishing cassavaproducer organizations to operatecassava drying plants and sell thedried cassava to animal feedfactories. This approach appearedattractive for two reasons: first, theresource-poor farmers in the areacould not individually afford toestablish cassava processinginfrastructures, whereas as anorganization they could do so; and,second, the cassava drying processwas proposed as a way to create aneffective floor price for cassava roots,so that if prices in the fresh marketwere high, farmers could sell intothese markets and make enoughprofit to pay off loans on the cassavadrying plants. Roots unsuitable forthe fresh market could be sold to thedrying plants, allowing them tooperate at a low level. Conversely, ifthe prices for cassava roots dropped,farmers could sell the roots to thedrying plants and still make a profit.

To test this model’s validitythrough a pilot project, the firstcassava natural drying plant to beoperated by farmers was establishedin Betulia, Department of Sucre, in1981. The farmers were aided byCIAT who already had expertise inthe cassava chipping and dryingtechnology brought from Asia.

Despite their total lack ofexperience and tradition in cassavaprocessing activities, the Colombianfarmers quickly assimilated andadapted the technology. The initialpromising results were used toformulate the project’s expansion,

339


which underwent two additionalphases: the semicommercial(1981-1983), and the replication orcommercial (1984 to date).

By 1991, about 150 cassavadrying plants were operating in theNorth Coast (Figure 2), of which105 were owned and operated bysmall-scale, cassava producers’associations and/or cooperatives.The remainder were exploited byprivate entrepreneurs who, during1987-1991, were rapidly participatingin the industry. The fast, widespreadadoption of diverse types ofcassava-drying in the region is nowmaking accurate monitoring difficult(Henry, 1992).

During 1991, these 150 dryingplants produced about 25,000 t ofdried cassava chips, corresponding to62,500 t of cassava roots—a demandrepresenting 6.6% of total cassavaproduced in the region in 1991 andaccounting for 5.7% of the total areaplanted to cassava. Project activitiesled to the rapid penetration of theColombian animal feed market withdried cassava chips.

Throughout the project’s span,cassava producers and processorsreceived important institutionalsupport, especially credit lines,technical assistance, and training.Important results were also obtainedin the area of improved cassavaproduction technology. The impact ofthe Colombian ICRDP can be bestassessed by considering the addedmonetary value of dried cassava’sannual production, foreign exchangesavings from decreased cereal importsfor animal feed, added employmentopportunities generated in rural areasby expanding cassava production andprocessing activities, and enhancedlinks with goods sectors and services.

The Economics Section of CIAT’sCassava Program estimated that, from1984-1991, the cassava sector innorthern Colombia benefited byalmost US$22 million (Gottret andHenry, 1994). These benefits resultedfrom integrating research to improvecassava crop management,processing, marketing, and consumerpreferences in the framework ofcassava-based, development projectswith strong farmer participation.

123456123456

123456712345671234567

123456123456123456123456123456123456

123456123456123456

123456123456123456123456123456123456123456123456123456123456

123456123456123456123456123456123456123456123456123456123456123456

123456712345671234567123456712345671234567123456712345671234567123456712345671234567

123456123456123456123456123456123456123456123456123456123456123456123456123456

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

180

160

140

120

100

80

60

40

20

0

Cass

ava

dry

ing

pla

nts

(n

o.)

1981 82 83 84 85 86 87 88 89 90

Year

81

8

22

37 41 46 50

150

76

Figure 2. Adoption of cassava drying plants in Colombia, 1981-1990, by farmers’ groups ( 123412341234 ) and private

plants ( ). (After Henry, 1992.)

340


Studies have also shown thatadoption of cassava productiontechnology components is significantlyhigher in areas with ICRDP activitiesthan it is in areas not so influenced(Gottret and Henry, 1994). Forexample, 93% of cassava producers inareas with cassava drying activitiesand strong institutional presenceadopted the cassava variety‘Venezolana’ in 1991. But in areas notdirectly influenced by ICRDP activities,only 48% of cassava producersadopted it (Gottret and Henry, 1994).

The main lesson of the Colombianproject was that farmers, whenencouraged to participate inresearching and solving their currentproblems and needs, becomeimportant partners for research-and-development institutions and makevaluable contributions to identifying,adapting, and evaluating alternativesolutions.

The Colombian project alsovalidated an original hypothesis of theICRDP model: that the integratedproject approach, by creating newmarkets and better prices for cassava,will increase farmers’ incentives toadopt improved productiontechnologies.

Finally, the project demonstratedthat small-scale farmer associationsare indeed a viable mechanism orvehicle for technology diffusion.

The Ecuadorean ICRDP

From its beginning in 1985, theEcuadorean ICRDP represented achallenge for CIAT in that theColombian project, successful as ithad been, demanded very highinstitutional costs, which had to bebrought down. The project in Ecuadorwas therefore conceived as both asocial and technical experiment,requiring specific institutional and

organizational arrangements andallowing farmer organizations,extension workers, and nationalresearch and extension staff to playnew roles (CIAT, 1992).

The project in Ecuador wasimplemented in a traditional,cassava-processing area of ManabíProvince, a seasonally dry, coastalregion, which accounts for 20%-30%of the national cassava production(MAG, 1990). In Manabí, familyfarming households have extractedcassava starch for over 100 yearswith little change in processingtechnology. The potential ofcassava-drying technologies had beenearly identified as a viable alternativefor promoting alternative uses andmarkets for the crop. But it was notuntil 1985 that conditions becameeconomically favorable to launch theICRDP in Manabí.

A favorable climate for cassavaprocessing and sun drying, excesscassava production, and apredominance of small-farmpopulation characterized the regionas “optimal” for the project.Farmers were organized into smallproducer-processor associationscalled APPYs (Asociaciones deProductores y Procesadores de Yuca).From the start, these associationsjoined, as a second-order farmerorganization, known as the Unión deAsociaciones de Productores yProcesadores de Yuca (UAPPY).

The UAPPY changed its legalstatus in 1992 to admit associationsof rural workers (ATAPYs) andbecame the Unión de Asociaciones deTrabajadores Agrícolas, Productores yProcesadores de Yuca (UATAPPY).This change allowed small-scalefarmers, lacking titles to their lands,and landless rural workers (such aswomen who could easily benefit fromprocessing-generated jobs) to legallyparticipate.

341


Begun as a marketingcommittee, the Union now includes17 associations and performs a varietyof functions including technicalassistance, credit, marketing,accounting, training, productdevelopment, and monitoring.Farmers meet annually asstockholders to evaluate their progressand make recommendations toUATAPPY leaders and other projectcollaborators.

Colombian farmer-processors werebrought to Ecuador to teach Manabífarmers the new chipping and dryingtechnology. These farmer-to-farmercontacts were later reinforced byManabí farmers visiting Colombia.They were able to see in actiontechnical, organizational, andoperational features of the Colombiancassava processing plants. From thestart, farmer processors playedan important role as promoters,technology transfer agents, teachers,and leaders of the project. Staff fromlocal agencies and CIAT joined andsupported farmers’ efforts.

The basic chipping technologyadopted was the same as that usedin Colombia. Drying trays, aCIAT-suggested technology, werequickly adopted as an intermediatestep toward building a cement dryingfloor. This allowed poorer farmergroups to get started quickly withfewer initial investment costs. Later,profits earned could be used to buildthe floor.

Project leaders and CIATresearchers assumed that the marketfor dried cassava in Ecuador would bethe same as that in Colombia: thebalanced feeds industry for poultryand livestock. Early in the project,serendipitously, it was discovered thatcassava was an ideal substitute forimported chemical agglutinants for thefeed pellets used by the Ecuadoreanshrimp industry. The scale of this

industry was such that demand forcassava flour could be more than8,000 t/year.

Transforming dried cassava chipsto flour for shrimp feed required newsteps in processing because roots hadto be peeled before drying. Peelingsoon became an important source ofadded income for member andnonmember families, especiallywomen, children, and elderly people,who usually had no other sources ofincome.

A different management systemwas also needed, in which theassociations produced dried chips andsold them to the Union. The Unionwas obliged to develop milling capacityand management, using portablehammer mills to grind the dried chipsinto flour. This process catalyzed theidea of developing a Union-ownedand administrated “DemonstrationCenter,” where new cassavaprocessing technologies could bedesigned, adapted, and tested, andtraining and demonstrations forfarmers could be held.

In 1993, the DemonstrationCenter became the “Central Plant”,reflecting its increasing roles intransformation, storage, andtransshipment activities. Trainingand research activities were takenover, to some extent, by specificfarmer associations, encouragingincreased participation.

In 1989, the shrimp industry inEcuador slumped: strong competitionfrom Asian producers and problemswith a shortage of larvae ponds cutshrimp production overnight,eliminating 95% of the demand forcassava flour. The Union reactedquickly, launching an all-outcampaign to identify other markets.The Demonstration Center made itpossible for farmers to rapidly adaptexisting products for new markets.

342


For example, whole-root cassavaflour was refined by passing it througha mechanical vibrating sifter, aprocess yielding a flour of the samegranular size as wheat flour. Thisrefined cassava flour begansubstituting wheat flour in fillers forresins used for making plywood, thuscapturing an important share of thismarket. Bran, a byproduct fromsifting, was also sold as a source offiber to livestock feed industries.

In 1989, farmers, collaboratinginstitutions, and CIAT learned avaluable lesson about the importanceof diversifying products and markets.Since then, the Union’s markets andproducts portfolio continued todiversify. Today, seven differentprimary products and four byproductsare produced and sold to sevendifferent market sectors (Table 2),reaching more than 40 buyers.

The Ecuadorean cassava project’sgrowth has been operational rather

than in number of processingorganizations. Initially fueled bystrong market demand and reasonablefunding for construction andoperational credit, the processingassociations’ expansion was veryrapid, from 2 in 1985 to 16 in 1988.By the end of 1988, a scarcity of donorfunds for construction and a rapidlyincreasing inflation combined to makepromoting the formation of newassociations much more difficult forthe Union. In 1992, there were17 associations in Manabí, with atotal of 320 members (Figure 3).

The Ecuadorean ICRDP differedfrom other ICRDPs by having theUATAPPY as agent of its members’growth and development. It hasmanaged and often carried out projectfunctions normally assigned tosupporting state institutions ornongovernmental organizations(NGOs), including handlingdevelopment funds. This has servedto strengthen and promote

Table 2. Market sectors and products in the Ecuadorean cassava project, 1989-1992.

Market sectors Products Total annual output (t)

1988/89 1989/90 1990/91 1991/92 1992/93

Shrimp feed and White industrial flour 574 982b 304b 631b

exports to Colombia

Shrimp feeda Whole industrial flour 1,100 304 258b 464b 127b

Plywood industry Refined whole industrial 200b 170b 292b

flour

Ice-cream cone Refined white food flour 33 6b 33b

factories

Cardboard industry Industrial starch 70 188b 57b 256b

(Ecuador and Colombia)

Bakeries, traditional Food starch 5 10 6b 9b 17b

and large-scale

Livestock feed Starch bagasse and 24 103b 29b 166b

flour bran

Total 1,105 1,015 1,743b 1,033b 1,522b

a. After 1990/91, most of the “whole industrial flour” was used for other livestock feeds, and not shrimp pellets.b. Includes starches or “bagasse” purchased by the Unión de Asociaciones de Trabajadores Agrícolas, Productores

y Procesadores de Yuca (UATAPPY) from private starch processors.

343


sustainability of the project after stateinstitutions and NGOs withdrew theirsupport as funds run out.

Another difference has been thedirect and active participation ofwomen from the start in the Unionand in all project activities, asproducers, processors, and managers.Today, three kinds of processingassociations exist: only men, onlywomen, and mixed. Women comprisenearly 33% of total membership.

The UATAPPY experience with theintegrated cassava project over thepast years has fully validated threeguiding principles that can beconsidered as part of the culture of theproject’s participants and the criteriafor their collaborating well:

(1) The transfer of technologies ismore rapid, efficient, and effectivewhen end-users are directlyinvolved and responsible.

(2) Farmers’ organizations areeffective intermediaries betweenfarmers and institutions and canbe used as efficient channels forproject services, credit, andinformation dissemination. Afarmers’ organization accumulatesexperiences and learning, thuscontributing to the growth,maturity, and ultimatesustainability of the farmers’group.

(3) Farmer organizations—not merelyreceiving project benefits butactively participating with farmers“owning” their research agenda—should be part of the institutionalstrategy of an ICRDP.Collaboration between farmerorganizations and supportinginstitutions in an ICRDP should beencouraged without creatingrelationships of dependence amongthem.

The Brazilian ICRDP

In 1989, the Kellogg Foundationapproved a 3-year grant (1989-1992)to CIAT and collaborating Brazilianagricultural research and technicalassistance institutions and farmerorganizations. The grant’s overallobjective was to support theintroduction of improved cassavaproduction, processing technologies,and appropriate organizationalschemes for institutions and farmergroups throughout the maincassava-growing areas of Ceará State,Northeast Brazil.

In this region, an estimated110,000 ha of cassava are harvestedyearly with a total output of almost1.2 million tons of cassava roots. Forcenturies, the main commercial outletfor production has been the casas defarinha. These are small communalprocessing units used to processcassava roots into a toasted flour ormeal known as farinha de mandioca.This flour is a staple product andsource of income, especially in therural sectors of Northeast Brazil.In Ceará State, an estimated14,000 casas de farinha producealmost 200,000 t of cassava flour peryear, representing about 65% of thestate’s cassava production.

Northeast Brazil’s extremelyvariable rainfall patterns cause widefluctuations in cassava yields and so

Figure 3. Expansion of cassava-dryingagroindustries in Ecuador, 1985-1991.

20

15

10

5

0

12341234123412341234

1234123412341234123412341234123412341234

123412341234123412341234123412341234123412341234123412341234

123123123123123123123123123123123123123123123123

123123123123123123123123123123123123123123123123

123412341234

123412341234123412341234123412341234123412341234123412341234

24

10

16 1617 17

1985 86 87 88 89 90 91

Cas

sava

dry

ing

pla

nts

(n

o.)

Year

344


the supply and prices of farinha demandioca vary greatly. Because themain income of the region’ssmall-scale farmers derives fromcassava flour, this situation createsinstability. Further, poor quality, thesmall scale of operation, andrudimentary cassava processingtechnology further contribute toestablishing commercial systemswhereby farmer groups are forced tosell their product at low prices.

The ICRDP’s strategy was to seek alarge, alternative, market that cassavacould enter, especially in good rainfallyears when excess cassava productionusually means low prices. Once theanimal feed market was focused, apilot project was established todevelop the local experience needed tostrengthen local institutional capacityfor implementing cassava-based ruraldevelopment programs with potentialto benefit targeted groups. Along-term objective was to developnational capacity to carry out similardevelopment programs in otherregions of Brazil.

Two factors significantly benefitedthe project: first, counterpart agenciesin the State had previously worked onrelating small-scale cassava farmingand processing. And, second, theState’s institutional setup includedtop-level administrators, policymakers, and local agencies’ staff whohad been exposed to similarexperiences in other countries. Theirparticipation was fundamental indefining the project’s organizationaland operational strategies.

A state cassava committee (CCC),created before the project, wasstrengthened and soon gained generalrecognition as the coordinating bodyfor project activities and all thoserelated to promoting and developingthe cassava crop in Ceará State.Regional cassava committees (RCCs)were established to decentralize

project activity coordination andintegrate the research and extensioncollaborating agencies in the project’sarea of influence.

The project’s expansion, in termsof number of farmer groups organized,was explosive, mainly becausenational and state governmentagencies strongly intervened to launchprograms of financial aid in the formof grants, thus allowing ruralcommunities to build cassava dryingagroindustries. From the 11 that hadalready existed at the project’s start,the number of farmer groups rose to158—about 75% were establishedduring the project’s last year (1991)(Figure 4).

The CCCs and RCCs played acrucial role in the task of approachingdifferent government agencies andprograms to obtain grants on behalfof the farmer organizations. At thesame time, both these committeeshad permanent access to projectfunds for assisting and supportingfarmer activities. Despite the adverseeconomic situation the countryfaced during the project’s span, thecommittees were very active inidentifying sources of financialsupport and channeling them towardtargeted groups. Project activities

123412341234

12341234

12341234

160

120

80

40

012341234

12341234123412341234

1234123412341234123412341234123412341234123412341234123412341234

1986 87 88 89 90 91

3 3 1122

40

158

ICRDP

Figure 4. Expansion of cassava-dryingagroindustries in Ceará, Brazil,1986-1991. (ICRDP = integrated cassavaresearch and development project.)

Year

Cas

sava

dry

ing

pla

nts

(n

o.)

345


could therefore be executed within theproposed goals. Total financialsupport from local agencies duringthe project’s 3 years was almost1 million American dollars (notincluding local staff salaries).

The project in Ceará followed animplementation model similar to thatof other ICRDPs, based mainly ontransferring and adapting availablecassava processing technologies. Thistook advantage of a strong extensionservice that allowed rapid cover of theregion’s main cassava-growing areas.

Project activities also includedsome production technology research.Initial results indicated that theadoption of improved technologycomponents would help increase theregion’s cassava productivity by asmuch as 50%. However, these resultsare suggestive only because farmersdid not have to pay the expenses.To what extent small-scale,poor-resource farmers would bewilling to invest in, for example,organic fertilizers or weed control,remains to be assessed.

The relationship between farinhade mandioca and dried cassava chipsas the two main commercial optionsfor cassava farmers in Cearádetermines the financial success ofthe cassava drying agroindustries theproject promoted. When marketprices for farinha de mandioca arelow, the cassava drying plantsfunction efficiently as an alternativemarket. Conversely, when prices forfarinha de mandioca are attractive,then finding adequate supplies of rawmaterial for the cassava dryingagroindustries becomes difficult.

Skewed land and farm-sizedistribution, plus climaticfluctuations, also strongly influencethe seasonal availability of cassavaroots and thus the performance of thecassava processing units.

With only 3 years ofimplementation, a complete ex postevaluation of the project’s impact isdifficult. The speed with whichcassava drying technologies areadapted for new regions and ruralcommunities can be assessed throughthe pilot project’s monitoring andevaluation system model (involvingbase data on 133 cassava-processingfarmer groups), a survey conducted atthe beginning of the project, andanother at its end. The assessmentcan be done in terms of increasednumber of drying agroindustries andregions of influence, continuousincrease in client numbers for driedcassava, and the degree ofstrengthening of the organizationalstructure implemented for bothinstitutions and farmer groups, which,by the project’s end, included stateand regional cassava committees andfarmer organizations.

A preliminary analysis of datafrom the two surveys indicates thaton-farm consumption and use ofcassava is changing. Farmers now sella share of their production to thecassava drying agroindustries, incontrast with the situation before theproject, when the casas de farinhawere the main commercial outlet.Farmers participating in the projectare starting to adopt the newprocessing technology. The newmarket has stimulated them totransform their patterns of cassavause and become more marketoriented.

Qualitative information availableon direct impact on communitywelfare, institutional support, and thegeneral environment indicates that thepilot project served as a vehicle toincrease community development ingeneral (organization, knowledge,employment opportunities, incomes),and to strengthen local institutionalsupport (technical assistance, workingcapital). However, the project’s impact

346


on cassava production andproductivity was rendered negligiblebecause of the lack of opportunitiesfor farmers to purchase or rentadditional land. Adoption of improvedproduction technology was slowamong the project’s beneficiaries.

The Ceará ICRDP proved thatsmall, cassava-based, farmerorganizations were attractive tocassava producers. These groups firsthad to improve their marketingschemes. Early success indicatespotential for consolidation throughstronger institutional commitmentto support efforts by farmerorganizations. Those cassava-basedagroindustries able to operate duringthe project helped create additionalemployment opportunities, openedalternative markets, stimulated localindustry, raised farmer incomes, andencouraged overall communitydevelopment.

Funds are now being sought fora second phase: to consolidate theresults obtained during the pilotproject and to demonstrate thesetechnologies and results to otherregions and farmer groups.

Benefits and Beneficiariesof the ICRDPs

Benefits generated by the ICRDPs arecaptured principally by farmermembers of the cassava-basedagroindustries (Gottret and Henry,1994). Members can receive benefitsfrom (1) a new market available fortheir cassava roots at more stableprices; (2) more employment (andtraining) opportunities in thecassava processing agroindustries;(3) value-adding second-rate cassavaroots which previously had no marketvalue and were basically written offbefore the introduction of cassavaprocessing; and (4) the annualshare of profits generated by the

cassava-based farmer organizations.This last benefit is available only toorganization members, whereas theother three apply to any member ofthe larger community within whichthe agroindustry operates.

The total income of farmermembers of the Ceará cassavaprocessing groups during 1989-1992reached US$163,887.00, of which37.3% corresponded to cassava rootsales, 10.0% to processing wages, and52.7% to sharing annual profits(Figure 5). Another source of benefitsthe project generated was captured bynonmembers responsible for selling61.6% of the 7,080 t of cassava rootsprocessed during the project. Incontrast, in the Ecuadorean project,farmer members of the cassava-basedagroindustries earned, over 6 years,an average annual income of US$225,whereas nonmembers earned US$89(Figure 6).

Regarding direct economicbenefits, for the Colombian ICRDP,almost 75% (US$16.2 million) of thetotal project benefits was estimated toaccrue to cassava farmers (producersand processors) (Gottret and Henry,1994). But considerable indirectbenefits have also been generated:backward linkages to several smallindustries supplying materials forconstructing and operating the dryingplants; forward linkages includeespecially the income-generating effectfrom increased rural incomes. Thiswill have a multiplier effect in thatincreased rural demand for goods andservices will boost urbanmanufacturing. As such, ruralagroindustries have a strong, positiveeffect on overall economicdevelopment.

The ICRDPs also represent animportant source of benefits forgroups such as women and landlessfarmers, who usually do not benefitsignificantly from projects. For

347


123456123456123456123456123456123456123456123456123456

123456123456123456

strengthening of community spirit.Increases in local income during thedry season resulted in increasedpurchases of foodstuffs and other itemsfrom local shops in rural communities,stimulating local economic growth. Insome Manabí communities, thecassava-processing activity lessenedmigration of men to other coastalregions to work in the bananaindustry. The cassava-processinginfrastructure can be used for othercommercial and cultural activities. Forexample, in Ecuador, cassava-dryingpatios are rented to dry other products(maize, castor beans, cocoa, rice).Associations hold community fiestas,charging entry to earn money. Thedrying patios make excellent dancefloors!

example, in the Ecuadorean project,US$15,000 was paid in 1990/91 forpeeling cassava roots, 80% of whichwent to poor, nonmember, women andchildren who peeled cassava as theirsole off-farm income. In the 1991/92processing season, this amount evenincreased to 90%. In Ceará, Brazil,the distribution of income earned byfarmers during the 3-year pilot projectwas 58.9% for smallholders, 32.4% forrenters, and 8.7% for sharecroppers.

Other important benefits passedto the community in which thecassava-based agroindustriesoperated. Among these were easieraccess to credit programs andtraining opportunities, integrationof institutional presence, and

123456123456123456123456123456

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

200

150

100

50

0123456789

123456789123456789123456789123456789123456789

123456789123456789123456789

US

$ (i

n t

hou

san

ds)

1989/90 1990/91 1991/92 Total1989-1992

52.7%

10.0%

37.3%19,869

95,277

48,741

163,887

Figure 5. Total incomes for cassava-processing group members, Ceará, 1989-1992. Sources of incomewere

123123123 = cassava sales; = processing wages;

123123123 = annual profits.

Year

1234567123456712345671234567123456712345671234567123456712345671234567

12345123451234512345

123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456123456

1234512345123451234512345

123456712345671234567123456712345671234567123456712345671234567123456712345671234567123456712345671234567

1234512345123451234512345

400

300

200

100

0

133

75

321

103

225

89

1985-88 1988-91 Total1985-1991

Periods

Figure 6. Income earned by members ( 123412341234 ) and nonmembers (

123123 ) of cassava-based agroindustries,

Ecuador, 1985-1991.

An

nu

al ave

rage

in

com

eper

farm

er (U

S$)

123456789123456789123456789123456789123456789123456789123456789123456789

348


Recommendations forSuccessfully ImplementingICRDPs (Lessons Learned)

The ICRDPs now under way in severalLatin American countries haveprovided a dynamic framework withinwhich CIAT’s Cassava Program hasbeen interacting with various nationalinstitutions, whether research- ordevelopment-oriented, and withfarmer groups. This interaction hasmade it easier to validate and adaptexisting production and postharvesttechnology, together with thetechniques developed for marketanalysis. Hopefully, these generalizedmethodologies for implementingICRDPs will be adaptable to differenteconomic conditions, farmingsystems, institutional capacities, andmarkets. Based on the experiencesthat the Cassava Program has builtup over the past years, some criticalfactors have been identified, whichneed to be addressed if ICRDPs are tobe successfully implemented.

Product and market development

Until now, for marketing cassavaroots, ICRDPs have depended on thetraditional market (humanconsumption) and a new market(animal feed). Recently, they havebegun to diversify considerably, bothconsolidating the markets for existingcassava products and creating newproducts for new markets. This, inturn, has forced attention onimproving market financialmanagement and quality control.The long-term viability of the modeldepends on the farmer organizations’ability to move their products into awider range of markets or to developa broader range of end uses for theproduct, especially those that canoffer a high margin of profitability(added value). This not only appliesto cassava but also to othercommodities produced by farmerorganizations.

In several communities, thecassava-based associations havemotivated the creation of day-carecenters, and encouraged thebuilding of roads and bridges,sponsored with government funds.In Ceará, the wives of ICRDPmembers started small, poultryfattening operations beside thecassava-drying floors to generatecomplementary income.

Types of Institutionsand Their Functions in

ICRDPs

The ICRDPs, in which differentactivities have to be developedsimultaneously (e.g., production,processing, marketing, organization,training, and monitoring), areintegrated by nature. Because theyare based on farmer organizations,they generate demand forsubstantial resources andcoordinating mechanisms from otherinstitutions. The organizationalstructure of any ICRDP must besufficiently flexible and adaptable toincorporate different farmerorganization schemes andinstitutional configurations. Table 3shows the range of institutionscurrently participating in projects inColombia, Ecuador, and Brazil andthe different functions eachperforms.

In Brazil, state institutionsplayed leading roles, whilesecond-order farmer organizationswere slow to form. In contrast, inColombia, the second-order farmerorganizations led thecommercialization activities andsome large-scale input buying. Butfew other activities, such asresearch, were coordinated. InEcuador, a wide range of institutionsplayed a multitude of roles, but theUATAPPY was the key player forvirtually all ICRDP functions.

349


Table 3. Types of organizations and their functions in integrated cassava research and developmentprojects (ICRDPs).

Function of organization Region in countrya

North Coast Manabí Ceará(Colombia) (Ecuador) (Brazil)

Agricultural research ICA INIAP EMBRAPAEPACE

Technical assistance ICA EMATERCE

Rural development DRI FODERUMA SUDENE

Credit Caja Agraria Banco Nordeste

Farmer organizations:

First order 180 groups 18 groups 165 groups

Second order ASOCOSTAANPPY UATAPPY COOPEMUBA

COPROMA

Nongovernmental FUNDIAGRO Esplar

International CIATACDI USAID CIAT

IBRDKellogg Foundation

Governmental:

National Min. of Agriculture Min. of Agriculture Ministry of Agriculture

Regional Sec. of Agriculture Sec. of Agriculture Sec. of Agriculture

Sec. of Industry andCommerce

a. Colombia: ICA = Instituto Colombiano Agropecuario; DRI = Fondo de Desarrollo Rural Integrado;ASOCOSTA = Asociación de Cooperativas de la Costa; ANPPY = Asociación Nacional deProductores y Procesadores de Yuca; FUNDIAGRO = Fundación para la Investigación y elDesarrollo de Tecnologías Apropiadas al Agro; ACDI = Agricultural Cooperative DevelopmentInternational.

Ecuador: INIAP = Instituto Nacional de Investigaciones Agropecuarias; FODERUMA = Fondo para elDesarrollo Rural del Ministerio de Agricultura; UATAPPY = Unión de Asociaciones de TrabajadoresAgrícolas, Productores y Procesadores de Yuca; USAID = United States Agency for InternationalDevelopment.

Brazil: EMBRAPA = Empresa Brasileira de Pesquisa Agropecuária; EPACE = Empresa de PesquisaAgropecuária do Ceará; EMATERCE = Empresa de Pesquisa, Assistência Técnica eExtensão Rural do Ceará; SUDENE = Superintendência do Desenvolvimento do Nordeste;COOPEMUBA = Cooperativa de Productores de Mandioca de Ubajara; COPROMA =Cooperativa de Productores de Mandioca de Açarau; IBRD = International Bank forReconstruction and Development (the World Bank).

Crop production technologyresearch

Developing and adopting cassavaproduction systems that will sustainor increase productivity and reducecosts are critical to the ICRDPs’success. For cassava to continuecompeting, more intensive farmpractices may have to be introduced,thus risking increased pressure on thenatural resource base. Research anddevelopment on suitable productionsystems must be initiated, continued,

and strengthened. This requiresintroducing adapted genetic materials,carefully exploring additionalalternatives to maintain and enhancesoil fertility, and adapting ecologicallysound, crop protection practices.

Sufficient evidence exists to provethat small, cassava-based, farmerorganizations can function as efficientand effective enterprises and, as aresult, as vehicles for adapting andtransferring production technology.The challenge is to make them

350


efficient and dynamic private sectorenterprises.

Interinstitutional coordination

Institutions. Interinstitutionalcoordination is important to bringtogether the expertise needed tosupport the farmer organizations inthe different areas and activitieshandled by the ICRDPs. At theirinception, these projects involvediverse activities, beyond the scope ofany single institution. Theinterinstitutional coordinationmechanisms that an ICRDP requiresare usually new to local organizations,who will need an adjustment period tofunction appropriately and efficiently.To ensure smooth coordination, oneinstitution should be designated as“coordinator” among the rest, andsufficient funds should be allocated.

In summary, successfulinterinstitutional coordination mustinclude at least (1) the identificationof a coordinating institution,(2) agreement on the necessaryfunctions of each participatinginstitution, and (3) development ofcoordinating mechanisms at project,regional, and national levels.

Farmer group, or organization,or enterprise? Small, cassava-based,organizations has proved attractive tocassava producers who rapidly buildtheir organizations. But first-orderfarmer organizations are usuallyexceptionally weak in businessmanagement and administration.Suitable instruments andmethodologies for improving theseskills are not always available, and ifthey are, their use is often hindered byvery low levels of education.

If the ICRDPs are to achieveautonomy in the medium term, thenthey must help form second-orderfarmer organizations that can(1) support their members with a wide

range of services, from marketing,through technical assistance, toapplied research, and (2) representtheir members in dialogue with othercollaborating institutes or withgovernment policy makers (creation oflobbying power). In Ecuador and, to alesser extent, in Colombia, farmersecond-order organizations are playingthese roles and giving authority andautonomy.

The interests of farmer,cooperative-based, agroindustriesmust be reconciled with those of smallor medium-sized entrepreneurialagroindustries. In the Colombianproject, conflicts have already arisen.

The organizations, includingcooperatives and associations, need tobe efficient, dynamic, and marketoriented to be commerciallysuccessful. The social objectives ofthese groups are seen principallyby the way profits are distributed.Long-term sustainability dependsheavily on commercial survival.

Human resource development

Poor human resource development is awell-known constraint to theimplementation of any rural program.Training and networking are twoimportant strategies to counteract it.

Training. Establishing ICRDPs inseveral Latin American countries hashighlighted the region’s deficiency ininstitutions and personnel specializedin postharvest research anddevelopment, including marketing.Thus, a great demand exists fortraining research and extensionpersonnel and farmers in such areasas cassava processing, cropmanagement, basic accounting,production technology, human andfinancial resource management,marketing, market analysis,monitoring, and evaluation.

351


Experiences accumulated invarious countries where ICRDPs havebeen implemented, especially Braziland Colombia, show that trainingactivities have been mainly orientedtoward building capacity among localagency staff rather than amongfarmers, given the class structure andorganizational profile of theirinstitutional environment. Ecuadorhas been an exception to thetendency: the UATAPPY andcollaborating institutions carried outfarmer training. Training strategiesfor technicians should link trainingand work, using current and realwork-related problems as the trainingissues, and work groups as the basictraining unit. The sharing of training,management, delivery, andparticipation has resulted in greatercollaboration among partnerinstitutions.

Educational and organizationalneeds of cassava producers are muchgreater than those of project staff.High rates of illiteracy and lack oforganizational skills (particularlythose related to handling funds,keeping records, organizing meetings)are among the major constraints toincreased farmer participation inICRDPs and to a more efficient,two-way information flow betweenthem and project staff.

The current farmer-trainingstrategies that local agencies andtechnicians use in most ICRDPs tendto include mainly formal training andmass communication activitiescentered on extending technologicalservices rather than on training andeducation. As such, these trainingmethodologies tend to be useful onlyfor those farmers with the neededskills. This results in segregating therest of the community, making it moredifficult to develop a broaderleadership base at the communitylevel. The Ecuadorean ICRDP,however, tried to improve this by

having an explicit UATAPPY trainingfunction. It designated a UATAPPYmember (farmer) to manage thisfunction and trained this person tocarry it out in a highly professionalmanner.

Networking. Forging links withinand among regions and countries isone important aspect of implementingICRDPs. At a regional or nationallevel, it is sometimes hard to achievethe interinstitutional andinterdisciplinary approach needed totranslate new or improved productionand postharvest technologies intocommercially viable activities. Theproject framework, within whichICRDPs are usually implemented,facilitates integrating several nationalinstitutions into a network thatprovides a forum for interchangingexperiences and methodologies and forresolving problems common acrossregions and projects.

The Cassava Program at CIATand its partners in many nationalinstitutions have developedmethodologies over the last 12 yearsthat have been operationally,economically, and technically viable.Regional and national networkingseems to be the best way of ensuringthat accumulated experiences andknowledge can be made available toother regions and countries facingsimilar problems and opportunities.

Monitoring and evaluation

Project monitoring and evaluation(M&E) has been an integral part of theICRDPs’ methodology from the start.Besides its use in defining potentialmarkets, research priorities and sites,and beneficiaries, it has provedessential for short-term decisionmaking in refining specific objectives,then undertaking appropriate actions.

During the early 1980s, an M&Esystem was designed to be carried out

352


at three levels, using differentmethodologies: (1) continuous updateof a database on farmer organizations;(2) an annual survey of a large sampleof collaborating farmers; and(3) intensively monitoring a subsampleof farmers (Bode, 1991).

For the first ICRDP in Colombia,the M&E system worked well in thebeginning, but as the project matured,the database updating and subsequentannual reports based on its databecame the only activities and outputsof the M&E system, with much of thedata underused. Furthermore, theannual report was circulated to only afew collaborating institutions, withinsufficient feedback to the farmerorganizations themselves. Themonitoring model was seen assuitable only for the pilot phase of acassava-based development project,being too static to evolve with theproject—different levels of projectmaturity required different emphasesand aspects for M&E.

An improved model of M&E wasdeveloped for the Ecuadorean andBrazilian projects. First, key to severalof the M&E limitations, was the model’sorganizational structure and execution,which had to be based “in house.” Thatis, the second-order farmer organizationhad to internally analyze the M&Esystem and coordinate its operation.Collaborating institutions should adoptonly technical assistance roles. Aneffective feedback of appropriateinformation is thus delivered in timelyfashion to relevant audiences.

Second, the M&E system shouldallow for the dynamics of the projectitself. Parameters of interest duringthe project’s early stages may not berelevant for its expansion phases.Adoption and impact studies need tobe included, but only at later stages.Different M&E activities thus becomeimportant as the project matures(Henry and Best, 1994). Table 4 showshow different monitoring activities areintroduced according to the project’s

Table 4. A modified monitoring and evaluation model for an integrated cassava research and developmentproject.

Activity Sourcea Pilot stage

Experimental Semicommercial

Short-term monitoring: Technical 1, 2 X X X Financial 1, 2 X X X Social 2 X X Commercial 2 X X X Institutional 2 X X

Long-term monitoring: Markets 2, C X X Models 2, C X

Adoption: Processing plants 2 X X Production technology 2, C X X Other technologies 2, C X X

Impact: On-farm/processing plant 2, C X X Community C X Aggregate C X

a. 1 = first-order farmer organizations; 2 = second-order; C = collaborators such as institutions, universities, andnongovernmental organizations.

SOURCE: Henry and Best, 1994.

Commercialstage

353


evolution. For example, marketstudies need to be conducted at theexperimental phase to suggest viable,potential markets for the project. Butmarkets are dynamic so the marketstudies need to be repeated later toensure a sustainable market potentialor, as in the case of the Ecuadoreanexperience, to look for product andmarket diversification opportunities(Brouwer, 1992; CENDES, 1993).

Another feature of the new M&Emodel is that the intensiveness of datacollection diminishes as the speed ofadoption increases.

The new M&E model has alreadyproved to be superior in that it is bothmore usefully effective and hasincreased efficiency in using resourceswhile contributing to the project’ssustainability. In Colombia, forexample, results of adoption andimpact studies have been fed back toresearch managers, scientists,second-order farmer organizations,policy makers, and donors for differentspecific uses. In Ecuador, moremarket studies have been recentlyconducted, which generated evidenceof potential demand for alternativecassava flour uses in nonconventional,industrial products (CENDES, 1993).In Brazil, cooperative-level processeddata have been fed back to farmerorganizations within a month,allowing them to assess their ownperformance and relate it with that ofother farmer groups.

Policy support and decisions

From their very inception, ICRDPshave been closely related to andaffected by policy decisions andsupport. For example, all countries intropical Latin America are netimporters of cereals and mostgovernments in the region have triedto supply this increasing demand forcarbohydrates through policyinterventions and subsidized

production credit. This has meantthat traditional starchy staples, suchas cassava, have had to compete withgrains at a substantial disadvantage.

Exploiting postharvestopportunities for root and tuber cropsis currently less of a technologicalproblem, given the extensive expertiseavailable. The central issue indeveloping cassava-based marketsand products is the economics of thewhole production and marketingsystem. This is directly affected bypolicy interventions oriented towardstrengthening the bargaining powerand the organizational levels ofcassava producers.

In the Colombian project, policyissues were present from the verystart. The pilot project was begun inan area where an on-going land reformprogram was operating: farmers werereceiving credit and technicalassistance aimed at increasingcassava production in the region.Farmer organizations even had accessto credit for cassava production andprocessing and for constructingprocessing infrastructure. TheGovernment controlled cereal importsand included dried cassava in thepolicy of minimum prices foragricultural products. This latterpolicy was first established, on asix-monthly basis, in 1990 by theMinistry of Agriculture.

Policy issues became even moreimportant during 1993/94 whendecreased import duties (a result ofColombia’s economic aperture) allowedhigh-quality cassava pellets fromIndonesia to be imported at “dumping”prices. This act set off a series ofhigh-level discussions that broughttogether representatives of governmentresearch and extension institutes, theprivate sector, second-ordercassava-processing organizations,and CIAT. They then discussed theframework, individual responsibilities,

354


and an action plan for a collaborative,long-term effort to optimize theeconomic sustainability of thecassava sector in the North Coast ingeneral and of the ICRDP inparticular.

In the Ecuadorean project, thelack of government intervention toprovide small-scale credit has beencrucial in impeding the establishmentof cassava-based agroindustries,preventing project activitiesexpanding to other potential regionsand cassava-producing areas.

In the Brazilian project, cassavafarmers benefited from policydecisions. Ten financial programsprovided grants that helped establishcassava-processing plants. Twocredit programs for cassavaproduction and processing, based onprice variation of cassava products,provided a certain stability of creditfor farmers within the country’shighly unstable economic situation,typified by inflation rates of 25%-30%per month.

Conclusions

Three key conclusions result from thecomparative analysis of the threeICRDPs (North Coast region,Colombia; Manabí Province, Ecuador;and Ceará State, Brazil).

(1) Integrating production, processing,and marketing research anddevelopment activities

The ICRDPs clearly demonstrate thatresearch and development must beintegrated if the cassava crop’s fullpotential is to be effectively realized.The intertwined relationships anddependencies of production,processing, and marketing make itinefficient and illogical forinstitutions—whether national orinternational—to work exclusively on

any one cassava activity in isolationfrom the others.

The ICRDPs provide an appropriatemechanism to bring together theseactivities in a context where severalkinds of institutions—including farmerorganizations—can collaborateeffectively.

For CIAT, as an internationalresearch center, the ICRDPs haveprovided a crucial testing-ground forlinking production and processingtechnologies, and for developingappropriate socioeconomic tools formarket and monitoring research. Thefeedback from the results has servedto shape priorities for future CIATresearch directions. To maintainrelevance to cassava farmer andprocessor needs, CIAT must preservestrong links with ICRDPs activities andan equally strong human and technicalresource capacity in the areas ofproduction, postharvest, andsocioeconomics. Partnerships andcollaborative arrangements betweenCIAT and national entities are a mustfor future activities.

Strengthening farmer organizationsand their links to research anddevelopment are critical objectives forthe future. The ICRDPs offer bothinternational and national institutionsa framework on which to buildcollaborative working arrangementswith farmers through theirorganizations.

(2) Providing important social andeconomic benefits

The ICRPDs fulfill this role to holdersof small and medium-sized farms andlandless rural workers in marginalfarming sectors with few alternatives.Cassava’s exceptional adaptability tosuch marginal areas makes it a naturalindicator for poorer households and anappropriate vehicle for organizingfarm-level, income-generating,

355


productive activities. The ICRDPsact as “magnets” for other types ofdevelopment efforts and can provide abase to anchor and integrate these,thus contributing to increased socialstability and economic growth.

(3) Farmer investment in improvedproduction technologies

The ICRDPs have clearly proved thatwhen increased value for the cassavacrop is created through identifying newmarkets and developing new productsto suit these markets, farmers willinvest in improved productiontechnologies. Providing an appropriateincentive for farmers to invest in theircassava production systems hasprofound implications for using newtechnologies to increase productivityand to induce resource sustainability.

Future Steps

Looking beyond the immediateconclusions drawn from the ICRDPs’current experiences, one can seeseveral important tasks yet to beaccomplished.

First, despite the many years ofcollaboration between nationalprograms in ICRDPs, there is relativelylittle consolidation of the experiencesand lessons learned from the individualprojects, and what has been written isnot yet widely available for public use.Most of the experiences remain lodgedin the minds of practitioners whodedicated considerable portions of theirprofessional careers to these projects.CIAT must make a concerted effort todocument these experiences, analyzethe results, and make them availablefor wider consumption.

Second, there is a crucial need tocouple this documentation withtraining programs that distill theICRDP methodologies from caseexperiences, and transform them into

appropriate training materials. These,in turn, will provide the vehicles bywhich others will learn how to planand implement ICRDPs in othercassava-producing regions in LatinAmerica, Africa, and Asia.Concomitantly, such materials needto be dynamic, that is, created in aformat that allows new lessons andexperiences from more recent projects(on all three continents) to be assessedand incorporated.

Third, the ICRDPs would gain timeand reduce duplication of negativeexperiences through networking andexchanging visits between projects andthrough training and technicalassistance between technicians andfarmers. But no structure exists tocontinue such exchange andcollaboration. Funding and leadershipare needed to create this structure.CIAT could contribute significantly byestablishing norms for suchinteractions to take place.

Technology generated by publicfunds and agencies must remain freelyaccessible in the public domain. At thesame time, private sector participationmust be encouraged and its interestsunderstood and accommodated in anequitable fashion. This will requireconsiderable international diplomacyand negotiation.

If a networking program is firstplaced within an existing,agroindustrial, regional network(e.g., Programa para el DesarrolloAgroindustrial Rural [PRODAR] inLatin America and the Caribbean), thenadministrative costs would be reducedand duplication of efforts prevented.The ICRDP experience would be passedto other productive sectors orcommodities that could benefit fromthis integrated approach. Likewise, theICRDPs would benefit from connectionsto other possible agroindustrialtechnologies that could diversifycurrent farmer organizations’ outputs.

356


Linking an ICRDP from Latin Americaand the Caribbean region with those ofsimilar interests in Africa and Asia maycreate further possibilities for internalgrowth, lessen duplication, and reducetechnology development lag time. Itcould create greater horizontalexchange across regions where similarcassava problems and opportunitiesexist. These efforts may encouragefarmer-to-farmer communication andassistance across large distances andperhaps enable cassava developmentto occur in areas where other, morecostly, institutional efforts have failed.

Finally, because cassava is oftengrown in marginal environments wherethe resource base is rapidly beingdegraded, ICRDPs offer an idealopportunity to explore with farmers thequestions and problems of long-termsustainability for cassava-integratedsystems. Farmer-processors who havelearned and earned the value that newmarkets can give their cassava cropshave an incentive to conserve theirresource base and ensure that itsproductivity will endure. Such farmersand their organizations can becomewilling collaborators in expanding theICRDP’s focus to a landscapeperspective where the longer termmanagement of cassava is but one partof a complex resource managementsystem.

Mature ICRDPs must now turntoward these more complex problemsand begin to focus attention on longerterm sustainability. Explicit attentionmust be directed to the systemicimpact of cassava production andprocessing, including work onproductive capability, water and wastemanagement, and relationships withcomplementary and competingsystems. If ICRDPs can indeed widentheir horizons and incorporate theseissues and problems, then they mayachieve a long-term, positive impact onthe lives of rural people depending oncassava.

References

Bode, P. 1991. Monitoring and evaluationsystems for cassava drying projects.In: Pérez-Crespo, C. A. (ed.).Integrated cassava projects.Working document no. 78. CassavaProgram, CIAT, Cali, Colombia.p. 214-246.

Brouwer, R. 1992. The cassava flour demandin the plywood industry inEcuador. Thesis research report.Department of Market Research,Agricultural University, Wageningen,the Netherlands. 108 p.

CENDES (Centro de Desarrollo). 1993.Estudio de mercado para conocer lademanda potencial de productoreselaborados de yuca. Unión deAsociaciones de TrabajadoresAgrícolas, Productores y Procesadoresde Yuca (UATAPPY) and CENDES,Quito, Ecuador.

CIAT. 1992. Cassava Program Report1987-1991. Working documentno. 116. Cali, Colombia. 477 p.

__________. 1993. Trends in CIATcommodities. Working documentno. 128. Cali, Colombia.p. 173-182.

Cock, J. H. and Lynam, J. K. 1990. Researchfor development. In: Howeler, R. H.(ed.). Proceedings of the 8thSymposium of the InternationalSociety for Tropical Root Crops(ISTRC), Oct. 30-Nov. 5, 1988,Bangkok, Thailand. CIAT, Bangkok,Thailand. p. 109-119.

Gottret, M. V. and Henry, G. 1994. Laimportancia de los estudios deadopción e impacto: el caso delproyecto integrado de yuca en laCosta Norte de Colombia. In: Interfaseentre los programas de la yuca enLatinoamérica. Working documentno. 138. CIAT, Cali, Colombia.p. 193-223.

Henry, G. 1992. Adoption, modification andimpact of cassava dryingtechnology: the case of the ColombianNorth Coast. In: Scott, G. J.;Ferguson, P. I.; and Herrera, J. E.(eds.). Product development for rootand tuber crops, vol. III. CentroInternacional de la Papa (CIP), Lima,Peru. p. 481-493.

357


__________ and Best, R. 1994. Impact ofintegrated cassava projects amongsmall-scale farmers in selected LatinAmerican countries. In: Ofori, F. andHahn, S. K. (eds.). Tropical root cropsin a developing economy: proceedingsof the Ninth Symposium of theInternational Society for Tropical RootCrops (ISTRC), 20-26 October 1991,Accra, Ghana. ISTRC, Wageningen, theNetherlands. p. 304-310.

__________; Izquierdo, D.; and Gottret, M. V.1994. Proyecto integrado de yuca en laCosta Atlántica de Colombia: Adopciónde tecnologia. Working documentno. 139. CIAT, Cali, Colombia.

Janssen, W. 1986. La demanda de yuca secaen Colombia. In: Best, R. and Ospina,P. B. (eds.). El desarrollo agroindustrialdel cultivo de la yuca en la CostaAtlántica de Colombia: Cuarto informesobre las investigaciones realizadas enapoyo al establecimiento de las plantasde secado natural de yuca, períodojulio 1984-junio 1985. ProyectoCooperativo Fondo de Desarrollo RuralIntegrado (DRI)-CIAT. CIAT, Cali,Colombia. Vol. 2, p. 41-50.

Lynam, J. K. 1978. Options for Latin Americancountries in the development ofintegrated cassava productionprograms. In: Fisk, E. K. (ed.). Theadaptation of traditional agriculture:socioeconomic problems ofurbanization. ANU DevelopmentStudies Centre monographs, no. 11.Australian National University,Canberra, A.C.T., Australia.p. 213-250.

__________. 1987. Cassava consumption inevolution in Latin America: staple orvegetable. International Food PolicyResearch Institute (IFPRI), Washington,DC. 38 p.

MAG (Ministerio de Agricultura), Departamentode Programación, Instituto Nacional deEstadísticas y Censos. 1990. Encuestade superficie y producción pormuestreos de áreas: Resultados de1990, vol. 1. Portoviejo, Manabí,Ecuador.

Pérez-Crespo, C. A. (ed.). 1991. Integratedcassava projects. Workingdocument no. 78. Cassava Program,CIAT, Cali, Colombia. 242 p.

358


CHAPTER 39

THE CASSAVA FLOUR PROJECT INCOLOMBIA:FROM OPPORTUNITY IDENTIFICATION TO

MARKET DEVELOPMENT

Carlos F. Ostertag, L. Alonso, Rupert Best*,and C. C. Wheatley**

Product developmentconcentrates on three main areas:first, the generation, evaluation, andselection of ideas for new products,in this case cassava based; second,the development of a productprototype and process design,accompanied by industrial and/orconsumer research; and, third,product presentation, that is, qualityspecifications, product name, andpackaging.

The term “integrated cassavaproject” describes a ruraldevelopment strategy, executed inthree phases by rural inhabitants, topromote the agroindustrialtransformation of cassava throughthe integration of production,processing, and marketing functionsand supported by governmental andnongovernmental organizations.

The research phase is in twoparts: national analysis, in which thenational economy, the commercialoutlook for cassava, and thepotential of ideas for newcassava-based products are studiedto select new products and a region.In the second part—regionalanalysis—the selected region isstudied in greater detail, especiallyregarding cassava production, farmerorganizations, and nearby markets,to select the best scenario for a pilotproject.

Abstract

The cassava flour project in Colombiabegan in 1984 with financing fromthe International DevelopmentResearch Centre (IDRC). It seeks toincrease the income of small farmersin cassava-growing areas by creatingan agroindustry focused on theproduction of cassava flour for humanand industrial consumption.

The following discussion of theproject “Production and marketing ofcassava flour in Colombia” outlinesthe underlying methodologicalframework, and describes the projectactivities executed during theresearch, pilot project, and expansionphase. Emphasis is given to the pilotproject.

Methodological Framework

The project uses integrated projectand product developmentmethodologies. During the researchstage, activities can be seen asbelonging to one or the othermethodology (Figure 1), but asblending in the pilot project.

* Cassava Program, CIAT, Cali, Colombia.** Centro Internacional de la Papa (CIP),


359

The Cassava Flour Project in Colombia:...

Figure 1. Outline of the integrated project and product development methodologies.

During the second phase—pilotproject—a pilot plant is establishedand operated semicommercially underreal market conditions to determinethe feasibility of the agroindustry.

In the final phase—expansion—the processing units are replicatedand the market for the product isexpanded to consolidate the newagroindustry.

Evolution of the CassavaFlour Project

Research phase

The objective of this first phase was todetermine the technical and economicconditions required for developing thecassava flour agroindustry in

Colombia. The main use anticipatedfor cassava flour was in thepreparation of a wheat and cassavacomposite flour for bread making.

The region selected, the NorthCoast—also known as the AtlanticCoast—is the main cassava-producingarea in Colombia, with the root grownmostly by small-scale farmers.Accordingly, the economy of cassavaproduction in the North Coast wasstudied, along with the wheat-millingand bread-making sectors. Theequipment for small-scale ruralprocessing of cassava was adaptedand developed. In addition, theinfluence of cassava varieties on thequality of cassava roots and derivedproducts was examined. Surveysamong consumers and bread makerswere conducted to evaluate the

Phase

I Research Integrated project Product development

National analysis IdeasRegional analysis Research

Project design

Plant construction

Plant operation

ExperimentalSemicommercial

Testmarket

Feasibility study

III Expansion Expansion phase

II Pilot project

360


acceptability of breads made fromcomposite flours of wheat andcassava.

The research concluded that thedevelopment of a cassava flouragroindustry was viable becausecassava flour could be sold at lowerprices than wheat flour, andconsumers found the composite breadacceptable. However, bakers saw ahigh risk in lowering the quality oftheir products by using cassava flour.

The decision was made tocontinue with the pilot project, on theunderstanding that alternativemarkets for cassava flour wereidentified.

Pilot project phase

In this phase a pilot cassava flourplant was set up and operated underreal market conditions to assess thefeasibility of establishing the newagroindustry. The following activitieswere carried out:

Adjusting and evaluatingproduction. A set of criteria (such asstability of farmer organizations orperformance of cassava crops) wasdetermined and used to select the sitefor the pilot plant at Chinú, Córdoba.The Cooperativa de Productores de losAlgarrobos (COOPROALGA) waschosen as the executing farmerorganization. The pilot plant wasdesigned, and a local civil engineeringfirm built it within 3 months. Most ofthe equipment and machinery wasmanufactured in a Cali workshop butthe metallic coal burner was availablecommercially. A well was dug tosupply the plant with water.

Workers and administrativepersonnel were selected and trained.A daily and weekly timetable ofactivities was drawn up and analmost year-round supply of freshcassava roots coordinated. In total,

42 t of dried chips were produced andtransported by road to be milled in acommercial wheat mill in Medellín.

An information system forproduction was developed andimplemented, and control parametersestablished. Specifications for thequality of raw material and sanitarycontrols were drawn up. Themicrobiological quality of the cassavaflour was monitored, and variablecosts of production closelysupervised.

Support research was conducted,with the collaboration of theUniversidad del Valle (UNIVALLE) andthe Natural Resources Institute (NRI),UK. Areas investigated included theimprovement of processingequipment, control of microbiologicalquality of cassava flour, developmentof a small-scale milling system forcassava chips, research on storage ofcassava products, and development ofmoisture-measuring equipment forcassava products.

Testing and demonstrating animproved cassava productiontechnology for the North Coast.Since 1989, 120 farmer-managedpre-production plots fordemonstration were established onthe North Coast with cassava-maizeand cassava-maize-yamcombinations; farmers weresupervised by an agronomist.

The recommended technologies,which improved maize and cassavayields, combined adjustments in theuse of preemergent herbicides,fertilization of maize and yam, moreintensive use of human labor, anduse of improved maize varieties.

Identifying markets for cassavaflour and product promotion. Asdescribed above, the focus on breadmaking was modified after theresearch phase. Market opportunities

361


were sought in other food industrycategories where cassava flour wouldhave an equal or better functionaladvantage or where it could besubstituted, partially or completely,for other flours or starches.

A market study was conductednationally among food companies ofdifferent sizes. The study firstfocused on products marketed andraw materials; then, flour sampleswere distributed for substitutiontrials; and, finally, feedback wasobtained on the trials and buyingintention was gauged. The studyshowed that potential markets forcassava flour included processedmeats, cookies, ice-cream cones,pasta, pastry, soup and sauce mixes.Cassava flour exhibits functionaladvantages in most of these products.More than 80% of the volume wouldbe destined to replace wheat flour.Assuming that cassava flour could besold for 10% less than wheat flourand that there would be adequatepromotion, the estimated mid-termmarket demand would be20,000 t/year.

The promotional effortconcentrated on Medellín, which hadmilling facilities and the largestsingle market detected in the study.Sixteen firms were visited and givenfree samples of flour. Thesubsequent trials were closelymonitored. Inferences from thisexperience were that themicrobiological quality was notacceptable to most companies, thatthe food industry was conservative,and that sales efforts would benefitfrom better technical information oncassava flour.

The flour developed wasyellowish white and contained about80% starch. Its granule size wassmaller than that of wheat flour. Itwas called “Yukaribe,” and packagedin polypropylene sacks, complete

with a graphic design. The flour waspriced at 15% below wheat flour.

Feasibility of the agroindustry:pilot project phase. A computerizedfinancial model of the pilot plant wasdesigned and updated periodically tomonitor production costs, plantefficiency, and profitability. At theend of the pilot project phase, thefeasibility of the cassava flouragroindustry was seen as follows:

(1) Technical feasibility. The artificialdrying process was inefficient; andthe microbiological quality ofcassava flour was substandard.

(2) Commercial feasibility. Additionaltechnical information wasrequired. The physicochemicaland microbiological qualitiesneeded improvement.

(3) Cooperative-managementfeasibility. Sales and marketingpersonnel were needed to handleproduct marketing.

(4) Economic feasibility. The financialrate of return (FRR), a profitabilityparameter, was calculated at 22%,which was considered low.

Expansion phase

The project could not proceed with aformal expansion phase because ofthe constraints described above, butsome action could be taken inpreparation for a future expansion.A hybrid pilot/expansion phase wasdeveloped to convert the pilot plantinto a commercial operation withimproved profitability.

Artificial drying costs were greatlyreduced by doubling heat generationand switching from coke to mineralcoal. This resulted in a shorterdrying period and improved flourmicrobiological quality.

The plant received a smallcassava-chip mill, developed jointlyby CIAT and UNIVALLE. The mill

362


consisted of a premilling componentthat reduced chip size and twocylindrical screens that alsofunctioned as mills. The output wasa first-grade flour (70%-85%extraction) and bran. In-plant millingreduced variable costs andcontributed toward satisfying localdemand for the product.

Members of COOPROALGA, thefarming cooperative managing thepilot plant, were trained in theadministration of small enterprises.

Developing and executing aplan for expanding the cassavaflour agroindustry in Colombia.Cassava flour was promoted amongthe North Coast food industries,especially meat-processing, cookies,and spices, with the eventualpenetration of the meat-processingsector.

However, to increase sales further,the marketing strategy was changedand cassava flour was promoted innonfood industries—especiallyadhesives and plywood—wheremicrobiological quality was lessimportant and higher market pricescould be obtained. Adhesivecompanies in major Colombian citieswere provided with samples ofcassava flour and, simultaneously,the Fundación para la Investigación yel Desarrollo de TecnologíasApropiadas al Agro (FUNDIAGRO)provided support in the developmentof cassava flour-based adhesives. Theadhesive markets in Cali andBarranquilla were penetrated,although industrial requirementsdemanded increased flour purity byreducing the extraction rate duringmilling.

The design of the prototypebuilding for the processing plant wasrevised to reduce costs and increaseperformance in accordance with thepilot project experience. Designing

involved a team of architecturestudents from UNIVALLE, supervisedby a member of the university staff.

Training materials, includingvideos and manuals on productionand management, were developed.

Feasibility of the agroindustry:expansion phase. By the end of1993, the feasibility status of thecassava flour agroindustry was seenas follows:

(1) Technical feasibility. Foodindustry: yeasts levels were toohigh. Adhesives industry: nolimitations.

(2) Commercial feasibility. Foodindustry: additional technicalinformation on cassava flour wasrequired by firms; physicochemicaland microbiological qualityrequired improvement; cassavaflour price is competitive againstwheat flour only in the NorthCoast. Adhesives industry:additional technical informationwas required by firms.

(3) Cooperative-managementfeasibility. Food and adhesiveindustries: sales and marketingpersonnel were needed to handleproduct marketing.

(4) Economic feasibility. Foodindustry: FRR was 26%.Adhesives industry: FRR wasabove 30%.

Conclusions

The major outputs of the cassavaflour project were:

(1) The development of an efficientsmall-scale system for cassavaflour production.

(2) Although members of theexecuting cooperative had beentrained to manage the plant, amajor priority was to improve thequality of the raw material used in

363


the plant, including industrialvarieties.

(3) Project feasibility is uncertain,because of high costs, deficientsupplies, poor quality of theregion’s cassava roots, andinsufficient entrepreneurialcapacity of the executingcooperative.

(4) To plan successful ruralagroindustrial projects, thefollowing points must beconsidered:

(a) the importance of theintegrated, or entrepreneurial,approach, encompassinginterventions in production toguarantee a sufficient supplyof low-priced, quality rawmaterial;

(b) the need to assign enoughfunds and time for productdevelopment and marketing;and

(c) the need to identify projectexecutors with entrepreneurialabilities.

Bibliography

Best, R. and Ostertag, C. F. (eds.). 1988. Theproduction and use of cassava flourfor human consumption: researchphase. Final report of a collaborativeproject (Oct. 1984-Oct. 1986). CIAT,Instituto de InvestigacionesTecnológicas (IIT), and theUniversidad del Valle, Cali, Colombia.85 p.

Ostertag, C. F. 1993. Plan de mercadeo paraharina de yuca (July 1993-June1994). Working document. UtilizationSection, Cassava Program, CIAT, Cali,Colombia.

__________ and Wheatley, C. C. 1992.Proyecto de producción ycomercialización de harina de yucapara consumo humano: Informe final,Fase de proyecto piloto (junio 1989-diciembre 1991). CIAT, Universidaddel Valle, and Fondo de DesarrolloRural Integrado (DRI), Cali, Colombia.

__________ and __________ (eds.). 1993.Production and marketing of cassavaflour in Colombia: expansion phase.Annual report of a collaborativeproject (Jan.-Dec. 1992). CIAT andFondo de Desarrollo Rural Integrado(DRI), Cali, Colombia. 40 p.

__________ and __________. (eds.). 1994.Producción y mercadeo de harina deyuca en Colombia: Fase deexpansión. Annual report of acollaborative project (Jan.-Dec. 1993).CIAT, Universidad del Valle,Fundación para la Investigación y elDesarrollo de Tecnologías Apropiadasal Agro (FUNDIAGRO), and Fondo deDesarrollo Rural Integrado (DRI),Cali, Colombia. 41 p.

364


CHAPTER 40

WOMEN AS PROCESSORS AND TRADERS

OF CASSAVA FLOUR:THE PHILIPPINE EXPERIENCE

D. L. S. Tan, J. R. Roa, and E. A. Gundaya*

areas, which comprise about15%-20% of land use in thePhilippines. About 500,000small-scale and marginal farmers relyon root crops for food security andsupplementary cash income. Becausethe traditional ways of consumingroot crops are few, optimizing theiruses depends on an expandedagroindustrial market. This meansfocusing on postproductiontechnology.

Studies in Asia and Africa haveshown that women farmers are largelyinvolved in postharvest activities,particularly selling and processing.Thus, the “Women in PostproductionSystems” (WIPS) Project” wasconceived to improve women’sefficiency in postproduction activitiesand increase household incomes.This was first spearheaded by theSoutheast Asian Regional Center forGraduate Study and Research inAgriculture (SEARCA), collaboratingwith the Philippine Root CropResearch and Training Center(PRCRTC), the National PostharvestInstitute for Research and Extension(NAPHIRE), and the Isabela StateUniversity (ISU) for root crops, rice,maize, and groundnuts. Fundingcame from the InternationalDevelopment Research Centre (IDRC).

Previous surveys in the Philippinesshowed that women are active as farm

Abstract

Earlier surveys revealed thatcassava chipping is traditionalamong women in Mabagon, acassava-growing village in thePhilippines. This paper discusses theexperiences gained from the pilotproject conducted there to increasethe women’s processing efficiency byimproving cassava chip and flourprocessing technologies and to assessopportunities for market expansion.

An association, consisting mostlyof women (16 out of 19 members), wasorganized and trained to operate thepilot plant and to promote and marketthe cassava flour produced. The pilotoperation began in October, 1991.The association now produces20-75 bags (at 20 kg/bag) of cassavaflour per month. These are sold tonearby bakeries, which use thecassava flour to prepare differentbaked goods. Plans for a full-scalecommercial expansion are alreadyunder way.

Introduction

Root crops are given high prioritybecause of their ubiquity in upland

* Philippine Root Crop Research and TrainingCenter (PRCRTC), Leyte, Philippines.

365

Women as Processors and Traders of Cassava Flour:...

helpers, sellers, and processors ofroot crops into native delicacies. Animportant finding was the indigenousprocessing of cassava into dried chipsin the islands of Leyte, Bohol, andthe Camotes, and Misamis OrientalProvince, for feed, food, and trade.Women were largely involved. Thisfinding provided an opportunity forintervention by the project.

A diagnostic survey of root-cropfarming households in LeyteProvince, and an inventory ofpossible technologies that could fitinto local systems, helped develop theidea of introducing components ofcassava flour technology (which usesdried chips) to women farmers in thearea around the town of Hindang,Leyte Province. The chips made inthis area are of relatively good qualityfor flour.

This paper serves as amethodological note on the PRCRTC’sexperience in cassava flourprocessing and commercialization,involving mostly women. This phase,which started in October, 1991, is anintegral component of the WIPSProject.

Objectives

The project has the followingobjectives:

(1) To introduce root-crop equipmentthat would improve the efficiencyof cassava flour processing in aselected community (i.e.,Mabagon village near the townof Hindang, Leyte Province).

(2) To test storage technology fordried chips and flour.

(3) To strengthen the capability ofbeneficiaries in organizationaland entrepreneurial skills.

(4) To assess the effects of theintroduced technologies onfarming households.

Project Site and Beneficiaries

The project site is Mabagon, a villagewith a little more than 100 farminghouseholds, situated about 3 km tothe northeast of Hindang. Copra (fromcoconut) and palay (from rice) arethe main agricultural products ofHindang’s hinterland, But most ofthe lands planted to these crops areowned by a few, relatively rich,landholders.

Small-scale farmers farm theuplands, 70% of them owning theland they till. A transect showed apredominance of upland cultivationof either sequential cropping orintercropping of cassava, sweetpotato,maize, vegetables (string beans,ampalaya, eggplant), and upland rice,with patches or fringes of taro andbanana.

Most men receive income byworking as hired hands (planting,harvesting, threshing) in the rice fields,selling upland cash crops (such as rootcrops, bananas, and vegetables), andraising livestock. Most women areengaged in selling various farmproducts and processing cassava, onthe farm, into dried chips for feed.

This mixed farming system, whereindigenous cassava processing playsan important role, was a promisingmatch for village-based, cassava-flourprocessing and “shovel” feed mixingfrom byproducts. More especially, thelocal system fitted the project’sparticular concern for gender roles infarming systems and for improvingwomen’s cassava postproductionactivities.

The Mabagon Root CropAssociation (MARCA), the coregroup of collaborators, consisted of19 farmers and processors: 3 men and16 women. It was formed after severalconsultations among the local peopleand was finalized during the general

366


innovations unknown. The projectwas integrated in the sense that amultidisciplinary team implemented it,coordinating the phasing of variouscomponents, both technical andsocioeconomic.

This approach is characteristicallysystems-oriented, interdisciplinary,participatory, and oriented by users’perspectives, local knowledge,practices, and norms.

Project implementation involvedcarrying out activities of differentcomponents designed in a stepwisebut interphased manner to effectivelytransfer the technology to afunctioning enterprise (Table 1).

MARCA: Its Progress to Date

Registration

By consensus, MARCA was firstorganized as an association andregistered with the Department ofLabor and Employment (DOLE). Thiswas partly because about half of themembers, at first, resisted acooperative registration, seeing aconflict of interest with an alreadyexisting cooperative in the village.Later, realizing the benefits of forminga cooperative, particularly that ofobtaining funds, MARCA membersvoluntarily agreed to registration as aprocessing cooperative.

With the registration, thecooperative met the requirements toreceive support from the CountrywideDevelopment Fund (CDF) for theprocessing facility. The registrationand fulfillment of other requisites(including articles of cooperation,constitution and bylaws, seminar, andeconomic survey) also brought supportfrom the Cooperative DevelopmentAuthority (CDA). MARCA also receivedassistance in registering with theDepartment of Trade and Industry

assembly in December, 1991.Membership was voluntary; interest,commitment, and availability for thegroups’ activities were prerequisites.Currently, MARCA’s registration as acooperative is in process, with allrequirements already met.

Technologies

The component technologies, pilottested for cassava flour processing,included:

(1) PRCRTC-developed,.village-level,cassava flour processing machines(chippers, modified tapahan dryer,grinder, and flour finisher).

(2) Storage for chips.(3) Technology for byproduct use, that

is, a neutral-scale “shovel”technology, in which cassava mealis mixed with other ingredients toproduce a feed for swine.

In the initial phase ofcommercialization (toward the middleof the second year), expansion ofmarket uses was explored. This led tonew food-processing productstechnologies and, therefore, newbakery products being introduced tobakeries and to MARCA (e.g., cacharon[a puffed product with various flavors],polvoron, and processing cassavasticks and chips from fresh roots).

Methodology

The project took an integrated processapproach, that is, a set of strategieswas developed to coordinate neededcomponents and was flexible enoughto allow redesigning as new orimproved methods were tested,refined, and disseminated through thetargeted beneficiaries. This approachwas essential for the project’s successat the village level, whereuncertainties were common and thereaction of people to introduced

367


Table 1. Components and methods used for processing and commercialization of cassava flour toMabagon village, Leyte Province, the Philippines, 1991.

Component Methods

Technical:Farmer-processors Training, processing trials(for both equipment and Informal team or group discussionsprocessing) Feedback

Participant observation

Bakers TrainingBaker-to-baker visits

Production: Study or field visitsFarmers Farmer-to-farmer visits

Organizational buildup and Participant observationentrepreneurship development On-the-job training (e.g., in recording, inventory-taking,

purchasing)Team buildup and group dynamicsSpecialized skills training (e.g., marketing, bookkeeping,

accounting, and keeping financial records)Technical assistance (e.g., registration)Advisory discussions or consultations, formal meetings

Market development Market research (unstructured, use of checklist-users’ survey,feedback, contacting other markets)

Researcher and farmer partnerships:(1) Expanding flour uses and testing markets (ready mixes

with packaging and product testing)(2) Institutional collaboration for promotion(3) Byproduct use: coarse-grained flour in feed mix for swine(4) Integrated enterprise scheme

Monitoring and evaluation Informal group discussions, field visits, meetings(MARCA team, local advisory group)

WorkshopsResident research assistant’s logbook and diaries

(DTI), Bureau of Food and Drugs(BFAD), and a local nongovernmentalorganization.

Group buildup and entrepreneurialdevelopment

Building up entrepreneurial skillsand strengthening the group wereessential components for operationalsustainability. These were carried outthrough:

Sessions on group dynamics.The first four sessions were designedto form values and attitudes that

would condition individual memberstoward effective group endeavor. Theremaining sessions concentrated onlearning entrepreneurial skills, thatis, planning, organizing, operationalmanagement, control, andevaluation.

Specialized training inentrepreneurship. Sessions weregiven in collaboration with the DTI.An initial, one-day, entrepreneurialappreciation session was conductedwith emphasis on marketing skills.Other sessions included bookkeepingand marketing skills.

368


On-the-job training. During fieldand monitoring visits, researchers gaveinformal consultations and discussionson business management, marketing,recording, inventory-taking, productquality control, and equipment andfacility maintenance. The residentresearch assistant gave a tutorial aspart of the post’s responsibility.

Workshops. Two workshops wereconducted at the PRCRTC in August1992 and March 1993. They weredesigned as learning exercises forMARCA members in presentation andanalytical evaluation. The firstworkshop included MARCA and theother agencies involved in the project.The second was an attempt by thePRCRTC to encourage the differentpilot project collaborators, andgovernmental and nongovernmentalagencies to interact and learn fromeach others’ experiences and to definetheir respective roles in this and otherdevelopment programs.

Changes in the operational scheme

Cassava flour processing follows adecentralized operation where roots arechipped and dried to a specified qualityby individual farmers, then sold to theplant. The chips are then milled andstored. MARCA distributes finishedflour on a per order basis, once a week.The schedule, organization, andmanagement of operational activitieswere made according to the members’time and ability.

Several changes resulted frommembers’ experiences and feedback.For example, the daily shift for athree-member processing team waschanged because some members wereunavailable for family and workreasons. Starting January 1993,MARCA hired a regular processor whowas paid P40.00 daily. The MARCAmembers took turns in assisting theregular processor. Each member waspaid P35.00 daily.

Developing an integrated enterprise

The viability of the cassava flourprocessing enterprise is expected toevolve only gradually because ofcompetition from wheat flour and thelag involved in learning to use cassavaflour. MARCA needed to engage inenterprises related to its flourprocessing operations to solve liquidityproblems, improve income, and enhancecapacity as a market strategy. Thus, in1993, the following operations wereintegrated into MARCA’s businessactivities:

Trading. MARCA buys freshroots from farmers and sells them to astarch plant situated about 80 km tothe north. Trading is carried out fromMay to December when cassava dryingis difficult because of the rainy season.Dried chips are bought during the dryseason (January to April). Tradingstarts in the last week of November.

Operating a cooperative store topromote flour and feed. In November1993, MARCA opened a village store(Mabagon is strategically located,serving mountain barangays) to sellflour, feed, and other cassava-basedfood products. Noncassava feeds forswine, such as the widely used basefeed, were also sold to promote theMARCA feed mix, which had a higherprotein content (14%-16% crudeprotein).

Market Development

Flour

The market target (25-30 bakeries, witha daily average total of eight bags, 10%cassava to wheat ratio) that wouldoptimize plant use (4 t flour/month)was not reached within the first6 months of operation. Only fourbakeries were regular users and eightothers were irregular, resulting in anaverage plant use rate of only 20%-30%.

369


The reasons for the slowpenetration of the market were, first,learning how to handle cassava flourdough requires time, if both ownerand baker are interested. If the priceincentive was too low or the learningtime too long, owners and bakerslost interest. Second, the number ofbakers trained unexpectedly fellbelow target for the first 6 months.Third, the existing market used flourat a lower rate than expected (i.e.,less than eight bags per day).Fourth, during the first year, thepilot project was still testing thestability of chip and flour quality, aswell as test marketing, and anintensive marketing campaign wasnot being pursued at the time.

Given the results of this phase,the following marketing strategieswere undertaken:

Visits and training.Consultation workshops,baker-to-baker visits, and trainingsessions were conducted to createthe market for cassava flour.

Twenty-five bakeries on Leyte Islandwere involved, from Maasin(about 45 km to the south) to Ormoc(about 80 km north). Visits betweenbakers gave them a chance toexchange ideas and experiences, andwere more effective in transferringtechniques and knowledge than wastraining. Because of the peculiaritiesin handling cassava-wheat compositeflour, further market expansion willtake time, as more training andexchanges are needed.

Market survey. The mainconstraints remained the lack ofskills in using cassava flour for breadand the lack of price incentive. In1993, cassava flour sales declinedby more than 50%. From theinformal market feedback, thenonsustainability of even the previousregular market may have been partlya result of the MARCA personnel’sinadequate marketing efforts(Figure 1). This meant that thetrainees needed intensive exposureand training during the remainingmonths.

Figure 1. Sales of cassava flour in Mabagon village, Leyte Province, the Philippines, 1992/93.

100

80

60

40

20

0Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June1992 1993

Nu

mber

of bags

370


Promotional posters for cassavaflour. Initially, 150 promotionalposters were produced with fundingfrom the PRCRTC’s extensionprogram. These are now distributedin Leyte and Samar Islands andSurigao Provinces, and at the tradeexhibits in the city of Cebu and at thePhilippine Council for Agriculture andResources Research and Development(PCARRD) at Los Baños (near Manila).

Testing other product ideas.Mixes with cassava-wheat compositeflour were developed for theconvenience food market. The500-g packages, with recipes and animproved packaging design, was readyfor promotion in November 1993.

Other markets explored. Adistribution chain of native foodproducts, with 150 outlets throughoutthe country and based in Manila,expressed interest in cassava flour.Samples of cassava flour, cassavameal (2 grades), and dehydratedcassava gratings were sent for testing.The cassava flour passed the qualitytest but the texture needed to be finer,to pass through a 60 mesh (down fromthe current 80). The dehydratedgratings were also acceptable and thefirm is interested in placing an initialorder. Grating equipment has beeninstalled for processing and marketingtrials for the first quarter of 1994. Thecassava meal was unacceptable forready mixes.

Processing other food productsfrom cassava flour

To expand the market for cassavaflour, MARCA members were trainedin food processing. Three trainingsessions on various food productsfrom cassava flour and fresh rootswere conducted for MARCA members,one at Visayas State College ofAgriculture (ViSCA) and two at thesite. Because of a more promisingmarket, MARCA decided to

concentrate on cacharon and polvoron.The equipment is made andprocessing trials are expected to startduring the first quarter of 1994.

Use of byproducts: cassava meal

Markets for cassava meal wereexplored. The flour is made into anative delicacy (ira-id) but the marketis limited. About 2,000 kg of cassavameal were also sold to the ViSCA feedmill at P3.50 per kg. But this marketis unstable.

A more promising venture is to usecassava meal in swine feed mix. Theration was market tested in outlets inthe nearby towns of Hindang andHilongos. The formula contains about15%-16% crude protein, as tested bythe PRCRTC Laboratory. The pricewas competitive, being only about 72%of that of the popular base feed forswine.

The MARCA feed mix is made frompre-mix, and meals of cassava, fish,copra, and ipil-ipil leaves. Localwomen and children supply the leaves(Table 2). Profitability ranged fromP0.50 to P0.75 per kg, depending onthe sources of protein used. Othermix combinations, including kohol (asnail), will be tried out to determinethe most efficient mix.

A plan is under way to promotethe feed mix by integrating intoMARCA’s operations the sale ofcomplementary feeds, such as theswine base feed, which farmerscommonly use.

Machine Evaluation andImprovement

Flour finisher

The first finisher brought to the sitewas a manually operated machinewith a capacity of 20-40 kg/h when

371


Table 2. Formula of a feed mix for swine, and costs. The mix was made by a cassava processingcooperative in Mabagon village, Leyte Province, the Philippines.

Ingredient or input Weight Crude protein content Cost (kg) (%) (P)

Cassava meal 52.10 1.0 208.40Copra meal 24.00 5.3 120.00Fish meal 9.00 5.4 90.50Rice bran 8.50 1.0 25.50Ipil-ipil leaf meal 4.15 0.9 12.45Golden snail 1.50 0.3 6.00Salt 0.50 2.50Afsillin 0.25 29.00Labor 30.00Bags 6.00Transport 10.00

Total 100.00 540.35a

a. Cost of feed mix per kilogram = P5.40; wholesale price = P6.50/kg; retail price = P7.00/kg.

tested at the PRCRTC. But when usedby the farmer-operator, it did notperform as expected. The flour cloggedthe screen, barely flowing out of thefinisher. This was withdrawn and atemporary, manual one used until anew one was designed and made.

Motor-operated, the new finisherhad a fan which forced the fine flour topass through the screen. Thismachine had a higher capacity thanthe old one: at least 50 kg/h in a singlepass. The cooperative’s members usedit until another, improved, machinewas made, based on farmer-operators’evaluations.

The design now used has two mainimprovements: the feeder hopper wasenlarged to accommodate a larger flourvolume, and a metering device wasmounted. These modificationsimproved the operation’s efficiency byremoving the tedium of manuallyfeeding and frequently stirring theground chips in the hopper. They alsoreduced finishing time by about 50%.

Portable chipper

A portable chipper introduced to thecooperative was expensive for the

members, having an estimated cost ofP500. In contrast, the andolan—alocal chipper made from a perforatedGI sheet mounted on a piece of wood—costs about P15.00 per unit. Thechipper was therefore not cost effectivefor individual households, becauseof the very small scale of homeprocessing. The portable chipper alsohad to be mounted for the operator’sconvenience.

However, the expected advantagesof the portable chipper are anincreased yield of chips andeliminated risk of abraded handswhile processing. Cost-sharing amonghouseholds may make the chippermore cost attractive.

Modified tapahan dryer

Dryers were used only during therainy season, when sun drying wasimpossible and orders for flour had tobe satisfied.

A dryer was constructed on siteand tested. In the first two tests, itwas too heavily loaded (463 and200 kg of fresh chips), and the mixingturned the chips brown and henceunsuitable for flour production. In the

372


third test, the chips were not mixedand their color was more acceptablefor flour production. In the fourthtest, the farmers evaluated the dryer,loading it with 190 kg of chips. Thechips were not mixed, and dried inabout 12 h. Their color was lighterthan in the third test.

Improving CassavaProduction Systems

Although the issue of environmentfriendly cassava production onsloping land was raised duringinformal discussions with farmers, itbecame a pressing concern withincreasing commercialization. Theissue of sustainability of cassavaproduction and processing systemsbecame integrated into the project.

Two groups of cassava farmersfrom Hindang visited farming-systemprojects in Matalom, where modelcontour farms were shown. Thefarms had cropping systems similarto those of Mabagon and othercassava-producing communities.Farmers discussed the benefitsof contour farming and thedisadvantages of irresponsiblefarming. After two visits, seven modelcontour farms were set up inHindang. These are still beingfollowed up.

Project Management:Monitoring and Evaluation

Interdisciplinary team approach

Active interaction among teammembers was tried informally duringfield visits and discussions, andformally through monthly meetings.Although the independentcontribution of each discipline wasvaluable, team members wereconstrained by having otherresponsibilities.

Resident research assistant

Living in the village enabled theresident research assistant toobserve social behavior and norms.The assistant had to observe,facilitate, train, and catalyze thefarmers, and monitor results inlogbooks or diaries. The assistantleft the village in July 1993 asfarmers took over the flourproduction project.

The local advisory group

This interagency group consisted oflocal government representatives,the DTI officer, a technician from theDepartment of Agriculture, andrepresentatives from the village andthe farmers’ group, MARCA, and thePRCRTC. The local advisory groupwas to build up local managementcapability and to continue assistingMARCA after the PRCRTC leftMabagon in March 1994.

Monitoring and evaluation

These were done, first, by the team,based on observations from marketsurveys, field visits, and notes andfeedback from the resident researchassistant; and, second, throughMARCA’s regular monthly meetings,and informal discussions with, andfeedback from, MARCA. Farmerparticipation was alwaysencouraged. Findings andobservations were used to plan,modify, and improve the executionof activities.

Some Impact Indicators

An important objective of the projectwas to assess the acceptability andadoptability of introducedtechnologies and their effects onprocessing and on farminghouseholds (Table 3).

373


Table 3. Acceptability and adoptability of introduced technologies for cassava flour production inMabagon village, Leyte Province, the Philippines.

Technology Acceptability and/or adoptability

Machines:

Pedal-operated chipper Power efficient, acceptable, but of limited use. Most processorsprefer the local andolan.

Grinder and finisher Acceptable, adopted. Easily learned. Modifications to enhancethe grinding and finishing capacities.

Modified tapahan dryer Acceptable, but not adopted. Costly to use. Sun drying withplastic mats produces better quality flour more efficiently.

Portable chipper Initial testing. Not acceptable. Design being improved.

Bulk storage in Acceptable. Adopted.polyethylene bags

Processing:

Flour processing Adopted. Needs market expansion and promotion.

Swine feed mix Adopted. Needs market expansion and promotion.

Food processing: cakes, Accepted. Not adopted. No facilities, and difficulties in operation.doughnuts, siakoy.a

Cacharon.b Acceptable. Promising market. In the process of fabricatingequipment.

Polvoron.a Acceptable. Adopted by individuals. Planned for enterprise.

a. Editors’ note: No description of this product was provided by the authors.b. Cacharon is a puffed product with various flavors.

Organizing and entrepreneurialskills

Markedly satisfactory performance wasobserved in terms of growth incooperation, improved attitudes,processing skills, and enterprisemanagement. It was also evident ingroup work, attendance at meetings,assemblies, workshops, participationin discussions, and carrying out ofassigned responsibilities in operationsand marketing.

Enterprise diversification

MARCA is learning to integraterelated enterprises to make operationsmore viable and profitable. An“entrepreneurial” culture at the villagelevel is gradually evolving.

For the first 5 months ofoperation, at 30% capacity, the netprofit per kg of fine flour was P0.61.With 90% capacity, this couldimprove to P1.50/kg. If cassavameal is included, net profit per kg is1.04 at 30% capacity and P1.79 atfull (Table 4). Financial statements(Jan.-Dec. 1992) show that, duringthe year’s operation, profitabilitydecreased because plant usedropped to 20% (see Appendix). Theenterprise’s profitability wasimproved with the use of cassavameal in feed mix. Profit per kgranged from P0.50 to P1.10,depending on the ingredients used,which, in turn, were chosen so toobtain a feed mix price that wasP2.50 lower than the popular, basefeed brand.

374


Table 4. Cost and returns for cassava flour per month per capacity use in a flour plant at Mabagonvillage, Leyte Province, the Philippines. Assumptions were: fine flour yield = 85%;full capacity = optimal flour production for 120-h week; cost of chips = P4.00/kg; price of fineflour = P8.00/kg; price of cassava meal = P3.50/kg.

Cost and return per capacity use 1.2 t/month 4 t/month(15 bags/week) at (50 bags/week) at

30% capacity full capacity

Flour 9,600.00 32,000.00(1,200 kg) (4,000 kg)

Less costs for:Chips 5,647.00 18,823.53

(1,411.75 kg) (4,705.88 kg)

Labor 1,577.45 3,154.90Electricity 135.95 543.80Marketing costs 271.60 1,086.40Bags 300.00 1,000.00Depreciation costs 937.00 1,405.50

Total costs 8,869.00 26,014.63

Net profit (fine flour)(per kg) 731.00 5,985.87

0.61 1.50

Cassava meal sales 741.16 2,470.58(211.76 kg) (705.88 kg)

Total monthly income 1,472.16 8,456.45

Overall net profit per kg 1.04 1.79

Cost per kg of fine flour 7.39 6.50

Multiplier effect on the community

Market and income. FromJanuary to December 1992, MARCAbought from the farmers a total of 20 tof cassava chips, costing, in total,P73,921.00. This contribution tofarmer income was substantial,compared with the chips marketbefore MARCA, in which only about5 t/year were sold, assuming anaverage of 3.5 bags/week. Farmers inMabagon, Himacugo, Katipunan, andBaldoza are some of MARCA’ssuppliers.

Improved chip quality. Thequality (whiteness, aroma, andbrittleness) of home-processed chipsimproved markedly, conforming to thedesired moisture content (12%-14%).Proof of this was that flour qualitystabilized.

Supplementary income viawage employment and othersocial benefits. Members derivesatisfaction from earning even aminimal wage by working at theplant. They also feel a sense ofachievement from learning newskills and being active in anenterprise—stimulated bymotivation, pride, and hope that notall government projects fail. Thebuildup of entrepreneurial spirit,strengthening local institutions, andimprovement in people’sorganizational performance andattitudes are benefits which aredifficult to measure. Yet they areessential for rural mobilization andgrowth. The processing plantbecame a source of prestige to thecommunity, and gave farmersconfidence that they could achieveeven more.

375


intensifying the farmers’understanding of cassava flour useand market expansion. Currently,commercial cassava flourproduction is not viable withoutintegrating the commercialization ofbyproducts, expanding end uses offlour as, for example, a rawmaterial in other products and foodprocessing, or trading relatedcomplementary products. Toencourage farmers to set up asuccessful enterprise thereforerequires a carefully integrated plan,in which each step, made small,simple, and focused, is graduallyintroduced. The step-by-stepprocess would help farmersunderstand that every product oradditional activity needs a minimalstandard of quality, stablesupplies, good service, andcompetitiveness.

A pilot and commercializationproject should therefore have theinstitutional or external support forinvestment in market research andpromotion because farmer groupsusually do not have the neededfunds to start up marketingactivities.

Institutional flexibility

The research team members mustbe sensitive if they are tosuccessfully collaborate inarranging development activities insuch a way that the project hastechnical, marketing, economic,and operational viability. From thebeginning, team members mustunderstand the need for suchsensitivity even in the planningprocess. The project’s successdepends on the ability of the teamand farmers to respond to changes,and modify plans and strategies toachieve objectives. Planning thenbecomes an iterative process. Butbeing responsive to uncertaintiescan be demanding on the team,

Conclusions andRecommendations

The approach

This exploited the skills of, not onlydifferent kinds of researchers, butalso of the beneficiaries themselves.The participatory process eventuallygave the farmer-processors a sense ofachievement: that throughinteractive, informal discussion withresearchers they could make effectivedecisions. Such “ownership” ofachievement is key to the buildup ofcapacities, which can only beachieved through gradual experientiallearning.

Local people’s involvement inproject management also eased thetask of team members who wereconstrained by time andresources. Local agencies sharedresponsibilities. Once the projectends, their familiarity with it willenable the locals to continuemanaging the established enterprise.

Defining the focus and makingthe project small and simpleintensified technology learning andenabled the processors to stabilizeproduct quality.

Participatory observation helpeddiscover social processes andinterrelationships, facilitating themodification or redesign of introducedtechnologies. Behavior and attitudesare central concerns in the process oftechnology transfer, preconditioningtechnology adoption. The project alsofacilitated on-the-job learning oftechnical and entrepreneurial skills.Placing a qualified resident researchassistant in the field significantlyfacilitated technology transfer.

Technology viability

The commercial viability of cassavaflour processing depends largely on

376


´

which demands, if not addressed,may cause delays or even failures.

This implies that some degreeof institutional flexibility is neededin teaming up and distributingworkloads to ensure thatresearchers have adequate time toundertake the responsibilitiesinvolved in a “commercialization”project. These responsibilities arebased on the implications of theintegrated process approach: acommitment to (1) usingparticipatory research methods,and (2) ensuring interactivelearning between farmers andresearchers.

The integrated approach alsoassumes the availability of aminimum, adequate, institutionalsupport to permit the exercise ofthe two basic responsibilities andprovide the support servicesneeded to make the project work.The policies of the participantinstitutions should thus be gearedtoward making the integratedapproach work.

Bibliography

Buvinic, M. and Rekha, M. 1990. Women andagricultural development. In: Eicher,C. K. and Stratz, J. M. (eds.).Agricultural development in the ThirdWorld. John Hopkins University Press,Baltimore, MD, USA. p. 290-308.

Cernea, M. M. 1991. Using knowledge fromsocial science in development projects.World Bank discussion papers, no. 114.World Bank, Washington, DC, USA.

Pérez-Crespo, C. A. (ed.) 1991. Integratedcassava projects. Working documentno. 78. Cassava Program, CIAT, Cali,Colombia. 242 p.

Pretty, J. 1993. Participatory inquiry andagricultural research. IIEDParticipatory Inquiry: notes for the192A course. 16 p.

Röling, N. Facilitating sustainable agriculture:turning policy models upside down.IIED-PAP N92, version 2. InternationalInstitute for Environment andDevelopment (IIED), London, UK.

Sands, D. M. and Kaimowitz, D. 1990. Thetechnology triangle: linking farmers,technology transfer agents andagricultural researchers. InternationalService for National AgriculturalResearch (ISNAR), The Hague, theNetherlands.

377


Appendix

Financial statements for the period January to December, 1992, of the MabagonRoot Crop Association (MARCA), a cassava processing cooperative at Mabagonvillage, near the town of Hindang, Leyte Province, the Philippines.

Statement A. Flour processing: the cost of goods manufactured (Jan.-Dec. 1992operation)

Item P

Materials usedBeginning raw material inventory (chips) 0Plus purchases 73,921.00

Cost of raw material available for use 73,921.00Less ending raw material inventory 1,196.00

Cost of raw material used 72,725.00Plus other raw materials used (fresh roots) 2,720.00

Total cost of raw materials used 75,445.00

Direct labor 14,570.00Factory overhead 12,816.05

Total manufacturing costs 102,831.95Less cost of goods in process inventory 0

Cost of goods manufactured P102,831.95

Statement B. Flour processing: income statement for Jan.-Dec. 1992.

Item P

SalesFlour 98,930.25Cassava meal 15,558.55

Damaged chips 1,365.50

Total sales 115,854.30

Cost of goods soldBeginning finished goods inventory 0Cost of goods manufactured 102,831.95

Total costs of goods for sale 102,831.95Less ending flour inventory 2,759.45

Cost of goods sold 100,072.50

Gross profit on sales 15,781.80

Less selling expensesDelivery personnel 900.00Packaging 1,366.00

Transportation 3,983.00

6,249.00

Net income from operation P9,532.80

378


Statement C. Chips trading operation: income statement.

Item P

Sales 7,017.65

Cost of goods soldBeginning inventory (chips for feed) 0Plus purchases 6,412.50

Cost of goods available for sale 6,412.50Less ending inventory of chips 235.50

Cost of goods sold 6,177.00

Gross profit on sales 840.65Less purchasing expenses 50.60

Net profit from operation P790.05

Statement D. Balance sheet: feed mix operation at end of December 1992.

Item P

Assets

Cash 533.45Accounts receivable 1,542.25Inventory 1,471.30

Total assets 3,547.00

LiabilityAccumulated profit 3,547.00

379


Statement E. MARCA balance sheet for the year ending December 1992.

Item P

Assets

CashFlour and chips operation 11,206.20Feed mix operation 533.45Registration fees 340.00Chipper rent collection 130.15

Bank account 33,437.15

Accounts receivableFlour and chips operation 2,031.50Feed mix operation 1,542.25

Inventory

Flour and chips operation 4,190.95Feed mix operation 1,471.30

Supplies and materials purchased

Drying mats 950.00Fluorescent tube 152.00

Buildings and warehouse 10,434.10

Electricity installation 333.65

Processing machines 50,958.65

Total assets 117,711.35

Liabilities

Accounts payable 789.75Accumulated depreciation 9,843.75Loans: SEARCAa 5,000.00

Others 23,497.60Processing machines 50,958.65

Owners’ equity

Labor capital raised 14,559.25Profit from operation 13,869.85

Less bank charges 75.00

Wall clock 189.50Registration expenses 93.00Drying mats 450.00

-807.50

Total liability and owners’ equity 117,711.35

a. SEARCA = Southeast Asian Regional Center for Graduate Study and Research in Agriculture.

380


CHAPTER 41

DEVELOPING THE CASSAVA FLOUR

INDUSTRY IN RURAL AREAS OF INDONESIA

A. Setyono*, Sutrisno*, and D. S. Damardjati**

Abstract

Increasing cassava production anddeveloping the technology for cassavaprocessing involve tackling problemsin such areas as technology,productivity, marketing pricestability, and production continuity.Once harvested, cassava isperishable, that is, the roots are ofacceptable quality for only a few days,creating a major problem for farmerswho are thus in a low bargainingposition. Cassava flour is one way ofovercoming this problem.

This study aims to (1) develop thecassava flour industry at three levels,(2) increase cassava’s added valueand thus farmer income, and(3) develop this industry in ruralareas.

Three levels of development of thecassava flour industry were attemptedin rural areas. The first was at thelevel of the individual farmer, wherefarmers’ activities included cassavaflour production, marketing ofcassava flour, and processing andmarketing flour-based products. The

* Sukamandi Research Institute for FoodCrops, West Java, Indonesia.

** Bogor Research Institute for Food Crops(BORIF), Bogor, Indonesia.

second was at the farmer group level,where farmers worked together toproduce cassava flour, market it, andprocess and market flour-basedproducts. The third was at thecooperative level, where thecooperative village unit (Koperasi UnitDesa) collects cassava flour fromfarmers and farmer groups, and thensells it to the retail trade and foodindustry, the feed industry, and otherconsumers.

Results indicated that marketingwas a major problem in CentralJava’s cassava flour industry.Cassava flour use and its processingtechnology have not yet developed.The cassava flour industry has littleknowledge of and experience withmarketing, which hindersdevelopment. The industry can bedeveloped in rural areas only throughthe cooperative system, or the farmergroup system, if given support indeveloping processing technology andhousehold industry, and in obtainingprocessing equipment and machinery.

Introduction

Cassava (Manihot esculenta Crantz) isthe most important staple food cropafter rice and maize in Indonesia. Atpresent, farmers cultivate cassava inalmost all areas of Indonesia, fromlowlands to highlands, in dry or wet

381

Developing the Cassava Flour Industry in Rural Areas...

climates, and under various soilconditions. Table 1 summarizes theuse of upland areas in Indonesia.Wargiono (1988) stated that theharvested area of cassava inIndonesia showed a decreasing trendof 0.7% per year. But since 1986, thetrend has increased slightly and, in1991, the total harvested area wasabout 1.3 million hectares, with atotal production of almost 16 milliontons (Table 2). Cassava’s productionrate from 1969 to 1985 was 2.05%per year (Affandi, 1986).

Thirteen of Indonesia’s provinces(Table 3) are major cassava-producingareas, each province having morethan 10,000 ha under cassava(Pabindru, 1989). The average yieldper ha is low at 10-12 t/ha. This canbe increased by introducing newtechnologies to farmers such asimproved varieties and culturalpractices. On a cassava estate ownedby a tapioca factory in Lampung,yields of 25-30 t/ha have beencontinuously achieved (Rusastra,1988).

Table 1. Summary of use of upland areas (ha) in Indonesia, 1984-1987.

Kind of upland

1984 1985 1986 1987

Cultivated(permanent basis) 8,327,282 8,091,282 8,377,480 8,761,476

Cultivated(temporary basis) 2,950,485 2,826,683 2,902,528 3,125,278

Unused oruncultivated 7,371,511 7,409,646 8,097,646 8,320,418

Total 18,649,278 18,327,611 19,377,654 20,207,172

SOURCE: SFCDP, 1990.

a. Preliminary data.

SOURCE: CBS, 1992.SOURCE: CBS, 1991.

1. 2. 3. 4. 5. 6. 7. 8. 9.10.11.12.13.

North SumatraSouth SumatraLampungWest JavaCentral JavaYogyakartaWest JavaBaliMalukuNorthern TerritoryWest KalimantanSouth SulawesiSoutheast Sulawesi

337.7 407.81,828.22,129.03,313.4 680.73,718.2 260.5 223.9 763.3 264.1 483.1 996.7

12.412.212.713.312.111.312.613.510.810.3 9.911.611.3

Table 3. Production and yield per hectare incassava-producing provinces ofIndonesia, 1991.

No. Province Production Yield rate(000 t) (t/ha)

Table 2. Harvested area, production, andyield rate of cassava in Indonesia,1984-1992.

Year Harvested Production Yield ratearea (000 t) (t/ha)

(000 ha)

1984

1985

1986

1987

1988

1989

1990

1991

1992a

1,350

1,292

1,170

1,222

1,302

1,408

1,312

1,319

828

14,167

14,057

13,312

14,356

15,471

17,117

15,830

15,955

10,221

10.5

10.9

11.4

11.7

11.9

12.2

12.1

12.1

12.3

Year

382


Problems of Developing theCassava Agroindustry

These are cassava’s association withlow social status; inadequatemarketing, postharvest handling,processing, and cultural practices;and low productivity.

Association with social status

Most Indonesians consider cassava asa food for those of low socioeconomicstatus. When income increases, thenconsumers switch from cassava torice. Cassava consumption per capitaper year has tended to decreasegradually in Indonesia, dropping from57.4 kg per capita in 1983, through51.0 kg per capita in 1988, to 43.1 kgper capita in 1990 (Table 4).

Marketing

Most cassava in Java is used forhuman consumption or starch(tapioca). Cassava farmers nearstarch factories usually sell freshroots directly to the factories, whilethose in remote areas tend to firstprocess cassava into gaplek, or driedcassava chips, and then sell tomiddlemen, who transport and sellthe chips to exporters or pelletingfactories in urban areas. Farmers in

Java sell about 50% to 90% of theirfresh cassava roots to traders ormiddlemen.

Farmgate prices of cassavafluctuate between Rp (rupiahs) 26and Rp 177/kg, according to locationand harvesting time (Tjahjadi, 1989).Cassava is also perishable, and oftencannot be processed or consumedimmediately after harvest. Theseproblems limit the flexibility ofcassava and force farmers into a lowbargaining position.

Postharvest handling andprocessing

Cassava is usually harvestedmanually, and may suffer severedamage if the roots are not carefullydug out of the ground. Rootsdeteriorate rapidly after harvest, andare bulky, making transportationdifficult and expensive.

According to the Indonesian foodbalance sheet data (CBS, 1992), totalcassava production in 1991 was15.8 million tons (Table 5). Of this,56.0% was consumed fresh or asgaplek; 15.7% was exported as gaplek(chips), pellets, and tapioca; and20.3% was used as raw material inindustries such as tapioca (starch)

Table 4. Average per capita consumption of major food crops in Indonesia, 1983-1990.

Commodity Per capita annual consumption (kg)a

1983 1986 1988 1989 1990 1991

RiceCassavaTapiocaGaplekb

SweetpotatoWheatMaizeSoybean

145.21 57.41 0.50 -

12.46 8.19 27.35 4.45

147.36 51.49 1.35 -

11.05 5.96 29.25 8.80

150.18 51.00 1.00 1.46 10.93 6.59 30.75 9.49

140.84 51.41 - -

11.04 6.93 26.81 8.80

150.05 43.07 - - 9.74 7.54 29.68 10.72

145.53 48.87 - - 9.61 7.71 28.73 11.01

a. - = data not available.b. Gaplek = dried cassava chips.

SOURCES: CBS, 1991; 1992.

383


Table 5. Trends in production and use of cassava in Indonesia, 1987-1991. No data were available forexports or nonfood industries. (Values in parentheses refer to percentages rounded off.)

Production or use Fresh roots (or equivalent) (thousands of tons)

1987 1988 1989 1990 1991

Total productionWasteManufactured for:

FeedFood industryFood consumption

14,3561,866 (13)

287 (2)3,401 (24)8,802 (61)

15,4712,011 (13)

309 (2)4,288 (28)8,863 (57)

15,8302,058 (13)

317 (2)5,781 (37)7,674 (48)

17,1172,225 (13)

317 (2)5,781 (37)7,674 (48)

15,8132,056 (13)

316 (2)4,583 (29)8,858 (56)

SOURCES: CBS, 1989; 1991; CBS, 1990, personal communication.

(7.9%) manufacture and nonfoodindustries (12.4%). Postharvestlosses are relatively high, about13.0%.

Cassava roots can be used invarious forms: fresh roots are cooked(boiled, roasted, steamed, or fried);fermented to produce tape; dried(either whole root, slices, or chips)to produce gaplek; extracted toproduce tapioca (starch); or thegaplek milled to produce flour.Gaplek can be kept as a food reserveor as animal feed. In villages of Java,most cassava is used for humanconsumption, and many traditionalproducts are produced for local andnational consumption.

Cassava as a marginal crop

Farmers tend to grow cassava withtraditional, sometimes inadequate,technology. Being a crop withunstable prices (a consequence ofundeveloped processing technology),cassava is often grown in fragile soilswith little or no investment infertilization.

Study Objectives

Our study aimed to (1) improvepostharvest handling of cassava by

farmers, (2) introduce and developcassava flour production and theprocessing of flour into otherproducts, and (3) develop householdand small-scale cassava processingindustries in a village of CentralJava.


Research took place in KejobongSubdistrict, Purbalingga District,Central Java Province, during1991-1993. It was conducted inthree phases: surveying, introducingcassava flour production andprocessing technology, and evaluatingthe development of the cassava flourindustry.

Surveying cassava postharvesthandling and processing

A survey was carried out inPurbalingga District, from May toJune 1991. Primary data wascollected from farmers on how theyhandled cassava after harvest,processed the roots, and marketedtheir products. A literature searchwas also conducted on cassavaproduction, area of land use, andcassava processing in PurbalinggaDistrict. From all these data, wechose the experimental site.

384


Introducing cassava flourproduction and processingtechnology

The second phase, that ofintroducing new cassava flourproduction technology andprocessing, was conducted fromAugust 1991 to March 1992.

Evaluating the development of thecassava flour industry

The development of the cassavaflour industry in KejobongSubdistrict was evaluated duringJune to September 1993. Theevaluations were at individual,group, and cooperative levels inrural areas, including cassava flourentrepreneurs. Production,processing, marketing, and otherproblems were also assessed.


Survey results

Socioeconomic conditions ofKejobong Subdistrict. Table 6summarizes land use in PurbalinggaDistrict. Land use in KejobongSubdistrict is divided as uplands(about 84%), lowlands (4%), anddegraded lands (12%) (Table 7).Altitudes range from 70 to 100 mabove sea level, climate is type A, andannual rainfall is 4,048 mm (Table 8).

Total population in KejobongSubdistrict was 66,712 (32,622 malesand 34,049 females). Most people inKejobong derived their income fromagriculture: 40% from food crops,fishery, and cattle raising; and 4%from other agricultural work. Therest worked in industry (7%); retail

Table 6. Summary of land use (ha), by subdistrict, in Purbalingga District, Central Java, Indonesia, 1983.

Subdistrict Lowlands Uplands Degraded lands Total

Bukateja

Kejobong

Kaligondang

Kemangkon

Purbalingga

Kalimanah

Kutasari

Bobotsari

Mrebet

Karangrejo

Karanganyar

Karangmoncol

Rembang

TotalPercentagea

2,103

361

1,121

2,311

798

2,928

2,377

1,275

1,607

1,295

2,414

1,609

2,007

22,204 25

2,137

7,811

3,932

2,203

676

1,049

5,833

1,953

3,182

10,785

4,422

4,419

7,152

55,55464

-

1,090

1,212

-

-

-

310

311

924

795

290

1,068

3,372

9,37211

4,240

9,262

6,265

4,514

1,474

3,977

8,520

3,539

5,713

12,875

7,126

7,096

12,531

87,130100

a. Values are rounded off.

385


Table 7. Summary of land use in Kejobong Subdistrict, Central Java, Indonesia, 1983.

Type of land use

(ha) (%)a

Lowlands:Technical 166 2Semitechnical 31 <1Simple 16 <1Rainfed 148 2

Total 361

Uplands:Building and garden 3,224 35Cultivated (temporary basis) 4,332 48Other 225 3

Total 7,781

Degraded lands 1,090 12Overall total 9,232


Harvesting, and postharvesthandling and processing

Harvesting. Two major varietiesof cassava are planted in Kejobong:the bitter ‘Klanting’ (90%) and thesweet ‘Darme’ (10%). Harvesting isusually by hand during August toNovember. Of 36 respondents inseven villages, 7 harvested thecassava themselves, 6 paid others toharvest, 4 harvested through thecooperative system, 17 had sold theharvest in advance, and 2 did notharvest.

Postharvest handling andprocessing. Postharvest handlingand processing of cassava have notyet developed in Kejobong, becausemost farmers sell cassava as freshroots. Only about 30% of farmersprocess cassava, producing suchtraditional goods as gaplek, tiwul, andcantir. Only 5% of farmer-processorsproduced tapioca (Table 11).

The farmers either sold peeledcassava to middlemen (50%) andretailers (8%), or were paid in advancebefore harvest (42%). Because thetotal potential capacity of factories

Table 8. Monthly rainfall and number of rainydays per month, averaged over 9 years,Purbalingga District, Central Java,Indonesia, 1981-1989.

Month Monthly rainfall Number of rainy(mm) days per month

JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember

TotalAv./month

481 484 436 418 236 181 137 105 256 336 493 483

4,046 337

19 20 19 19 12 10 8 7 11 15 18 20

178 15

(4%); transport (<1%); government(1%); and others (<1%). About 2% ofemployed were not reported (Table 9).

Kejobong is the major center ofcassava production in PurbalinggaDistrict. Production was 43,671 t in1986 (Table 10).

Area

386


Table 9. Population by livelihood, Kejobong Subdistrict, Central Java, Indonesia, 1986.

Source of livelihood People Remarks

(no.) (%)a

Agriculture:Food crops, fishery, andcattle raising 26,614 40Other agricultural work 2,392 4 Agricultural laborer

Industry and Services 4,995 7 Entrepreneur or employerRetail trade 2,895 4Transportation 133 <1Government 814 1 Functionary, laborer, army

worker, pensionerOthers 378 <1Unreported 1,538 2

Total employed 39,759 60Total population 66,712


Table 10. Total production (t) of food crops in Purbalingga District, Central Java, Indonesia, 1986.

Subdistrict Lowland Upland Maize Soybeans Groundnuts Sweet- Cassavarice rice potato

BukatejaKejobongKaligondangKemangkonPurbalinggaKalimanahKutasariBobotsariMrebetKarangrejoKaranganyarKarangmoncolRembang

Total

18,023 2,604 7,810 23,850 8,118 28,694 16,046 8,958 15,670 3,073 19,281 5,095 9,804

167,026

172,134 -

1,082 - -

1,931 - - -

385 - 91

5,640

13314,435 925 1,898 226 921 6,805 1,636 1,285 2,357 1,092 226 1,052

32,991

151 67 143 426 79 285 143 106 142 130 96 53 80

1,901

221 239 221 256 103 241 672 35 263 - - - 3

2,254

- - - 34 - 29 714 126 49 435 53 - 224

1,664

2,04743,671 4,513 970 34 21

5,657 1,223 1,824 3,244 3,044 176 5,422

71,846

Table 11. Percentage (values rounded off) ofrespondents (farmers) who processcassava, and their products, KejobongSubdistrict, Central Java, Indonesia.

Product Respondents

Gaplek (dried cassava chips) 17Cantira 8Cendola 3Tiwula 3Tapioca 5Cassava flour 0

Total 31

a. Editors’ note: No description of this product wasprovided by the authors.

was equivalent to 12,092 t of freshcassava (i.e., 2,423 t of tapioca)(Table 12), the total cassavaproduction of 71,846 t could not beprocessed. Consequently, cassavaprices fell, fluctuating according toretailer or tapioca factory. In 1989,the price of cassava ranged fromRp 25 to Rp 30/kg.

But after cassava flour processingwas introduced, the price of cassavarose from Rp 50-60/kg in 1990 to apeak in 1991 and 1992 at Rp 90-115,

387


Table 12. The capacity of tapioca factories in Purbalingga District, Central Java, Indonesia, 1989.

Manufacturer Subdistrict Started Potential Operationaloperations capacity capacity

(t/year) (t/year)

LamukPandansariWanakusumaSribumikaryaTridaya

Total

KejobongKejobongKejobongKemangkonBukatiga

19831987198819881988

8 75 720 720 900

2,423

8 60 600 600 900

2,168

and dropped slightly to Rp 70-80 in1993. Two years ago (i.e., 1991/92),the production of cassava flour wasnot profitable, because the cassavaflour price ranged from Rp 350 toRp 400/kg.

Cassava flour productioninvolves several processing steps:peeling, washing, slicing or raspingor chipping, pressing, drying, andmilling (Figure 1). These newtechnologies were introduced, bydemonstration, to farmers, farmergroups, and members of the VillageUnit Cooperative, or Koperasi UnitDesa (KUD). These people assessedthe technologies and were thentrained in their use. Cassava flour isused to substitute wheat flour in themaking of such foods as pancakes,cookies, cheese sticks, and putu ayu.

The cassava flour industry wasdeveloped in rural areas as athree-level system (Figure 2). Onthe first level, the individual farmerproduced cassava flour, andprocessed and marketed it himself.On the second level, the farmergroup performed these activities.On the third level, the KUD not onlyproduced cassava flour, but alsoreceived it from farmers or farmergroups, and then sold it to foodindustries and middlemen.

Cassava Flour Productionand ProcessingDevelopment

The objectives of this developmentproject were to (1) encourage theproduction and processing ofcassava flour by individualfarmers and farmer groups, and(2) increase the added value offresh cassava by processing it intoflour. The project was conductedin three phases.

Introducing cassava processingequipment

Processing equipment wasdemonstrated to farmers to arouseinterest in cassava flourproduction, encourage anincreased working capacity andproduct quality, and promote thedevelopment and manufacture ofprocessing equipment in ruralareas. The equipment introducedand demonstrated included slicers,carvers, graters, and presses.

The Kejobong KUD, inparticular, received, from theGovernment, equipment with adaily capacity to process 10 t ofcassava roots into flour.

388


Cassava flour production

Flour production would help sellcassava when prices are low anddemand from tapioca factories is alsosmall.

In 1992, individual farmersproduced about 300 kg of cassava

flour, which was processed tocookies and pancakes. Farmergroups produced 1 t of cassava flourand cooperatives 5 t. They sold it tofood industries and middlemen.The major problems of cassava flourproduction at all three levels werefound in marketing.

Peeledcassava

Press

Figure 1. Cassava flour production in Kebojong Subdistrict, Central Java, Indonesia. (KUD = VillageUnit Cooperative.)

Mesra 2(rasper/slicer, used

by KUD)

Triguna(slicer/chipper/rasper, used by

individual farmer)

Apessi (dryer)

Mill

Packer

OR

Mesra 1(rasper/slicer, used by

farmer group)

OR

389


Food technology and cassava flouruse

Marketing and food technology areimportant in the successfuldevelopment of a cassava flourindustry. Successful marketing isinfluenced by product utility andrequisites for quantity and quality.Developments in food technologyincreases opportunities for marketingcassava flour. Outlets for the flourare food, chemical, and otherindustries, household consumption,and traders (middlemen, retailers,and exporters). Such developmentsin food marketing and technologyhave yet to arrive in KebojongSubdistrict.

Assessing the Development ofthe Cassava Flour

Agroindustry

Developing the cassava flouragroindustry in rural areas wasexpected to extend cassava marketingand agroindustrial development inrural areas, and to increase farmers’income. The project was evaluatedfrom May to September 1992.

The price of fresh cassava duringMay and June 1993 in Kejobongranged from Rp 50 to Rp 60/kg, andthat of cassava flour ranged fromRp 250 to Rp 350/kg. In May andJune 1992, the typical farmerproduced 200 kg of cassava flour,

Fresh cassavaprocessing

Cassava flourproduction

Cassavaproducer

III. Cooperative(KUD)

Cassava flourproduction

Fresh cassavaprocessing

Foodprocessing

Con

sum

er

Entrepreneur

Export

Processor

Foodprocessing

Tapiocafactory

Cassava flour production

Figure 2. Developing an agroindustry based on cassava flour production in rural areas of Indonesia.(KUD = Village Unit Cooperative.)

I. Farmer

II. Farmergroup

390


selling it to food industries. Thetypical farmer group produced500 kg of cassava flour, selling it tofeed industries. The KUD, however,did not produce cassava flour, notfinding it profitable.

The price of cassava rootsincreased from Rp 70-90/kg inAugust-October 1993 toRp 95-105/kg in November 1993.Farmers and farmer groups thereforefound cassava flour productionunprofitable. Other farmers usedcassava for feed and food. Thetypical farmer would use5-10 kg/day of dried cassava to feedsheep and 40-50 kg/day for cantirproduction.

The problem

Despite the low prices (Rp 50 toRp 60/kg) farmers said they had noproblem marketing cassava roots.In contrast, producers explainedthey had problems marketingcassava flour because it is a newproduct for which food processingtechnology has not yet beendeveloped, and of which fewconsumers know much about.

Summary

(1) Farmers, farmer groups, andKUDs had no experience inmarketing cassava flour.

(2) Processors and consumers lackedinformation on cassavaprocessing technology and flouruse.

(3) Appropriate cassava flourprocessing technology must bedeveloped if the cassava flourindustry is to develop in ruralareas.

(4) Further research is needed oncassava flour use.

References

Affandi, M. 1986. Agricultural developmentin Indonesia. Central ResearchInstitute for Food Crops (CRIFC),Bogor, Indonesia.

CBS (Central Bureau of Statistics). 1989.Food balance sheet in Indonesia,1989. Jakarta, Indonesia.



Pabindru, M. 1989. Government policy inproduction of cassava in Indonesia.In: Proceedings of a national seminaron the Effort to Increase the AddedValue of Cassava. AgricultureFaculty, Padjadjaran University,Bandung, Indonesia.

Rusastra, I. W. 1988. Study on aspects ofnational production, consumptionand marketing of cassava. Agric. Res.Dev. J. (Indones.) 7:57-63.

SFCDP (Secondary Food Crops DevelopmentProject). 1990. Vademekum Palawija2. Ubikayu dan Ubijalar. (Maize,cassava and sweet potato). DirektoratJendral Peranian Tanaman Pangan.SFCDP and United States Agency forInternational Development (USAID).Jakarta, Indonesia.

Tjahjadi, C. 1989. Utilization of cassava asraw material of foods. In: Proceedingsof a national seminar on the Effort toIncrease the Added Value of Cassava.Agriculture Faculty, PadjadjaranUniversity, Bandung, Indonesia.

Wargiono, J. 1988. Agronomic practices inmajor cassava growing areas ofIndonesia. In: Howeler, R. H. andKawano, K. (eds.). Cassava breedingand agronomy research in Asia:proceedings of a regional workshop,Rayong, Thailand. CIAT, Cali,Colombia. p. 185-204.

391


APPENDICES

393

Appendix I: List of Participants

APPENDIX I

LIST OF PARTICIPANTS1

Argentina

de Fabrizio, SusanaJefe, Laboratorio de Microbiología de

AlimentosUBA/FCEYNPab. II Piso 3Ciudad UniversitariaCasilla Postal 1428Buenos Aires

Tel.: (54-1) 7820529Fax: (54-1) 3313272

Austria

Van Zanten, LeonardTechnical OfficerJoint FAO/IAEA DivisionRIFP.O. Box 200, A-1400Vienna

Tel.: (43-1) 23601617Fax: (43-1) 234564

Belgium

Pierreux, FrédéricIPESAT du HainautAth

Brazil

UNESP

Bicudo, Silvio JoséProfessor Assistente

Cereda, Marney PascoliInvestigadora

Chuzel, GerardInvestigador, CIRAD/SAR-UNESP

Takitane, Isabel CristinaProfessor Assistente

Vilpoux, OlivierDepartamento de Tecnología

Facultad de Ciências AgronómicasUNESPCaixa Postal 237Fazenda Experimental LageadoCEP 1860 Botucatu, São Paulo

Tel./Fax: (55-149) 213438, 213883Telex: 0142107

Cabello, ClaudioProfessor AssistenteUNESPRua Luiz Edmundo Coube, 1Caixa Postal 47317033-360 Bauru, São Paulo

Tel.: (55-142) 302111Telex: 142-312 FEBU

Garcia, Edivaldo AntonioProfessor AssistenteFaculdade de Medicina Veterinaria e

ZootecniaUNESPCaixa Postal 50218618-000 Botucatu, São Paulo

Tel.: (55-149) 213883 R.185Fax: (55-149) 213883 R.180

Morães, IracemaProfessor TitularUNESPRua Cristovão Colombo, 2265Caixa Postal 13615054000 S. José do Rio Preto, São Paulo

Tel.: (55-192) 244966 ext. 64Fax: (55-192) 410527

1. Most acronyms are explained in the “List ofAcronyms and Abbreviations Used in Text,”p. 402.

394


Others

Amante, Edna ReginaVice-Chefe, Departamento/Coordenadora

de ExteniasUFSCRodôvia Adman, Gonzaga, Km. 03 Itacorubi88034-001 Florianópolis, Santa Catarina

Tel.: (55-482) 344888Fax: (55-482) 342014

Cintra, Odorico L.Gerente GeneralFecularia Mon SaguRua Dom Aquino No. 506-Centro79025000 Campo Grande, Mato Grosso do

SulTel.: (55-67) 3824502, 7215225Fax: (55-67) 3824502

Demiate, Ivo MottinProfessor

Wosiacki, GilvanProfessor Titular

UEPGPraça Santos Andrade, s/nCaixa Postal 9928410-340 Ponta Grossa, Paraná

Tel.: (55-422) 252121Telex: 442 242 UEPG BRFax: (55-442) 237708

Lorenzi, José OsmarIACAv. Barão de Itapura 1481Caixa Postal 2813.001.970 Campinas, São Paulo

Tel.: (55-192) 419057

Mello, FabioPesquisadorUEPGPraça Santos Andrades, s/nCaixa Postal 992/99384.100.000 Ponta Grossa, Paraná

Tel.: (55-422) 252121 R.164Telex: 442 242 UEPG BRFax: (55-442) 237708

de Morães, Flávio FariaProfessor da Graduação e Pós-graduaçãoDepartamento de Ingeniería QuímicaUEMAv. Colombo 369087020.900 Maringá, Paraná

Tel.: (55-442) 264004Telex: 442198Fax: (55-442) 222754

Ospina, BernardoCIAT-EMBRAPAEMBRAPA/CNPMFRua EMBRAPA, s/nCaixa Postal 00744.380 Cruz das Almas, Bahia

Tel.: (55-75) 7212120Fax: (55-75) 7211118

Sarmento, Silene BruderProfessor AssistenteESALQAv. Pádua Dias 9Caixa Postal 913400 Piracicaba, São Paulo

Tel.: (55-194) 294150

Soccol, Carlos RicardoProfessorUFPRCentro Politécnico/Jardim das AméricasCaixa Postal 1901181531-970 Curitiba, Paraná

Tel.: (55-41) 3662323 R.285Fax: (55-41) 2660222

Takahashi, MarioInvestigadorIAPAREstação Experimental de ParanavaiCaixa Postal 56487701-970 Paranavai, Paraná

Tel.: (55-444) 231157Fax: (55-444) 231607

Canada

Edwardson, BillSenior Program SpecialistEnvironment and Natural ResourcesIDRC250 Albert StreetP.O. Box 2500KIG 3H9 Ottawa, Ontario

Tel.: (613) 2366163 Ext. 2215Telex: 0533753Fax: (613) 5677749E-mail: [email protected]

Fitzpatrick, DennisProfessor and Head of DepartmentFoods and Nutrition DepartmentUniversity of ManitobaR3T 2N2 Winnipeg, Manitoba

Tel.: (204) 4748080Fax: (204) 2755299E-mail:

[email protected]

395


China

Jin Shu-RenResearch and ManagementNCTDCGuchen Road 4-2Nanning, Guangxi

Tel.: (86-771) 552730Fax: (86-771) 562417

Colombia

PROPAL S.A.

Gutiérrez Herrera, MeyerIngeniero de ProcesoPROPAL S.A.Apartado Aéreo 4412Cali, Valle del Cauca

Tel.: (57-2) 4425757 Ext.323Fax: (57-2) 4425769

Meléndez Santacruz, GuillermoGerente, Area Departamento TécnicoPROPAL S.A.Planta No. 2Caloto, Cauca

Tel.: (57-9282) 82133

Or

Apartado Aéreo 4412Cali, Valle del Cauca

Tel.: (57-2) 4425757

UNIVALLE

Aguinaga, AsunciónJefe, Sección de Ciencia y Tecnología de

Alimentos

Castañeda Andrade, Jesús DavidJefe, Departamento de Diseño y

Procesos de Manufactura

Duque Santa, WaldoDirector, Especialización Maquinaria

Agroindustrial

Fernández, AlejandroProfesor TitularSección de Ciencia y Tecnología de

Alimentos

Moreno Santander, Martín AlonsoProyecto Almidón Agrio CIAT-UNIVALLE

de Stouvenel, Aida RodríguezProfesora

Sánchez Rodríguez, JaimeProfesor

UNIVALLEApartado Aéreo 25360Cali, Valle del Cauca

Tel.: (57-2) 3307285, 3393041ext. 133

Fax: (57-2) 3302479

Alazard, DidierInvestigador, ORSTOM

Raimbault, MauriceInvestigador, ORSTOM

UNIVALLEApartado Aéreo 32417Cali, Valle del Cauca

Tel.: (57-2) 6682594Fax: (57-2) 6682757

Others

Barberi Ramos, Julio DaríoGerenteProductos La NiñaApartado Aéreo 1422Pereira, Risaralda

Tel.: (57-963) 325269, 228073

Blanco Arango, HéctorGerente, Proyecto Carbohidratos y

CelulosasARPISOL LTDA.Carrera 100 No. 42A-20Apartado Aéreo 80827Santafé de Bogotá, D.C.

Tel.: (57-91) 4158025/29Fax: (57-91) 4150830

Cifuentes Arara, PorfirioCarretera PanamericanaMondomo, Cauca

Tel.: (57-928) 299085

Duque Vargas, AmparoControl de la Contaminación Recursos

HídricosCVCCarrera 56 No. 11-36Apartado Aéreo 2366Cali, Valle del Cauca

Tel.: (57-2) 3396671

396


Durán Restrepo, María VictoriaUnión Europea-Delegación de la Comisión

EuropeaCalle 97 No. 22-44Santafé de Bogotá, D.C.

Tel.: (57-91) 2369040, 2564828

Figueroa Sánchez, Francisco JoséPresidente

Sánchez Arrieta, Carlos AlbertoAsesor Técnico

Sarria Nuñez, HelberthGerente Regional/Vicepresidente

FUNDIAGROCarrera 54 No. 55-127, Oficina 905Barranquilla, Atlántico

Tel.: (57-958) 411306

García Millán, ArbeyJefe de Planta“RICOPAN” Panadería El PorvenirCarrera 2C No. 30-35Apartado Aéreo 5384Cali, Valle del Cauca

Tel.: (57-2) 4445544

Gómez Botero, ClaudiaDirectora, Investigación y DesarrolloRICARONDO S.A.Calle 31 No. 2-80Apartado Aéreo 4842Cali, Valle del Cauca

Tel.: (57-2) 4422637

González de Duque, Olga LucíaGerente, Departamento Técnico y

Desarrollo

Reyes Madriñan, Francisco JoséAsesoría y Mantenimiento

Colombiana de Almidones yDerivados S.A.

Calle 16 Norte No. 6N-21Cali, Valle del Cauca

Tel.: (57-2) 6681287

Guzmán Roa, Néstor GonzaloGerente de Marca

Rubiano Mejía, Luz ElenaJefe, Investigación y Desarrollo de

Nuevos Productos

Companía Nacional de Levaduras,LEVAPAN S.A.Avenida de las Américas No. 40-81Santafé de Bogotá, D.C.

Tel.: (57-91) 2684299, 2683651Fax: (57-91) 2681983

Idárraga, Gloria AmparoJefe, Investigación y DesarrolloProductos Yupi S.A.Calle 70 No. 3N-74Cali, Valle del Cauca

Tel.: (57-2) 6644330Fax: (57-2) 6644379

Jaramillo Diaz, HebertJefe del AreaUniversidad Autónoma de OccidenteCalle 9B No. 29A-67Cali, Valle del Cauca

Tel.: (57-2) 5565444

Lozano, Alvaro FigueroaGerente, Investigación y Desarollo

Quintero Muñoz, Claudia I.Asistente, Investigación y Desarrollo

Industrias de Maíz, MAIZENA S.A.Carrera 5A No. 52-56Apartado Aéreo 6560Cali, Valle del Cauca

Tel.: (57-2) 4474853, 4470914Fax: (57-2) 4477477

Mejía Gómez, Jaime ArturoGerente OperativoProductos Alimenticios Crunch S.A.Carrera 43A No. 61 Sur 152, Local 116Sabaneta, AntioquiaApartado Aéreo 8561Medellín, Antioquia

Tel.: (57-94) 2880254Fax: (57-94) 2882913

Ruíz Cabrera, RicardoCoordinador, Departamento de

AgroindustriaCETEC-SEDECOMDiagonal 26A No. 26-94San Fernando, Cali, Valle del Cauca

Tel.: (57-2) 5564809

397


Sadovnik, AlejandraAdministradora

Sadovnik Sánchez, Hardy AlfonsoGerente

YUCA LTDA.Avenida 5 Norte No. 51-05Cali, Valle del Cauca

Silva Bernal, John EdgarGerente, SIMARKSCalle 90 No. 40-82Apartado Aéreo 251789Santafé de Bogotá, D.C.

Tel.: (57-91) 6101028

Zambrano Sarmiento, Francy MagdalenaIngeniera de AlimentosDiagonal 103 No. 57-49Santafé de Bogotá, D.C.

Tel.: (57-91) 7110021, 535944

Costa Rica

Blanco-Metzler, AdrianaJefe, Unidad de Tecnología NutricionalINCIENSAApartado 4, Tres Ríos

Tel.: 506 799911Fax: 506 795546

Boucher, FrançoisDirector EjecutivoIICA/PRODARCasilla Postal 55-2200, Coronado

Tel.: 506-290222Fax: 506-294741, 292659Telex: 2144 IICA CRE-mail:

FBoucher@UCRVM2@Bitnet

Laboucheix, JeanDelegado para América Latina y el CaribeRepresentante del CIRAD ante el IICA200 m. al Sur de la Iglesia de San PedroFrente cost. Oeste Esc. RooseveltApartado 1127-2050 San PedroSan José

Tel.: (506) 255972Fax: (506) 250940

Ecuador

Caballero Vera, Hernán HumbertoEstudiante

Delgado Castro, PlinioPromotor

Ruiz Chévez, VicenteDirector Técnico

UATAPPYCalle Olmedo y 9 de OctubreCalderón, Porto Viejo, Manabí

Tel./Fax: (593-4) 637240

Cantos Sornoza, Enrique JuvanTécnico de Producción

Palacios Delgado, John HerbidsonAdministrador General

FACEKm. 101/2 Vía Manta-MontecristiApartado 13-5-4821Montecristi, Manabí

Tel.: (593-4) 606399Fax: (593-4) 606109

Carpio, CeciliaAsistente de InvestigaciónInstituto de Investigación TecnológicaEPNAndalucia s/n y VeintimillaQuito, Pichincha

Tel.: (593-2) 507138Fax: (593-2) 507142

Egüez, CarlosCoordinador, Programa de YucaFUNDAGROMoreno Bellido 127 y AmazonasApartado 17-16-219Quito, Pichincha

Tel.: (593-2) 220557, 220533Fax: (593-2) 507442

Evans, CodyManager of Operations

Evans, Edward

Compañía Ganadera ManipiliApartado 17-01-3214Quito, Pichincha

Tel.: (593-2) 2570177, 2570768Telex: 22368 ADMINEDFax: (593-2) 2570964

Intriago Vera, Solanda ElinaGerenteATAPY-San VicenteCalderón, Porto Viejo

Tel.: (593) 637240

398


Poats, Susan V.Anthropologist,CIAT-FUNDAGRO ProjectMoreno Bellido 127 y Mariano de JesúsCasilla 17-16-219Quito, Pichincha

Tel.: (59-32) 220533Fax: (59-32) 507422

Ruales, JennyProfesor PrincipalEPNIsabel La Católica y VeintimillaApartado 17012759Quito, Pichincha

Tel.: (593-2) 507138Fax: (593-2) 507142

Verdesoto Medrano, Lécide UvaldinoCoordinador, Proyecto de Yuca de

EsmeraldasUAPPY-EsmeraldasAvenida Olmedo entre 10 de Agosto y

RocafuerteEsmeraldas, Esmeraldas

Tel.: (593-2) 713869

France

Brauman, AlainORSTOM213 Rue Lafayette75010 Paris

Tel.: (33-1) 48037777

della Valle, Guy Rene NoelIngeniero de Investigación, Tratamientos

Físicos de AlmidonesCentre de recherche agro-alimentairesINRABP 52744026 Nantes

Tel.: (33) 4067 8000Fax: (33) 4067 8005

Giraud, EricIngénieurORSTOM2051 Avenue du Val de MontferrandBP 504534032 Montpellier, Cedex 1

Tel.: (33) 67617400, 67617575Fax: (33) 67547800

Griffon, DanyDeputy Director, Program and Development

Sautier, Denis Pierre-JeanChercheur

Zakhia, NadineChercheur

CIRAD/SAR2477 Avenue du Val de MontferrandBP 503534032 Montpellier, Cedex 1

Tel.: (33) 67615707Fax: (33) 67414015

Hebert, Jean PaulENSIA/SIARC1101 Avenue AgropolisBP 509834032 Montpellier, Cedex 1

Tel.: (33) 67617051Fax: (33) 67410232

Mestres, ChristianCIRAD/CA73 rue Jean François Bretan34032 Montpellier, Cedex 1

Pourquié, JacquesProfesseurINA-PG16 Rue Claude Bernard75231 Paris, Cedex 05

Tel.: (33-1) 44081830Fax: (33-1) 44081700

Ghana

Safo-Kantanka, OseiCrop Science DepartmentUniversity of Science & TechnologyKumasi

Tel.: (233-51) 53519Fax: (233-51) 3137

Honduras

Santos Sosa, SamuelResponsable, Proyecto de Yuca (Utilización)Fundación Friedrich EbertColonia Humaya 2a. Calle No. 2401Apartado No. 1701Tegucigalpa

Tel.: (504) 332753Fax: (504) 332800

399


India

Moorthy, NayaranaSenior ScientistCTCRITrivandrum, Kerala

Tel.: 91-448554

Indonesia

Damardjati, Djoko S.DirectorBORIFJL. Tentara Pelajar 3ABogor 16111

Tel./Fax: 62-251-333440

Kenya

Mbugua, SamuelTeaching and ResearchDepartment of Food Technology & NutritionCollege of Agriculture & ET SciencesUniversity of NairobiP.O. Box 29053Nairobi

Tel.: 254-2-632211, 632141,632401

Fax: 254-2-630172

Malawi

Saka, J. D. K.Senior Lecturer in ChemistryChemisty DepartmentChancellor CollegeUniversity of MalawiP.O. Box 280Zomba

Tel.: (265) 522222, 52327Telex: 44742 CHANCOL MIFax: (265) 522046, 523021

Mexico

Guyot, Jean-PierreHead, Research ProgrammeORSTOMCicerón 60911530 México DF

Tel.: (55-5) 6807688

Monroy Rivera, José AlbertoProfesor InvestigadorInstituto Tecnológico de VeracruzCircunvalación Norte e IcazoApartado 1420Veracruz

Tel./Fax: (52-29) 345701

Pereira Pacheco, Fabiola EstherResponsable, Laboratorio de Ciencia de los

AlimentosFacultad de Ingeniería QuímicaAv. Juárez No. 421 Cd. IndustrialCasilla Postal 1226-A97288 Mérida, Yucatán

Tel.: (52) 460981Fax: (52) 460994

Nicaragua

Bosche, PaulDirector EjecutivoYuca Centroamericana S.A., YUCASANo. 10 Planes AltamiraApartado A-179Managua

Tel./Fax: (505-2) 784915

Briceño Lovo, Milton MarceloDirector, Control Industrial

Miranda Astorga, Lester JoséGerente General

Yuca Centroamericana S.A., YUCASAKm. 31 Carretera vieja a León

Tel./Fax: (505-2) 784915

Nigeria

Bokanga, MpokoBiochemistIITAPMB 5320, Ibadan

Tel.: (234-22) 400300Telex: 31417 TOPIP NG

Paraguay

Hüg de Belmont V., Carlos A.Administrador de Planta/Asistente de

ProducciónFUNDAINAyolas 451, Edificio Capital, 4o. PisoAsunción Central

Tel.: (595-21) 442-518/9Fax: (595-21) 442-520

Peru

Espínola de Fong, NellyFisiólogaCIPAvenida La Universidad s/nCasilla 5969, Lima 12

Tel.: (51-14) 366920Fax: (51-14) 351570

400


Rodríguez Zevallos, Antonio RicardoJefe, Centro de Producción AgroindustrialUniversidad Nacional “Pedro Ruiz Gallo”Ciudad UniversitariaLambayeque, Lambayeque

Tel.: (51-074) 282787

Salas Domínguez, SoniaJefe, ProyectosCáritas del PerúCalle Omicrón 492Callao

Tel.: (51-14) 640299Fax: (51-14) 642595

Salas Valerio, Walter FranciscoJefe, Departamento Ingeniería de AlimentosUNALMAv. La Universidad s/n, La MolinaCasilla Postal 456Lima

Philippines

Tan, Daniel LeslieAgricultural EngineerPRCRTC-ViSCABaybay, Leyte, Zip Code 6521-A

Fax: (63-2) 588692

Tanzania

Mlingi, Nicholas L. V.Senior Food ChemistTanzania Food and Nutrition CentreP.O. Box 977Dar es Salaam

Tel.: (255-51) 29621, 74107Fax: (255-51) 44029

Thailand

Maneepun, SaipinResearcherIFRPDKasetsart UniversityP.O. Box 170Bangkok 10400

Tel./Fax: (66-2) 561 1970

Titapiwatanakun, BoonjitAssistant ProfessorKasetsart UniversityBangkhen, Bangkok 10903

Tel.: (66-2) 561 3467Fax: (66-2) 561 3034

Trinidad and Tobago

Badrie, NeelaFood Technology UnitDepartment of Chemical EngineeringUWISt. Augustine, Trinidad

Tel.: (1-809) 6632001/2007Fax: (1-809) 6624414

United Kingdom

Trim, David S.Agroprocessing Group

Wenham, June ElizabethPost-Harvest Horticulture

NRICentral AvenueChatham Maritime, Kent ME4 4TBEngland

Tel.: (44-0634) 880088Telex: 236907/8 LDNFax: (44-0634) 880066/77

Venezuela

INDELMA C.A.

Bustamante Valero, Lizardo de JesúsGerente, Planta Industrial

Contreras Vega, José LuisSuperintendente, Almidones Modificados

INDELMA C.A.Km. 1 La Encrucijada vía TurmeroTurmero, Aragua

Tel.: (58-04) 4632320

Perdomo Ramos, María TeresaRepresentante TécnicoINDELMA C.A.Av. Lecuna Esquina de PetiónEdificio El Aguila, San Agustín del Sur122 Caracas

Tel.: (58-2) 5762133Telex: 21308 ARCIA VCFax: (58-2) 5736712

401


Others

Faroux, GerardDelegado Regional de Cooperación

Países AndinosEmbajada de FranciaCalle Madrid con Av. TrinidadLas Mercedes-CaracasApartado 62324, Caracas 1060-A

Tel.: (58-2) 9937448Fax: (58-2) 9935256

González Parada, Zurima MercedesProfesor/InvestigadorICTA-UCV/FACCalle Suapure Lomas de Bello MonteApartado 47.097, Caracas 1041-A

Tel.: (58-2) 7524403Fax: (58-2) 7523871

Pacheco Cedeño, Simón JoséPresidenteAsociación de Productores de YucaTransversal 11 No. 417, Urb. Fundemos IMaturín, Monagas

Tel.: (58-91) 512550Fax: (58-91) 412441

Vietnam

Dang Thanh HaLecturerDepartment of Agricultural EconomicsUniversity of Agriculture and ForestryHo Chi Minh City

Tel.: (84) 8-966780Fax: (84) 8-231541

Zambia

Namposya, RebeccaFarming Systems NutritionistFarming Systems Research TeamAdaptive Research Planning TeamMt. Makulu Research StationP/Bag 7, Chilanga

Tel.: (26-01) 278514Fax: (26-01) 213927

CIAT

Apartado Aéreo 6713 Cali, Colombia

Tel.: (57-2) 4450000Fax: (57-2) 4450073

Alarcón Morante, FreddyAssistant

Ayala Aponte, Alfredo AdolfoResearch Assistant

Bellotti, AnthonyEntomologist

Best, RupertLeader, Cassava Program

Bonierbale, MeridethGeneticist

Brabet, CatherineVisiting Researcher

Dufour, DominiqueCassava Utilization SpecialistCIRAD/SAR-CIAT

El-Sharkawy, MabroukPhysiologist

Henry, GuyEconomist

Jones, DebbieChemical Engineer

López Alarcón, John MarioStudent

Mosquera Palacio, LilianaResearch Assistant

Murcia, LuzmilaTechnician

O’Brien, Gerard MichaelFood Scientist

Orozco Marmolejo, OswaldoTechnician

Ostertag, Carlos FelipeAssociate

Pérez Valdés, DiegoResearch Assistant

Salcedo Bonilla, Enna ElenaTechnician

Sánchez, TeresaResearch Assistant

Scowcroft, WilliamDeputy Director General

Thro, Ann MarieCoordinator, CBN

402


APPENDIX II

LIST OF ACRONYMS AND ABBREVIATIONS

USED IN TEXT

Acronyms

AACC American Association ofCereal Chemists, USA

ABAM Associação Brasileira dosProdutores de Amido deMandioca, Brazil

ABES Associação Brasileira deEngenharia Sanitária,Brazil

ACDI Agricultural CooperativeDevelopment International,Colombia

AFST Association of Food ScienceTechnology, India

ANPPY Asociación Nacional deProductores y Procesadoresde Yuca, Colombia

AOAC Association of OfficialAnalytical Chemists, USA

APPYs Asociaciones deProductores y Procesadoresde Yuca, Ecuador

ASOCOSTA Asociación de Cooperativasde la Costa, Colombia

ASTM American Society forTesting and Materials,USA

ATAPYs Asociaciones deTrabajadores Agrícolas yProductores de Yuca,Ecuador

BFAD Bureau of Food and Drugs,the Philippines

BNH Banco Nacional deHabitação, Brazil

BOI Board of Investment,Thailand

BORIF Bogor Research Institute forFood Crops, Indonesia

CA Département des culturesannuelles (CIRAD)

CBN Cassava BiotechnologyNetwork, based in Colombia

CBS Central Bureau ofStatistics, Indonesia

CDA Cooperative DevelopmentAuthority, the Philippines

CDF Countrywide DevelopmentFund, the Philippines

CENDES Centro de Desarrollo,Ecuador

CEPAGRO Centro Estadual dePesquisa Agronómica,Brazil

CERAT Centro Raizes Tropicais(UNESP)

CETEC Corporación para EstudiosInterdisciplinarios yAsesorías Técnicas,Colombia

403

Appendix II: List of Acronyms and Abbreviations Used in Text

CETESB Companhia de Tecnologiade Saneamento Ambiental,Brazil

CGPRT Center for Research andDevelopment of CoarseGrains, Pulses, Roots andTuber Crops in the HumidTropics of Asia and thePacific, Indonesia

CIP Centro Internacional de laPapa, based in Peru

CNPMF Centro Nacional dePesquisa de Mandioca eFruticultura (EMBRAPA)

CONAB Companhia Nacional deAbastecimento, Brazil

COOPEMUBA Cooperativa deProductores de Mandiocade Ubajara, Brazil

COOPROALGA Cooperativa deProductores de losAlgarrobos, Colombia

COPROMA Cooperativa deProductores de Mandiocade Acarau, Brazil

CORAF Conférence desresponsables derecherche agronomique enAfrique de l’Ouest et duCentre

CPC Corn Products Company,USA

CRIFC Central Research Institutefor Food Crops, Indonesia

CRVZ Centre de recherchevétérinaire et zootechnique(DGRST)

CTCRI Central Tuber CropsResearch Institute, India

CVC Corporación AutónomaRegional del Valle delCauca, Colombia

DANE DepartamentoAdministrativo Nacional deEstadística, Colombia

DEMSA Derivados del Maiz, S. A.(private company), Peru

DEPD Department of EconomicPlanning andDevelopment, Malawi

DGRST Direction générale de larecherche scientifique ettechnique, Congo

DOLE Department of Labor andEmployment, thePhilippines

DRI Fondo de Desarrollo RuralIntegrado, Colombia

DTI Department of Trade andIndustry, the Philippines

EEC European EconomicCommunity, now the EU

EMATER Empresa de AssistênciaTécnica e Extensão Rural,Brazil

EMATER-CE Empresa de Pesquisa,Assistência Técnica eExtensão Rural do Ceará,Brazil

EMBRAPA Empresa Brasileira dePesquisa Agropecuária,Brazil

ENSAM Ecole nationale supérieureagronomique deMontpellier, France

ENSBANA Ecole nationale supérieurede biologie appliquée à lanutrition et l’alimentation,France

ENSIA Ecole nationale supérieuredes industries agricoles etalimentaires, France

EPACE Empresa de PesquisaAgropecuária do Ceará,Brazil

EPN Escuela PolitécnicaNacional, Ecuador

ERS Economic ResearchService (USDA)

404


ESAL Escola Superior deAgronômia de Lavras,Brazil

ESALQ Escola Superior deAgricultura “Luiz deQueiroz”, Brazil

EU European Union, oftenknown as the EC or EEC

FAC Facultad de Ciencias (UCV)

FACE Fundación AdelantoComunitario, Ecuador

FAO Food and AgricultureOrganization of the UnitedNations, Italy

FCA Faculdade de CiênciasAgronômicas (UNESP)

FIBGE Fundação InstitutoBrasileiro de Geografia eEstatística, Brazil (alsoIBGE)

FODERUMA Fondo para el DesarrolloRural del Ministerio deAgricultura, Ecuador

FUNDAGRO Fundación para elDesarrollo Agropecuario,Ecuador

FUNDAIN Fundación Paraguaya deApoyo a la Agroindustria,Paraguay

FUNDIAGRO Fundación para laInvestigación y elDesarrollo de TecnologíasApropiadas al Agro,Colombia

GATT General Agreement onTariffs and Trade, EU

GBSA Laboratoire demicrobiologie et biochimieindustrielles of theUniversité Montpellier II,France

GNCTDC Guangxi Nanning CassavaTechnical DevelopmentCenter, China (alsoNCTDC)

IAC Instituto Agronômico deCampinas, Brazil

IADS International AgriculturalDevelopment Service, NewYork, USA

IAEA International AtomicEnergy Agency, Italy

IAPAR Instituto Agronômico doParaná, Brazil

IARCs International AgriculturalResearch Centers of theConsultative Group onInternational AgriculturalResearch (CGIAR), USA

IBGE see FIBGE

IBPGR International Board forPlant Genetic Resources,now IPGRI

IBRD International Bank forReconstruction andDevelopment (also knownas the World Bank), USA

IBSRAM International Board of SoilResources andManagement, Thailand

ICA Instituto ColombianoAgropecuario

ICH International Child HealthUnit, Sweden

ICMSF International Commissionon MicrobiologicalSpecifications for Foods,UK

ICONTEC Instituto Colombiano deNormas Técnicas

ICTA Instituto de Ciencia yTecnología Agrícolas,Guatemala

IDRC International DevelopmentResearch Centre, Canada

IFRPD Institute of Food Researchand Product Developmentof Kasetsart University,Thailand

405


IFS International Foundationfor Science, Sweden

IIAP Instituto deInvestigaciones de laAmazonía Peruana

IICA Instituto Interamericanode Cooperación para laAgricultura, Costa Rica

IIED International Institute forEnvironment andDevelopment, UK

IITA International Institute ofTropical Agriculture,Nigeria

INA-PG Institut nationalagronomique,Paris-Grignon

INCIENSA Instituto Costarricense deInvestigación y Enseñanzaen Nutrición y Salud,Costa Rica

INDELMA C.A. Industrias del Maíz C.A.,Venezuela

INIA Instituto Nacional deInvestigación Agraria, Peru

INIAP Instituto Nacional deInvestigacionesAgropecuarias, Ecuador

INN Instituto Nacional deNutrición, Venezuela

INRA Institut national derecherche agronomique,France

IPESAT Institut provenciald’enseighement superieuragronomique et techniquedu Hainaut, Belgium

IPGRI International Plant GeneticResources Institute, Italy

ISNAR International Service forNational AgriculturalResearch, the Netherlands

ISU Isabela State University,the Philippines

ITCF Institut technique descéréales et des fourrages ofthe Céréaliers du France

KUD Koperasi Unit Desa (VillageUnit Cooperative),Indonesia

MAE Ministère des AffairesEtrangères, France

MAG Ministerio de Agricultura,Ecuador

MARA Ministério da Agricultura eReforma Agrária, Brazil

MARCA Mabagon Root CropAssociation, thePhilippines

MIPRE Ministerio de laPresidencia, Peru

MOAC Ministry of Agriculture andCooperatives, Thailand

MOI Ministry of Industry,Thailand

NAPHIRE National PostharvestInstitute for Research andExtension, the Philippines

NCTDC see GNCTDC

NRI Natural ResourcesInstitute, UK

ORSTOM Institut français derecherche scientifique pourle développement encoopération, France

PCARRD Philippine Council forAgriculture and ResourcesResearch and Development

PMO Prime Minister’s Office,Tanzania

PRCRTC Philippine Root CropResearch and TrainingCenter

PRODAR Programa Cooperativo deDesarrollo AgroindustrialRural, Costa Rica

406


PRONAA Programa Nacional deAlimentación, Peru

PROPAL S.A. Productora de Papeles S.A.,Colombia

SEARCA Southeast Asian RegionalCenter for Graduate Studyand Research inAgriculture, the Philippines

SECTI Secretaría Ejecutiva deCooperación Internacional(MIPRE)

SEDECOM Servicio de Desarrollo yConsultoría para elSector Cooperativo y deMicro-Empresas, Colombia

SFCDP Secondary Food CropsDevelopment Project,Indonesia

SIARC Section industries agricoleset alimentaires des régionschaudes (ENSIA)

SUDENE Superintendência doDesenvolvimento doNordeste, Brazil

TAPPI Technical Association of thePulp and Paper Industry,New York

TDRI Thailand DevelopmentResearch Institute

TDRI Tropical Development andResearch Institute, UK

TPPIA Thai Pulp and PaperIndustries Association

TTFITA Thai Tapioca FlourIndustries TradeAssociation

TTTA Thai Tapioca TradeAssociation

UAM Universidad AutónomaMetropolitana, Mexico

UAPPY Unión de Asociaciones deProductores y Procesadoresde Yuca, Ecuador

UATAPPY Unión de Asociaciones deTrabajadores Agrícolas,Productores yProcesadores de Yuca,Ecuador

UBA Universidad de BuenosAires, Argentina

UCV Universidad Central deVenezuela

UEM Universidade Estadual deMaringá, Brazil

UEPG Universidade Estadual dePonta Grossa, Brazil

UFPR Universidade Federal doParaná, Brazil

UFSC Universidade Federal deSanta Catarina, Brazil

UNALM Universidad NacionalAgraria “La Molina”, Peru

UNESCO United Nations Education,Scientific, and CulturalOrganization, France

UNESP Universidade EstadualPaulista, Brazil

UNIVALLE Universidad del Valle,Colombia

USAID United States Agency forInternationalDevelopment, USA

USDA United States Departmentof Agriculture

UST University of Science andTechnology, Ghana

UTC Université de Tecnologiede Compiègne, France

UWI University of the WestIndies, Trindad andTobago

ViSCA Visayas State Collegeof Agriculture, thePhilippines

407


WAG Water Activity Group of theEuropean Cooperation inthe Field of Science andTechnical Research(investigates water contentof substances)

WHO World Health Organization

WIPS “Women in PostproductionSystems,” collaborativeproject in the Philippines

Abbreviations

ALAB Amylolytic lactic acidbacteria

APC Aerobic plate count

ATP Adenosine triphosphate

aw Water activities (indetermining foodisotherms)

BA Brabender amylograph

BE Starch-branching enzyme(gene responsible for thecross linkages that formamylopectin)

BMP Bread-making potential

BOD Biochemical oxygendemand

BU Brabender viscosity units(for starches)

CAP Common AgriculturalPolicy of the EU

CCF Chlorinated cake flour

cDNA Complementary DNA

CF Control fermentation

cfu Colony-forming unit

CG Cyanogenic glucosides

c.i.f. (C.I.F.) Cost, insurance, andfreight

CNP Total cyanogenic potential

COD Chemical oxygen demand

COO- Chemical symbol ofdouble-bound carbon

d.b. Dry basis

DEAE Diethylaminoethyl (used inenzyme analysis)

DM Dry matter

DNA Deoxyribonucleic acid

DSC Differential scanningcalorimetry

D.W. Devon-Watson estimate(statistics)

ECU European currency unit

EU Enzyme units

f.o.b. (F.O.B.) Free on board

FRR Financial rate of return

fwb Fresh weight basis

g Gravitation constant (incentrifuging)

G.A.B. The Guggenheim-Anderson-De Boer modelused for describing foodisotherms in equations

GBSS Granule-bound starchsynthase (gene responsiblefor amylose synthesis)

GDP Gross domestic product

g.f.b. Glucose-fermentingbacteria

GNP Gross national product

GPI-ID Indice geral de preços-demanda interna (deflator,in economics = generalprice index-internaldemand)

hab Habitant

408


HCN Hydrogen cyanide, andsometimes used to expresscyanide content incassava

HDPE High-density polyethylene(used for packaging)

HFCS High fructose corn syrup,also known as isoglucose

HFS High fructose syrup

HPLC High-performance liquidchromatography

ICRDPs Integrated cassava researchand development projects

IQR Interquartile range(statistics)

kDa Kilo Dalton (measure ofmolecular weight)

l.a.b. Lactic acid bacteria

l.f.b. Lactate-fermenting bacteria

M Molecular weight

MDF medium density fiber board(timber)

MG Minais Gerais, state ofBrazil

MPN Most probable number (amethod of enumeratingmicroorganisms)

MRS de Man-Rogosa-Sharpe agarmedium

MS Modified starches

MSG Monosodium glutamate

MTBE Methyl tertiary butyl ether

NaCl Common salt or sodiumchloride

NGC Nonglucosidic cyanogens

NGFI Nongrain feed ingredients(relating to CAP)

NGOs Nongovernmentalorganizations

OD Optical density

OVL Organic volume load

p Pressure (used inphysicochemicalmeasurements)

PCA Plate count analysis (forestimating microbialpopulations)

PCR Polymerase chainreaction

PDA Potato dextrose agarmedium

PE Pectinesterase (pectinpectylhydrolaseE.C. 3.1.1.11)

PG Polygalacturonase (poly(1,4-α-D-galacturonide)glycanohydrolase,E.C. 3.2.1.15)

PGL Polygalacturonate lyase

pKa Negative logarithm ofequilibrium constant forassociation (used formeasuring acidification)

PNPG p-nitrophenol-β-D-glucopyranoside (achromogen)

pO2 Partial oxygen pressure

PVC Polyvinyl chloride

R&D Research and development

RAPD Random amplifiedpolymorphic DNA

RFLPs DNA restriction fragmentlength polymorphisms

RMS % Relative root meansquare error

rpm Revolutions per minute

RVA Rapid Visco Analyzer(equipment for measuringstarch viscosity profiles,which are expressed in“RVA units”)

409


SDS Sodium dodecyl sulfate

SDS-PAGE SDS-polyacrylamide gelelectrophoresis

SF Sterile fermentation

TC Total cyanogen content

TR Taxa de reajuste (afterreadjustment, ineconomics)

UMS Unmodified or nativestarches

UV Ultraviolet (radiation)

V Volt

VFA Volatile fatty acids

VSS Volatile suspended solids

VUC Village union cooperative,Indonesia

w/v Weight by volume

w.b. Wet basis

w/w Weight-to-weight ratio

WSI Water solubility index

xiii

Contents

CIAT Publication No. 271CIAT Publication No. 271CIAT Publication No. 271CIAT Publication No. 271CIAT Publication No. 271

CIRAD/SARCIRAD/SARCIRAD/SARCIRAD/SARCIRAD/SAR andandandandandCIAT’s Communications UnitCIAT’s Communications UnitCIAT’s Communications UnitCIAT’s Communications UnitCIAT’s Communications Unit

Editing: Elizabeth L. de PáezAnnie Jones

Editorial assistance: Gladys Rodríguez

Translation(six chapters): Lynn Menéndez

Deborah Jones

Photography: Dominique DufourCIAT Photographic Section

Production: Graphic Arts Unit, CIATAlcira Arias (layout)Jorge Gallego (layout)Julio C. Martínez (cover design)

Printing: Impresora Feriva S.A., Cali

Documents

The Centre de coopération internationale en recherche … · 2017-12-17 · The Centre de coopération internationale en recherche agronomique pour le développement (CIRAD) is a