THE ECOLOGICAL SOCIETY OF AMERICA Frontiers Ecology...Dec 07, 2011 · `THE ECOLOGICAL SOCIETY OF AMERICA Frontiers inEcology and the Environment Can tropical farmers reconcile subsistence

` T H E E C O L O G I C A L S O C I E T Y O F A M E R I C A

Frontiers in Ecologyand the Environment

Frontiers in Ecologyand the Environment

Can tropical farmers reconcile subsistenceneeds with forest conservation? TThhoommaass KKnnookkee11**,, BBaallttaazzaarr CCaallvvaass11,, NNiikkoollaayy AAgguuiirrrree22,, RRoossaa MMaarrííaa RRoommáánn--CCuueessttaa33,, SSvveenn GGüünntteerr33,, BBeerrnndd SSttiimmmm33,,MMiicchhaaeell WWeebbeerr33,, aanndd RReeiinnhhaarrdd MMoossaannddll33

548

wwwwww..ffrroonnttiieerrssiinneeccoollooggyy..oorrgg © The Ecological Society of America

Currently, there are heated debates on the role of mar-kets as forest “lifesavers” (Fearnside 2001; Knoke etal. 2008a; Potvin et al. 2008). However, ongoing discus-sions on reducing emissions from deforestation and forestdegradation (REDD), conservation of non-market valuessuch as biodiversity, and integration of deforestation intothe international carbon markets by means of paymentsfor ecosystem services (PES) are condemned to fail, both

socially and economically, if the local people’s needs arenot taken into account in that debate. The problem istwo-pronged: economic on the one hand and philosophi-cal on the other. Economically, implementing strategiesto avoid deforestation through PES may be expensive.Global estimates of the compensation costs for forestpreservation amount to US$5–10 billion annually, if70% of CO2 emissions associated with current land-usepractices are to be avoided (Stern 2006). To give credi-bility to the ongoing REDD efforts, realistic financialoptions must be found (Potvin et al. 2008). Philo-sophically, debates are suffering from what – in the con-text of biological conservation – could be defined as the“people versus parks dilemma” (Schwartzman et al. 2000;Terborgh 2000), where values such as wilderness andpristine (untouched) ecosystems prevail over the subsis-tence needs of local people. In this debate, it is doubtfulwhether conservation values can be harmonized with anytropical land-use system. However, even if it were possi-ble to mobilize substantial PES to compensate people fornot engaging in activities leading to deforestation, wecannot exclude humans from the landscape. It is unlikelythat tropical farmers would remain idle while receivingPES; they need an economic system that keeps thememployed in productive and sustainable activities that donot result in habitat destruction (Knoke et al. 2008a).Research should therefore be focused on identifying andevaluating new production systems and alternative man-agement techniques (Bennett and Balvanera 2007).

We investigated a quantitative concept within ecologi-cal–economic farm diversification (EEFD) that is capableof convincing farmers of the benefits of sustainable prac-tices and of halting deforestation. Our objective is to

CONCEPTS AND QUESTIONS

Can tropical farmers reconcile subsistenceneeds with forest conservation? TThhoommaass KKnnookkee11**,, BBaallttaazzaarr CCaallvvaass11,, NNiikkoollaayy AAgguuiirrrree22,, RRoossaa MMaarrííaa RRoommáánn--CCuueessttaa33,, SSvveenn GGüünntteerr33,, BBeerrnndd SSttiimmmm33,,MMiicchhaaeell WWeebbeerr33,, aanndd RReeiinnhhaarrdd MMoossaannddll33

If tropical farmers cannot be provided with sustainable land-use systems, which address their subsistenceneeds and keep them gainfully employed, tropical forests will continue to disappear. We looked at the abil-ity of economic land-use diversification – with reforestation of tropical “wastelands” as a key activity – tohalt deforestation at the farm level. Our ecological–economic concept, based on land-use data from thebuffer area of the Podocarpus National Park in southern Ecuador, shows that stopping deforestation after 10years is possible without violating subsistence demands. Tropical, farm-level diversification may not onlyreduce total deforestation by 45%, but also increase farmers’ profits by 65%, because the formerly unpro-ductive wastelands have been returned to productive land use. We therefore conclude that a “win–win” sce-nario is possible: the subsistence needs of people can be reconciled with conservation objectives. However,inexpensive microcredits (at interest rates below 6%) and experience on alternative land-use opportunitiesmust be offered to farmers.

Front Ecol Environ 2009; 7(10): 548–554, doi:10.1890/080131 (published online 11 Mar 2009)

1Institute of Forest Management, Center of Life and Food SciencesWeihenstephan, Technische Universität München, Freising,Germany *([email protected]); 2National University ofLoja, Ciudadela Universitaria, Loja, Ecuador; 3Institute ofSilviculture, Center of Life and Food Sciences Weihenstephan,Technische Universität München, Freising, Germany

IInn aa nnuuttsshheellll::• Tropical conservation values are seen as incompatible with

livelihood demands; payments for ecosystem services are sup-posed to avoid destructive activities like deforestation

• However, there are inherent socioeconomic problems in choos-ing land-management options in tropical landscapes

• Alternative ecological–economic land-use systems may help tosolve these problems, by restoring natural resources in conjunc-tion with tropical forest conversion and product diversification

• These reduce farmers’ risk of losses and the demand for land;this concept applies particularly to subsistence farms located inmontane tropical ecosystems, where slow succession ofoverused, degraded, and abandoned lands toward secondaryforests is characteristic

• Inexpensive microcredits, the establishment of local timbermarkets, the transfer of experience with alternative land-useoptions to farmers, and the allocation of higher value to stand-ing forests are important preconditions for sustainable land use

T Knoke et al. Reconciling subsistence and forest conservation

counter the following hypothesis: “As long as people usetropical landscapes for their livelihood, they will con-tinue to convert native forests into alternative uses, as itis profitable for them” (Tschakert et al. 2007). After test-ing this hypothesis, we identified the basic conditionsunder which sustainable, non-destructive tropical landuse would be possible.

� Background

We developed the EEFD concept (WebModel 1) as amulti-partner project on biodiversity and sustainablemanagement in a mega-diverse, montane, southEcuadorian ecosystem (Beck et al. 2008), located nearthe Andean Podocarpus National Park. Here, and else-where in the tropics, farmers customarily convert tropi-cal forests into new farmland to replace degraded – andultimately abandoned – pastures (sensu Silver et al.2000). These abandoned lands (“wastelands”) are usuallyleft unproductive, and this is followed by a very slow suc-cession toward shrub vegetation and secondary forest(Paulsch et al. 2001). However, these overused anddegraded wastelands can be reintroduced into the eco-nomical system, and thus serve as a powerful means todiversify land management.

� Economics of land-use options

Our model shows that, for a small farm (30 ha) including10 ha of previously degraded wastelands, the traditionalland-development strategy would lead to deforestation of18.3 ha over the course of one generation of farmers (40years; Figure 1a). This type of land management providesrelatively constant net revenues (Figure 2a) when a spe-cific probability of pasture abandonment is given(WebTable 1). The risk of future yield losses due to pas-ture degradation promotes further deforestation; further-more, after the establishment of new pastures, the con-version of tropical forests continues (after 10, 20, and 30years), to compensate for yield losses as a result of pastureabandonment.

The financial consequences of three possible land-useoptions (reforestation of abandoned pastures, cattle pas-ture, and selective logging in natural forests; Figure 3)seem to confirm pasture as the most economically attrac-tive option for the farmer (Figure 4). However, the uncer-tainty of the financial estimates, expressed by their stan-dard deviation (WebTable 1), is by far the highest for thisoption. High uncertainty of financial consequences leadsto the clear-cutting of larger new areas, to decrease poten-tial risk. Management options that reduce the variability

(a) 30

25

20

15

10

5

0

Land

are

a (h

a)

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

YearWasteland Active pasture Natural forest (without use)

(b)30

25

20

15

10

5

0

Land

are

a (h

a)

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

YearReforestation Wasteland Active pasture Selective logging Natural forest (without use)

FFiigguurree 11.. Allocation of land-use activities to the available farm area under (a) a classical single-use pasture system as compared with(b) ecological–economic farm-level diversification.

(a)

100 000

10 000

1000

100

Net

rev

enue

(US

$)

Average

Minimum (0.1-quantile)

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39

Year

(b)100 000

10 000

1000

100

Net

rev

enue

(US

$)

Average

Minimum (0.1-quantile)

FFiigguurree 22.. Average expected and minimum net revenues under (a) a classical single-use pasture system as compared with (b)ecological–economic farm-level diversification. Note the logarithmic scale of the y-axis.

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39

Year

549

© The Ecological Society of America wwwwww..ffrroonnttiieerrssiinneeccoollooggyy..oorrgg

Reconciling subsistence and forest conservation T Knoke et al.

550


of financial flows and thus the risk of losses should there-fore contribute to the prevention of deforestation andimplementation of sustainable land use.

� Addressing subsistence and restoration needs

Modeling sustainable land use should address farmers’subsistence needs under the given risks, and this shouldreplace the often-assumed pure profit maximization offarmers (eg Carpentier et al. 2000). As a result, the EEFDmodel suggests an immediate land-use diversification thatprovides multiple products, ie agricultural commoditiesand timber obtained by selective logging (Figure 1b).Diversification stabilizes the uncertain net revenues andreduces the demand for land, which is necessary to pro-vide subsistence net revenues. Agricultural and timbermarkets are uncorrelated (Lönnstedt and Svensson2000); the price of timber may be moderate or high whenthe price of milk is low and vice versa, resulting in aneffective compensation of market price fluctuations.Moreover, forestry production is independent of agricul-tural yields. Our model took these effects into account byintegrating market price fluctuations, correlations andpasture degradation, and changes in dairy productivity, aswell as uncertainties associated with sustainable harvestunder selective logging and fire damage (WebTable 1).

The combination of pasture and selective logging(Figure 2b) produces higher minimum net revenues peryear than does the single-use pasture system (Figure 2a),although the conversion of tropical forests is only 8.7 ha

instead of 10 ha under the single-use pasture system.Modeling minimum net revenues is adequate to address thesubsistence needs of the tropical farmer, whom we assumedto be risk adverse (Pichón 1996). For this farmer, it is notenough to achieve subsistence net revenues only on aver-age; instead, it is necessary to obtain them every year. Giventhe data in WebTable 1, we estimated probability distribu-tions of yearly net revenues for various combinations of dif-ferent land uses. We defined those net revenues as the min-imum that can be achieved with a probability of 0.9. Theresults show that the requirement of achieving subsistencenet revenues is better fulfilled by the diversified land-usesystem, combined with a reduction in land demand.

The key activity in halting deforestation is the accumula-tion of “new” natural resources through wasteland reforesta-tion. For every unit area of forest converted into pasture,the farmer must reforest a corresponding area of wasteland.To this end, farmers can use the productive native treespecies Andean alder (Alnus acuminata) (Figures 3, 5),adapted to the landscapes of south Ecuador and also to awide range of areas in Latin America. This nitrogen-fixingspecies is widely recommended for agroforestry or reforesta-tion in Ecuador because of its fast growth, soil improvementcapability, tolerance to diverse environmental conditions,and multiple uses (Dunn et al. 1990). The timber ofAndean alder can be used for a variety of purposes, includ-ing firewood, packing boxes, or construction; the tree alsohas palatable, nitrogen-rich leaves and can be used as asource of “energy fodder”. It resprouts vigorously after har-vesting (Grau and Veblen 2000) and can be managed bymeans of a coppice system. The first commercial harvestscan be expected from age 10 years onwards (Dunn et al.1990). Andean alder achieves fire resistance from age 5onwards, and this is strong enough to survive the low-inten-sity fires typical of open woodland sites (Grau and Veblen2000). Furthermore, the aesthetic appearance of the land-scape is only slightly affected by plantings of Andean alder(Figure 3), and biodiversity levels in Andean alder planta-tions may be high, even when compared with those in nat-urally regenerated forests (Murcia 1997).

Reforestation makes intuitive sense as well, because thefarmer reintegrates unproductive areas into areas of pro-ductive use and simultaneously achieves restorationeffects, given that Andean alder accumulates nitrogen inthe soils. Increased levels of carbon sequestration areanother ecological benefit of Andean alder plantations.From the perspective of the farmer, however, the maineffect of the reforested areas is the compensation of yieldlosses from pasture degradation as a result of revenuesfrom thinning and final harvesting operations.

� Financial consequences of halted deforestationand mixed land use

The results of our model (WebModel 1) suggest that halt-ing deforestation after 10 years is possible, when Andeanalder fulfills diversification requirements (replacing tim-

FFiigguurree 33.. Andean alder (Alnus acuminata) plantationsurrounded by natural forest and cattle pasture.

B S

timm


ber from the natural forest), first by providingthinning revenues and – after a period of 20years (rotation) – by harvesting the finalcrops and growing a second rotation ofAndean alder. Despite some calculated lossesof Andean alder through fire (Figure 1b andWebTable 1), reforestation areas diversify theland-use portfolio effectively and compen-sate farmers for losses due to pasture degrada-tion and market price fluctuations of agricul-tural products. After 40 years, 8.3 ha oftropical forest per farm will have been con-served under the optimized concept of EEFD,as compared with that under the classical sys-tem, and valuable natural resources will beavailable, as a result of the Andean alderplantations. During the 40-year time frame,land devoted to Andean alder can be used tore-establish agriculture, while the areasdegraded by pasturing can be reforested.

Following the optimized management pathunder EEFD (WebModel 1), farmers canachieve a 65% increase in profit from theirland (US$20 680 ± 2260, discounted at arisk-free 5%, versus US$12 560 ± 2560 withthe classical system), with deforestationbeing limited to a maximum of 10 ha per farm. The farmprofit at risk (profit that is achieved with a probability of0.9), which we actually maximized (WebModel 1), is109% greater for the EEFD approach (US$16 970 versusUS$8100 for single-use pasture). Our model shows con-siderable peaks in revenue in year 20, 30, and 40 (Figure2b), when final crops of Andean alder plantations can beharvested. These financial results should be attractive tofarmers who are not necessarily interested in conserva-tion, because – from their point of view – fulfilling subsis-tence needs and increasing profits are of primary impor-tance. Although these results remain stable underchanging revenue coefficients, as expected, the discountrate was critical to the success of this sustainable land-management system (Pearce et al. 2003).

Non-linearities as a result of the introduction of uncer-tainty led us to use non-linear programming to solve theproblem of diversified farm management (WebModel 1;Knoke and Moog 2005). To estimate coefficients anduncertainties, we used data on montane tropical forests,reforestation, and pasture under site conditions in theAndes provided by the German-Ecuadorian multi-part-ner DFG (Deutsche Forschungs-gemeinschaft) researchgroup, summarized in WebTable 1.

� Conclusions

Our results are similar to those of Grafton et al. (2007), inthat we were able to show a feasible win–win scenario;that is, it is possible to reconcile conservation objectivesand the subsistence needs of local people. However, ade-

quate land-use concepts and ecological–economic mod-els are prerequisites to achieve this outcome. Risk inte-gration into the model leads to realistic results andacceptability for farmers. Economic analysis often endswith the comparison presented in Figure 4. The compar-atively high net revenues obtained from cattle pasturesseem to support this option as the only attractive land-use alternative. However, ecological–economic farmmodeling, focused on subsistence revenues and long-termeffects, shows the economic benefits of diversificationand reforestation of abandoned wastelands. We suggestthat the absence of quantitative experience of the avail-able land-use options and of land-use combinations is animportant factor in the currently accelerating deforesta-tion. Benjamin et al. (2008) also suggest that lack ofknowledge about different land-use options that facilitatereforestation could explain why there is so little reforesta-tion of abandoned farmland, even in developed coun-tries. Next steps will include transferring the results tofarmers and the establishment of a demonstration farm inour south Ecuadorian research area. Our model, based onexperimental data, contradicts the original hypothesisthat human subsistence and tropical forest conservationcannot be reconciled. Farmers can be provided with sus-tainable and profitable alternatives.

Resource managers, ecologists, and consultants canapply the EEFD modeling approach to analyze and opti-mize land-management opportunities and strategies fromthe farmers’ perspective, taking into account individualpreferences, requirements, and capacities. Decisions onland management would, however, remain with the

FFiigguurree 44.. Annualized net revenues (annuities) of land-use options and theirstandard deviation (black bars). The sum of annualized net revenues,discounted by 5%, forms the net present value (NPV) of land-use options(NPV is the sum of discounted actual net revenues, with the net revenuesfluctuating from year to year in the cases of reforestation and pasture). We referto a 20-year period for this example.

140

120

100

80

60

40

20

0Ann

ualiz

ed n

et r

even

ues

(US

$ ha

–1yr

–1)

Reforestation ofabandoned pasture

Cattle pasture Selective logging

551



552


farmer, who has to be convinced mainly by the economicarguments. We have to estimate biophysical and financialcoefficients for different regions. Also, the land-useoptions discussed here do not represent all possibleoptions (eg bee keeping, nurseries of native tree seedlingsand saplings, non-timber forest products, home gardens,etc). Our example therefore represents a first step in ana-lyzing and modeling options to reconcile subsistence andconservation demands.

Finally, we provide some general conditions andrequirements, based on our results, to achieve sustainableland use in the tropics:

(1) Alongside the conversion of tropical forests, alterna-tive natural resources must be provided (in our case,by means of reforestation of wastelands). This followsa basic requirement of sustainable development,whereby exploited natural resources must be replacedby other natural resources (Daly 1990). Without suchreplacement, depletion of existing natural resourcescannot be justified from a sustainability perspective.This is increasingly important in abandoned landscharacterized by slow recovery, such as the pastures inmontane southern Ecuador.

(2) To further improve the productivity of agriculturallands, it is important to establish alternative land uses(for example, forestry on so-far-unused wastelands),which do not compete economically with the stand-ing natural forests. Relying on agricultural intensifi-cation alone is problematic, as exemplified by palm-

oil plantations in lowland tropical ecosystems.Intensified production on pastures would reducedemand for land only under one specific set of cir-cumstances: the farmer has to stop production imme-diately when subsistence net revenues are achieved.This does not seem very realistic; rather, the activitiesof farmers are limited by the availability of workers,who are normally family members (Tschakert et al.2007). If (in our example) we double the net revenuesfrom pasture to simulate intensification, the farmercould earn substantially more money during the first10 years of farm management. This alone wouldenable payment of additional workers from outsidethe farmer’s family, great enough to increase thedeforested area by 5.2 ha. It is likely that the farmerwould carry out extensive deforestation, because thiswould increase short-term profitability. In conclusion,therefore, we believe that improved agricultural pro-ductivity will probably increase deforestation(Carpentier et al. 2000).

(3) The economic diversification of farmlands has thepotential to reduce land demand. Crop diversificationis particularly important for subsistence farmers withsmall land holdings, because farmers with large landareas may have other alternatives to balance out therisks (eg financial investments). Whereas concentrat-ing on a single crop exposes farmers to diverse risks (eginternational market fluctuations, crop losses frominclement weather, pests, or diseases), having a diver-sity of products provides farmers with a form of eco-nomic insurance, should the market for any particularproduct decline (Pichón 1996). In spite of the obviousparallels with financial diversification (of one’s per-sonal investments) and diversification of natural assets(of crops by farmers), the virtues of diversification aregenerally not acknowledged with regard to natural sys-tems (Figge 2004). Thus, the use of systematic diversi-fication strategies – based on quantitative, theoreti-cally well-founded optimization approaches – is onlyslowly becoming incorporated into ecosystem manage-ment. Only a few examples exist, from fisheries(Edwards et al. 2004), grassland (Koellner and Schmitz2006), and forest (Knoke and Seifert 2008) manage-ment. However, to our knowledge, no application ofthe economic diversification concept to reduce tropi-cal deforestation has been published so far.

(4) A high interest rate is detrimental to any long-terminvestment and thus to sustainable land use. Theapplied discount rate, if too high, can accelerateunsustainable land use (Pearce et al. 2003). The dis-count rates applied to natural systems should not bemuch above 5%. In our approach, we not only dis-counted all net revenues by 5%, but also assumedreforestation to be financed at a rate of 5% (Web-Table 1). Credit repayment after 20 years, includingaccumulated interest, should be secured, giving aprobability of 0.9 that the value of the timber

FFiigguurree 55.. Tree of the native species Andean alder (Alnusacuminata) at age 8. The measured stand already represented avolume of 75 m3 ha–1.

B C

alva

s


timber of natural forests will probably be harvested tocompensate for these losses. Allocating more value to thestanding natural forests (Hartman 1976) would increasethe probability of their continued survival. According toour financial comparison of land-use options (Figure 4),the additional revenue from natural forests should be atleast US$40 ha–1 yr–1 to offset the financial advantages ofpasture management (ie the cost of retaining the tropicalforest under selective logging). However, it is debatablewhether this amount would be enough, since it does notrepresent a financial incentive for farmers to choosereforestation (Knoke et al. 2008b). Given the possiblecost of retaining tropical forests, the economic value ofthe ecosystem services they provide is huge (more thanUS$2000 ha–1 yr–1; Costanza et al. 1997). Unfortunately,this value (so far) is only theoretical. Future researchshould focus on sustainable, non-destructive, and market-based uses of natural forest products (eg seeds of nativetree species to generate saplings for reforestation, etc).Farmers might view reforestation as a more attractiveoption if they were offered earlier positive net revenues.In our model, the farmer would not obtain this added rev-enue until the 10th year; a solution involving mixedcrops could perhaps solve this problem.

We need to offer farmers alternative options to tropicaldeforestation. The sustainable use of landscapes should beseen as a practical way of guaranteeing social and ecologicalsustainability, not only in the tropics, but in any ecosystem.Ecologists and natural resource managers should use ecolog-ical–economic concepts to support the development ofmore sustainable, alternative land-management systems.

� Acknowledgements

We are grateful to the Deutsche Forschungsgemeinschaftfor financial support, to the members of the researchgroup FOR 816, whose research initiative made the studypossible, and to the Alexander von HumboldtFoundation for supporting the research of RMR-C.

� ReferencesBarbier EB. 2007. Frontiers and sustainable economic develop-

ment. Environ Resour Econ 3377: 271–95.Beck E, Makeschin F, Haubrich F, et al. 2008. The ecosystem

(Reserva Biologica San Francisco). In: Beck E, Bendix J,Kottke I, et al. (Eds). Gradients in a tropical mountain ecosys-tem of Ecuador. Ecol Stud 119988: 1–13.

Benjamin K, Bouchard A, and Domon G. 2008. Managing aban-doned farmland: the need to link biological and sociologicalaspects. Environ Manage 4422: 603–19.

Bennett EM and Balvanera P. 2007. The future of production sys-tems in a globalized world. Front Ecol Environ 55: 191–98.

Carpentier CL, Vosti SA, and Witcover J. 2000. Intensified produc-tion systems on western Brazilian Amazon settlement farms:could they save the forest? Agr Ecosyst Environ 8822: 73–88.

Costanza R, d’Arge R, de Groot R, et al. 1997. The value of theworld’s ecosystem services and natural capital. Nature 338877:253–60.

Daly HE. 1990. Toward some operational principles of sustainabledevelopment. Ecol Econ 22: 1–6.

obtained through reforestation will exceed the repay-ment amount at year 20 (WebModel 1). A 5% inter-est rate is very low. So far, farmers’ access to inexpen-sive credit is limited in Ecuador. Commercial credithas an average interest rate of about 12% (BancoNacional de Fomento), whereas credit for this type ofinitiative may be even more expensive. Agriculturalcredit can be obtained for 5%, but this is limited to aperiod of 5 years, US$5000, and requires securities.This type of credit is not available for establishingtree plantations. In order to support sustainable land-management concepts, policy makers need to estab-lish local and inexpensive microcredits for farmers. Itis also crucial to establish personal contact betweenfarmers and members of the credit institute (Yunus2003). Furthermore, this type of credit should beextended mainly to women, since they are seen asparticularly reliable. Under these preconditions,repayment rates of 99% have been reported (Yunus2003). The provision of inexpensive credit, being atype of subsidy, could perhaps be supported by PESand might be preferable to pure subsidies, because iffarmers need to repay the credit they will be moti-vated to care for reforested areas, since this willdeliver the repayment amount in 20 years.

(5) Well-established markets for forest products are a fur-ther precondition for the success of the land-use con-cept described here. So far, the principal reforestationspecies (Andean alder) has been successfully used inmany Latin American countries (Dunn et al. 1990;Murcia 1997). In central and northern Ecuador, mar-kets already exist for its timber, and these markets canalso be established in southern Ecuador. Establishmentof a timber market could be accelerated by the govern-ment, by permitting the use of firewood or construc-tion wood within appropriate legal frameworks.

(6) Barbier (2007) identified agricultural expansion as afundamental feature of economic development in poorcountries. Yet, such development only generates smallamounts of additional net revenues, and these smallsums are not being reinvested in other sectors, leadingto inherently unsustainable development. Reforest-ation may be able to generate sufficient extra money toallow farmers to engage in additional economic activi-ties and may therefore help to achieve both sustainableland use and substantially improved quality of lifeamong farmers. As mentioned above, the Andeanalder plantations can generate substantial amounts ofmoney. Whether this is used for reforestation of aban-doned land or to re-establish pasture, there will still bea considerable amount left over, which can be saved orinvested in other parts of the economy.

One of the limitations of our approach is that the stand-ing natural forest has no value for the farmers, other thanas insurance. If reforestation or pasture re-establishmentefforts perform more poorly than expected, the standing

553



554


Dunn WW, Lynch AM, and Morgan P. 1990. Benefit–cost analy-sis of fuelwood management using native alder in Ecuador.Agroforest Syst 1111: 125–39.

Edwards SF, Link JS, and Rountree BP. 2004. Portfolio manage-ment of wild fish stocks. Ecol Econ 4499: 317–29.

Fearnside PM. 2001. Saving forests as a global warming counter-measure: an issue that divides the environmental movement.Ecol Econ 3399: 167–84.

Figge F. 2004. Bio-folio: applying portfolio theory to biodiversity.Biodivers Conserv 1133: 827–49.

Grafton RG, Kompas T, and Hilborn RW. 2007. Economics ofoverexploitation revisited. Science 331188: 160.

Grau HR and Veblen TT. 2000. Rainfall variability, fire and veg-etation dynamics in neotropical montane ecosystems innorth-western Argentina. J Biogeogr 2277: 1107–21.

Hartman R. 1976. The harvesting decision when a standing foresthas value. Econ Inq 1144: 52–57.

Knoke T and Moog M. 2005. Timber harvesting versus forestreserves-producer prices for open-use areas in German beechforests (Fagus sylvatica L). Ecol Econ 5522: 97–110.

Knoke T and Seifert T. 2008. Integrating selected ecologicaleffects of mixed European beech–Norway spruce stands inbioeconomic modelling. Ecol Model 221100: 487–98.

Knoke T, Stimm B, and Weber M. 2008a. Tropical farmers neednon-destructive alternatives. Nature 445522: 934.

Knoke T, Hildebrandt P, Klein D, et al. 2008b. Financial compen-sation and uncertainty: using mean-variance rule and sto-chastic dominance to derive conservation payments for sec-ondary forests. Can J Forest Res 3388: 3033–46.

Koellner T and Schmitz OJ. 2006. Biodiversity, ecosystem func-tion, and investment risk. BioScience 5566: 977–85.

Lönnstedt L and Svensson J. 2000. Return and risk in timberland

and other investment alternatives for NIPF owners. Scand JForest Res 1155: 661–69.

Murcia C. 1997. Evaluation of Andean alder as a catalyst for therecovery of tropical cloud forests in Colombia. Forest EcolManag 9999: 163–70.

Paulsch A, Schneider R, and Hartig K. 2001. Land-use inducedvegetation structure in a montane region of southernEcuador. Die Erde 113322: 93–102.

Pearce D, Putz FE, and Vanclay JK. 2003. Sustainable forestry inthe tropics: panacea or folly? Forest Ecol Manag 117722: 229–47.

Pichón FJ. 1996. The forest conversion process: a discussion ofthe sustainability of predominant land uses associated withfrontier expansion in the Amazon. Agric Human Values 1133:32–51.

Potvin C, Guay B, and Pedroni L. 2008. Is reducing emissionsfrom deforestation financially feasible? A Panamanian casestudy. Clim Policy 88: 23–40.

Schwartzman S, Moreira A, and Nepstad D. 2000. Arguing tropi-cal forest conservation: people versus parks. Conserv Biol 1144:1370–74.

Silver WL, Ostertag R, and Lugo AE. 2000. The potential for car-bon sequestration through reforestation of abandoned tropi-cal agricultural and pasture lands. Restor Ecol 88: 394–407.

Stern N. 2006. The economics of climate change. London, UK:HM Treasury.

Terborgh J. 2000. The fate of tropical forests: a matter of steward-ship. Conserv Biol 1144: 1358–61.

Tschakert P, Coomes OT, and Potvin C. 2007. Indigenous liveli-hoods, slash-and-burn agriculture, and carbon stocks in east-ern Panama. Ecol Econ 6600: 807–20.

Yunus M. 2003. Banker to the poor: micro-lending and the battleagainst world poverty, 2nd edn. London, UK: Aurum Press.


WebModel 1. Ecological–economic farm-level diversification (EEFD)

VaR Value at risk: present value of all future management activities under careful assumptions. Given the expected coefficients and uncertainties,VaR is achieved or exceeded with a probability of 1 – � = 0.9.

�–1 Inverse of the normal distribution function

1-� Probability with which a financial figure is achieved or exceeded and � as the quantile of the distribution, ie the accepted error probability, here � = 0.1

E[.] Expected value (average)

R̃F Estimated present value of all future farm activities: sum of discounted net revenues, net present value

�̃ Estimated standard deviation

l Land-use activity

L Set of land-use activities (or types) (l: pasture, reforestation, selective logging, wasteland, natural forest)

t Point in time

T Set of all points in time (t = 0, 1, 2, ..., 40)

q Discount factor, here q = 1.05

r~ Estimated yearly net revenue from one single land-use activity per hectare

a Decision variable, area in hectare of a land-use activity allocated to an available land-cover type

kl,j Correlation coefficient between yearly net revenues of land-use activity l and j

rmin Minimum farm net revenues achievable every year, maximized iteratively

aReforestation Reforested wasteland area

aNew_pasture Newly established pasture, tropical forest conversion area

tNew_pasture Points in time when new pasture establishment was modelled

IReforestation Initial investment to establish Andean alder reforestation, credit financed and to be paid back (inclusive compound interest) after 20 years; repayment amount to be achieved with a probability of 0.9

V~

Stumpage value of Andean alder plantation

T Knoke et al. – Supplemental information

Supplemental information T Knoke et al.

wwwwww..ffrroonnttiieerrssiinneeccoollooggyy..oorrgg ©© The Ecological Society of America

WebTable 1. Input coefficients and sources used to derive them

Land-use Uncertaintyoption Item Coefficient References (coefficient of variation) References

Pasture Merchantable timber 21.3 m3 ha–1 Computed acc to own data, Günter 30% Günter et al. (2008)volume from initial et al. (2008); Leischner (2000)forest clearing

Timber price US$68.5 m–3 Departamento Forestal (2005) 10% Knoke and Wurm (2006)

Establishment US$550 ha–1 Own calculation 10% Own estimation and Pichónof grasses (1996)

Logging costs 2.62 days m–3 x Carpentier et al. (2000) and authors’ 10% Own estimation and PichónUS$10 day–1 own estimation (1996)

Revenues from US$140 ha–1 yr–1 Own data from interviews 36% Probability of abandonmentmilk production estimated sensuCattle opportunity US$92 ha–1 yr–1 Olschewski and Benitez (2005) Abandonment possible from Beck et al. (2008) and obser-costs, maintenance year 10 onwards, linearly vation of Günter (unpub-

increasing probability lished), milk productivity andof being abandoned price uncertainty sensu(Pa) from year 10 (Pa=0) Benitez et al. (2006)to year 40 (Pa=0.75)

Reforestation Establishment costs US$1000 ha–1 Own calculation and data from Sapling survival Weber et al. (2008)interviews probability 0.8

Stand growth 10.5 m3 ha–1 yr–1 Fehse et al. (2002); Weber et al. (2008); 10% Fehse et al. (2002) and Roman-(up till year 20) stand volume for age 8 adjusted with Survival probability 0.81 Cuesta, unpublished data on

own measurements up till age 20 fire probability (deduced from MODIS-Terra 2000–2006 hotspots)

Stumpage price 7 (year 10) to Olschewski and Benitez (2005) and 10% Knoke and Wurm (2006)US$25 m–3 own estimation(year 20)

Selective Forest structure, 1.55% basal area Günter, unpublished data; Oesker 26% Oesker et al. (2008)logging growth growth, 2.6 mm et al. (2008); Günter et al. (2008)

yr–1 diameter increment

Sustainable, 0.75 m3 ha–1 yr–1 Computation based on unpublished 30% Günter et al. (2008)merchantable harvest data on forest growth and mortality

provided by Günter (unpublished)

Timber price US$68.5 m–3 Departamento Forestal (2005) 10% Knoke and Wurm (2006)

Logging costs 2.62 days m–3 x Carpentier et al. (2000) 20% Own estimationUS$10 day–1

All land uses Discount rate 5% Pearce et al. (2003)

Coefficient of kPasture, forestry=0.0, Lönnstedt and Svenssoncorrelation for net (2000) and own estimationrevenues kReforestation, selective logging=0.8

T Knoke et al. Supplemental information


WebTable 1. Continued

� ReferencesBeck E, Makeschin F, Haubrich F, et al. 2008. The ecosystem

(Reserva Biologica San Francisco). In: Beck E, Bendix J,Kottke I, et al. (Eds). Gradients in a tropical mountainecosystem of Ecuador. Ecol Stud 119988: 1–13.

Benitez PC, Kuosmanen T, Olschewski R, et al. 2006.Conservation payments under risk: a stochastic dominanceapproach. Am J Agr Econ 8888: 1–15.

Carpentier CL, Vosti SA, and Witcover J. 2000. Intensified pro-duction systems on western Brazilian Amazon settlementfarms: could they save the forest? Agr Ecosyst Environ 8822:73–88.

Departamento Forestal. 2005. Evaluacion de los RecursosForestales Mundiales 2005: Ecuador. Rome, Italy: FAO.

Fehse J, Hofstede R, Aguirre N, et al. 2002. High altitude tropicalsecondary forests: a competitive carbon sink? Forest EcolManag 116633: 9–25.

Günter S, Cabrera O, Weber M, et al. 2008. Natural forest man-agement in neotropical mountain rain forests – an ecologicalexperiment. In: Beck E, Bendix J, Kottke I, et al. (Eds).Gradients in a tropical mountain ecosystem of Ecuador. EcolStud 119988: 347–59.

Knoke T and Wurm J. 2006. Mixed forests and a flexible harveststrategy: a problem for conventional risk analysis? Eur J ForestRes 112255: 303–15.

Leischner B. 2000. Holznutzung und Holzmarkt in den Provinzen

Loja und Zamora-Chinchipe, Süd-Ecuador. UnpublishedDiploma-thesis. Munich, Germany: Technische UniversitätMünchen.

Lönnstedt L and Svensson J. 2000. Return and risk in timberlandand other investment alternatives for NIPF owners. Scand JForest Res 1155: 661–69.

Oesker M, Dalitz H, Günter S, et al. 2008. Spatial heterogeneitypatterns – a comparison between gorges and ridges in theupper part of an evergreen lower montane forest. In: Beck E,Bendix J, Kottke I, et al. (Eds). Gradients in a tropical moun-tain ecosystem of Ecuador. Ecol Stud 119988: 267–74.

Olschewski R and Benitez PC. 2005. Secondary forests as tempo-rary carbon sinks? The economic impact of accounting meth-ods on reforestation projects in the tropics. Ecol Econ 5555:380–94.

Pearce D, Putz FE, and Vanclay JK. 2003. Sustainable forestry inthe tropics: panacea or folly? Forest Ecol Manag 117722: 229–47.

Pichón FJ. 1996. The forest conversion process: a discussion of thesustainability of predominant land uses associated with fron-tier expansion in the Amazon. Agric Human Values 1133: 32–51.

Weber M, Günter S, Aguirre N, et al. 2008. Reforestation ofabandoned pastures: silvicultural means to accelerate forestrecovery and biodiversity. In: Beck E, Bendix J, Kottke I, et al.(Eds). Gradients in a tropical mountain ecosystem of Ecuador.Ecol Stud 119988: 431–41.

Documents

THE ECOLOGICAL SOCIETY OF AMERICA Frontiers Ecology...Dec 07, 2011 · `THE ECOLOGICAL SOCIETY OF AMERICA Frontiers inEcology and the Environment Can tropical farmers reconcile subsistence