291

Chapter 18

Soil Fauna and Sustainable Land Use inthe Humid TropicsP. Lavelle C Gilot1 C. Fragoso' and B. Pashanasf

'Labomtoite d'Ecologie des Sols Tropicaux, Universite ParisVIIORSTOM, 72 route d'Alllnay, 93143-Bondy Cedex,France: 2lnstituto de Ecologia, Kalapa, Ver., Mexico:3EstaciOll Experimental SanRaman, Yurimaguas, Loreto,Peru

Introduction

It has been long claimed that soil fauna are key actors in sustainable landuse. Long after Aristoteles caIled earthworms the 'intestine of the earth',Darwin (1881) said that'It maybedoubted whether there are manyotheranimalswhich have played so important a part in the history of tile world, as haue tfleselowly organized creatures'. Since that time, considerable efforts have beenmade to substantiate these statements and extend them to the whole soilfauna community. Their importance and role in energy cycling has beenwidely studied during the IBP Programme (Petersen and Luxton, 1982).These studies reinforced the feeling that they had significant impacts onmajor soil processes. During the last decade, considerable efforts have beenmade to describe and quantify these effects and assess the influence of landuse on their communities (see e.g. Veeresh et al., 1991; Andren et al., 1988and syntheses by Lee, 1985; Anderson and Flanagan, 1989; Lavelle et al.,1992b). .

Processes whereby soil fauna may affect the dynamics of soil fertilityhave been described and quantified, mostly at the 'micro' and 'rneso' scalesat which these organisms operate. These studies gave a clear understandingof the processes involved; none the less, the lack of experiments at the scaleof a farmer's plot have not so far aIlowed the evaluation of the exact roleof these processes, and the potential to manipulate them to improve soilfertility.

This chapter gives an overview of the effects of land management prac­tices on soil fauna communities in tropical soils. The effects of soil fauna onprocesses relevant to sustainability are discussed, and finally, recent resultsof manipulative experiments are presented.

© 1994 CAB INTERNATIONAl. SoilResilience and Sustainable Land Use(eds D.J. Greenland and l. Szabolcs)

292

Soil Fauna Communities

P. Lal/elle et al.

Soil fauna comprise a large variety of organisms with contrasted sizes andadaptive strategies. The abundance and composition of their communities,and hence their impact on soil processes, vary greatly depending on vege­tation and land use practices.

Invertebrates in the soil system

Adaptive strategies

Microfauna comprise hydrobiont invertebrates which live in free soil waterand water films that cover soil particles; they are usually either micro­predators of microorganisms and other microinvertebrates, or plant para­sites. Their average size is less than 0.2 mm. Protozoa and nematodes are themain representatives of this group. The spatial range of their activities is ofmicro- to millimetres. They usually fonn food webs of micropredators witha significant impact on nutrient cycling, e.g. in the rhizosphere (e.g,Trofymow and Colernan, 1982; Setala et al., 1991a);

Mesofauna comprise microarthropods and small Oligochaeta Enchy­traeidae, Average length ranges from 0.2 to 2 mm. They are typical inhabi­tants of litter systems where they feed on litter and microorganisms. Nonethe less, they may also colonize the whole profile of soil, although withreduced densities. They may have a significant impact on litter comminutionand dispersal of fungal spores (Persson et al., 1980; Swift and Body, 1985).

Macrofauna operate at much larger scales of time and space. They arelarge-sized invertebrates that may disrupt the soil and modify its structurethrough their movements and feeding behaviour. They are 'ecosystemengineers' (Stork and Eggleton, 1992) that may transport and mix soil andorganic residues in the whole soil profile and create diverse and conspicuousstructures, e.g. mounds, galleries and soil aggregates.

These groups adapt differently to the three main constraints that soilorganisms face, i.e.. feeding on relatively poor quality resources, resistingoccasionally unfavourable microclimatic conditions and moving in thelimited and discontinuous pore space of soil. Small invertebrates have amuch higher resistance to environmental stress than larger ones. They areunable to develop mutualistic associations with microflora to easily digestorganic resources. As a consequence, they rely mainly on predation or plantparasitism. On the other hand, large invertebrates have better abilities todevelop external ('exhabitational' sensu Lewis, 1985) or internal ('inhabit­ational') mutualistic associations with microflora. These associations are effi­cient at using low-quality resources, e.g. woody or humified material,

Soil Fauna and Sustainablr Land U~(' 11/ ltn: I/WllIlJ Tropi«s

Although these invertebrates may escape fmm unfavourable microclimaticconditions by building suitable structures for their shelter and movement.they arc limited 10 non-extreme conditions.

SOI/loodwl'bs

Food webs in the soil arc fundamentally based on relationships of invi-rtebratcs with microorganisms. These relationships art' organized at diff,'rl'ntlevels which depend on the size of the organisms (Fig. 11'.1).

Microorganisms directly exploit resources: they comprise the first level.Microfauna mainly act as predators of mirrotlora: this is the only kind ofinteraction that they can develop since their small size does not allow anysignificant mutualism.

Mesofauna comprise a mixture of microbe and fungal grazers andinvertebrates, which rely on the 'external rumen' feeding strategy. Theycreate suitable conditions for microbial life in their faecal pellets and anintense microbial activity results in the dig"sti"n of part of the undigl'stedorganic matter. By reingesting their faecos at that stage. invertebrates maymake use of assimilable organic matter, and possibly feed, at least partly, onthe microbial biornass,

IncreasedturnoverofSOMand

nutriantsEarthworms

Termites

FIg.18.1. Relationships among soil microtlora and soil fauna acooceptual model. For explanations seetext.

294 1'.Lauel!e et al.

Macrofauna rely on mutualistic interactions with microorganisms toextract assimilable compounds from the decomposing materials. They mayuse the external rumen strategy. Some of them, especially the termites, andearthworms, have developed with microorganisms facultative or obligateinternal mutualist systems of digestion. This is the case with earthwormswhich stimulate the activity of microflora ingested with soil by adding largeamounts of water (up to 100%) and mixing significant amounts of readilyassimilable organic compounds (Barois and Lavelle, 1986). These additionsand the intense mixing in the gut trigger intense microbial activity. In theposterior part of the gut, organic matter has been digested and the wormmay absorb part of the assimilable compounds. Termites have even moresophisticated inhabitational (5<'11511 Lewis, 198:;) mutualist relationships withmicroflora: in the lower termites obligate associations with protozoa havebeen described.

In soils where climatic constraints do not allow large invertebrates tolive, microbial activity is mainly regulated by micro foodwebs (e.g. Sestedtand james, 1987). When these constraints are released, large invertebratesbecome dominant regulators of the microorganisms that they influencedirectly, as explained above, or via predation on micro and mesofauna. Thebehaviour of earthworms feeding partly on protozoa (Piearcc and Phillips,1980; Rouelle et al., 191\5) and nematodes (Dash et al., 1980; Yeatos, 1981) isan example of this type of interaction. In these conditions, micro foodwebsare still an essential component of the soil system, but they tend to restricttheir activities to specific microsites (e.g. root tips in the rhizosphere).Macrofauna operate at much larger scales of time and space, and behave asecosystem engineers (Stork and Eggleton, 1992). When present, effects ofmicroorganisms and smaller invertebrates are largely dependent on theirphysical activities (mixing litter and soil, building structures and galleriesand aggregating the soil) as well as metabolic activities (utilization of theavailable organic resources, development of mutualist or antagonist relation­ships).

IR the humid tropics there are seldom strong climatic or l'daphic limit­ations to colonization of the soil by macroinvcrtcbrates. They becomepredominant regulators of soil biological activities. Voluntary or involuntarymodifications of their communities thus affect other soil organisms withinbiological systems of regulation. These systems link key macroinvertebratcs(e.g. termites or earthworms) to roots, to the whole microflora. and to meso­fauna living in the part of the soil that they affect by their activities. Foursuch systems have been identified, i.e. the rhizosphcre, the litter system, thedrilosphere (earthworms) and the termitosphere (termites).

Soil Fauna a"d Sustainable La"d Useill /lIe H"mid Tropics

Soil macro-invertebrate communities and types of land use

Generaleffects

295

Types of land use deeply affect the composition and abundance of soilmacrofauna communities. Microclimate and food resources together withapplication of pesticides are major factors that affect the diversity andabundance of communities. The initial disturbance linked to the clearance ofthe original ecosystem rapidly eliminates a large number of speciesespecially the ones with narrow niches and slow population turnover rates(Fig 18.2).

Annual cropping is the most detrimental practice as rqprds soil faunacommunities. Within a few weeks of initial cultivation. biomass dramaticallydecreases. Traditional cultivation techniques in general may be more conser­vative and in some areas, local fauna may better resist the disturbance.None the less, the average biomass recorded in such environments (ea. 10 gfresh weight m") is more than two to four times as low as average valuesmeasured in original forest and grassland ecosystems.

Pastures often have a much higher overall biornass of marroinverte­brates than the original ecosystems. This is mainly due to the proliferationof populations of one or two indigenous or exotic earthworm species, and

SAVANNAS

t-

FORESTS

TREE

~38.01 9

ANNUALCROPS

e5.12 9

~32.06 9

~PASTURES

73.20 9

o Earthwormso Termites!ll AntslZl Coleopteraill ArachnidaEl Myriapoda[] Ot hers

Fig. 18.2. Structure ofsoil macrofauna communities inmajor types ofland use inthe humid tropics.

2l)(, P. Lal'clle et al.

the local development of Coleoptcra larval communities, e,g, in CentralAmerican pastures (Villalobos and Lavelle. 1990). In pastures of tropicalAmerica, Australia ,lnll Asia two species of earthworm, ronloscolexrorcthruru« and r"/II/'!'I'l'l'Iillla do"xala, may build a large biornass of 1­.t t ha I fn'sh weight. III rc'gions of natural savannah in Africa, localspecies adapt naturally to pastures and their biomass may be enhanced bygrMing (Kouassi. 1l)~7). However. overgrazing may result in the declineand disappearance of earthworm communities (Castilla and Sanchez,unpublished).

Finally, peronnial crops and agroforestry systems generally have largernacrof.iuna biornasscs which Illay be higher than in the original ecosystem,Tree plantations with legume covers may be suitable environments for bothindigenous forest species and exotic colonizers, as they offer abundant anddiverse resources for dl'composers, As a result, these ecosystems often havediverse and high biorn.i-,-, of macroinvcrtebrates and represent types of landuse which sustain both acceptable levels of biodiversity and overall activity,

rill' cl/erge/lc bottle neck

These systems, however. may not be able to sustain active macro­invertebrate communities for indefinite periods of time, Sampling conducted

ForestEl Liner-feeder earthworms• Endogeic earthworms

• Wood feeder termites• Soil feeder termites• Fungus growing termites

I!il ColeopteraE:a Diplopoda

I::l Others

17,9 9 m-'•Rubber 30 years

Rubber 20 years

Al.,;"Rubber 5 years •

~. )iI "" ".',m'·l~:,·~W

84,1 gm"

li-e~

65,0 gm" 74,lgm-'

Fig, 18,3, Soil macrofauna communities inHevea plantations ofdifferent ages and intheoriginal torest(GIIOI etai, incress).

Soi! Faunaalld Sustaillable Lalld Use ill the Humid Tropics 297

in Hevea plantations of different ages showed that communities changewith time and tend to be severely depleted after 30 years. At Bimbresso(Cote d'Ivoire), soil macrofauna biomass was large in the original forest(74,2 g m- 2) ; earthworms comprised most of this biomass, but termites andmyriapods were significant elements in this community (Fig. 18.3). In the5-year-old plantation, biomass was still high, but composition was different.Termites, especially xylophagous termites, were overdominant. In 10- and20-year-old plantations, biomass of macroinvertebrates was sustained athigh levels (65.0 and 74.1 g m-2 respectively) but xylophagous termites hadnearly disappeared. They had been replaced by large earthworm com­munities, especially of large endogeic species feeding on soil organic matter.After 30 years the community was highly depleted as was the production ofrubber. These results demonstrate the dynamic changes in the communityas the Hevea plantation ages. It is likely that this community had beenmaintained for 20 years by the large energy input from the wood of trunks,branches and large roots in the early years of the plantation. Xylophagoustermites first exploited these resources. Earthworms came later in' thesuccession to feed on organic material transformed by the termites. After 20years, the initial flux of energy faded and organic resources sustained alower biomass. Changes in the herbaceous cover may also explain part ofthe observed variations, During the first few months, a legume coveriPueraria phase%ides) had been maintained. This cover was rapidly replacedby a poor herbaceous stratum which was reduced to a minimum as thecanopy closed.

These results clearly emphasize the need for significant amounts ofenergy sources to maintain active soil macrofauna communities,

Tilecolonization issue

Energetic deficiencies may not be the only process whereby communities aredepicted, In some circumstances species which might adapt to the newlycreated conditions may not be able to colonize because they are not presentand/or do not have the ability to invade the site rapidly. This is especiallytrue for earthworms and termites as not all the existing functional groupsare present in a given biogeographical area. Such a situation gives scope fora wide range of manipulative experiments. In temperate soils, inoculation ofwell-adapted earthworm species has been implemented in a number of situ­ations, e,g. for the improvement of pastures in New Zealand, and Australia(Syers and Springett, 1984), reclamation of degraded sites (Curry and Boyle,1987) or improvement of polder soils (Hoogerkamp et al., 1983). In soils ofthe humid tropics critical targets might be to introduce species withexpected favourable effects on the soil function in low input agriculturalsystems (termites),

~qs I' l.aivtlc pi al.

Manipulative Experiments: Introduction of EndogeicEarthworms in Low Input Agricultural Systems

In Pew, Mexico and C,'!l' uIvoirc. situations have been ide-ntified where theintroduction uf earthworms was feasible. They comprise traditional systemsbased on slash-and-burn agriculture. Research has been conducted whichdemonstrates that carctullv chosen adapted species may build up sizeable1"'l'ulations These popul.uions affvct plant production and soil fertility asas't'ssed in terms of soil l>rf;anic matter (SOM) and nutrient conservation..ind ",'nsl'n'ati"" of soil I'hv,il'al properties. Most of the results presentedhen' have been obtained ,11 Yurimuguas (Peru) in acid Ultisols under a tradi­t ional rot.uion with no it'rtJii'ers. or continuous maize cropping fertilized,11tl'r the third en)ppin~ n', le

Evperimental design and ~<'neralmethodology

Soil monoliths nO ern in di.irnetcr and 50 cm dvcp were isolated from thes"i! of a 20-year-old sc'cl,,,d.uv forest and a nylon mesh was used to preventany movement of earthworms After clearing and burning the forest, theseunits were cleaved of native earthworms by an application of carbofuradanto lTl'ate 'no earthworm situations. After six weeks, earthworm populationswe-n- introduced. The sp,'ei,'s chosen, Ponloscoll'X corethrurus is very wide­'pr""d in all di-turbod ,oil, of thl' humid tropics; it has a large tolerance fora wide range of cdaphic f"elms, except for drought. It has also the ability tobuild numerous populations rapidly when placed in favourable conditions(I avclle cl ul., ]9/\7). At Yurimaguas. ]20 such experimental units werein sta llcd. 10/\ of them Wl'rl' cropped tu a traditional rotation system with nofl'rlilizers; 12 wvre croppvd to maize with fertilizers from the fourth,.,.oppin~ cycle onward-. Six diffcn-nt treatments were applied, i.e. thre-eIl'vels of ,,,.~anic trc',ltnH'nts (no llr~anic residucs) (C); organic residuesproduce-d on thv cxpcrinu-nt.i! unit (CR), and CRV i.c. CR + k~ume gre-enmanure of C'/llre"c",a 11/111,.omr1'WI/ (2.~ t ha J dry weight), with or withoutearthworms.

In 'with earthworm' treatments. a rclativclv low initial (36 gm-I)biomas« \\',b introduced. II was believed that biornass would rapidly reach,111 l'quilibrium rl'J'H',,'nting thv actual carrying capacity of the system fortill' sp",il's. TIH' expcr inn-nt was dl'si~nl'd to have six successive croppingcycles. At each harvest. ,'bove~round production was measured in all units.'Internal' parameters. i.e characterization of SOM, earthworm and rootbiomass. bulk density, infiltration and aggregation were measured on threecvpcrirncntal units whu h were destroyed for sampling. During the growthperiod. soil water content and the decomposition rates of organic residueswere monitored. Method, arc basically the ones proposed in the Tropical

.'J

i"'"....f. . . ~""~;,C\~'J

;, ~

'1.~

""1.:"

",~:\~

" ...-"@.,

~~.)

.~.~

Soil FalOJllllnd SIl~/llillab1t' La"d l.IS(' in the Humid Tropics 2YY

Soil Biology and Fertility programme (TSBF).In a nearby plot. 20 units were allocated to continuous maize cropping to

monitor dynamics of SOM using natural 13C labelling, by shifting from apure C3 vegetation (forest) to a pure C4 maize culture to sustain significantproduction in the system. Fertilizers (lOO N, 20 P, 20 K) were applied fromthe fourth cropping cycle onwards to sustain production.

At I.amto, similar experiments were installed with a native species(Mills<lIlia a/loma/a) under continuous maize cropping in a forest soil, andcontinuous yam cultivation in a savannah soil.

&t.tblishment of esrthworm populations

The establishment of introduced populations was not equally successful atall sites. At Larnto (Cote d'Ivoire), populations were sustained at very lowbiomass (3-4 g m"? after four crops). There was clear indication from aspecific experiment that the limitation to earthworm settlement was the lackof food (Gilot et al., 1992).

At Yurimaguas, populations actively reproduced in all treatments. Theyresponded significantly to organic inputs, with the lowest biomass in treat­ments without residues and highest biornass in treatments receiving eithercrop residues alone or crop residues + legume green manure (Fig. 18.4).

100

f ~j. .

I \--0- C

---.-- a:l

E --------CRV

Cl 60

IIIIII

E40

0iD

20

00 , 2 3 4 5

D 89 A 90 At90 F 91 At91 J 92

Cropping periods

~1" Changes inbiomass ofPontoscolex corethrurus inatrad~ionaJ cropping rotation submitted to_dII"ent organic treatments atYurimaguas (' indicate significant differences with other treatments atIt...dale). rC: no organic inpuls; CR :application ofcrop residues produced atthe ~e ;CRY :croplIiduII+2.5 t ha-I legume green manure).

Effect of earthwormsx Time

P. lm'<'il,'l'l al.

o I I I I I I -., I I I I

5

Worm-

43

Crop n°2

\Ill )

3

I+41 %

bI r-""'"

2'I

coJ::. I a.::::."0ID

>=

Fig. 18,5, Effect ofeanhworm aciv.nes on grain production atfive successive harvests irrespective oforganic treatments.

l'nlp prlldudillll w.r-, ""I,linl'd at acceptable levels according to loccllst.mdards with grain production in the range 0.1'-2.4 t ha' I. There were.... ignifil".1nl cffl'l.."ls uf l"Hlh\\~H"Il\S UIl crllp produ .....t ion.

Crain produrtinn ",IS inrTl'as('d on th.: aVl'rcl~l' hy 27% during the first

I'Vl' croppiru; cyell'S (':'l~ I~.~). A m.ixirnurn increase of 71'% was obtained atth« Sl'rllnJ harve-st. \\ Ill'l"l'.1.....1 siy.llifil·.lnt dl'cn'a~l' ut --1-.1% \t\'(1S observed at

Ihl' fifth harve-st. TIlt' ru-xl lJ'llP will show whether this dl'rTl',l.Sl' indicated asignificant inve-rsion llt 111l' trend, l1I' simply was the consequence of accidentall'Ifl'l'h PI' l'arthwllfll1s, l'.g. on water availability at cl critic-al slag<' uf plantgrO\dh '1 Ill' I'l'Spl1nSl' W,lS higl",r in Irl'atnll'nts without fl'sidul'S ('f' ](,'Y..)

nr n'n,jving crop rl':-'Idlll'''' + Il'gun1l\ gn'l'n manure (+ Jh'%J) than in tn-.it­nu-nt-, with nllp fl'sidlll" llnly (+ 1'%).

In the plot with nllltinllllUS maize, the tlvt..'rLlMl' increase l)f productiondue to till' introduction oi e-arthworms amounted to ]JO%. This effect wasl'specially spcrtacular ell the second harvest with production of 0,1' t andJ.2 t ha I in trcatrncnt s 'without' or 'with' earthworms. After threesue<'l',SSiVl' rropping l'yd,'s, production had dramatically decreased in bothsl'sll'lllS, Thcn-t ore. I" maintain the nop, fl'rlilizers were applied, At thetiith cmp, production was twice as high in the earthworm treatments as int hv no-c-arthwurm trcatrncnt s (Fi~. 1H.h).

~I,.. 50c:CO

C. 40~c:o 30°fl

::::l'020e0-c: 10

'co...~ 0

Soil Faunaand SustainableLandUse i,. the Humid Tropics

Application 01fertilizers

301

o Worm­ID Worm+

cs cs CS CSTreatments

CS

fig. 18.6. Effect ofearthworm inoculation on grain production inacontinuous maize crop atYurimaguas.CS, surface application ofstubble.

These results indicate that: (i) earthworms can prolong and increaseproduction; and (ii) their positive effects may not be explained simply interms of the improvement of the nutrient supply to plants. since their effectremains significant when fertilizers are used to correct nutrient deficiencies.

ERects of eerthworms on SOM and nutrient dynamics

After five cropping cycles, earthworm activities had not prevented losses ofsoil organic matter (Fig. 11'.7). Nevertheless, the quality of organic matter asassessed by particle Sill' fractionation had significantly changed (Feller,1979). At Lamto. the proportion of SOM in the coarse fraction (> 50 IJ.m)decreased less in the earthworm than in the no-earthworm treatment. Onthe other hand, more organic matter from the fine fractions seemed to havebeen mineralized (Fig. 11'.8). Earthworm activities did not help maintainmicrobial biomass as this was significantly lower (by 23%) in treatmentswith earthworms.

Neither did earthworm activities impede the depletion of nutrients. Aprogressive return towards acid pH and high levels of Al saturationoccurred in all treatments. None the less, "N from labelled legume greenmanure applied at the soil surface was better recovered by plants in thepresence than in the absence of earthworms. This demonstrated significantqualitative differences in nutrient cycling, oriented towards a more efficientuse of available nutrients.

2 3 4 5

Crop n°

Worm +Worm

1,9

,J A 0,14

1,7

r-, X'\ 0,13

\,6

~ "~ \1\" ~."a a1,4

0,11a

1.3

1.20 1 2 3 4 5

Crop n"

Fig. 18.7. Variation ofCand Ncontents ofthe (}-10cmstrata with time and treatments.

[] 0-2Jlm

o 2-50Jlm

IJ 50-250Jlm

11 250-2000~

12

10

-Iec 8~

C

C 6

'"C0 4""U

2

0Initial - worm + worm

20 months cultivation

Fig. 18.8. Parution ofthe soil organic maner between granulometric fractions Inthe maize plot allheimtJaltime and after the founh crop.

303SoilFauna and Sustainable LandUse in ti,e Humid Tropics

Cluutges in /fOilphysical propertie«

t,'J•~

r'~

All soil physical properties measured were significantly affected by earth­worm activities. Rough calculations based on assessments of earthwormbiomass and on the results of past laboratory experiments gave an estimateof ISO t ha-I year-' dry soil ingested, equivalent to 10-15% of theupper 10 cm of soil. P. coreihrurus ingests small aggregates and organicdebris of a size smaller than the size of the mouth (i.e. up to 2 mm). Theeffects on soil aggregation were highly significant after five crops at Yuri­maguas. In 'with earthworm' treatments, the relative proportion of large(> 10 mm) and intermediate (2-5 mm) aggregates was significantlyincreased, whereas the reverse effect was observed in the absence of earth­worms. Earthworms clearly increased soil macroaggregation and they

~ counteracted the trend for soil disaggregation observed in the conventional>.~.. system.~.•: These changes of macroaggregation had effects on other global physicalF' parameters. In the presence of earthworms, bulk density was significantly. increased; infiltration rate was first decreased, and then returned to values

o I J J I , I

o 1 2 3 4 5Crop n°

fig. 18.9. Changes ofinfiltration with time. Data with different leners indicate signfficant differences a\agIVen lime. Bulk density was significantly higher than the initial value attime 3inthe earthworm treatment.OrganiC treatments had no effect.

3114 P. LIlI'e/leet a!.

similar to those of control treatments, after the earthworms had increasedtheir production of surface casts and created a macropore system at the soilsurface (Fi~. It'IJ).

Discussion

Invertebrates arc major determinants of soil processes in tropical eco­sysn-rus. Whereas inSl'd pests have been actively fought, the potential forbeneficial use of insect activities has not been taken fully into considerationin the design of rnanogenu-ut practices. Current research demonstrates thatpractice'S which anrnhilato the activities of soil fauna are unlikely to besustainable in the long term.

Soil fauna communities have contrasting reactions to changes inducedbv human land manage-ment. Their abundance and diversity arc indicatorsof till' quality of soils expressed in terms of soil organic matter, nutrientconte-nt, and physical properties such as bulk density, porosity and waterre~ime. Annual crops generally support depleted communities, withespecially low earthworm populations, whereas termites arc usually less.1lieded. Pcrvnnial cultures kg. sugar cane). pastures. and plantations (e.g.oil palm with a le~ull1" cover) ~l'nprally have less diverse communities thanthe ori~inal ecosyste-m. but biomass is often higher due to colonization byperegrine earthworm sppcips, and persistence of key native species.However, as the system degrades a deplcrion of the soil macrofauna occurs.

The abundance of soil fauna communities is significantly affected by theavailability of suitabh- food resources and by their diversity. This is thereason why communities MP much depleted in annual crops which havenon-permanent and little-developed root systems. and limited inputs ofur~anic residucs. In some casps, the activity of soil fauna may be sustainedfor some time at Ihe e'pl'n,,' of l'ner~etic sources which had been accumu­lated in the original ecosystem. Examples of such sources arc the deadwoody rn.itcrial in n-ccntlv deforested areas or coarse particle siz« fractionsin SOM rvscrvcs. Our obse-rvations suggest that some termites or earth­worms mav live on resources with high lignin contents and/or highC: nutru-nt ratio. Thvn-Ioro. there is scope for the use in carefully designedconditions of wastes like sawdust or coil' (c.g. Pashanasi cl al., 1442). Beforeusing such practice'S, till' timing and placement of these residues will need tobc' addfl'sSl'd to prove-ut axvnchrony between nutrient availability and plantre-qui rcnu-nt S.

I'rvhuunarv l'\lh'l imcnt-, conducted [or two Yl'lHS on f ivi- ~lICCeSSiVl'

{"I"tlpping l'yell'S h.iv« ~;i\'l,1t sorm- insight on the l,lh·(t of IlHHlospecitic

f'opul'ltiol1s of endog,,;c CMthwurm rurnrnunities on traditional agricultural-vstcrn«. Ouring th.it f'l'ri'ld there was no evidence that the depletion ofSl )1\1 and the nutrient reserves of the soi] were decreased by introduced

Soil Faunaand SustainableLand Use ill tileHumid Tropics 305

earthworms. Nevertheless, qualitative differences in SOM, as assessed interms of particle size fractions, were different and indications of better useof nutrients by plants were obtained. Soil fauna activities in general mightresult in the long term in the appearance of a new state of equilibrium whichmight be characterized by lower amounts of SOM, but with a much fasterturnover rate, as a result of both increased mineralization of the organicmatter in the supposedly more recalcitrant fine soil fraction, and the relativeprotection of coarse fractions. The duration of our experiments was tooshort to reach a new state of equilibrium of the system.

Earthworms had major effects on soil physical properties. The specieschosen significantly increased macroaggregation in the upper 10 cm. Bulkdensity was significantly higher in the presence of earthworms (+ 15%).Infiltration was first decreased up to the stage when earthworms started todeposit surface casts and open macropores at the soil surface. Significantchanges were observed in water regime, with increased flooding or droughtat very moist or dry periods. Not all earthworms will have such effects(Case nave and Valentin. 1988). In wet savannahs of Cote d'Ivoire, Blanchart(1990) has described complementary effects of large earthworm species,which can transform a significant proportion of the soil microaggregatesinto larger macroaggregates, and smaller-size species splitting up these largeaggregates into smaller ones which are mixed with root litter. Earthwormcasts and other faunal structures often have high structural stabilities andthe overaccumulation of these structures may alter physical properties in thesoil. Complementary biotic or abiotic processes which regulate the dynamicsof aggregation are thus necessary. These may include assemblages ofspecies from different zoological groups acting simultaneously or intemporal successions (David, 1988; Blanchart, 1990).

References

Anderson, J.M. and Flanagan, P. (19H9) Biological processes regulating organicmatter dynamics in tropical soils. In Colernan, D.C.. Oades, T. and Uehara, G.(eds.) Dyllamics of Soil Organic Matler ill Tropical Ecosystems, NifTAL project,Universityof Hawaii, Honolulu, pp. 97-125.

Andren, 0., I'austian, K. and Rosswall, T. (19H8) Soil biotic interactions in the func­tioning of agroecosystems. Agriculture, Ecosystems and Environment 24, 57-65.

Barois, J. and Lavelle, 1'. (1986) Changes in respiration rate and some physioco­chemical properties of a tropical soil during transit through Pontoscolex core­thrurus (Glossoscok-cida, Ohgocha-ta). S,,1i Biology alld Biocl,emistry 18(5), 539­541.

Blanchart, E. (1990) Ri,le des vers de terre dans la formation et la conservationde lastructure des sols de la savane de Lamto (Cote d'Ivoire). These UniversiteRcnnes I., 27H pp.

306 P. Lavclleet al.

Blanchart, E. and Spain, A. (1989) R61e des vers de terre dans I'elaboration et laconservation de la structure des 5015 de savane. In: Lavelle, P. (00.) ProcessuslJiolosique' et Ferlilil': tIu Sol tIalls Ics Savanes HlIIllides de COIe d'lvoire. RecherchesFondall1L'l'lnJf$ ,'I J\/'I,liqlll'I''' sur 11' RMedes Vcr" de Terre, Ministere de l'Environne­mont (SRETIE), Paris, pp. 26-43.

Casenave, A. and Valentin, C. (1988) Les Elal" tic Surface de la ZOIlI' Sllhtlie/ll'I'. Ll'/lr11I1)u"IIce Sill' f'llI/iltrati",1. Rapport CEE·ORSTOM, 202 pp. ORSTOM, Bendy,France.

Curry, 1.1'.and Boy!'" K.E. (1987) Growth rates, establishment and effect on herbageyield of introduced earthworms in grassland on reclaimed cutover peal. BioloSYalld Fertility01 Soil, 3, 95-98.

Darwin, C (1881) TIlt' Fornuuion of Vegela/lle Mould Tllroug" nie J\cti,'" of Wor",,, "'illlO/l"crlllltioll" Oil Ilrfir Ilal'its. Murray, London.

Dash, M.C., Senapati. B.K. and Mishra, Cc. (1980) Nematode feeding by tropicaleurthworms. Oikos 34, 322-328.

David, ).1'. (1988) 1."5 peuplements de Diplopodes d'un massif Ioresticr temp,'rp sursols acides. These Universite Paris VI. 225 pp.

Feller, C. (1979) Une methode de fractionnement granulometrique de la rnatiercorganique des sols: application aux sols tropicaux a texture grosstere, trespauvres en humus. Cahiers de I'ORSTOM. Serie Pedologie 17(4), 339-346.

Gilot, C, Lavelle. 1'., Kouassi, ['h. and Cuillaume, G. (1992) Biological activity ofsoils in Hevea sands of different ages in Cote d'lvoire. Acta Zoologica Fellllica(inpress). .

Hoogcrkamp, M.. Ro~a~r, H. and Eijsackers, H.).1'. (1983) Effect of earthworms onArassland Oil 1'l'C('"lly reclaimed polder soils in the Netherlands. In: Satchell. j.E.(ed.) Earlln"onll Ecology: [tom Darwill '" Ver",iCIIllllre. Cnapman and 1-1 all,London, New York, pp. 85-105.

Hunta. V.. Setalii, H. and Haimi, I. (1988) Leaching of Nand C from birch leaf litterand raw humus with special emphasis on the influence of soil fauna. Soil Biologymill Biochcmistn; 20, 875-878.

Kouassi. 1'.K. (191'\7) Etude comparative de la macrofaune cndogee d'ecosysterncsguinccns naturcls et transforrnes de Cote d'Ivoire, Doctoral de 3cllle cycle,Unlverslte dAbidjan, 115 pp.

Lavelle. 1'. (1988) Earthworm activities and the soil system. Bio/"S.v /lIId Fertility"f S"i/,;6, 237-251.

Lavclle, 1'., Barois, I., Cruz, C, Hcrnandez, A., Pinoda, A. and Range). 1'. (1987)Adaptative strategies of I'olll"scolex corellrruI'lIs (Clossoscolecidar, Ollgochaita), aperegrine geophagous earthworm of the humid tropics. Biol"SY alld Fertility "fSoils5,188-194.

Lavelle. P., Alcgre. ).. Barois. I., Fragoso, C, Cilot. C, Conzalez. C, Kanyonyo kaKajondo. Martin, A., Melendez, G., Moreno, A and Pashal1asi, B. (1992a) COII­serration oi Soil Fertility ill Low IIIIJIII Agricultural Systems t>f till' HUll/id TropicsIlyMalli,",latiIlS [arlllllw,,, Com""l11ities. CCE-STD2 programme. Final reportproject n" TS2 • 0292-F (EDB).

Lavello. 1'., Spain. A.V.. IllandlMI, E.. Marlin, A. and Martin, S. (1992b) The impactof soil fauna on the properties of soils in the humid tropics. In: Myths and Sd,-"ce"t' Soil"r 1111' Tropic». Soil Science Society of America, special publication, 2'J, 157­185.

Soil faulla and Suslaillable Lalld Use ill 11,,' Humid Tropics 307

i~~-

LavelIe, P., Dangerfield, M., Fragoso, C.; Eschenbrcnner. V., Lopez-Hernandez, D.,Pashanasi, 8. and Brussaard, I.. (1993) The relationship between soil macrofaunaand lropical soil fertility. In Swift, M.J. and Woo mer, 1'. (eds) Tropical Soil BioloXYand F",li/ily, John Wiley, New York (in press).

Lee, KE. (1985) Earlllll'Orms: Thl'" [c%gy a",1 Rc/atioll''';ps ",iI" Soils and Lalld Use.Academic Press, New York.

Lewis, D.H. (19H5) Symbiosis and mutualism: crisp concepts and soggy semantics,In: Boucher. D.II. (ed.) The BIOlogy of Mutl/alism, Croorn Helm, Beckenharn,p. 29.

Pashanasi, B., Melcndez. C .. Szott, I.. and Lavelle, P. (1993) Effect of inoculationwith the endogcic earthworm POlllosco!rx corell"u",s (Clossoscolccidac) on Navailability, soil microbial biomass, and the growth of three tropical fruit treeseedlings in a pOI experiment. Soil Biologyand Biochemislry 24, 1655-1660.

Persson, T., Baath, E., Clarholm, M., Lundkvist, H., Soderstrorn. B.E. and Sohlcnius,B, (19HO) Trophic structure, biornass dynamics and carbon metabolism of soilorganisms in a srots pine forest. [cologica/ Bulletins 32,419-459.

Petersen. H. and Luxton, M. (1982) A comparative analysis of soil fauna populationsand their role in decomposition processes. Oikos 39, 287-388.

Piearce, T.C. and Phillips, M.J. (1980) The (ate of ciliates in the earthworm gut: an invitro study. Microbia/ Ecology5,313-319.

Rouelle, ). (1983) Introduction of amoeba> and Rhizobium japl'"i,u", into the gut ofEise"ia [octida (Sav.) and LumbriCUS terrcstris I.. In: Satchcll. J.E. (ed.) Earll",'orlllEc%iN Chapman and Hall, London, pp, 375-381.

Rouelle, J.. l'ussard, M., Randriamamonjizaka, J.L., Loquet, M. and Vinceslas. M.(1985) lnteractions microbiennes (Bacteries, Protozoaircs). alimcntation d,'s versde terre "I mineralisation de la rnatiere organique. Bulletill of Ecoh>SY 16, 83-8!l.

Seastedt, T.R.T.T.e. and jarnes. S.W. (1987) Experimental manipulations ofarthropod, nematode and earthworm communities in a north American tangrassprairie. Pedobi%sia 30, 9-17.

Setala. H. and Huhta. V. (1991) Soil fauna increase Betula pcndula growlh: laboratoryexperiments with coniferous forest floor. [c%XY 72, 665-671.

Sdala, H., Tynisrnaa. M., Martikaincn, E. and Huhta, V. (1991) Mineralization of C.Nand P in relation to decornpo ....e-r community structure in coniferous forest soil.Pedobiologia 35, 285-2%.

Stork, N.r. and Egg!<'lnn, 1'. (1992) Invertebrates as determinants and indicators ofsoil quality. Ag:';nll'ure, Ecosys/£'l/J> alld EllvirOlIll/"1I1 (in press).

Swift, M.J. (1486) Tropical soil biology and fertility (TSBF): inter-regional researchplanning workshop. Report of the third Workshop of the decade of the Tropics/TSBF program. fjlt'ltlgy IlIlerllatlll/lal, Special Issue 13, 68 pp.

Swift, M.). and Boddy, I.. (1985) Animal-microbial interactions in wood decompo­sition. In: Andcrson, J,M. Ravner. A.D.M. and Walton, D.W.H. (eds) Allimal­Microblal/llleradi,'"s. Cambridge University Press, Cambridge, pp. H9-131.

Syers. J.K. and Springett, J.A. (1984) Earthworms and soil fertility. Plant alld Soil 76,93-104.

Trofymow, J.A. and Colvman, D.e. (!lJH2) The role of bactcrivorous and fungivnrousnematodes in cellulose and chitin decomposition in the context of a root (rhizo­sphere) snil ""neLTtual model. In Frerkman, D.W. (ed.) Nematode> ill Soil Eco­'ysll'",'. Universrtv of Texas Prl'ss, Austin, Texas, pp. 117-137.

P. La,'cl!c et a1

Veeresh, G.K., Rajagopal. D. and Viraktamath, CA. (l'JY1) Adl'allccs ill Mallagcmenlalld COlIscr1'alioll of Soil Faulla. Oxford & IBH Publishing, New Dehli, Bombay,Calcutta.

Villalobos, F.J. and Lavelle, P. (1Y90) The soil coleoptera community of a tropicalgrassland from Laguna Verde. Veracruz (Mexico). Rt'l'ul' d'Ecologie cl de Biologicdu 501 27, 73-93.

Yeates, G.W. (lY81) Soil nematode populations depressed in the presence of earth­worms. Pedobiologia 22, 191-196.

Lavelle P., Gilot Cécile, Fragoso C., Pashanasi B.

Soil fauna and sustainable land use in the humid tropics.

In : Greenland D.J. (ed.), Szabolcs I. (ed.) Soil resilience and sustainable land

use. Wallingford : CAB, 1994, p. 291-308.