Article Robert A. Rice and Russell Greenberg Cacao Cultivation Biological Diversity and the Cacao (Theobromacacao) is a crop of the humidlowland tropics produced largely by small-scale producers and often on farms with a canopy of shade trees. Where a diverse shaded canopy is used, cacao farms support higherlevels of biologicaldiversitythan most othertropical crops. A host of viral and fungal diseases, loss of soil fertility,and numerous socioeconomic problems facing producers,often makes cacao productionlocallyunsustainable. Continuedclearingof new lands threatens biodiversity.Moreover,new frontiersforcacao expansionare rapidly disappearing. Such problems can be addressed by increasingthe long-termproductivityof existingcacao farms and restoringabandoned lands. Improvedshade management offers guidance along this path. Institutionsinvolved with cocoa should establish collaborations with groups concerned with development, environmentalprotection, and most importantlyproducers themselves to pursue a program of research, extension and policy initiatives focused on the ecologically and economicallysustainable cacao productionon farms witha diverse shade canopy. Conservation of est species-to commercialshade where other tree crops are interspersed among planted shade trees and the cacao, to monocultural,specialized shade, where the shade is dominated by one or a few tree species or genus (genera).Some indigenous shade systems are truly diverse agroforests.However, in most plantedsystems where a multitudeof shade species is found (up to 30-40 in some planted systems), one or a few species generally comprise the "backbone"shade component in which other fruitingand timberspecies are inserted.Such backbonespecies, usuallyfast-growing,nitrogen-fixinglegumes, includeErythrina spp., Gliricidiasepium, Cassia, and Inga spp. In some areas,cacao is grownunder,or intercroppedprimarilywith, fruitandfast- Figure1. Cacao shade managementsystems, showing shade gradientand canopy heights. Rustic cacao planted beneath thinned primary or old secondary forest 1 {; a INTRODUCTION As the world faces alarmingrates of tropical forest loss, some agricultural systems offer a glimmer of hope. In particular, agroforestry,wherecrops are cultivatedin associationwith trees, provides some of the ecological benefits of naturalforest while allowing farmersto make a living off theirland (1-3). However, only in agroforestrysystems that involve the productionof globally tradedand economically importantcommodities, such as cocoa (Theobroma cacao), can we realistically expect that enough land and people will be affected to make a difference. The environmentalbenefits thatcacao farmsmay confer depend upon the managementsystem. On one hand, we find a low-input and generally low-yield system that requires few, if any, agrochemicals. These farms often incorporateremnant forest shadeinto the farmmanagementsystem. In contrast,are the systems characterizedby high-input,high-yield farms, often with little or no shade. These are extremes of a continuumand between them we find managedagroforestrysystems with planted shade trees and moderatelevels of chemical inputs. We will develop the argument that systems using diverse shade offer the greatest potential for long-term production, biodiversityconservation,and environmentalprotection.We begin our discussion of how different shade cacao systems influence tropicalbiological diversity, then provide a brief overview of patternsof production,and finally explore ways in which the patternof productionmight be changed to optimize biological diversity, cacao production,and the livelihood of the millions of small farmerswho grow the crop. 1 3 . .. i ....... . ........... ~~~~~~.. w 3 frarn 2 t : I-.:.- I .. . .. -.-.-2_ ... .2.- \ , , 11 ....... , ...... .r........ f ~~~~~~~. 8 meters 16 Shade ~~~~~~Planted ranges from traditional polyculturethrough commercialto a specialized, single-species shade 8 0 meters SHADESYSTEMS Shade managementsystems in cacao form a gradient(Fig. 1), analogousto thatseen in coffee (3, 4). Rusticcacao management, widespreadin the humid portion of West Africa and local in Latin America, is characterizedby the planting of cacao under thinnedprimaryor older secondaryforest (5). Plantedshade systems vary widely, from traditionalpolyculturalsystems-multiple species of plantedshadetrees with occasional remnantforAmbio Vol. 29 No. 3, May 2000 8 Technified Cacao no shade component o meters C Royal Swedish Academy of Sciences 2000 http://www.ambio.kva.se 167 growing timbertrees. Zero-shadecacao (6) cultivation,without shade, is common in Malaysia and becoming more widespread in parts of Colombia and Peru. These three basic systems contributevery differentlyto the conservationof biological conservation. DIVERSITYIN CACAOFARMS BIOLOGICAL Three fundamental questions concerning the relationship between the cultivationof shade-growncacao and biodiversityare: i) How much of the original tropical forest diversity is maintained on cacao farms? ii) How does this compare with other agriculturalsystems?iii) Whatis the role of differentcacao managementsystems in the maintenanceof tropicaldiversity? Rustic Cacao A limitedbody of work upholdsthe notionthatcacao farmswith a diverse shade canopy supportgreaterbiological diversity,particularlyof forest-dependentorganisms,thanothercash crop systems in the lowland tropics (7-15). However, even the most diverse rustic shade farmsare highly modified comparedto naturalforest-combining forest generalist and early successional species (7, 16, 17); tropicalforest specialists are often missing. For example, one study found only two out of seven forestAnolis lizard species in rusticplantations in Costa Rica (18). In Brazil's cabruca system, large-bodiedterrestrial mammals and larger primates were among the groups underrepresentedin the cacao. Rustic cacao bird assemblages lackedmany of the specializedunderstoryspecies, and supported low abundancesof foliage-gleaning insectivorousbirds, understory frugivores, and relatively high abundances of canopy frugivoresand omnivores (19). Rustic cacao farms form highly modified tropical forest systems. The change to the understory is obvious, and the canopy is also dramaticallyaltered.For example,in the cabrucamanagementsystem of Bahia,Brazil,trees and shrubsare thinned to 10% of their original abundanceand most lianas removed (19), which substantiallyreduces plant diversity. More importantly,canopy regenerationis eliminatedor the ecological conditionsfor regenerationare not met, and eventually the naturallyoccurringtrees are replacedby plantedtrees often nitrogen-fixing legumes, trees that provide optimal shade,or plantsthatproduceuseful fruits,wood, and otherproducts. Management that removes or modifies specific microhabitats(mistletoe, epiphytes, etc.) may cause critical changes in overall diversity.The importanceof specific microhabitats to ants and otherinvertebrateshas been underscoredin the re- search of Room (8, 20) and Young, (21, and ref. therein). At the landscape scale, cacao farms suffer from fragmentationand isolation, representingone land-use type within a mixed mosaic of uses. This alone leads to the loss of the original forest species and the increase in habitatgeneralists-organisms usually associatedwith disturbedor recoveringhabitats. Zero and Planted Shade Systems Althoughlittle researchhas examinedbiodiversityin zero-shade cacao systems, ongoing studies of bird communities in the PeruvianAndes have shown that zero-shadecacao supportsa few early successional species (Greenberg,unpubl.data);further,it is reasonable to assume from work in coffee that cacao farms using few or no shade tree species supportlower levels of biological diversity and relatively few forest-dependentorganisms. Some work suggests that planted shade systems are less diverse than rustic systems (2). Room (17) speculated that the abovegroundant diversityof Ghanaiancacao farmswas substantially greaterthanthatof New Guineaat least partlybecause the former are small rustic plantationsand the latter largerplanted shadefarms.From studies of the ant-mosaics,thereis some suggestion thatin plantationsof structurallysimple shade,fewer ants are able to dominatethe habitatmore thoroughly(22). Greenberget al. (23) found that the bird diversity of cacao plantationsin the Gulf Lowlandsof Tabasco,Mexico, supported only a modest abundanceand diversity of birds, even compared with other agriculturaland disturbedhabitats.These farms supported almost no resident forest birds, although migratoryforest species were quite common. They speculated that the lack of bird diversity may, in part, be a result of the lack of small bird-dispersedfruitin the cacao understoryand the shadecanopy. However, the Tabasco cacao region has little forest and the lack of nearby naturalforest may have played an importantrole as well. It is likely that a numberof forest organismscan occur in cacao, but requireforest for partof theirlife cycle. A small amount of data (9-13, 19) sustain the concept that rustic cacao systems close to naturalforest-particularly large tractsof forest-support substantiallygreaterdiversity of forest birds and mammals than those systems isolated from naturalhabitats. Young (21) has made a similar argumentfor invertebrates,including some of the pollinatorsfor cacao. In fact, some noteworthyendemic vertebratesfound in rusticcacao farms,i.e., the recentlydiscovered Pink-legged Graveteiro,Acrobatornisfonsecai and the endangeredGolden-headedLion Tamarin,Leontopithecuschrysomelis, seem also to require forest patches and use cacao as a secondary habitat(19, 24). The implicationsof this proximity effect are unclear.Cacao farms may be populationsinks, unableto supportpopulationsof many forest organismsin the long run,but benefiting from periodic emigration from source forest habitats.On the positive side, mobile forest organisms may requireresourcesnot present in a protectedforest patch on a seasonalbasis. This is particularly true for altitudinally migrating birds and insects where forest remains only at higher elevations. Moreover, cacao farms may provide a suboptimal buffer habitat thatwould promotepopulationstability in the optimal forest habitat. Finally, cacao farmsmight provide corridorsor stepping stones of ac- Figure 2. Majorcacao producing countries, hectares harvested. Thousandsof hectares *1200 * 0 168 600 120 C Royal Swedish Academy of Sciences 2000 http://www.ambio.kva.se Ambio Vol. 29 No. 3, May 2000 forest lands, 1.3% of all forest converted to agriculturallands, and 0.5% of all lands convertedfrom primaryforest to agriculture or secondaryand degradedforests (27). Cacao's contribution to deforestationvaries considerablybetween regions (Table 3). In particular,13.4% of the original forestlandsof Cote d'Ivoire are now in cacao plantations.Almost one-sixth of the cleared lands is in cacao farms. Cacao historiansgenerally concur that the expansion of "cacao frontiers"has often occurredat the expense of forestlands. During recent decades, when cacao area expanded upwardsof 2.1 mill. ha worldwide, COted'Ivoire in West Africa, and Malaysia and Indonesiain SoutheastAsia, accountedfor some 82% of thatgrowth.Duringthe 1970s and 1980s, approximately90% of the cacao expansionin Cote d'Ivoire took place in forestlands (Ruf, pers. comm.). In Malaysia, the advanceof the cacao frontier into forested areas characterized60% of the new production zones overall;however, whereasin Sabahand Sarawak8090% of the expansion occurredin forests, in peninsularMalaysia, where producersopted to place cacao beneath extant coconut groves, only about20% of the growth took place in forested lands. In Indonesia, and especially in Sulawesi, the 1980s and 1990s saw about 50% of the cacao area appearingin formerly forestedareas(Ruf, pers.comm.). The directconnectionbetween cacao and deforestation,however,must be viewed within the dynamic of other agriculturalactivities and logging interests. For instance, the alluvial soils of the Indonesian coastal cacao region were previously planted to other crops. In this particular CACAOPRODUCTIONAND DEFORESTATION instance, cacao displaced those crops due to the rich soil and Cacao productionshows rapid expansion in recent decades. In favorablerainfallpatterns.Moreover, cacao often moved in af1970, the global cacao harvestwas only 1.5 mill. tonnes on 4.1 ter logging activity in Indonesia. Thus, approximately50% of mill. ha. By 1995, 3.0 mill. tonnes of beans (25) was harvested the cacao expansiontook place at the expense of forest,with only from 6.57 mill. ha. This representsa doubling of cacao produc- half of that (25%) being the direct result of small producersadtion and a 50% increasein areacultivatedin only 25 years. Cur- vancing into forested zones (Ruf, pers. comm.). rent productionis concentratedin African, and to a lesser degree Asian countries(Fig. 2; Table 1), particularlyCOted'Ivoire and Ghana, each with more than a mill. ha harvested in 1996 SMALL-SCALEPRODUCERS With a few regional exceptions, such as Brazil and Malaysia, (26). How has this increaseimpactedthe scarce remaininglowland small-scale producersform the backbone of the industry.Even humid tropical forests? On a worldwide scale, cacao accounts when the average size of holdings is large, such as in Brazil, for only (and very approximately)0.3% of the original tropical those farms at the small end of the size spectrumgreatly outnumber larger plantations.Cameroon and Ghana have typical distributions for farm size, with a predominance of holdings ranging in size from less than 1 ha to 5 or 10 ha (see Table 2). Table1. Cacaoarea harvestedand percentof In Nigeria, the averagefarm size is 1.7 ha, in Ecuadorit reaches globalarea represented,by country,1996. 4.5 ha, and in Cote d'Ivoire, farms average 2.8 ha (Ng, pers. Source: FAO,1996 (26). comm.). Brazil, by contrast,has farms larger than 1000 ha and %Global Area farmsaverageabout28 ha. Yet, for nearlyall countries,the small Cacao Area (000s ha) Country end of the farm-size spectrumdominatesthe cacao landscape. 32.7 2150 C6te d'lvoire In general, small- to medium-sized farms are both more pro18.25 1200 Ghana ductive (pods per ha) and more efficient (pods per dollar input) 688 10.47 Brazil 6.54 430 Nigeria than largercorporateholdings (28). Moreover, small producers 360 5.48 Cameroon accomplishthis despitegovernmentpolicies favoringlargerplan5.32 350 Ecuador 332 5.05 Indonesia tationsdatingback to colonial times in Asian and Africancoun3.12 205 Malaysia 2.08 137 DominicanRepublic tries (29). Governmentsin Java and East Sumatra,for instance, 124 1.89 Colombia delivered the tripodof production-land, labor, and capital-to 91 1.38 Mexico 88 1.34 PapuaNew Guinea large producersin efforts to keep estate productioncosts com65. 0.99 Venezuela petitive. Exportregulationsin Cameroonfavoredlarge holdings. 60 0.91 Guinea Equatorial 0.55 36 Peru And in the Dutch East Indies, state authoritiesaccused small0.38 25 Togo holdersof spreadingdreadedpests into estateproperties,a charge 0.37 24 Sao Tome 0.3 20 Costa Rica thatresultedin the uprootingof small producers'tree stock. Suit0.25 16.6 Philippines 0.16 10.5 able cacao lands were monopolized by largerholdings in EcuaUganda 0.07 4.6 Guatemala dor in the late 19th century. Even so, small producersaround 0.06 4 Panama 0.03 2.1 Honduras the worldhave remainedcompetitivedespite the institutionalob0.02 1.5 Nicaragua stacles placed before them. Large-scale plantersgenerally paid 0.01 0.4 El Salvador 2.27 149.2 Others more for their land, used more expensive labor, and had higher fixed and working capital costs comparedto their smallholder Note :Totalworldarea harvestedin 1996 was 6574000 ha; not all producingcountriesshown in chart. brethren(29). The tendency to favor plantationproductionover smallhold- ceptable habitat for animals dispersing between small forest patches.Shade cacao is being promotedas a corridorforest and buffer zone crop in the Talamancanmountainregion in Costa Rica (13). Other areas have been identified where cacao could be used in restorationefforts, as well, in particularas part of a reforestationaims in Vietnam. In summary,a greaterdiversity of tropical forest organisms occurs in shaded cacao plantationsthan in most other lowland tropical agriculturalsystems. Rustic plantations incorporating naturalforest shade trees are probably the best in this regard. However, to the degree that these rustic systems are not stable, they may not provide habitatin the long term. Cacao grown under plantedshade may provide the best long-termprotectionfor some tropicalforestbiodiversity.Systems thatincorporatea high diversityof trees with animaldispersedand pollinatedfruitsand flowers, along with retainingepiphytes, lianas, and mistletoes, will supportthe greatest diversity. Research on the ecological and agronomiccharacteristicsof trees, as well as an evaluation of theireconomicvalue, may help establisharraysof species that, when used, will optimize the economic value of farms and biodiversityconservation.Cacao grown closer to forest supports a greaterbiodiversity-perhaps forminga bufferhabitatfor mobile organisms.Cacao farms could be concentratedin the buffer zone of existingreservesor used to form corridorsbetween small forest reserves. Ambio Vol. 29 No. 3, May 2000 C Royal Swedish Academy of Sciences 2000 http://www.ambio.kva.se 169 ings, as well as Europeanplantersover local residents, derives largely from unquestionedbelief by people in governmentagricultural institutions in western "scientific" growing practices (29). Part of this belief system has held-even until the 1980s in the case of Malaysia-that advantages from economies of scale could be obtained with larger productionunits. Malaysia provides a textbook case of plantationproduction.With at least 85% of its productioncoming from holdings larger than 40 ha (30) upon which high levels of costly agrochemicalsare applied, we should expect some ecological and social consequencesnow commonly attributedto monoculturalsystems (31-33). The Malaysianproblemsinvolve high productioncosts (labor, agrochemicals,etc.), the aging of the tree stock, and shortages of labor(34, 35). Regularlyscheduledsprayings,whetherneeded or not, increase production costs. The cost of labor has also climbed as the work force opts for higher industrial Table2. Structureof the cacao sector InCameroon,Ghana,and Brazil,showing numberof farmsand hectareaccordingto size category of holding. wages. Furthermore, prices have dropped in the 1990s and other crops Cameroon Ghana Brazil have shown better relative returns;esSize category Farms ha Farms ha Farms Hectares pecially oil palm. <0.1 14 200 600 In general, small producers tend to 0.1-0.5 92 300 22 100 have lower per unit production costs. 0.6-1.0 57 700 43 700 <1.0 (164 200)? (66 400)? 119 850 98 963 Peasantsdepend upon and benefit from 1.1-2.0 65 500 95 000 87 529 159 258 family laborworkinglong hoursand re2.1-3.0 24 100 57 800 2.1-4.0 84 535 276 532 ceiving scant pay. But there are other 3.1-5.0 14 400 54 900 reasons for the greater efficiency of 4.1-8.0 40 330 258 694 <5.0 401 1350 small growers. Smaller farmersare po>5.1 8200 75 700 5.0-10.0 1031 5674 sitioned to have an intimateknowledge 8.1-20.0 13 323 182 398 of theirplots. This knowledge and care10.0-20.0 2475 19 998 20.0-30.0 2386 25 261 ful tendingis ideal for a crop like cacao, 20.1-40.0 37 382 1175 a crop often characterizedby small plots 30.0-40.0 27 512 1848 40.1-100.0* 258 190 062 5030 129 711 of 1000 to 3000 cacao trees. As one 100.1-200.0 2043 107 979 200.1-400.0 Malaysian cacao research manager 962 81 777 400.1-1000.0 482 62 020 stated, "Cocoa is like horticulture:the >1000.1 122 23 738 planter must almost know each tree" Total 276 400 349 800 347 000 1 203 280 16780 485 380 (30). This small-scale approachcan reNotes: ? Thissize categoryforCameroonis the sum of all smallerones; this intermediatesum has of in incrediblyhigh yields. For examsult course not been added to the total.* this size categoryis ">40"forGhana;also size categoryrelatesto ple, smallholdersin Sulawesi obtain up totalfarmarea whereas "hectares"for Brazilreferonlyto cacao area cultivated. Source: RecensementAgricole 1984 Republicdu CamerounMinisterede L'Agriculture Volume1 to 2000 kg ha-' of marketable beans; (Cameroon);Cocoa Services Division unpubi.data obtainedfromDr.BeatricePadi (Ghana);CEPLAC figures rarely achieved by commercial data providedby Dr.KeithAlger(Brazil). estates (28). Individual pioneering farmers have often been responsible for the new expansions of much of the world's cacao area. Whenever prices Table 3. The geography of the global cacao cycle. increase,and sometimeseven when they do not, pioneeringpeasCountry Stage in Comments ant producersturnto cacao as a survivalstrategy.With low barcacao cycle riers to entry and the crop's ability to produce something with C6te d'lvoire middle* littlelandavailableforexpansion;about relativelylittle strenuouslabor,it has been classified as an "easy 20%of tree stock >30 yrs old;labor crop" (28). In forested areas, producersbenefit from untapped difficultiesfewer, now thatworkersare comingin fromneighboringcountries; soil fertility-augmented by the burning of aboveground low incidenceof pest and disease; 66% of cacao unshaded biomass-and the lack of weeds. From the standpointof a small Ghana some landavailablein southwest;neariy late* grower, it is more attractiveand profitableto go into a forested halfof tree stock >30 yrs old;labor limitedbutfunctional;considerable area,clear the forest for new planting,and producethereinthan losses fromdisease and pests; most of to replantareasalreadydevoted to cacao. The fundamentalforce the cacao area is managedwithshade behind this attractionresides in differentialyields and labordeNigeria late littlelandforexpansion;60%of tree mands. A newly cleared forested area produces, at least for the stock >30 yrs old;no laborshortage; severe problemswithBlackPod, which first few years, 15%to 25% higher yields than a replantedarea reduces productionup to 70% Cameroon late average landarea availablefor (Matlick, pers. comm.). Moreover, the labor involved in clearexpansion;neariyhalfof tree stock >30 ing primary forest for cacao versus replanting is nearly half. yrs old;acute laborshortages except in newest productionareas; high incidence Ruf (36) found in Cote d'Ivoire that forest clearing and prepaof disease and pests rationrequired86 days ha-' of labor, comparedto 168 days unMalaysia middleto late littleto no landleftforexpansion;tree stock quiteyoung;laborshortageon der a replantingregime. peninsula Indonesia Brazil early late considerablelandavailablefor expansion;veryyoungtree stock;plenty of labor,except on Sulawesi,where access to workershas become difficult; potentialfutureproblemswithCacao Pod Borer virtuallyno landavailableforexpansion; 50%of tree stock >30 yrs old;labor supplygood now, but mobileand able to seek higherwages in othersectorspossible long-termshortages;severe disease problemswithWitches'Broom Sources: Ruf,(3);Taylor,(35);John Lunde,pers.comm. pendingliberalization schemes aimedat reducing/eliminating govemment pricecontrolscould bringhigherpricesto growers,thus reinvigorating the cacao sector forsome period. 170 CACAOCYCLES As in any agriculturalcommodity, cocoa suffers periodically from the crisis of overproduction.These wavelike variationslead to global price fluctuations.Cyclical fluctuationsin the international price translate,in turn,into boom/bustcycles at the level of production.When prices are high, one or more production zones boom onto the global scene. As supply creeps up, prices fall. Simply stated, prices follow a supply and demand model, whether real or perceived. However, as with other major commodity pricing,the variablesfeaturedin the complex price equation all relatein some way to futuresupply, and include weather ?DRoyal Swedish Academy of Sciences 2000 http://www.ambio.kva.se Ambio Vol. 29 No. 3, May 2000 and yield forecasts, changing consumptionhabits, social conditions in producingcountries,disease and pest problems,and institutionalinterventions.Historical data on cocoa show a pricing cycle of about 25 years (36). The development of genetic stock that can toleratemarginalconditions, in conjunctionwith policies that have allowed for their disseminationwith human migration,are also factors favoring boom/bust cycles in cacao (6). Focusing on the behaviorof small cacao producers,Ruf and Zadi (6) present a profile of the regional progressionof cacao systems. Placementwithin the model depends not merely upon how long cacao has been associatedwith a region, but also upon the managementtechnology used, available labor, and exploitable forest regions (Table 3). Regions newly opened to cacao show boom periods, during which the harvestedarea,the numberof producers,and exports, all increase as prices remain relatively high. Over a period of 15 to 20 years or more, the aging process of the cacao holdings proceeds, characterizedby falling yields that can be linked to increasingdisease and pest problems,and decreasingsoil fertility. Ecological changes are accompaniedby shifts in humandemography:migrationof landless people slows; emigrationmay begin; the farmpopulationages (36). Establishinga cacao farm involves the use of shade trees to help form an erect habit and provide a windbreakfor young cacao trees (5). As the cacao matures,shade is often removed or reduced; shade reduction leads to increased short-termyields, but yields may decline dramaticallyunless chemical fertilizers are applied (37). The "zero-shade"system is a recurrenttemptation, capturingthe imagination and hopes of small growers, extension agents, and policymakers.The final stage of the cocoa cycle often finds growers replanting shade to create an agroforestrysystem. Diversificationis a response to the falling cocoa yields and diminished income brought on by continued shadeless conditions. At this stage of the cocoa cycle, farmers may shift entirely to the productionof anothercommodity crop (see ref. 21 for discussion of shifts between cacao and banana productionin CentralAmerica). Because of predictablychanging ecological and social conditions, regionalcacao productionis often unstable,moving to agriculturalfrontiers, areas of primarytropical forest. Although cacao zones of long-termstabilitydo occur (Trinidad,Ecuador), the overall patternof instability and the search for new lands predominatesand needs to be addressed. First, the continued clearing of agriculturalfrontiersis a threatto biodiversityconservation. Second, concentrationof activity at particularfrontier regions makes the global cacao crop vulnerableto the invasion of new diseases or pests. And third, there are few forest frontierregions left in the world. litter from trees provides mulch and a supply of organic matter for the soil. This, in turn,can increase aeration,infiltration,and drainage,as well as result in a slow and steady release of nutrients into the mineralsoil. The decaying litterprovidesresources for a greaterdiversity of soil and litter organisms, which may be critical in nutrientbreakdownand cycling (41). In particular, fungal symbionts such as vesicular-arbuscularmycorrhizae play a key role in the nutrientbudget of moist forest understory plants (42) and cacao has long been known to harborsuch fungi (43). Dependingupon the species involved, shade trees can also fix atmosphericnitrogenand hold it within the soil layer. However, any shade tree can potentiallycompete with the cacao for nutrients,and wood or fruit productsremoved from the shade component representnutrientsleaving the agroforestrysystem (38). Young cacao plantsbenefit from the protectionof shadetrees and the influence shade has on growth form. One study has shown that shade also promotesgreaterlong-termproductionof older cacao plants with low levels of fertilization (37). Shade trees also serve to lessen the winds at or neargroundlevel within the cacao agroforestrysystem.Perhapsmoreimportantis the protection shade trees can provide from windbornespores of fungal diseases. We have long known that shade reducesthe spread of coffee leaf rust (Hemileia vastatrix)spores within plots (44). Evans (45) demonstratedthe benefitsof shadeanddisease spread in cacao, at least in the cases of Witches Broom (Crinipellis perniciosa) and Frosty Pod Rot (Moniliophthoraroreri). Much more work is needed to untanglethe relationshipbetween shade and disease control. Economic Value of Shade Trees In agroforestrysystems, additionalplant species mixed with the targetcrops can provide an arrayof valuableproducts(46-48). In COted'Ivoire, cacao producersmake use of some 27 mostly wild forest species as shade, 13 of which (48%) provide fuelwood and medicine, 11 of which (41%)offer food products, and 6 of which (22%) are used in construction(49). Researchers (50) showed that low-inputcacao systems that include commercialfruit productionfare quite well duringtimes of low cocoa prices.The break-evencocoa price for these virtuallychemical-free holdings is just over 50% of the price needed to break even in cacao withoutfruittrees. Shade trees themselves may also be valuable as timber (38, 51, 52). These trees have low maintenancecosts (41). Trials in coffee show that laurel (Cordia alliadora), plantedat a density of 100 trees ha-', yields up to 6 m3 ha-' yr-1.Earningsfrom timber could compensate for lost coffee yields of 17% when coffee prices are high, of 33% when prices are intermediate,and up to 100% of any lost productionwhen prices are low; likewise, we assume similarearningscould apply to cocoa. BENEFITSOF ENVIRONMENTAL Carbonsequestrationis a poorly assessed potential value of SHADEDSYSTEMS shadetrees, althoughNewmark(53) has been exploringthe posThe impact of cacao cultivation on biodiversity will be mini- sibility of using emissions tradingto providesmall farmerswith mized if productionis focused on already cleared lands, if ex- additionaleconomic incentives for shade tree establishmentand isting cacao farms show greaterlong-term stability, and if the protection. Some work (15) found mature (40-year-old) cacao farmsthemselvessupportgreaterlevels of biodiversity.Improved agroforestrysystems in Cameroonto be fixing carbonat around shade managementcan addressall three of these problems.The 154 tonnesha-';cacao systems in place for 15 and 25 years show greatestlong-termincentives for promotingthe managementof averagecarbonamountsof 111 and 132 tonnes,respectively.Ala diverse shade canopy can be found in the ecological and ag- though these sequestrationvalues are less than for primaryforronomic services providedby the shade itself (38). Many of the est (307 tonnes ha-'), they are generally greaterthan for annual long-termeconomic advantagesof shade alreadyexist and need crops. Based on fallow periods of 7 year and < 3 years. 106 and greaterrecognition. Some of the potential value to be obtained 87 tonnes ha-' of carbon,respectively,are sequestered. from shade will requireresearchfocused on shade management The Role of Shade in Pest and Disease Control sections. Agrochemicalsarecommonlyused to controlpests, diseases,and Shade Trees, Soil Protection, and Crop Health weeds, in cacao holdings (32). In additionto the health and enShade trees provide protectionfrom the physical impact of pre- vironmentalproblems this may engender, the chemicals themcipitation and hence can reduce soil erosion (39-41). The leaf selves often induce resistance in target species. Capsids Ambio Vol. 29 No. 3, May 2000 ?DRoyal Swedish Academy of Sciences 2000 http://www.ambio.kva.se 171 (Miridae),small plant-suckinginsects thatattackcacao pods and are the vectors of viral disease, developed resistance to aldrin and lindanein the early 1960s in Ghana.Severalauthoritieshave suggested thatthe removal of shade and the sprayingof insecticides are two major contributingfactors in the developmentof pest species in cacao (5, 7). The use of shade can lessen a farmer'sdependenceon chemicals. Shade usually inhibitsweed growth(38), and has also been shown to help control outbreaksof at least one of the important fungal diseases (Witches' Broom, Crinipellisperniciosa). In addition, shade reduces the activity of capsids (54,55). Shade is thoughtto affect the physiology of the cacao plantand the physical environmentas well, which affects disease and pest organisms. MacVean (56) has suggested that natural predators of nematodes, which damage plant roots, are dependentupon leaf litter createdprimarilyby shade trees in coffee farms. The role that mycorrhizaland endophyticfungi might play in conferring disease resistance is an active and promising area of research (Herre,pers. comm.). A central hypothesis in need of much more investigation is that organismsare less likely to reach pest or epidemic proportions in the presence of more complex predatorassemblages. Such research is related to the "enemies hypothesis" (57-59), which posits thatincreasedvegetationaldiversityof polycultures supportsa relatively greaterabundanceand diversity of predators and parasitesof herbivorousinsects. This explains why herbivore populationsare often lower in the same species of plant grown in a polyculturethan in a monoculture.Ratherthan manipulatespecific predator-preyrelationships,as some have done (60), manipulationof the cacao habitatin orderto retainthe coevolved ecological relationshipscharacteristicof naturalforest should be the first approachto be taken to prevent disease or pest problems. Much more research is needed to address the role of an increasingly diverse predatorycommunityassociated with shade. Trophicrelationshipsare often complex. Ants, for instance, are the most abundantarthropodpredatorsin many cacao farmsand the dominantant species can be importantpredators.But they also defendhomopterans,e.g. mealybugs,which themselvescan be pest species. Furthermore,in discussing the role of increased biological diversity in controlling pest and potential pest populations,it shouldbe noted thatbufferhabitatsmay also provide resourcesfor pests themselves. In particular,mammalsand birds, e.g. squirrelsand woodpeckershave been noted to damage cacao pods (61). SHORT-TERMINCENTIVES Many of the ecological services providedby shade management play out over time, yet the decisions must be consideredin light of short-termprofits for farmers.If there is a short-termcost of reduced productionassociated with shade management,it may need to be addressedby additionalmarketand institutionalincentives. Market Incentives Recent yearshave seen the developmentof the shadecoffee concept, followed by a numberof efforts to bring such a productto the market.The goal has been both to rewardproducersthrough a premium and provide consumers with a product with highly verifiable standards.Conceivably, a similar programcould be established for cocoa producers.Among the myriad issues are who pays for and what institutionsadministerthe certification program,of course, how large can the marketpremiumbe and who realizes the increase? Because no shade certificationfor cacao currentlyexists, and since the organic standardsfor cacao include a shade component (62), certified organic ("ecological" in Europe) cocoa is probablythe best surrogatefor marketableshade cacao. Organic 172 certification confers an added value to the product. However, certifiedorganiccocoa productionis in its infancy and currently representsa minusculefractionof total world productionof cocoa. Estimates now put total certified organic area at 8000 ha worldwide,coming mostly from LatinAmerica.Its marketshare within the United States is less than 1% (USD 15.4 million out of USD 13 billion in the US). In Europe,organic cocoa weighs in with about USD 18.2 mill. in sales (Daniels, pers. comm.). Still, as in the case of coffee, organic cacao is growing rapidly. Between 1990 and 1996, organicfoods in generalgrew by 20% a year (63). The organicchocolatemarketgrew by 36% in 1996/ 97, with companies like the Organic Commodities Project reporting a doubling in sales within the first six months of 1998 (Daniels, pers. comm.). Other Financial Incentives Given thatno developedconsumerdemandnow exists for shade cocoa, the emergence of a price premiumin the near future is doubtful. But there are other ways to encourage shade grown cocoa. Growersproducingcocoa on farmswith shadethatmeets specific criteria related to biodiversity could receive higher prices, inexpensive credit, and relevant extension services through nationally or internationallymanaged environmental funds. In addition,a tax on agrochemicalscould act as a disinReferences and Notes 1. Greenberg,R., Bichier, P. and Sterling,J. 1997. Bird populationsin rustic and planted shade coffee plantationsof easternChiapas,Mexico. Biotropica29, 501-514. 2. Perfecto, I. Rice, R.A., Greenberg,R. and van der Voort, M. 1996. Shade coffee as a refuge for biodiversity.BioScience 46, 598-608. 3. Rice, R.A. and Ward, J. 1996. Coffee, Conservationand Commercein the Western Hemisphere.SmithsonianMigratoryBird Center/NaturalResourcesDefense Council: Washington,DC. 4. Moguel, P. and Toledo, V. 1999. Biodiversity conservationin traditionalcoffee systems of Mexico. Conserv.Biol. 13, 1-12. 5. Wood, G.A.R. and Lass, R.A. 1985. Cocoa. Longman,New York. 6. Ruf, F. and Zadi, H. 1998. Cocoa: From Deforestation to Reforestation.Paper presented at the FirstSustainableWorkshopon SustainableCocoa Growing,PanamaCity, Panama,March30-April 2, 1998. 7. Leston, D. 1970. Entomologyof the cocoa farm.Ann. Rev. Entomol. 15, 273-294. 8. Room, P.M. 1971. The relative abundanceof ant species in Ghana's cocoa farms. J. Anim.Ecol. 40, 735-751. 9. Estrada,A., Coates-Estrada,R. and Merritt,Jr.,D. 1993. Bat species richnessand abundance in tropical rain forest fragments and in agriculturalhabitats at Las Tuxtlas, Mexico. Ecography 16, 309-318. 10. Estrada,A., Coates-Estrada,R. and Merritt,Jr.,D. 1994. Non flying mammalsand landscape changes in the tropical rain forest region of Los Tuxtlas, Mexico. Ecography 17, 229-241. 11. Estrada,A., Coates-Estrada,R. and Merritt, Jr.,D. 1997. Anthropogenic landscape changes and avian diversity at Los Tuxtlas, Mexico. Biodiv. Conserv.6, 19-43. 12. Estrada,A., Coates-Estrada,R.. MerittJr.,D., Montiel, S. and Curiel,D. 1993. Pattern of frugivorespecies richness and abundancein forest islands and in agriculturalhabitats at Los Tuxtias, Mexico. In: Fleming, T.H. and Estrada,A. (eds): Frugivores and Seed Dispersal: Ecological and Evolutionarv Aspects. Kluwer Academic Publ. Dordrecht,The Netherlands,pp. 245-257. 13. Parrish,J., Reitsma,R. and Greenberg,R. 1998. Cacao as Cropand ConservationTool: Lessonsfrom the TalamancaRegion of Costa Rica. Paperpresentedat the First SustainableWorkshopon SustainableCocoa Growing,PanamaCity, Panama,March30April 2, 1998. 14. Power, A. and Flecker, A.S. 1998. Agroecosystemsand biodiversity.Paperpresented at the First Sustainable Workshop on Sustainable Cocoa Growing, Panama City, Panama,March30-April 2, 1998. 15. Gockowski, J., Nkamleu, B. and Wendt, J. 1998. Implicationsof Resource Use Intensificationfor the Environmentand Sustainable TechnologySystemsin the CentralAfrican Rainforest. InternationalInstitute of Tropical Agriculture-Humid Forest Station, Yaound6,Cameroon. 16. Larison,B. and Smith, T.B. 1996. A Surveyof Montaneand LowlandBirdsfrom Three Sites on the Island of Bioko, Equatorial Guinea. Reportto the governmentof Equatorial Guinea. 17. Room, P.M. 1975. Diversity and organizationof the groundforagingant faunasof forest, grasslandand tree crops in Papua,New Guinea.Aust. J. Zool. 23, 71-89. 18. Andrews, R.M. 1979. Evolution of life histories:A comparisonof Anolis lizardsfrom matchedisland and mainlandhabitats.Brevoria454, 1-52. 19. Alves, M.C. 1990. The Role of Cocoa Plantations in the Conservationof the Atlantic Forests of SouthernBahia, Brazil. M.S. Thesis. Univ. of Florida.Gainseville, Florida, USA. 20. Room, P.M. 1972. The constitutionand naturalhistory of the fauna of the mistletoe Tapinathusbangwensis (Eng. and Krause)growing on cocoa in Ghana.J. Anim.Ecol. 41, 519-535. 21. Young, A. 1994. The Chocolate Tree. SmithsonianInstitutionPress, Washington,DC, USA. 22. Majer,J.D. 1993. Comparisonof the arborealant mosaic in Ghana,Brazil, PapuaNew Guineaand Australia-its structureand influenceon arthropoddiversity.In: Hymenoptera and Biodiversity.LaSalle, J. and Gould, I. (eds). CAB International,Wallingford, England,pp. 115-141. 23. Greenberg,R., Bichier, P. and Cruz, A. 2000. The impact of bird insectivory on herbivorous arthropodsand leaf damage in some Guatemalancoffee plantations.Animnal Conserv.(in press). 24. Pacheco, J.F., Whitney, B.M. and Gonzaga. L.P. 1996. A new genus and species of furnariid(Aves: Furnariidae)from the cocoa-growing region of southeasternBahia, Brazil. WilsonBull. 108, 397-606. 25. WorldTradeOrganization,1995. World Cocoa Production ? Royal Swedish Academy of Sciences 2000 http://www.ambio.kva.se Ambio Vol. 29 No. 3, May 2000 centive toward habitatcontaminationand alleviate worker exposure to toxic chemicals. The taxes, in turn,could go into the aforementionedenvironmentalfunds that reward growers for land stewardshipmeeting biodiversity criteria.Finally, the cocoa industryitself could invest more of its researchand development funds to supportsustainablecacao initiatives as a way of ensuringstable supplies long into the future(64). versity and act as a more effective refuge for some tropicalforest organisms than alternativelowland tropical crops, particularly annualcrops and cattle pasture.Therefore,the approachto improvingthe role thatcacao cultivationhas in biodiversityconservationshould be two-pronged.First, in each cacao-growing region, programsshould be establishedto replantabandonedor failing cacao holdings using diverse shadethatis useful to farmers and supportswildlife, as well as to protectremainingforest lands. Second, in regions of new cacao production,farmsshould CONCLUSIONS be establishedon alreadydeforestedlands so that cacao would Cacao cultivation has a complex effect on global biodiversity. provide a mode of reforestation,and particularefforts shouldbe Because of continued increasing demands for chocolate, the made to incorporatecacao as a buffer zone crop for established amount of cacao produced has doubled and the area of cacao forest reserves and parks. If traditionalcacao is to be actively promotedratherthanconcultivationincreasedover 2.5 mill. ha in the past 25 years.Much of the expansion of cacao productionhas taken place in areas tinuing to be simply the result of laissez-faire neglect, then a of primaryforest in West Africa and on Borneo and Sulawesi. majorresearchand extension effort is required.A comprehenEven if demand for cacao were constant, cultivation for cacao sive programneeds to be establishedthat will increasethe longwould continue to involve the clearing of new lands because of term productionof small shade-grownfarms in a mannerthat the dynamic cacao cycles describedearlierin this paper.Cacao protectsbiodiversityand addressesthe needs of small-scale cais one of many crops that form the first step in a naturalpro- cao farmers(65). To ensureadequateand sustainedfunding,such a programmust receive backing from the diverse stakeholders gression from forest to agriculturaland degradedlandscapes. However, cacao cultivation plays a relatively small role in with interestsin cocoa, producergroups,traders,manufacturers, cacao farms etc., as well as from institutions concerned about global biotropicaldeforestationon a global scale. Furthermore, with diverse shade have the potentialto supportgreaterlocal di- diversity,ruralcommunities,and sustainableagriculture. 26. FAO. Variousyears. AgricultureProduction Yearbook.United Nations Food and AgricultureOrganization,Rome, Italy. 27. Cacao area from FAO figures; deforestationfigures from Myers, N. 1991. Tropical forests: presentstatusand futureoutlook ClimateChange 19, 3-32. 28. Ruf, F., Yoddang, J. and Ardhy, W. 1995. The spectacularefficiency of cocoa smallholders in Sulawesi: why? until when? In: Cocoa Cycles: The Economics of Cocoa Supply.Ruf, F. and Siswoputranto,P.S. (eds). WoodheadPublishing,Ltd. Cambridge, England,Ch. 17. 29. Clarence-Smith,W.G. 1997. Cocoa plantationsin the Third World, 1870s-1914: the politicaleconomy of inefficiency.In: TheNew InstitutionalEconomicsand ThirdWorld Development.Harris,J., Hunter,J. and Lewis, C. (eds). Routledge, London, pp. 157171. 30. Ruf, F. 1993. Comparisonof cocoa productioncosts in seven producingcountries:C6te d'Ivoire, Ghana,Nigeria,Cameroon,Malaysia,Indonesia,Brazil. ThePlanter 69, 247262. 31. Buttel, F.H. 1990. Social relations and the growth of modern agriculture. In: Agroecology. Carroll,C.R., Vandermeer,J.H. and Rosset, P.M. (eds). McGraw-Hill, New York, USA, pp. 113-145. 32. Soule, J., Carre,D. and Jackson, W. 1990. Ecological impact of modem agriculture. In: Agroecology. Carroll, C.R., Vandermeer,J.H. and Rosset, P.M. (eds). McGrawHill, New York, USA, pp. 165-188. 33. Pimentel, D. and Wen Dazhong 1990. Technological changes in energy use in U.S. agriculturalproduction.In: Agroecology. Carroll,C.R., Vandermeer,J.H. and Rosset, P.M. (eds). McGraw-Hill,New York, USA, pp.147-163. 34. Laird,S.A., Obialor,C. and Skinner,E.A. 1996. An IntroductoryHandbookto Cocoa Certification:A Feasibility Studyand Regional Profile of WestAfrica. RainforestAlliance, New York. 35. Taylor, M. 1998. The WorldCocoa Situation.Paperpresentedat the IntemationalForum in Cocoa, Lima, Peru,28-29 October 1998. 36. Ruf, F. 1995. In: Cocoa Cycles: The Economics of Cocoa Supply. Ruf, F. and Siswoputranto,P.S. (eds). WoodheadPublishing,Ltd., Cambridge,England,Ch. 1. 37. Ahenkorah,Y. Akrofi, G.S. and Adri, A.K. 1974. The end of the first cocoa shade and manurialexperiment at the Cocoa Research Instituteof Ghana. J. Horticult. Sci. 49, 43-51. 38. Beer, J. 1987 Advantages, disadvantagesand desirable characteristicsof shade trees for coffee, cacao and tea. AgroforestrySyst. 5, 3-13. 39. Gligo, N. 1986. Agriculturay Medio Ambienteen AmericaLatina. EDUCA, San Jos6, Costa Rica. (in Spanish). 40. Rice, R.A. 1990. TransformingAgriculture:The Case of Coffee Leaf Rust and Coffee Renovation in Southern Nicaragua. Doctoral Dissertation. University of Califomia: Berkeley, CA, USA. 41. Beer, J., Muschler,R., Kass, D. and Somarriba,E. 1997. Shade managementin coffee and cacao plantations.AgroforestrySyst. 38, 139-164. 42. Janos, D.P. 1980. Vesicular-arbuscularmycorrhizaeaffect lowland tropicalrain forest plant growth.Ecology 61, 151-162. 43. Laycock, D.H. 1945. Preliminaryinvestigationsinto the functions of the endotrophic mycorrhizaof Theobromacacao L. Trop.Agricult. (Trinidad)22, 77-80. 44. Ward,H.M. 1882. On the life-history of Hemileia vastatrix.J. LinneanSoc. London: Botany 19, 299-335. 45. Evans, H.C. 1998. Disease and Sustainabilityin the Cocoa Agroecosystem.Paperpresented at the FirstSustainableWorkshopon SustainableCocoa Growing,PanamaCity, Panama,March30-April 2, 1998. 46. Farrell,J.G. 1987. Agroforestrysystems. In: Agroecology: The Scientific Basis of AlternativeAgriculture.Altieri, M.A. (ed.). Westview Press, Boulder, Colorado, USA, pp. 149-158. 47. Nair, P.K.R. 1990. Agroforestry:an approachto sustainableland use in the tropics. In: Agroecology and Small Farn Development.Altieri, M.A. and Hecht, S.B. (eds). CRC Press, Boca Raton, Florida,USA, pp. 121-136. 48. Belsky, J.M. 1993. Household food security, farm trees, and agroforestry:a comparative study in Indonesiaand the Philippines.Human Organiz.52, 130-141. 49. Herzog, F. 1994. Multipurposeshade trees in coffee and cocoa plantationsin C6te d'Ivoire. AgroforestrySyst. 27, 259-267. 50. Duguma,B., Gockowski,J. and Bakala,J. 1998. Small Cocoa (Theobromacacao Linn.) Cultivationin AgroforestrySystems of Westand CentralAfrica: Challenges and Opportunities. Paper presentedat the First SustainableWorkshopon SustainableCocoa Growing. PanamaCity, Panama,March30-April 2, 1998. Ambio Vol. 29 No. 3, May 2000 51. Heuveldop,J., Fassbender,H.W., Alpizar,L., Enriquez,G. and Fiulster,H. 1988. Modelling agroforestrysystems of cacao (Theobromacacao) with laurel(Cordiaalliodora) and poro (Erythrinapoeppigiana) in Costa Rica. II: Cacao and wood production,litter productionand decomposition.AgroforestrySyst. 6, 37-48. 52. Somarriba,E., Beer,J. andBonnemann,A. 1996.Arbolesleguminososy maderablescomo sombraparacacao:el concepto.SerieTecnica.InformeT6cnicoNo. 274. SerieGeneracionv Transferencia de Tecnologka No. 18.CATTE, CostaRica.(InSpanish). Turrialba, 53. Newmark,T. 1998.CarbonSequestration and Cocoaproduction:FinancingSustainableDeWorkvelopment byTradingCarbonEmissionCredits.PaperpresentedattheFirstSustainable shopon SustainableCocoaGrowing,PanamaCity,Panama,March30-April2, 1998. 54. Gerrard,B.M. 1964. Insects associatedwith unshadedTheobromacacao. Ghana.Proc. Conferenceon Mirids and OtherPests of Cocoa. West African Cocoa ResearchInstitute, Nigeria, pp. 101-1I11. 55. Padi, B. and Owusu, G.K. 1998. Towardsan IntegratedPest Managemenltfor Sustainable Cocoa Production in Ghana. Paperpresentedat the First SustainableWorkshop on SustainableCocoa Growing,PanamaCity, Panama,March30-April 2, 1998. 56. MacVean, C. 1992. Causas y naturalezadel mal de viniasenicafetos de Guatemala. Institutode Investigaciones,Universidadde Guatemala,Guatemala.(In Spanish). 57. Root, R.B. 1973. Organizationof plant-arthropodassociations in simple and diverse habitats:the faunaof collards (Brassica oleracea). Ecol. Monogr.43, 95-124. 58. Altieri, M.A. and Smith, L. 1986. The dynamicsof colonizing arthropodcommunities at the interface of abandoned,organic and commercial apple orchardsand adjacent woodlands.Agricult.Ecosyst. Environ.16, 29-43. 59. Russell, E.P. 1989. Enemies hypothesis: a review of the effect of vegetationaldiversity on predatoryinsects and parisitoids.Environ.Ecol. 18, 590-599. 60. Khoo, K.C. 1992. The influenceof Dolichoderusthoracicus(Hymenoptera:Formicidae) on losses due to Helopeltistheivora(Heteroptera: Miridae),blackpod disease,andmammalianpests of cocoa in Malaysia.Bull. Entom.Res. 82, 485-491. 61. Bhat, S.K., Nair, C.P.R. and Mathew, D.N. 1981. Mammalianpests of cocoa in South India. TropicalPest Mgmt27, 297-302. 62. IFOAM 1996. Basic Standardsfor Organic Agricultureand Processing and Guidelinesfor Coffee,Cocoa and Tea. IntemationalFederationof OrganicAgriculturalMovements, Thorley-Theley,Germany(FRG). 63. Murphy,K. 1996. Organic food makersreap green yields of revenue. The New York Times,October26, p. B 1. 64. Yoon, C.K. 1998. Chocoholics take note: beloved bean in peril. The New YorkTimes, May 4, p. 1. 65. For a comprehensivelist of recommendationsfor such a program,see "The First Intemational Workshopon SustainableCocoa Growing",available in the "Cocoa Corner"of the SmithsonianMigratoryBird Center'swebsite at <www.si.edu/smbc>. 66. We thank the following people for providing us with information for this paper: Stephanie Daniels, Development Coordinatorat the Organic Commodity Project in Cambridge,MA.; Alan Herre,researcherat the SmithsonianTropicalResearchInstitute, PanamaCity, Panama;B.K. Matlick, consultantin cacao industry;FranqoisRuf, researcherat CIRAD, stationedin Cote d'Ivoire;John Lunde, SustainableCocoa Specialist for Mars, Inc.; Ed Ng, CommercialStaff Officer for Mars, Inc. We also appreciate the financial supportprovided by Mars, Inc. during the researchand writing of this paper, and the help of all participantsin the First SustainableWorkshopon SustainableCocoa Growing,PanamaCity, Panama,March30-April 2, 1998. 67. First submitted25 May 1999. Accepted for publicationafterrevision 13 Jan. 2000. Robert A. Rice is a geographer and policy researcher at the Smithsonian MigratoryBird Center (SMBC)in Washington, DC, USA, where he studies land-use changes related to bird habitat. E-mail:rarice@igc.org Russell Greenberg, Directorof SMBC,is an ecologist who has conducted avian research in the USA, Russia, and throughout LatinAmerica. E-mail:anfbird@erols.com Their address: Smithsonian MigratoryBird Center, National Zoological Park,Washington, DC-20008,USA 0 Royal Swedish Academy of Sciences 2000 http://www.ambio.kva.se 173