´ lvarez, Juan Mario Lazcano and Blanca Leo´n Jon Fjeldsa˚, Marı´a D. A
Illicit Crops and Armed Conflict as Constraints on Biodiversity Conservation in the Andes Region Coca, once grown for local consumption in the Andes, is now produced for external markets, often in areas with armed conflict. Internationally financed eradication campaigns force traffickers and growers to constantly relocate, making drug-related activities a principal cause of forest loss. The impact on biodiversity is known only in general terms, and this article presents the first regional analysis to identify areas of special concern, using bird data as proxy. The aim of conserving all species may be significantly constrained in the Santa Marta and Perija´ mountains, Darie´n, some parts of the Central Andes in Colombia, and between the middle Maran˜o´n and middle Huallaga valleys in Peru. Solutions to the problem must address the root causes: international drug markets, long-lasting armed conflict, and lack of alternative income for the rural poor.
INTRODUCTION The humid forests of the tropical Andes region support the richest floral and faunal assemblages in the world (1), with remarkable local aggregates of endemic and rare species. These forests are therefore a top priority for conservation (2–4). Unfortunately, large forest areas are becoming severely degraded or transformed, especially in the east-Andean foothills (5), the Colombian Andes (6), and lowlands of southwestern Ecuador (7). In many districts, political and economic power is in the hands of armed groups that generate instability and often serve to secure land for unscrupulous investors (8). This even happens in areas that are formally protected (9–11) and where management for conservation is also hampered by lack of resources. One of the most important and directly measurable land degradation impacts associated with armed conflict pertains to the expansion of illicit cash crops (8, 11–13), most prominently coca, Erythroxylum coca and Erythroxylum novogranatense. Coca is now grown mainly below 700 m in the Andean forelands. In addition, marijuana, Cannabis sativa, is planted in the submontane zone, and opium poppies, Papaver somniferum, around 2000–2700 m (14, 15). Traditionally, coca was grown for local consumption in the Andean valleys, especially in the Yungas of La Paz in Bolivia and in Peru. These thinly spread crops had a minor ecological impact compared with the much larger areas that were cleared for cash crops, such as coffee and citrus. New coca-growing areas were established in the sub-Andean lowlands to satisfy international demands, first in districts in Bolivia and Peru, where considerable forest clearing had already taken place for agriculture and ranching. However, the activity soon expanded to the Amazonian agricultural frontier districts and was displaced to Colombia in the 1990s by internationally financed antidrug campaigns (15). Marijuana and poppy fields were hidden in the least accessible Andean forests (11–13, 16). The incentives for growing these crops are primarily economic: harvesting can take place several times per year, prices are high, purchase is guaranteed, and cash advances are offered by traffickers. However, the growers may also be forced Ambio Vol. 34, No. 3, May 2005
at gunpoint (8, 13, 17), and today, the expansion of illicit crops is strongly linked with the presence of armed conflict. Paramilitary groups have mostly secured the interests of the traffickers, while many left-wing guerrilla groups have pursued some environmental policies amid the conflict (7, 8, 17; S. Hvalkof, N. Krabbe, and P. Salaman pers. comm.). However, as the conflict dragged on, guerrillas have turned to coca for finance (8, 18). In Peru, the leftist guerrilla movement Sendero Luminoso became stewards for investors in timber and coca in the Selva Alta. Despite official assertions that they have been defeated, Sendero still reigns locally in San Martı´ n, Hua´nuco, Pasco, and Junı´ n. In contrast, coca cultivation in Bolivia does not seem to be strongly controlled at gunpoint, except locally. Coca-growing areas consist of a matrix of agricultural lands (for coca as well as food crops for the growers) and secondary scrub, with only scattered trees and little wildlife, as the growers often hunt to supplement the protein-deficient diet of bananas, manioc, and rice (19, 20). Soil erosion is often severe because coca is often grown on sandy soils and is subject to frequent defoliation and understory clearing (21). Contamination of soils and watersheds with herbicides and from chemicals used for the local coca-paste production is also common (12, 22). Eradication campaigns using aerial spraying with glyphosate (manufactured and marketed by Monsanto as Roundupt) in Colombia and manual eradication in Peru and Bolivia force traffickers and growers to constantly relocate and clear more forest (23–26). Although the impact is poorly studied, fumigation from the air may affect large areas, especially when done at night (12, 27). Altogether, two or three times more land is affected than the area of illicit crops per se (16) and the total loss of natural forest habitats may amount to 3.0 million hectares in Colombia (11) and 2.3 million hectares in Peru over the last two decades (15, 23). Very little is known about the environmental impacts of armed conflict (8, 12, 13, 28, 29), cultivation of illicit crops (11, 12, 21, 23, 27), or the aerial spraying to destroy these crops. In particular, there are very few hard data to judge the effects on biodiversity. Thickets and forest fragments in the principal coca-growing district in the middle Huallaga valley in Peru had a very impoverished avifauna—with almost total lack of genuine forest species and potential game birds—in spite of local initiatives to resist the coca growers (30). Despite the dearth of evidence, we may assume that coca-growing areas lose substantial parts of their biodiversity and are unlikely to maintain viable populations of forest-dependent wildlife. In this article, we adopt a regional perspective, as we try to identify where the growing of illicit crops will constrain the efforts to prevent species extinction. For a large-scale assessment, we compared information about where illicit crops are grown with the most detailed regional dataset of species distributions. These data comprise only birds (1, 31, 32), but we assume that this mirrors the diversity in other organismal groups at broad geographical scales (33). Because there are competing interests determining land use, it is important that conservationists provide precise and fully accountable priority plans that identify the smallest possible set
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
of areas holding populations of all species (34–36). This is achieved most efficiently by using the principle of complementarity (35, 36). When relevant data are available, conservation needs can also be balanced against sociopolitical costs, based on population or development potential (37). We will examine here where conservation plans will be constrained by conflicts in land use associated with illicit crops. We do not find it particularly relevant to examine the role of existing protected areas, as these are not well placed for protecting the most unique species aggregates (31, 32) and are not respected by those involved with illicit crops (9–11). Our analysis is designed to reveal i) whether the conservation aims can be achieved outside the areas where illicit crops are grown and ii) where the aims cannot be reached unless conservation actions can be taken in spite of the conflicts that will be faced.
METHODS Data on Areas of Armed Conflicts and Coca Production Our main sources were internationally available documents (14, 15, 27, 38) providing data about illicit crops (for Colombia, see also 18; for Peru, 21, 38–41; for Bolivia, 42, 43). It should be noted that official documents are few and sometimes distorted as a security measure. Published maps have low resolution and even satellite images are often misleading, as the illicit crops may be hidden among other crops. The most serious problem, though, is the mobility of the coca front, as military actions and eradication displace cultivation. This fuzziness means that the data represent approximations and that we need to identify possible conflict zones in addition to those with well-documented problems. The latter areas (conflict level 1) are those where illicit crops have already been grown on a significant scale, often under armed control. The other category (level 2) are those areas where the areas of illicit crops may still be small but where expansion of armed groups can be predicted as different parties in the conflict fight to gain control (8, 18). Although much of the biodiversity may be found in other elevation zones than where the coca is grown, conservation actions may still be hampered by security problems. For the sake of this analysis, we therefore assume that conservation actions will be difficult in grid cells of which .20% is recorded as conflict area. For Colombia, we plotted conflicts by entire municipalities, regardless of the actual extent under cultivation. Presence of illicit crops and land purchases by traffickers was initially overlain and measured on the forest cover map using ArcView 3.1 (11) but was converted, for this analysis, to a 15 0 geographical grid (see below). Because municipalities in the Amazon are very large compared with the Andes, their areas do not represent the habitat loss proportionally, and municipalities with scattered coca fields in the Colombian Amazon lowlands were therefore considered as moderately threatened (level 2). For Peru, we recorded as level 1 areas where coca production is well consolidated, including areas that became seriously degraded in the 1980s and those where Sendero Luminoso still operates. We excluded some indigenous territories with strong autonomy (e.g., Gran Pajonal), where coca growers have been prevented from settling (41). Areas with less threat (level 2) were hard to define because of conflicting evidence. For Bolivia, the extent of areas with intensive coca cultivation in the Chapare region is well documented, but other areas are difficult to define because the coca fields are scattered among other crops and existing maps of coca zones in the Yungas are inaccurate (42, 43). These areas were recorded as level 2.
Bird Data For a general assessment of the biodiversity values of coca areas, we used a database of all 2900 resident South American nonmarine birds in a 18 3 18 grid (1, 31, 32; see Fig. 1A). For the more specific conservation priority analysis, we used a bird database for the Andean region and adjacent forelands in a 15 0 3 15 0 grid (32; see Fig. 1B). This latter dataset was initially developed only for highland birds (which are well documented and relatively easy to map in detail because of constraining topography and habitat gradients). This dataset was expanded to include range-restricted lowland birds (those representing the lower 25% of range sizes for all resident South American land birds), the distributions of which are generally well documented (4). This dataset includes almost all the threatened birds of tropical western South America (44), and extra maps were added for the 10 lowland species that are threatened but more widespread. Altogether, the high-resolution dataset comprises .280 000 in-grid-cell records of 987 species (80 000 confirmed records and the rest conservative interpolation). A minimum set of complementary target areas that contain all species were identified using a heuristic search algorithm with a redundancy backcheck (45). The results of such analyses will, in practice, cater for the widespread lowland birds that were not included in the fine-resolution database, and they will be very robust for areas of strongly aggregated (nested) endemism in the Andes. In order to secure viable populations and some flexibility, we provide an area selection plan where every species is represented in five 15 0 grid cells (an aim that cannot be achieved for species of which the range falls within ,5 cells). The effectiveness (number of species covered) and efficiency (species per area) was monitored by a stepwise procedure: First a minimum set covering all species in one cell was identified, then a new minimum set after blocking the first set, and so on until five representations were achieved. The procedure was repeated after blocking the defined conflict areas, thereby allowing a precise record of how—and where—the conflict may constrain an optimal conservation plan.
RESULTS Where the Conflicts Are The conflict areas are identified in Figure 4 (240 15 0 cells with conflict level 1 in Fig. 4A and 423 cells after adding areas with conflict levels 2 in Fig. 4B). In Colombia, the cultivation was displaced after ca. 1995 from the Guaviare to the Yarı´ and Putumayo in the southeastern lowlands and most recently to the Andean region and the Pacific slope, the Choco´ biogeographic region (18). Because of the rapid development of coca cultivation, large areas on the Pacific slope (Darie´n area and central and southern Choco´) and in Serranı´ a de San Lucas (17) were classified as level 1, even though the extent of clearing is poorly documented. In Peru, the coca center has been in the middle and upper Huallaga valley, which a decade ago provided the raw material for 60% of the world’s illicit cocaine (46). Within the Tingo Marı´ a and Tarapoto areas (Fig. 4A), very little natural habitat is left, but coca growing is now on the retreat. Other seriously affected areas are in the Palcazu-Pichis district and in areas with logging and colonization along the Huancayo-Satipo-La Merced-Pucallpa road and along the lower Apurı´ mac river. Data about the intensity and location of illicit crops in the areas between Tarabuco and the middle Maran˜o´n River are contradictory. The traditional (indigenous) coca cultivation in the submontane zone of Cuzco has become more intense.
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
Ambio Vol. 34, No. 3, May 2005
Figure 1. Variations in avian species richness in northwestern South America, the warmest red colors representing the highest numbers of species. A: All birds, in a one-degree grid. B: Endemism (defined as number of species representing the lower quartile—or 25%—of range sizes of all 2900 resident nonmarine South American birds, in a 15 0 grid. C: A near-minimum set of conservation options that allows all species to be protected in five areas (filled symbols showing 122 irreplaceable areas, half-filled symbols flexible choices). The most unique areas have been marked yellow, with numbers of goal-essential species (covered nowhere else in the minimum set). Note that the right map omits areas south of 208S, where coca does not grow.
In Bolivia, the main coca-growing area has been the subAndean Chapare zone in Cochabamba, particularly in the Isinuta-Etetazama area in the northwest, where government control has been very difficult. Satellite images were published in 1995 (20), but since then, some expansion has taken place to the north and west. Less intense cultivation (level 2) lies in the peripheral and eastern parts of the Chapare and some parts of the lower yungas of La Paz, some of them traditional production (for local consumption), others for external markets. In these districts, the coca crops cover relatively small patches in citrus- and coffee-growing districts. Ecuador has generally been free of illicit crops (apart from some marijuana), but because of the ongoing military intervention in Colombia, there is now a strong influx of refugees south of the Putumayo River in the northeast. Traffickers have started to purchase land there and established coca laboratories in the Lago Agrio area, the center of Ecuador’s oil industry. The areas with illicit crops overlap only marginally with the network of formally protected areas in Bolivia and Peru. In Colombia, we noted significant overlaps in Sierra Nevada de Sta Marta and Iroka and Socorpa in Perija´ Mountains in the north; Paramillo and Munchique in the west; Las Hermosas, Huila, and Purace´ in Cordillera Central; and Los Picachos, Macarena, and La Paya in the southeast (11). Overall, formal protection is not relevant in relation to the investors in illicit crops. Ambio Vol. 34, No. 3, May 2005
Bird Distributions The species richness pattern for South American birds (Fig. 1; 31, 32) is characterized by a very strong concentration along the humid Andean slopes toward the Amazon basin and a somewhat lower species richness—and much lower endemism—in the Amazon lowlands (Fig. 2). The species richness is moderate but endemism is high, with marked local aggregates in the humid parts of the Andes, especially in the Choco´, southwestern Ecuador, Cordillera Central in northern Peru, the Cuzco area, and the Bolivian Yungas (Fig. 1B). In general, the endemism peaks on higher elevation than where coca is grown (47), but conservation actions may still be difficult because of armed conflict. Using the coarse-scale database, grid cells overlapping .20% with coca areas contain populations of, altogether, 1950 bird species (67.3% of all South American resident land birds) or 1962 (67.7%) if coca areas in Acre, Brazil, are also included. As illustrated by the red symbols in Figure 2, illicit crops are produced through the entire range of variation in species richness and endemism except for the most species-poor parts. In the upper right quartile in Figure 2 (high species richness/high endemism), 39% of all grid cells are classified as conflict areas—against 20.4% in areas with few species/high endemism, 5.2% where there are many species/low endemism, and 8.2% with few species/low endemism. Very similar patterns would emerge if we replace endemism with numbers of threatened species (see 44). Peru and Colombia
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
Figure 2. Distribution of areas affected by illicit crops in the spectrum of variation of avian species richness and endemism in Colombia, Ecuador, Peru, and Bolivia (excluding arid zones, which are irrelevant in this context). Each point refers to one 18 grid-cell; large red points represent grid-cells where .20% of the area is affected by conflicts related to illicit crops. Parts of the dot swarms representing Amazon lowlands and Andean slopes on the Amazon interface are indicated. Horizontal and vertical lines are medians. Note that the endemism score 1.0 corresponds to the mean value for all grid cells in South America.
both have 133 threatened and near-threatened species, against 108 in the more politically stable intervening country of Ecuador (44). This is noteworthy, considering the low government interest in forest protection in Ecuador and the extreme forest loss in the southwestern region in that country (7).
Minimum Sets of Conservation Targets Figure 1C shows a network of conservation targets, defined as the minimum number of cells needed for five representations of all 987 species (which we consider a realistic requirement for maintaining viable populations). Out of 442 15 0 cells, 122
Figure 3. The effectiveness (percentage of all species) and efficiency (species per grid cell) of consecutive minimum sets where all species can be represented in five grid cells (see text). Blue dots: No constraints (other than limitations caused by species of which the range fall within ,5 cells). Yellow diamonds: Same after blocking areas of intense conflict (Fig. 4A). Red dots: Same after blocking also the areas of moderate conflict (Fig. 4B).
(27.6%) are irreplaceable, the others flexible, which means that alternative choices exist, although this may require a larger total area. A minimum set of 89 areas is needed for a single representation of all species. Of these, the most important are Sta Marta Mountains in the far northwest (two cells with 9 and 12 goal-essential species; see legend to Fig. 1), Me´rida Mountains in Venezuela (9 spp.), the mountains of Tolima in central Colombia (8 spp.), Mindo west of Quito in Ecuador (9 spp.), and Cocapata at the Cochabamba-La Paz boundary in Bolivia (11 spp.). If these 89 areas are rejected as conservation targets, then an alternative set of 97 cells will cover 979 species (missing species being single-cell endemics). The third choice is 99 cells with 962 species; fourth choice, 98 cells with 950 species; fifth choice, 99 cells with the remaining 936 species. The change in effectiveness and efficiency appears in Figure 3. When 234 cells with significant (level 1) conflict are blocked (Fig. 4A), five representations require 456 cells, of which 135 (29.6%) are irreplaceable. A minimum set of 91 cells is needed for a single representation of 984 species. Thus, by slightly modifying the configuration of the conservation network, it is still possible to protect all species except for two single-cell endemics, Eriocnemis mirabilis and Henicorhina negreti, at Cerro Munchique, Colombia (see 48 for current situation at this site). However, it is also a problem that 10 endemics of the humid submontane zone of the northeast slope of the Sta Marta Mountains have their core habitat in the conflict zone (4). The first alternative set comprises 100 cells with 971 species; third set, 98 cells and 956 species; fourth set, 91 cells and 938 species; fifth set, 100 cells and 927 species. The loss of effectiveness (Fig. 3) is mainly due to limited opportunities for multiple representation of Sta Marta endemics, but significant constraints appear also in the Perija´ Mountains along the Venezuelan border and in the northern parts of Cordillera Occidental (especially Paramillo) and in the Farallones of Cali and southern Cordillera Central in Colombia and in two areas in northern Peru (encircled areas in Fig. 4A). Adding 192 cells of moderate (level 2) conflict (Fig. 4B), 390 cells are needed for five representations. Of these, 129 (33.1%) are irreplaceable and fewer species can obtain five representations (Fig. 3): a minimum set for a single representation is 91 cells covering 983 species; second choice, 96 areas with 968 species; third choice, 96 areas with 951 species; fourth choice, 88 areas with 933 species; fifth choice, 96 areas with 925 species. Although the flexibility of area selection is now seriously constrained, it is still possible to make a conservation plan that is efficient in terms of species numbers per area. This is because loss of conservation opportunities in some districts can be compensated by relocating the conservation priorities within the Choco´ and Cordilleras Occidental and Central, in Santander (between Perija´ Mountains and the main part of Cordillera Oriental), Colombia, in the Cordillera Central of San Martı´ n (P.N. Rı´ o Abiseo and adjacent highlands), and in the high cordilleras of Cuzco, Peru. It is fortunate that, in the Cordillera Oriental of Colombia, the most biologically unique parts are the humid escarpments just west of the Bogota´ and Ubate´ Plateaus and not the coca districts on the east slope. Along the Cordillera Central, the endemism is concentrated in Tolima (with P.N. Los Nevados) rather than in the conflict areas further north and south. In the Darien, endemism peaks along the Tacarcuna hills rather than the coca zone to the southwest.
Areas of Particular Concern In the Sta Marta mountain, the conservation situation may be very difficult for the submontane endemics but better for those of the highlands that lie in indigenous and guerilla territories. In the Perija´ Mountains, conservation of Pyrrhura caeruleiceps and
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
Ambio Vol. 34, No. 3, May 2005
Figure 4. Minimum sets of conservation areas after blocking areas of coca and opium poppy production (shaded areas; A) showing those with high conflict and adding areas of less intensive conflict (B). Geographical names mentioned in the text are given. Areas where the optimal conservation plan was significantly constrained by the conflict and where alternatives could not be found are marked with red. Some species can be lost in these areas unless actions for forest conservation are undertaken despite the conflict.
Pauxi pauxi gilliardi may be seriously constrained while the highaltitude endemics still find safe areas (Angelo Viloria pers. comm.). Conservation options are also increasingly constrained for the narrow endemics of southwestern Darien (Crypturellus kerriae and especially Psarocolius cassini), for Pyrrhura subandina (Paramillo), and some Choco´ endemics (particularly for the above-mentioned E. mirabilis and H. negreti). Some conflict is also evident in Snı´ a de San Lucas (at least for Crax alberti), at the head of the Magdalena Valley (Tinamotis osgoodi, Atlapetes fuscoolivaceus, and Scytalopus sp.) and on the eastern Andean slope (Grallaria kaestneri). In Peru, conflicts may become serious around the bend of the Maran˜o´n River and the middle Huallaga as coca and marijuana areas expand into the intervening area, with biologically important sites in Cordillera Cola´n, Abra Patricia, and Moyobamba (species of concern being Xenoglaux loweryi, Grallaria blakei, Grallaricula ochraceifrons, Picumnus steindachneri, Herpsilochmus parkeri, Zimmerius vallerejoi, Hemitriccus cinnamomeipectus, Poecilotriccus luluae, Henicorhina leucoptera, and Loddigesia mirabilis in the adjacent Utcubamba basin). The coca districts in the Upper Huallaga and Pachitea-Aguaytia and other areas in Pasco and Junı´ n do not seem to raise particular conservation concerns. The coca-related activities in Cuzco in Peru are just below (north of) the core area for endemic species in Cordilleras Vilcanota and Vilcabamba, but conservation options may still be somewhat constrained for Amazilia viridicauda, Grallaria Ambio Vol. 34, No. 3, May 2005
erythroleuca, and Thryothorus eisenmanni. In Bolivia, many species of the lower yungas slopes have a poor conservation status (44), but it is difficult to associate this specifically with coca. Clearing of new areas for coca could seriously constrain conservation options for Pauxi unicornis and Terenura sharpei. Even fairly widespread species may be affected if large proportions of their ranges lie in armed-conflict zones. This is the case for the two large green macaws, Ara ambigua and Ara militaris, and possibly for Crypturellus duidae (Macarena Mountains population), Odontophorus atrifrons (Sta Marta to Santander), Columba oenops and Pyrrhura peruviana (northern Peru), and western Colombian populations of Chloropipo flavicapilla, Dacnis hartlaubi, Penelope perspicax, and the north Colombian endemics Chlorochrysa nitidissima, Clytoctantes alixii, Leptotila cassini, Pionopsitta pyrilia, and (especially) Habia gutturalis.
DISCUSSION The range of environmental impacts from illicit crops, trafficking, and violence is wide, and we deal here only with a spatial component of this complexity. Traditional farmers in Bolivia state that they clear for coca because only then can they earn enough cash on their small forest clearings to sustain their families, but the environmental effects of not having this option have not been studied (see 13, though). What is clear is that illicit crops represent a serious environmental threat when producing for international markets. We therefore focus
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
primarily on areas where the habitat destruction is significant and where conservation (including in reserves under government management) becomes difficult because of armed conflict. We need now to consider whether our approach is suitable and the data are good enough to justify our identification of critical areas. The complementarity analysis did not include widespread lowland birds, but inclusion of these would probably not have altered the configuration of the minimum set. Other taxonomic groups, such as plants, may show more complex and local patterns of endemism in the Andean foreland ridges (23), but the coarse-scale pattern may be similar overall (33, 47). Fortunately, the areas of frontier agriculture in the lowlands of eastern Colombia are not particularly rich in bird species (Fig. 1A) or unique (Fig. 1B). This is not meant to devaluate such forests (49) but rather to emphasize that most species are sufficiently widespread that the regional species pool is likely to persist in spite of local habitat loss. Our analysis may, however, underestimate the conservation conflicts in Peru, where local endemics may exist that are taxonomically neglected (see 50) or yet undiscovered, for instance, along the tectonically active Altos de Iquitos and Fitzcarrald and Serra do Moa in the upper Amazon Basin (51). Undiscovered species can also be expected in the unexplored parts of Snı´ a San Lucas between the Nechı´ and Magdalena Rivers (52), where conflicts are accelerating (17, 52). Overall, the conservation problems related to illicit crops are local and mainly in the Andes region. The conservation efforts that are being made in the Sta Marts Mountains must be given high priority, and conservation initiative, against difficult odds, are also needed in the Perija´ Mountains and in northern Peru (highlands between Bagua and Tarapoto). Fortunately, two of the endemics of the latter area have now been found in the largest remaining tract of intact foothill forest in the Cordillera Azul National Park further south (53). Conservation conflicts may also arise if the coca areas expand deeper into the sub-Andean zone of Pasco and Junı´ n or up the valleys of Cuzco, but conservation priorities are not yet constrained by armed conflict there. The situation in Bolivia is not particularly alarming today but may rapidly change. The recent coca eradication has caused strong anger among the local campesino communities, who receive little support for alternative income. In the Chapare region, this now means a shift toward hiding new coca fields (in remote sites and among other crops) in the Arepucho-Chuquiloma (where traditional coca growing is now legal) and the Bulo-bulo areas inside the Carrasco National Park. We also fear that new coca-growing areas could be established in the enormous lowland forests near the Cochabamba-La Paz border (in the Isiboro Se´cure N.P. and toward the Sur Yungas of La Paz, where there are several narrow endemics, including some undescribed species (54). Several logging concessions have been given in this district, and these could rapidly open up for coca.
What Kinds of Action? The traditional approach to conservation by fences and fines does not seem realistic in areas of armed conflict, and it seems more relevant to work with local communities and emphasize the urgent need for social and economic reforms (55, 56). Experiences from the Chapare region suggest that long-term subsidies to alternative crops are required, as the local smallholders otherwise shift back to growing coca as soon as the external support stops. In some areas, the key to forest conservation may be to support the local autonomous indigenous communities and Procesos de Comunidades Negras in the Choco´. Some of these are environmentally conscious, well organized, and oppose the timber and coca interests, as demonstrated by the Pichis revolt in Peru in 1990. For example, indigenous communities in the upper Amazon area are strongly
against coca, which they associate with colonists and Sendero Luminoso. So far, national governments and US-funded antidrug programs have provided little support to poor rural areas. Although Bolivia has an excellent written strategy for sustainable development (57), the recent neoliberal privatization policies have not promoted sustainability. In Colombia, similar reforms prompted the dismantling of institutions that provided technical and financial support to small growers, followed by the collapse of local production systems (58). In private hands, and sometimes even public hands, the income generated by using cheap labor to harvest and export natural resources ends in personal bank accounts and is not invested in education, capacity building, and improved production systems that would add value to the harvested natural resources. Drug trafficking, fueled by poverty and lack of opportunity and aided by systemic corruption, represents an extreme case of this economic model (59). Of the many actors in illicit production, the national government is constrained by bi- and multilateral alliances (e.g., United Nations Convention against the Illicit Traffic of Narcotic Drugs and Psychotropic Substances, World Trade Organization, etc.) that limit the range of political and economic alternatives. Even international institutions acknowledge that the context of economic liberalization and structural adjustments magnify the economic challenges posed by illicit crops (60). As a consequence, the current European Union strategy (61) suggests economic alternatives as a solution to the illegal-trade conundrum. This is in direct conflict with the US Drug Elimination Act of 1998, which provides economic support to forced eradication and to national armies fighting narco guerillas. This ‘‘war on drugs,’’ an admitted failure in US foreign and domestic policy (62), ignores fundamental capitalist principles: as long as there are markets for drugs, military intervention achieves little but stopping the production locally and temporarily, while prices are artificially increased (as seen recently in Peru) and finance is provided for alternative supply routes leading to never-ending habitat loss. Thus, market incentives continue to drive production into environmentally sensitive areas (S. Vehkama¨ki pers. comm. (63)). It is worth noting that the development of illicit-crop production and marketing escalated mainly in areas with long-lasting armed conflict. This is evident in Colombia (18) and Peru, although the historical unraveling has been unique in each case. One valid question would be whether armed conflict is a cause or a consequence of illicit-crop production: the world’s largest producers of illicit crops, Afghanistan, Colombia, and Myanmar (and Peru until recently) have all been mired in internal conflicts for decades (M.D. A´lvarez and E. Alvarez pers. comm.). If widespread violence is a precondition for the expansion of illicit crops, Peru’s temporary success at controlling coca production might be the result of the government’s justice adjudication and territorial gains against Sendero Luminoso in the 1990s and not the other way around. Conversely, efforts to control illicit-crop production in Colombia would have to change emphases from eradication and interdiction to the increase in coverage of the justice system and the reinsertion of armed groups into productive civilian life. Effective eradication would, then, also entail economic programs to facilitate the transition to production of legal crops among growers and to provide employment to combatants and displaced populations. The instability inherent to drug markets and the inevitable relocation of production to uncharted territories can only be reduced by reigning in demand and controlling the money-laundering operations throughout developed and developing countries. While the solutions to these problems are in the hands of politicians, we hope that our
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se
Ambio Vol. 34, No. 3, May 2005
analysis will guide conservationists to concentrate their efforts where it is most critically needed. References and Notes 1. Rahbek, C. and Graves, G.R. 2001. Multiscale assessment of patterns of avian species richness. Proc. Nat. Acad. Sci. 98, 4534–4539. 2. Mittermeier, R.A., Myers, N., Thompsen, J.B., da Fonseca, G.A.B. and Olivieri, S. 1998. Biodiversity hotspots and major tropical wilderness areas: approaches to setting conservation priorities. Conserv. Biol. 12, 516–20. 3. Olson, D.M. and Dinerstein, E. 1998. The global 2000: a representation approach to conserving the Earth’s most biologically valuable ecoregions. Cons. Biol. 12, 502–515. 4. Stattersfield, A.J., Crosby, M.J., Long, A.J. and Wege, D.C. 1998. Endemic Bird Areas of the World: Priorities for Biodiversity Conservation. BirdLife Conservation Series No. 7. BirdLife International, Cambridge, UK. 5. Stotz, D.F. 1998. Endemism and species turnover with elevation in montane avifaunas in the neotropics: implications for conservation. In: Conservation in a Changing World. Mace, G.M., Balmford, A. and Ginsberg, J.R. (eds.). Cambridge University Press, Cambridge, UK, pp. 161–180. 6. Etter, A. 1998. Mapa General deEecosistemas de Colombia. Instituto Alejandro von Humboldt: Bogota´. (In Spanish). 7. Best, B.J. and Kessler, M. 1995. Biodiversity and Conservation in Tumbesian Ecuador and Peru. BirdLife International, Cambridge, UK. 8. A´lvarez, M.D. 2003. Forests in the time of violence: conservation implications of the Colombian war. J. Sust. Forestry 16, 49–70. 9. Rodriguez, M. 1998. La Reforma Ambiental en Colombia: Anotaciones para la Historia de la Gestio´n Pu´blica Ambiental. Tercer Mundo Ed., Bogota´. (In Spanish). 10. World Commission on Protected Areas. 2003. World Database on Protected Areas. IUCN UNEP, Gland, Switzerland. 11. A´lvarez, M.D. 2002. Illicit crops and bird conservation priorities in Colombia. Conserv. Biol. 16, 1086–1096. 12. A´lvarez, M.D. 2001. Dan˜os ambientales por cultivos ilı´ citos y procesamiento de drogas prohibidas en Colombia. Acodal 193, 7–16. (In Spanish). 13. A´lvarez, M.D. 2001. Could peace be worse then war for Colombia’s forests? The Environmentalist 21, 305–315. 14. United Nations Drug Control Programme. 2000. Mapa Localizacio´n de Cultivos Ilı´citos en Colombia, Censo 2000. Divisio´n Nacional de Estupefacientes, Direccio´n General Antinarco´ticos de la Policı´ a Nacional/UNDCP, Bogota´. (In Spanish). 15. United Nations Office for Drug Control and Crime Prevention. 1999. Global Illicit Drug Trends 1999. United Nations Publication E99 XI15. United Nations, New York. 16. Cavelier, J. and Etter, A. 1995. Deforestation of montane forests in Colombia as a result of illegal plantations of opium (Papaver somniferum). In: Biodiversity and Conservation of Neotropical Montane Forests. Churchill, S.P., Balslev, H., Forero, E. and Luteyn, J.L. (eds.). The New York Botanical Garden, Bronx, pp. 541–549. 17. Da´valos, L.M. 2000. The San Lucas mountain range in Colombia: how much conservation is owed to the violence? Biodiv. Conserv. 10, 69–78. 18. Reyes, A. 1999. Especial: 35 an˜os de conflicto. Lectura Dominicales. El Tiempo (Bogota´) 17 October, 2-11. (In Spanish). 19. Kempff, N. 1986. Recursos Fauna Silvestre. En Jornadas Santa Cruz 2000. Comite ProSanta Cruz, Santa Cruz, Bolivia. (In Spanish). 20. Henkel, R. 1995. Coca (Erythroxylum coca) cultivation, cocaine production, and biodiversity loss in the Chapare region of Bolivia. In: Biodiversity and Conservation of Neotropical Montane Forests. Churchill, S.P., Balslev, H., Forero, H. and Luteyn, J.L. (eds.). The New York Botanical Garden, Bronx, pp. 551–560. 21. Dourojeanni, M. 1989. Impactos ambientales del cultivo de la coca y la produccion de cocaina en la Amazonia Peruana. In: Pasta Ba´sica de Cocaı´na. Leon, F.R. and de la Mata, R.C. (eds.). CEDRO, Lima, pp. 281–299. (In Spanish). 22. Schulberg, S. 1994. Efectos ambientales y ecologicos de la industria de conversio´n de coca. In: Drogas, Violencia y Ecologı´a. Anicama, J. (ed.). CEDRO, Lima, pp. 344–361. (In Spanish). 23. Young, K.R. 1996. Threats to biological diversity caused by coca/cocaine deforestation in Peru. Environ. Conserv. 23, 7–15. 24. Vargas, M.R. 1999. Fumigacio´n y Conflicto: Politicas antidrogas y deslegitimacio´n del estado en Colombia. Tercer Mundo, Bogota´. (In Spanish). 25. Gonza´lez-Posso, D. 2000. Coca, deforestation and food security in the Colombian Amazon region. Unasylva 51, 32–36. 26. Forero, J. 2001. In the war on coca, Colombian growers simply move along. The New York Times, 17 March: A1. 27. Velaidez, R. 2000. Impacto de los cultivos ilicitos y las fumigaciones aereas con glifosato sobre el medio ambiente. In: Cultivos Ilicitos en Colombia. Castan˜eda, D., Castro, M.V. and Medina, A.A. (eds.). Bogota´: Ediciones Uniandes, Bogota´, pp. 149–155. (In Spanish) 28. Martin, P.S. and Szuter, C.R. 1999. War zones and game sinks in Lewis and Clark’s West. Conserv. Biol. 13, 36–45. 29. Dudley, J.P., Ginsberg, J.R., Plumptre, A.J., Hart, J.A. and Campos, L.C. 2002. Effects of war and civil strife on wildlife and wildlife habitats. Cons. Biol. 16, 319–329. 30. Schjellerup, I. (ed.) 2001. Wayko-Lamas-La Gente y la Biodiversidad. DIVA Tech. Report 9. Museo de Historia Natural de la UNMSM, Lima. (In Spanish). 31. Fjeldsa˚, J. and Rahbek, C. 1998. Continent-wide diversification processes and conservation priorities. In: Conservation in a Changing World. Mace, G.M., Balmford, A. and Ginsberg, J.R. (eds.). Cambridge University Press, Cambridge, UK, pp. 139–160. 32. Fjeldsa˚, J. 2001. Cartografiar la avifauna andina: una base cientı´ fica para establecer prioridades de conservacio´n. In: Bosques Nublados del Neotro´pico. Kappelle, M. and Brown, A.D. (eds.). INBio, Costa Rica, pp. 125–152. (In Spanish). 33. Balmford, A. 2002. Selecting sites for conservation. In: Conserving Bird Biodiversity. Norris, K. and Pain, D.J. (eds.). Cambridge University Press, Cambridge, UK, pp. 74–104. 34. James, A., Gaston, K.J. and Balmford, A. 2001. Can we afford to conserve biodiversity? BioScience 51, 43–52. 35. Pressey, R.L. and Nichols, A.O. 1989. Efficiency in conservation evaluation: scoring versus iterative approaches. Biol. Conserv. 50, 199–218. 36. Pressey, R.L., Humphries, C.J., Margules, C.R., Vane-Wright, R. and I and Williams, P.H. 1993. Beyond opportunism: key principles for systematic reserve selection. Trends Ecol. Evol. 8, 124–128. 37. Williams, P.H., Moore, J.L., Kamden Toham, A., Brooks, T., Strand, H., D’Amico, J., Wisz, M., Burgess, N., Balmford, A. and Rahbek, C. 2003. Integrating biodiversity priorities with conflicting socio-economic values in the Guinean-Congolian forest region. Biodiv. Conserv. 12, 1297–1320. 38. PNUFID-UNOPS. (no year). Mapa de Coca/Bosques Tropicales. Lima. (In Spanish). 39. Urrelo, G.R. 1995. Impacto ecolo´gico del cultivo de la coca en la amazonı´ a peruana y el mundo. In: Foro Internacional Coca, Ecologı´ a y Desarrollo en el Tro´pico Sudamericano. Univ. Nac. Agraria de la Selva (UNAS), Tingo Marı´ a, Peru´, pp. 171–178. (In Spanish). 40. Articles in the Lima newspaper Caretas by journalist E. Zileri. 41. Picasso, B.M., Ya´nez, B.C., Brackelaire, V. and de la Torre, F. (eds.) 1997. Tierras y Areas Indı´genas en la Amazonı´a. Tratado de Cooperacio´n Amazo´nica, Lima. (In Spanish).
Ambio Vol. 34, No. 3, May 2005
42. Salm, H., Lorini, J. and Liberman, M. 1990. Evaluacio´n Ecolo´gica del Cultivo de la Coca en los Yungas de La Paz. Estudio de Impacto Ambiental. Centro de Estudios Ecolo´gicos y de Desarrollo/Liga de Defensa del Medio Ambiente (US Agency for International Development, CEEDI/LIDEMA/USAID-Bolivia, La Paz. (In Spanish). 43. Martı´ nez, J.A. (ed.) 1999. Atlas Territorios Indı´genas en Bolivia. Situacio´n de las Tierras Comunitarias de Origen (TCOs) y proceso de titulacio´n. CIDOP-CPT, La Paz. (In Spanish). 44. BirdLife International. 2000. Threatened Birds of the World. Lynx Edicions and BirdLife International, Barcelona and Cambridge, UK. 45. Williams, P.H. 1998. Key sites for conservation: area-selection methods for biodiversity. In: Conservation in a Changing World. Mace, G.M., Balmford, A. and Ginsberg, J.R. (eds.). Cambridge University Press, Cambridge, UK, pp. 211–249. 46. Boucher, D.H. 1991. Cocaine and the coca plant. Bioscience 41, 72–76. 47. Kessler, M., Herzog, S.K., Fjeldsa˚, J. and Bach, K. 2001. Species richness and endemism of plant and bird communities along two gradients of elevation, humidity, and land use in the Bolivian Andes. Divers. Distrib. 7, 61–77. 48. Mazariegos, L.A. and Salaman, P. 1999. Rediscovery of the colourful puffleg Eriocnemis mirabilis. Cotinga 11, 34–38. 49. Bates, J.M. and Demos, T.C. 2001. Do we need to devalue Amazonia and other large tropical forests? Divers. Distrib. 7, 249–255. 50. Isler, M.L., Isler, P.R. and Whitney, B.M. 1999. Species limits in antbirds (Passeriformes: Thamnophilidae): the Myrmotherula surinamensis complex. Auk 116, 83–96. 51. Kaliola, R, Puhakka, M. and Danjoy, W. 1993. Amazonia Peruana. Universidad de Turku, Turku, and ONERN, Lima. 52. Cuervo, A.M., Salaman, P.G.W., Donegan, T.M. and Ochoa, J.M. 2001. A new species of piha (Cotingidae: Lipaugis) from the Cordillera Central of Colombia. Ibis 143, 453– 468. 53. Alverson, W.S., Rodrı´ guez, L.O. and Moskovitz, D.K. (eds.). 2001. Peru´: Biabo Cordillera Azul. Rapid biological Inventories 02. The Field Museum, Chicago. (In Spanish). 54. Herzog, S.K., Fjeldsa˚, J., Kessler, M. and Balderrama, J.A. 1996. Ornithological surveys in the Cordillera Cocapata, depto. Cochabamba, Bolivia, a transition zone between humid and dry intermontane Andean habitats. Bull. Brit. Orn. Club 119, 162–177. 55. Oates, J.F. 1999. Myth and Reality in the Rainforest: How Conservation Strategies Are Failing in West Africa. California University Press, Berkeley. 56. DFID, EC, UNDPB, World Bank. 2003. Linking Poverty Reduction and Environmental Management. World Bank, Washington, DC. 57. Bolivia 1994. Plan General de Desarollo Econo´mico y Social de la Repu´blica. El Cambio para Todos. Ministerio de Desarollo Sostenible y Medio Ambiente, La Paz. (In Spanish). 58. Robledo-Castillo, J.E. 1999. Neoliberalismo y desastre agropecuario. Deslinde 25, 32–49. (In Spanish). 59. Rocha, G. 1999. Estado del medio ambiente en latine America y el Caribe: la calidad de vida. In: Drogas, Violencia y Ecologı´a. Anicama, J. (ed.). CEDRO, Lima, pp. 373–384. (In Spanish). 60. UNDCP. 2000. Alternative Development in the Andean Area. The UNDCP Experience. UNDCP, Vienna. 61. European Parliament 2001. Resolution on Plan Colombia and Support for the Peace Process in Colombia. B5-00871. 1 Feb. 2001. 62. Lazare, D. 2001. A battle against reason, democracy and drugs: the drug was deciphered. NACLA Report on the Americas 35, 13–17. 63. Fjeldsa˚, J., Lambin, E. and Mertens, B. 1999. Correlation between endemism and local ecoclimatic stability documented by comparing Andean bird distributions and remotely sensed land surface data. Ecography 22, 63–87. 64. This analysis was done as part of the European Union-funded BioAndes Project (ERBIC18CT980299), which focused on how biodiversity conservation in the Andes was affected by recent political shifts. 65. Acknowledgments: A nonexhaustive list of people who have helped in this pursuit: the man in the Wellington boots, G. Andrade, L.G. Baptiste, J. Bates, A.M. Bejarano, S. Carrizosa, ADR, EDA, M. Leo, M.V. Llorente, C. McIlwaine, G. Martin, A. Millington, C. Padoch, N. Peluso, M. Pinedo, A.L. Porzecanski, A. Reyes, M. Rodrı´ guez, K. Redford, E. Sanderson, V. Tafur, and K. Willett. Sources for bird distributions are mentioned by (64) and the principal recent addition is data from Colombia from P. Salaman. Søren Hvalkof, Mariela Leo, Andy Millington, P. Salaman, B. Torres, and S. Vehkama¨ki are thanked for inspiring discussions. Mikko Pyhala provided information from Peru and from European Union policy meetings. 66. First submitted 3 July 2003. Revised manuscript received 20 Jan. 2004. Accepted for publication 23 Jan. 2004.
Jon Fjeldsa˚ is professor of biodiversity at the Zoological Museum, University of Copenhagen, where he works mainly with large-scale variation in biodiversity and proactive conservation planning. His address: Zoological Museum, Universitetsparken 15, DK-2100 Copenhagen, Denmark. [email protected]
Marı´a D. A´lvarez is a student at Columbia University in New York City. Her work on environmental consequences of armed conflict and illicit crops has been reviewed in Trends in Ecology and Evolution and New Scientist. Her address: Columbia University MC 5557, New York, NY 10027-5557, USA. Juan Mario Lazcano is a socioeconomist, studying relationships between macropolitical change and land use around Carrasco National Park, Bolivia. His address: Centro de Biodiversidad y Genetica, Universidad Mayor de San Simo´n, Cochabamba, Bolivia. Blanca Leo´n is a plant taxonomist. Her research interests are neotropical fern systematics, biogeography, and conservation, with a focus on the humid montane forests of Peru. Her address: Museo de Historia Natural UNMSM, Av. Arenales 1256, Apto. 14-0434, Lima-14, Peru.
Ó Royal Swedish Academy of Sciences 2005 http://www.ambio.kva.se