• Open Access

Wildlife decline in Cambodia, 1953–2005: exploring the legacy of armed conflict

Authors


Correspondence
Colby Loucks, World Wildlife Fund, 1250 24th St. NW, Washington, DC 20037, USA. Tel: +202-778-9671; fax: +202-293-9211. E-mail: colby.loucks@wwfus.org

Abstract

Armed conflict is a social phenomenon often detrimental to wildlife and wildlife habitat, but the legacy of armed conflict for wildlife in post-conflict settings remains unexplored. We explore the effects of armed conflict on wildlife in eastern Cambodia, particularly the relationships among wildlife abundance, habitat loss, technological change, trade, governance, and local livelihood strategies. Based on ordinal-scale measures, both relative wildlife abundance and species richness declined from pre-1953 to 2005, with the sharpest declines occurring during the 1970s. These declines are consistent with three synergistic social processes: proliferation of guns; emergence of wildlife trade for external markets; and government policies mandating hunting by local villagers. Armed conflict officially ceased in 1991, but conflict-induced changes in livelihoods and continuing non-local demand for wildlife have fostered further wildlife declines. Our study demonstrates that the legacy of conflict for wildlife can be profound and destructive. To address post-conflict challenges more effectively, conservation must be integrated within broader peace-building processes, including disarmament, demobilization, and reintegration of combatants.

Introduction

Armed conflict commonly occurs in areas of high biodiversity. Since World War II, the world has witnessed 152 armed conflicts, including 127 civil wars in 71 counties (Fearon & Laitin 2003). These civil wars have caused social, political, and ecological upheaval in many biologically diverse countries. For example, 64 of the 206 countries harboring priority ecoregions for biodiversity conservation globally (Olson & Dinerstein 1998) have experienced civil war in the past 60 years (Fearon & Laitin 2003). Thus, armed conflict represents an important—but often underappreciated—sociopolitical phenomenon shaping biodiversity and biodiversity conservation.

Armed conflict is often detrimental to wildlife and wildlife habitat (Dudley et al. 2002; McNeely 2003; de Merode et al. 2007). Negative impacts result from direct military activities (e.g., habitat loss from widespread herbicide use; wildlife mortality from landmines); resource use intended to support military efforts (e.g., logging to raise funds for arms; hunting to feed soldiers); opportunistic poaching; direct targeting of conservation professionals by armed factions; and secondary effects of the direct ecological and social impacts of armed conflict (e.g., pest outbreaks resulting from ecological disturbance; increased exploitation of wildlife by displaced peoples) (Dudley et al. 2002; McNeely 2003). In the Democratic Republic of Congo, for example, armed conflict led to a 500% increase in sales of protected wildlife species and a 200% to 800% increase in poaching (de Merode & Cowlishaw 2006; de Merode et al. 2007). Colombia presents a more complex picture: civil war displaced conservation-minded local communities and encouraged conversion of lowland forests, but simultaneously limited the expansion of human settlements and conversion of intact forests in mountain regions (Dávalos 2001). Similar benefits to wildlife have been observed in demilitarized zones between warring states, such as historically “unoccupied” buffer areas (Keeley 1996; Martin & Szuter 1999) or contested administrative regions (e.g., Indian-administered Kashmir; Binoo 2006). These and other observed conservation benefits of conflict (e.g., reduced exploitation through declines in capital-intensive resource extraction, disarmament of local populations) tend to be inadvertent and accidental (McNeely 2003).

The legacy of armed conflict for wildlife in post-conflict societies remains unexplored. To fill this knowledge gap, we examine the impact of armed conflict on wildlife and wildlife exploitation in the tropical dry forests of eastern Cambodia, a global biodiversity hotspot with a complex political history (Kiernan 1996; Olson & Dinerstein 1998; Chandler 2000; Myers et al. 2000). Since 1900, Cambodia has been governed by seven distinct regimes, ranging from French colonial rule through authoritarian control to today's constitutional monarchy (Kiernan 1996; Chandler 2000) (Table 1). Armed conflict emerged in our study area in the late Sihanouk era (1953–1970), intensified during the Lon Nol (1970–1975) and Pol Pot (1975–1979) regimes, and persisted into the late 1990s. While the devastating impacts of armed conflict on Cambodian society are well known (Chandler 2000; Blunt & Turner 2005), we investigate the previously undocumented effects of conflict on Cambodian wildlife. We focus on the relationships among wildlife abundance, habitat loss, technological change, trade, governance, and local livelihood strategies from pre-1953 to 2005. Our findings highlight the enduring effects of armed conflict for wildlife in Cambodia, suggesting a pattern likely to repeat itself in other post-conflict biodiversity hotspots.

Table 1.  Species and political eras used in wildlife abundance and use interviews
SpeciesPolitical era
  1. aSpecies likely extirpated from the study area in the past 50 years.

  2. b“Red herring” species used to assess the accuracy and validity of our survey instrument (see methods).

  3. cSpecies whose historic range potentially overlaps with our study area, but were not identified as present in the interviews.

 1. Asian elephant (Elephas maximus)a1. French: before 1953
 2. Tiger (Panthera tigris)2. Sihanouk: 1953–1970
 3. Leopard (Panthera pardus)3. Lon Nol: 1970–1975
 4. Kouprey (Bos sauveli)a4. Pol Pot: 1975–1979
 5. Banteng (Bos javanicus)5. People's Republic of
 6. Gaur (Bos gaurus) Kampuchea-State of
 7. Eld's deer (Cervus eldii)a Cambodia (PRK-SOC):
 8. Sambar deer (Cervus unicolor) 1979–1993
 9. Hog deer (Axis porcinus)a6. Kingdom of Cambodia
10. Red Muntjac (Muntiacus muntjak) (KOC): 1993–Present
11. Sun bear (Helarctos malayanus) 
12. Dhole (Cuon alpinus) 
13. Black-shanked douc langur (Pygathrix 
 nigripes) 
14. Asiatic jackal (Canis aureus) 
15. Malayan pangolin (Manis javanica) 
16. Sarus Crane (Grus antigone) 
17. Lesser adjutant (Leptoptilos javanicus) 
18. Siamese crocodile (Crocodylus siamensis)a 
19. Javan rhinoceros (Rhinoceros sondaicus)b 
20. Khting vor (Pseudonovibos spiralis)b 
21. Serow (Naemorhedus sumatraensis)b 
22. Clouded leopard (Neofelis nebulosa)c 
23. Large Indian civet (Viverra zibetha)c 
24. Asiatic black bear (Ursus thibetanus)c 
25. Giant ibis (Thaumatibis gigantea)c 

Materials and methods

As follow-up to an earlier rapid geographic assessment (Oul & Cheam 2005), we measured local abundance and use of 25 wildlife species (Table 1). We used local ecological knowledge (Berkes et al. 2000; Turner & Berkes 2006) of hunters in the Sre Chis commune, a group of six ethnically diverse villages in northeast Cambodia with a resource-dependent relationship to the local environment dating to the 1930s (Figure 1) (Oul & Cheam 2005). We compared these data with remotely sensed land cover change data and with primary and secondary source information regarding wildlife use, gun availability, wildlife trade, local and national governance, local livelihood strategies, and other social data. Through “analytic narrative” (Bates et al. 1998), we used the available mix of qualitative and quantitative data to explain observed patterns and trends in wildlife abundance, richness, and use.

Figure 1.

Study area in eastern Cambodia.

Study site

The Sre Chis commune is an isolated collection of six villages encompassing approximately 50 km2 of residential and agricultural land in eastern Kratie province, Cambodia (Figure 1). In the 1930s, several Kraol families first settled the dry dipterocarp forests of the area. The resident population grew during Sihanouk and Lon Nol regimes, suffered from starvation under Pol Pot, then resumed gradual growth to its 2005 level of approximately 618 families (∼3,600 people) (Oul & Cheam 2005). Residents of the commune are primarily Kraol, but also include Thmon, Phnong, and Khmer ethnicities (Oul & Cheam 2005). Though traditional management regimes governed life within Sre Chis during its early years, state administration strengthened throughout the Sihanouk period and dominated local governance during the Pol Pot regime. A hybrid of state and traditional governance reemerged with the fall of the Pol Pot in 1979, with village chiefs and commune leaders sharing authority. Historically, rice cultivation, fishing, subsistence hunting, and collection of nontimber forest products were the primary livelihood strategies for local villagers, but hunting (for subsistence and income) is now paired with rice cultivation (for subsistence) as the predominant livelihood strategies (respondent interviews; Oul & Cheam 2005). Unique among the villages in this region, Sre Chis inhabitants were not translocated during the Pol Pot era, which preserved the multi-generational relationship between Sre Chis residents and their local environment (Oul & Cheam 2005).

Biodiversity abundance and use

During French colonial rule, the only comprehensive wildlife survey was Wharton's (1957) study of kouprey (Bos sauveli) ecology. Civil unrest and armed conflict precluded other wildlife surveys until the early 1990s (Martin & Phipps 1996), making population estimates for Cambodian wildlife largely unavailable until after 2000 (see Table S1).

In lieu of direct field observations throughout our study period, we interviewed a purposive sample of expert hunters (n= 27; age 35–85 years) in seven mixed-age focus groups during the spring of 2005. Each group deliberated upon the questions posed and provided a collective response. Interviews drew upon hunters' local ecological knowledge (Berkes et al. 2000; Steinmetz et al. 2006; Turner & Berkes 2006), a reliable method for collecting accurate data on wildlife abundance and use (Huntington 2000; Jones et al. 2008). We used ordinal scales to measure species abundance (four categories), distance traveled when hunting (four categories), and intended use (three categories) for 25 species (Table 1). We collected data across six political eras, ranging from pre-1953 to 2005 (Table 1).

Cambodian staff from World Wide Fund for Nature (WWF) (D.K. & C.N.) conducted interviews in Khmer, using photographs of each species or images of defining characteristics (e.g., horns) as a visual aide. Interviews proceeded species-by-species for all questions, rather than by political eras. We selected species based on potential endangerment, rarity, or use in trade. We intentionally excluded common species such as wild boar (Sus scrofa). To test the validity and accuracy of the responses, we included species whose historic distribution did not overlap with our study area (serow (Naemorhedus sumatraensis)); were extirpated from this region before the time frame of this study (Javan rhinoceros (Rhinoceros sondaicus)); or are fabricated species (khting vor (Pseudonovibos spiralis)). Consistent with accurate scientific knowledge of local fauna, respondents reported all of these “red herring” species as “absent” for all time periods. Finally, we chose to categorize time by political era, rather than by calendar years, because political eras were more easily referenced by focus group participants. Our analyses focused on the 18 species historically observed in the region (Table 1).

We measured species abundance in four categories of relative abundance: many individuals, some individuals, few individuals, and no individuals (see Table S2 for results). We measured distance traveled to find a given species based on distances provided to us by the villagers we interviewed. Villagers termed “near” as any distance less than 4 km from the commune; “moderate” as any distance between 4 and 10 km from the commune; and “far” as any distance beyond 10 km from the village. If villagers did not report finding a species beyond 10 km from the commune, we identified it as “not found” (Table S3). The distances traveled did not include distances traveled to get beyond village boundaries or agricultural areas immediately adjacent to the villages. We used repeated measures ANOVA to assess change in species abundance and distance traveled across time periods.

Finally, we collected information on whether the species hunted was traded locally within the commune, not traded at all, or if it was sold to external traders (Table S4). We define “external trade” as products delivered to consumers who are not members of the commune (e.g., soldiers, Khmer Rouge elites, wildlife traders).

Social data

In addition to interviewing local hunters regarding biodiversity abundance and use, we also interviewed them regarding current and past occupations and livelihood strategies, awareness of nearby wildlife sanctuaries, and household demographics. The complementary rapid geographic assessment (Oul & Cheam 2005) queried key informants regarding village history and governance; education and health; infrastructure; occupations, livelihoods, and income; natural resource use and management; and threats to biodiversity. We also reviewed published and unpublished secondary sources to determine trends in social factors known to shape wildlife populations, including technological change (i.e., gun availability), wildlife and environmental policies, wildlife trade, and livelihood strategies.

Land cover change

To measure land cover change, we compared a 1:1,000,000 scale vegetation cover map of Cambodia that was derived from 1958–59 aerial photographs (Legris & Blasco 1972) with a 2002 land cover classification we developed based on Landsat ETM imagery. To identify the land cover classes for 2002, we obtained Landsat ETM images from February 13, 2002. To assist with image classification, we also obtained GIS data on roads, villages, protected areas, and rivers. We used a 90-meter resolution SRTM digital elevation model (DEM) data to identify ridges, hills, or higher elevation areas, which have a greater likelihood of containing semi-evergreen or mixed deciduous forest types due to better soils and greater water retention. We obtained Landsat TM imagery from December 21, 1990 to backstop our analysis and provide additional information on land cover from earlier in the dry season. We also collected ground-truth data from field visits in January, February, and May 2005, and used the expert knowledge of WWF-Cambodia field staff to provide input on the classification. We used standard supervised classification techniques to assign land cover classes to each of 250 initial unsupervised classes. Due to limited access and resources to get into the remote portions of the study area, we were unable to collect a sufficient number of ground-truth points across the entire study area to conduct a statistically meaningful accuracy assessment of our land cover classification.

Results

Over the >50 years included in our study, wildlife species richness and abundance declined persistently. Ordinal-scaled mean species abundance declined continuously over time (Figure 2A, repeated-measures ANOVA F(7, 119) = 21.952, P < 0.001), with the greatest rate of change occurring between the Lon Nol and Pol Pot eras. Across the study period, species abundance declined in 14 of the 18 species, and the number of species represented in the “many” or “some” categories declined from 16 to 6. Five of the 18 species historically observed in the region have been locally extirpated: Asian elephant (Elephas maximus), kouprey, Eld's deer (Cervus eldii), hog deer (Axis porcinus), and Siamese crocodile (Crocodylus siamensis) (Tables S2–4). Mean travel distance for hunting increased across political eras (Figure 2B, repeated-measures ANOVA F(7, 119) = 10.348, P < 0.001), with the greatest increase occurring between the Sihanouk and Lon Nol eras. Across the study period, the mean hunting distance category for all species increased from “4–10 km” to “greater than 10 km.”

Figure 2.

Trends in wildlife abundance, travel distance, and trade for 18 assessed species in the Sre Chis commune. (A) Mean species abundance; (B) mean distance traveled; (C) total extant species, locally traded species, and externally traded species. Means and standard error are reported for (A) and (B). Based on repeated-measures ANOVA, means with different letters show evidence of a difference. Data correspond to political-era mid-points.

The number of species that Sre Chis villagers traded to external markets increased during the 1970s from one, gaur (Bos gaurus), to eight species—Asian elephant, banteng (B. javanicus), Eld's deer, hog deer, tiger (Panthera tigris), leopard (P. pardus), sun bear (Helarctos malayanus), and gaur (Figure 2C, Table S4).

These observed declines in wildlife richness and abundance cannot be explained by habitat loss. Intact forests within 10 km of the Sre Chis commune villages declined only 9.6% from 1959 to 2002 (Table 2), an average of 0.22% per year, well below the south and southeast Asia region's 0.83% per year forest loss rate from 1990 to 2000 and Cambodia's 1.1% per year forest loss over the same time period (Food and Agriculture Organization 2005). Previous analyses of land cover change in this region found only a 2.9% decline in forest cover between 1990 and 2002 (0.24% per year) (Loucks 2007).

Table 2.  Land cover classes and change for the study area. The 2002 land cover analysis is based on classifications of Landsat ETM imagery. The buffer distances chosen correspond to the distance classes used in the species use and abundance interviews
Land cover4 km buffer area (km2)% of total area10 km buffer area (km2)% of total area
  1. aNatural grasslands-scrub, water, and wetlands from 2002 land cover analysis were combined with intact forest to assess land cover change.

  2. bDegraded forest from 2002 land cover analysis were combined with village-agricultural mosaic to assess land cover change.

1959 Land cover assessment (Legris & Blasco 1972)
 Intact forest62.667.1373.688.5
 Village–agriculture mosaic30.732.9 48.511.5
2002 land cover analysis
 Intact forest54.358.2331.678.6
 Village–agriculture mosaic37.440.1 87.820.8
 Degraded forest 0.7 0.7  1.2 0.3
 Grasslands–scrub 0.4 0.4  0.5 0.1
 Water–wetlands 0.5 0.5  1.0 0.2
1950–2002 land cover change
 Intact foresta−7.4−7.9−40.5−9.6
 Village–agriculture mosaicb 7.4 7.9 40.5 9.6

Observed declines in wildlife richness and abundance are consistent with three synergistic social processes associated with armed conflict in our study area: proliferation of guns and their use by local villagers and military personnel to hunt wildlife; emergence of wildlife trade for external markets, including military forces; and Khmer Rouge policies that mandated commercial hunting by local villagers (Table 3; respondent interviews; Oul & Cheam 2005). During the French and Sihanouk eras (pre-1970), guns were either illegal or reserved for the urban elite (Wille 2006); Sre Chis villagers used crossbows and traps to hunt wildlife for household consumption (respondent interviews; Oul & Cheam 2005, 10) and, on a small scale, for valuable animal parts (e.g., ivory, skins, horns; respondent interviews). However, as armed conflict among governmental forces, the Khmer Rouge, and the Vietnamese engulfed Cambodia in the 1970s and 1980s, government policies to restrict gun availability were largely unsuccessful. During the 1970s, the Khmer Rouge gave guns to local hunters and paid them to hunt wildlife (C. Bruce, WWF—Cambodia, unpublished data). Eastern Kratie province was the scene of particularly heaving fighting during the PRK-SOC era: “weapons were everywhere,” which “led to local people and soldiers using guns for illegal hunting of wildlife” (Oul & Cheam 2005, 9). By 1991, an estimated 319,500–462,500 weapons were stockpiled nationally, with 136,000–200,000 soldiers and 250,000 militia trained in their use (Wille 2006). Though peace accords were signed in 1991, weapons remained in widespread circulation and armed conflict continued locally, creating opportunities for hunters and traders to exploit highly valued wildlife for external markets (respondent interviews; Oul & Cheam 2005, 9). Khmer Rouge insurgents in Kratie province were reintegrated into society by the end of 1998 and, in 2000–2001, local authorities began to remove guns from general circulation, causing declines in wildlife hunting (Oul & Cheam 2005, 9–10, 25). Highly valued wildlife are still hunted for external markets, however, using traps, crossbows, axes, dogs, and guns that villagers have hidden or borrowed from police and soldiers in a nearby commune (respondent interviews; Oul & Cheam 2005, 10, 23).

Table 3.  Timeline of gun policy and availability, hunting and environmental policies, wildlife markets and trade, and livelihoods for the Sre Chis commune in Cambodia
 French: Before 1953Sihanouk: 1953–1970Lon Nol: 1970–1975Pol Pot: 1975–1979People's Republic of Kampuchea-State of Cambodia (PRK-SOC): 1979–1993Kingdom of Cambodia (KOC): 1993–Present
Sre Chis Livelihoods (respondent interviews; Oul & Cheam 2005)Primary: Rice cultivation
Secondary: Fishing, non-timber forest product (NTFP) collection, subsistence wildlife hunting
Primary: Rice cultivation
Secondary: Fishing, NTFP collection, subsistence wildlife hunting, raising livestock
Primary: Rice cultivation,
Secondary: Commercial and subsistence wildlife hunting, fishing, raising livestock, NTFP collection
Primary: Rice cultivation, commercial and subsistence wildlife hunting
Secondary: Fishing, raising livestock, NTFP collection
Primary: Rice cultivation, commercial and subsistence wildlife hunting
Secondary: Fishing, raising livestock, NTFP collection
Primary: Rice cultivation, commercial and subsistence wildlife hunting
Secondary: Fishing, raising livestock, NTFP collection
Hunting Policy (Oul & Cheam 2005, 9)   Sre Chis designated as a "hunting area"-Dissolution of “hunting area” designation
-Hunting wildlife outlawed (1988)
 
Wildlife Markets & Trade (Kiernan 1996; Oul & Cheam 2005)Exclusively localPredominantly localLocal & NationalLocal, National, and InternationalLocal, National, and InternationalLocal, National, and International
Gun Policy (Oul & Cheam 2005; Wille 2006)Nationally: Strict licensing; hunters could possess a single gunNationally: Legal guns strictly regulatedNationally: Legal guns strictly regulatedNationally: Private gun ownership eliminated Sre Chis: 2000-2001 local authorities remove guns from general circulation
Nationally: In 2005 Cambodia revises gun law, strictly limiting civilian ownership of guns.
Gun Availability (Oul & Cheam 2005, 9, 25; Wille 2006; respondent interviews; C. Bruce, WWF-Cambodia, unpublished data)Nationally: Gun ownership rare due to laws and expenseNationally: Restricted to urban elite within government; Increasing number of illegal guns in rural societyNationally: Legal guns restricted to soldiers and urban elite; Increasing number of illegal guns in rural societySre Chis: Khmer Rouge gave guns to hunters; paid them to hunt
Nationally: Firearms prevalent among Khmer Rouge political cadre and soldiers; gun supply increases throughout country
Sre Chis: "weapons were everywhere"
Nationally: guns stockpiled (319,500-462,500); firearms widespread throughout Cambodia
Sre Chis: villagers borrow guns or use hidden guns to hunt wildlife; "relatively easy access to guns"
Nationally: Guns collected & destroyed; legal firearms limited, but found throughout Khmer society (22,000-85,000) through blackmarket trade
Environmental Policies     -National protected area system created
-Ministry of Environment created
-National Environmental Action Plan adopted
-International conventions ratified (CITES, CBD, Ramsar, World Heritage)
-Forestry Law ratified
-National Biodiversity Strategy and Action Plan produced

Concurrent with the proliferation of guns across Cambodia, external wildlife markets emerged in Sre Chis (Table S4). Indeed, while prohibiting private gun ownership, the Pol Pot regime simultaneously traded wildlife for guns and other military supplies (Kiernan 1996). The 1970s saw a six-fold increase in the number of species hunted for external trade in our study area, a level of external trade that has remained consistent throughout subsequent political eras. The nature of external trade has shifted, however, from military camps and frontline soldiers under Pol Pot (respondent interviews) to regional and international markets (Oul & Cheam 2005, 18, 25). By at least 2005, traders were buying wildlife for external markets from Sre Chis hunters on a daily basis (respondent interviews). Products for sale included antlers, horns, dried meat, (tiger) bone, and tortoises (Oul & Cheam 2005; 10, 16, 23).

The proliferation of guns, combined with emerging wildlife markets and new government policies, led to profound changes in local livelihood strategies in the 1970s. Rice cultivation was the primary livelihood strategy in Sre Chis prior to 1970; fishing, collection of nontimber forest products, and hunting wildlife were secondary sources of subsistence (Table 3) (respondent interviews, Oul & Cheam 2005). During the Lon Nol era, with the emergence of external wildlife markets, wildlife hunting for income became an important secondary livelihood strategy. Wildlife hunting gained its current position alongside rice farming as a principal livelihood strategy in Sre Chis during the Pol Pot regime, when policymakers divided commune residents into work groups, including “hunting teams,” to provide food locally, for military camps, and for the front lines (respondent interviews; Oul & Cheam 2005, 9). Hunting targeted large mammals, including elephant, guar, banteng, sambar, and Eld's deer (respondent interviews). After the fall of the Pol Pot regime in 1979, hunting remained a principal livelihood strategy and source of income for Sre Chis residents (respondent interviews; Oul & Cheam 2005, 9). Today, species hunted for external trade are less abundant and found further from the Sre Chis commune (χ2(3, n= 18) = 31.21, P < 0.001) than those species that were never hunted for external trade (χ2(3, n= 18) = 9.00, P < 0.029). Asian elephant, banteng, pangolin (externally traded); red muntjac (locally traded); and sarus crane (not traded) have declined in abundance since 1979 (Table S4).

Discussion

Our findings are consistent with previous studies demonstrating negative impacts of armed conflict on wildlife resulting from resource use intended to support military efforts (Dudley et al. 2002). Unlike other studies on armed conflict and conservation, however, our research highlights the social shifts induced by conflict: widespread technological change (i.e., proliferation of guns); emergence and growth of external wildlife trade; and policy changes (i.e., establishment of “hunting teams”). These changes led to the emergence of commercial hunting as a new and perhaps lasting primary livelihood strategy, which drove wildlife declines during the conflict period and continues to threaten wildlife in the post-conflict era. Moreover, the potential for post-conflict recovery of wildlife is hindered not only by depauperate wildlife populations (Duckworth & Hedges 1998; Leimgruber et al. 2003; McShea et al. 2005) and the prevalence of commercial wildlife hunting (Singh et al. 2006), but also by the conflict-induced loss of administrative capacity and functional governance systems. Cambodia has recently adopted several national policies that directly or indirectly conserve wildlife (e.g., establishing a protected area system, implementing a national gun confiscation program, developing a national biodiversity strategy) (Table 3) (Oul & Cheam 2005), but these post-conflict conservation efforts have met with limited local success. By elucidating social, economic, and political mechanisms through which armed conflict can drive wildlife declines, we lay the scientific foundation for addressing these threats to biodiversity. Because shifts in livelihood strategies may endure long after conflict ends, these secondary impacts may represent a greater threat to wildlife over the long term than the direct and immediate effects of conflict itself.

Addressing the complex, destructive legacy of armed conflict represents a daunting challenge. As a rule of thumb, we suggest that conservation investments in post-conflict societies should be integrated within and support broader peace-building efforts targeting combatants, noncombatants, civil society organizations, and the state. Current United Nations (UN) guidelines governing disarmament, demobilization, and reintegration (DDR) of combatants scarcely mention conservation of natural resources, despite the fact that many UN DDR operations occur in countries where highly valued wildlife, timber, minerals, and metals have been the cause of conflict or a vehicle to sustain the activities of armed factions (e.g., Liberia, Democratic Republic of Congo) (UN DDR 2008). In the 1990s, for example, UN DDR efforts in Angola failed partly because ex-combatants accessed diamond deposits to remilitarize and reinitiate fighting (Dzinesa 2007). In our study area, DDR of Khmer Rouge forces was not complete until 5–8 years after the original peace agreement of 1991, hindering local implementation of national governance reforms and precluding intensive conservation activities by civil society organizations for several years (Oul & Cheam 2005). Thus, as noted previously, Sre Chis villagers in post-conflict Cambodia used hidden and borrowed weapons to hunt endangered wildlife illegally and had extremely limited awareness of nearby state-sponsored wildlife sanctuaries or laws regulating hunting.

Well-designed DDR processes have the potential to benefit both broader peace-building efforts and biodiversity conservation objectives. Conservation provides an additional justification and possible source of resources for disarmament activities, which eliminate weapons as a resource for ex-combatants to use to hunt wildlife illegally and also prevent noncombatants from borrowing weapons for the same purpose. If disarmament and demobilization permits legitimate authorities to begin to (re)assert themselves in post-conflict societies, conservation organizations can help these legitimate authorities by building capacity within government agencies and civil society; promoting rule of law; encouraging participatory and transparent decision making; and supporting other activities that foster good governance within the conservation sector and beyond. As part of these broader moves for policy reform and good governance, post-conflict reconciliation processes may present windows of opportunity for progressive reforms of conservation laws and policies (McNeely 2003).

Moreover, conservation organizations can play a valuable role in the reintegration of ex-combatants, an area where DDR efforts often fail (Dzinesa 2007). First, as environmental technical experts and watchdogs, conservation organizations can help to ensure that post-conflict reintegration efforts do not encourage ex-combatants to engage in unsustainable natural resource exploitation, which—in weak states—can take often illegal forms that undermine peace, stability, and rule of law (and, of course, contribute or exacerbate biodiversity loss). Similarly, large-scale capital investments designed to absorb ex-combatants into the labor force (e.g., plantation forestry, industrial agriculture) may foster environmental degradation and marginalization of impoverished rural noncombatants. World Bank and UN Development Programme trust funds that support reintegration could, however, finance extensive environmental restoration that would serve multiple purposes: employment of both ex-combatants and noncombatants, enhanced delivery of ecosystem services to resource-dependent communities, critical habitat for wildlife, and reduced wildlife trade by providing alternative sources of income. Similarly, transboundary “peace parks” may represent an opportunity for both peace-building and biodiversity conservation (McNeely 2003; Ali 2007; Donald et al. 2007), though protected areas can sometimes exacerbate social conflict (West & Brechin 1991, Brechin et al. 2003; Blundell & Christie 2007).

Capacity-building at the community level may merit special emphasis in post-conflict societies, especially in biodiversity-rich rural areas. In these remote and perhaps unstable regions, weak states will rarely have the capacity to develop and sustain support for biodiversity conservation, making local involvement and community support critical to conservation success (Dudley et al. 2002). Participatory community-based natural resource management initiatives can empower local actors and strengthen local governance regimes, absorb ex-combatants into the labor force, and provide legal economic opportunities for ex-combatants and noncombatants alike. Taken together, these actions may limit the influence of external wildlife markets and other factors on local wildlife abundance. Indeed, it is these communities that have lived through war, and are thus best suited to know what it will take to create a path that sustains both society and biodiversity in peace.


Editor : Dr. Kendra McSweeney

Acknowledgments

We thank the villagers and hunters who took part in the surveys and gave of their time and knowledge. We thank T. Ricketts, M. Marschke, T. Christie, C. Bruce, C. Bradshaw, K. McSweeney, and three anonymous reviewers for comments that greatly improved the article. This study was partially funded by the National Aeronautics and Space Administration (NASA). M.B.M. was funded by the John D. and Catherine T. MacArthur Foundation.

Ancillary