Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Human–wildlife conflicts and management options in Bangladesh, with special reference to Asian elephants (Elephas maximus)

Human–wildlife conflicts and management options in Bangladesh, with special reference to Asian... International Journal of Biodiversity Science, Ecosystem Services & Management Vol. 6, Nos. 3–4, September–December 2010, 164–175 Human–wildlife conflicts and management options in Bangladesh, with special reference to Asian elephants (Elephas maximus) A.H.M. Raihan Sarker and Eivin Røskaft* Department of Biology, Norwegian University of Science and Technology (NTNU), Trondheim, Norway In this study, we explored the interaction between humans and Asian elephants (Elephas maximus). We analysed local mitiga- tion techniques and propose possible measures to reduce human–elephant conflicts. For four protected areas in Bangladesh, we investigated how people interact with elephants and other crop-damaging species. Crop-damaging species differed from site to site, but elephants were generally the most frequent offenders. Although some people were able to control crop dam- age caused by other wildlife, damage caused by wild elephants was mostly uncontrollable. Forest villagers were somewhat tolerant of damage caused by other wildlife but they were not tolerant of damage caused by wild elephants. Tolerance of crop damage caused by wildlife other than wild elephants tended to increase with distance of respondents from protected areas. Non-tolerance of crop damage caused by wildlife was more common in southeast Bangladesh than in the north. Habitat destruction, caused by high population growth and poverty, appeared to be a major cause for increasing human–elephant conflict. There is an urgent need for adoption of an umbrella strategy (e.g. estimate the size of elephant populations, develop landscape-scale conservation plans and create forest buffer zones under community-based natural resource management schemes) in the use of conflict mitigation as a conservation tool for Asian elephants. Keywords: human–elephant conflict; Bangladesh; attitudes; elephants; compensation 1991; Kangwana 1995; O’Connell-Rodwell et al. 2000). Introduction Elephants that traditionally lived in forests now frequently Human–wildlife conflict (HWC) is common where wildlife come out in groups or as scattered individuals and cause and human populations coexist and share limited resources financial and social losses (Sukumar 1989; Thouless 1994). (Schwerdtner and Bernd 2007), and has become a consider- The most frequently discussed losses are crop damage by able problem throughout the world (Wang and Macdonald elephants and injury or death of people (Hoare 2000). 2005). Conflicts are more intense where livestock holdings Elephants migrate seasonally to follow the accessibility of and agriculture are important parts of the rural livelihood water and favoured foods (Sukumar 1989). Their migra- (Ogra and Badola 2008). Crop raiding, property dam- tion patterns frequently expand not only beyond park age, livestock depredation and human casualties are the and reserve boundaries, but beyond national boundaries most common forms of conflict with wildlife (Inskip and as well (Hart and O’Connell 2000). This is true, for Zimmerman 2009). The most extreme form of human– example, for the Teknaf Game Reserve in Bangladesh wildlife conflict occurs when human lives are lost (Løe and the Alaungdaw Kathapa National Park in Myanmar, and Røskaft 2004; Packer et al. 2005; Gurung et al. 2008) South Asia (Personal communication, Mr. Nurul Kabir, and retaliation against wild animals follows (Treves and ACF, Bangladesh Forest Department). Therefore, the con- Karanth 2003). Retaliation, also for protection of livestock sequences of conflicts between humans and elephants are and agricultural crops, threatens the survival of wildlife not only a key concern for conservation, but also a major that comes into conflict with humans (Mishra et al. 2003). socio-economic and political issue because elephants move Large herbivores and carnivores are particularly affected great distances (Sukumar 1989). by this conflict and are either critically endangered or Elephant disturbances usually take place between dusk rapidly declining as a result (Woodroffe and Ginsberg and dawn and are strongly seasonal, corresponding with 1998). Human–elephant conflict is a key example of such crop harvesting periods (Sukumar 2004; Osborn and Hill an interaction. 2005). Farms adjacent to the park boundary are likely The human–elephant conflict arises from a range to be raided, especially during seasons (June–August and of direct and indirect negative interactions between October–December) when favoured foods (e.g. paddy) are humans and elephants, which can create negative atti- at the optimal stage of growth (Sukumar 1990). Conflicts tudes regarding the conservation of elephants (Røskaft are usually more severe close to protected areas (PAs), et al. 2007). Human pressures on elephants caused by which act as elephant refuges (Hoare 2000). Inevitably, the poaching and conflict over resources, as well as efforts to sharpest conflicts and greatest losses seem to occur where adjust the effects of elephants on vegetation and crops, villages are close to park borders (O’Connell-Rodwell et al. are widespread throughout Asia and Africa (Sukumar *Corresponding author. Email: roskaft@bio.ntnu.no ISSN 2151-3732 print/ISSN 2151-3740 online © 2010 Taylor & Francis DOI: 10.1080/21513732.2011.554867 http://www.infomaworld.com International Journal of Biodiversity Science, Ecosystem Services & Management 165 2000). Human–elephant conflicts arise in rural areas of Materials and methods Asia, posing significant problems to local communities, Study areas elephants and wildlife managers (Sitati et al. 2003). The This study was conducted in three PAs in south- conservation of the Asian elephant (Elephas maximus) east Bangladesh: Teknaf Game Reserve (TGR), Chunati is of both national and international significance because Wildlife Sanctuary (CWS), a reserve forest adjacent to it is considered to be a flagship species by international Dulahazara Safari Park (RF 2) and, in north Bangladesh, conservationist groups (e.g. the Asian Elephant Specialist a reserve forest in Sherpur district (RF 1) that lies between Group, AsESG) for the ecosystem it inhabits (Nath and the political border of Bangladesh and India (Figure 1). The Sukumar 1998). Therefore, resolution of human conflict TGR (IUCN category VI: a protected area managed mainly with elephants is a major concern and a high priority for for the sustainable use of natural ecosystems, IUCN 1994) conservation of elephants whose range includes countries was established in 1983 under the Cox’s Bazar (South) in Asia. ◦  ◦ Forest Division in southeast Bangladesh (21 04 N, 92 The elephant is a ‘Critically Endangered Species’ 09 E; total area 116.15 km ). TGR is part of a linear hill in Bangladesh, and it is included in Appendix 1 of range that gently slopes to rugged hills and cliffs; most the Convention on International Trade in Endangered of the streams are seasonal and dry up during the off- Species of Wild Fauna and Flora (CITES) and in the monsoon season. The CWS (IUCN category V: a protected Third Schedule of the Bangladesh Wildlife Preservation area managed mainly for landscape/seascape conserva- (Amendment) Act, 1974. Human settlement in forested tion and recreation, IUCN 1994) was established in 1986 areas, absence of tree camouflage inside the forest and and is managed by the Wildlife and Nature Conservation reductions in availability of food are considered to be Division. CWS is situated in southeast Bangladesh (21 the primary causes for the decrease in number of ele- ◦  2 58 N, 92 04 E, total area 77.64 km ). The sanctu- phants in the forests of Bangladesh (Khan et al. 2004). ary area is generally hilly with shallow to deep gul- Southeast Bangladesh is considered to be a primary lies and gentle to steep slopes, traversed by numerous conservation area for Asian elephants that provides a streams. RF 2 (IUCN category 1b: a protected area man- refuge for the existing population. An estimated 196– aged mainly for wilderness protection, IUCN 1994) is 227 elephants range over a fragmented landscape in surrounded by the Fasiakhali Range of the Cox’s Bazar Bangladesh (Sukumar 2006). However, the natural habi- ◦  ◦ (North) Forest Division (21 40 N, 92 04 E, total tat for the elephants in Bangladesh is rapidly diminish- area 27.1 km ) and was established in 1897. The topog- ing because more and more land is being cleared for raphy is undulating, and there are some natural peren- cultivation and development of infrastructure. Moreover, nial streams. RF 1 (IUCN category 1b) includes parts human encroachment on PAs in southeast and north ◦  ◦ of Jhenaigati (25 16 N, 90 01 E) and Nalitabari sub- Bangladesh has forced elephants to forage in non-protected ◦  ◦ district (25 13 N, 90 08 E) of Sherpur district and areas, thus increasing the disturbance caused to local was established in 2006. The total area of the RF 1 people. Therefore, with the dwindling populations of is 5.0 km , and it comprises a scattered secondary Sal Asian elephant, conflict between humans and elephants forest. has also become a progressively more serious issue in Bangladesh. Human–elephant conflict is a complex problem and Respondents a highly emotive issue that cannot be mitigated through reliance on a single mitigation technique (Sitati and In the early 1950s, the Bangladesh Forest Department Walpole 2006). Each field requires a specific deterrent at TGR, CWS and RF 2 allowed local landless and strategy. A range of mitigation techniques has proved poor people to live in these areas to protect the for- to be more effective in driving away raiders, but this est. We defined these inhabitants as local people because reportedly becomes less effective over time (Parker they were born in the areas and hold property rights and Osborn 2006). This is because, crop-raiding ele- from their ancestors. Similarly, indigenous tribal peo- phants soon learn to ignore deterrents and develop resis- ple who resided around RF 1 before the 1980s were tance to crop protection measures (Bandara and Tisdell also defined as local people. People who migrated to 2002). the study areas from the beginning of the 1980s to Ensuring the survival of Asian elephants in Bangladesh the present day and settled without government approval calls for an adequate and comprehensive understanding were defined as immigrants. Immigrants were reha- of their raiding patterns. This study aims to explore the bilitated around the study areas from other districts interactions between humans and wildlife in four PAs under the protection of various political and social in Bangladesh. An underlying assumption is that most groups and were often considered as strong potential human–wildlife interactions vary not only with season, but vote banks for political parties. However, in this study, also with distances of respondents’ homes from PAs. We we defined both local people and immigrants as ‘vil- also examined local mitigation techniques and their effec- lagers’. Sarker and Røskaft (Forthcoming) describe the tiveness, and propose possible remedies for reducing the demographic and socio-economic characteristics of the conflict between humans and elephants. respondents. 166 A.H.M. Raihan Sarker and E. Røskaft RF 1 RF 2 CWS TGR Source: Bangladesh Forest Department, 1999. Figure 1. Map showing the location of the four PAs studied in Bangladesh (RF 1 = Reserve forest in north Bangladesh, CWS = Chunati Wildlife Sanctuary, RF 2 = Reserve forest in southeast Bangladesh and TGR = Teknaf Game Reserve). Sampling and data collection sub-districts across the four study sites. We followed a structured questionnaire to reduce bias in the responses We prepared household maps in each PA using a GPS of respondents, while standardised questionnaires were up to a distance of 1 km from the park boundary. The used to ensure the reliability, generalisability and valid- demographic features of each village in the study area ity of the responses. A period of 45 days was spent in were collected from the local Union Council. We admin- each of the eight sub-districts, Teknaf (TGR), Lohagara istered a total of 388 standardised, structured and semi- and Banshkhali (CWS), Chokoria, Ukhia, and Ramu (RF structured questionnaires to 193 males and 195 females 2), Jhinaighati and Nalitabari (RF 1). The questionnaire over the age of 18 years. The participants were selected included socio-demographic variables and a set of closed randomly, and completion of questionnaires was facilitated (no/yes/don’t know) and open-ended questions related to through ‘face-to-face’ communication tools. The respon- perceptions of conflicts between humans and elephants dents were from 99 villages under 11 unions and eight International Journal of Biodiversity Science, Ecosystem Services & Management 167 (Box 1). Open-ended questions were asked at the end between perceptions and different independent variables of the structured portion to collect as much informa- were investigated using step-wise linear regression. tion as possible within the limited time (approximately 45 min per respondent) of the interview. Usually, house- hold heads (mostly males) were interviewed; however, in Results their absence, any member (mostly housewives) willing to Wildlife and crop losses participate was interviewed. The respondents were cho- Crop raiding by wild animals was an important issue in the sen for interview and categorised on the basis of their study areas, as almost two-thirds of respondents (64.1%) monthly income (e.g. <US$ 70 per month and >US$ 70 reported it as a significant cause of crop loss, while over per month), distance of home from a PA (e.g. <0.5 km, one-third (35.9%) reported natural disasters as most impor- 0.5–1.0 km and >1.0 km), land-holding status (e.g. 0 acre, tant (χ = 104.5, df = 3, N = 388, p < 0.0001). The 0.01–0.50 acre, 0.51–1.00 acre and >1.00 acre), occu- species that were responsible for crop damage differed sig- pation (e.g. farmer; other occupations such as business- nificantly from one site to another (χ = 129.0, df = 12, man, housewife, teacher, student, service provider, political p < 0.0001), but elephants were reported more frequently leader, etc. and daily labourer), education (e.g. illiterate or than other species (Table 1). Wild boar (Sus scrofa) ranked literate) and residential status (e.g. local or immigrant). second most likely species to be responsible for crop losses In this study, we defined both daily labourers and other around the PAs (Table 1). Southeast Bangladesh (i.e. CWS, occupations as ‘non-farmers’. TGR and RF 2) was more vulnerable to crop damage by wild elephants and other wildlife than north Bangladesh (i.e.RF1)(Table1). Box 1. Questions presented to respondents to assess their perceptions of human–elephant conflict Period of crop damage A. Which wild animals damage crops on your farm or homestead garden? Respondents claimed that wild elephants caused damage to B. Is crop-raiding by wild animals seasonal? crops mainly during the monsoon (June–September) and (Monsoon/post-monsoon/winter/all year) post-monsoon period (October–December), but this var- C. Are you able to control crop damage caused ied considerably between study sites (χ = 72.5, df = 9, by wildlife? (no/yes) p < 0.0001) (Table 2). Most of the crop damage occurred D. How do you look upon crop damage caused in north Bangladesh (i.e. RF 1) during the post-monsoon by wildlife? (non-tolerant/tolerant) period. In southeast Bangladesh (i.e. CWS, TGR and RF E. How do you deter wildlife from your field 2), crop damage caused by wild elephants was higher dur- and garden? ing the monsoon season and most frequent in and around RF 2. Only a few respondents claimed that wild boar F. Is this effective? (no/yes/do not know) (χ = 57.7, df = 9, p < 0.0001) and porcupine, Hystrix G. Do you think that human–elephant conflict indicus (χ = 10.5, df = 3, p < 0.015) raided crops at is increasing in your area? (no/yes) H. What is the main reason for the increase in any time of the year, but this varied considerably between conflict between humans and elephants? study areas (Table 2). There was no relationship between crop raiding birds and the period of crop damage across the PAs (χ = 7.4, df = 3, p < 0.060) (Table 2). Data analyses Ability to control crop damage Data were analysed using SPSS Version 16.0 (SPSS, Chicago, IL, USA). Perceptions were related to many Less than half of respondents (41.0%, N = 388) reported demographic variables. The significance level was set at that they were able to control crop damage caused by dif- p < 0.05. Differences between perceptions were tested ferent wildlife species, but this varied significantly between 2 2 using Pearson’s chi-square (χ ) test. The relationships study areas (χ = 87.8, df = 3, p < 0.0001). Among Table 1. Percentages of people who responded concerning crop damage caused by different wild animals in each of the PAs. Wildlife responsible for crop damage CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) None 6.5 10.1 60.0 11.4 35.8 Elephant 51.6 49.3 33.0 60.2 43.6 Wild boar 38.7 31.9 6.0 18.2 16.0 Porcupine 3.2 7.2 1.0 6.8 3.6 Birds 0.0 1.4 0.0 3.4 1.0 Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. 168 A.H.M. Raihan Sarker and E. Røskaft Table 2. Percentages of people who responded concerning the period of crop damage caused by different wild animals in each of the PAs. Wildlife responsible and the period of crop damage CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) Wild elephant None 51.6 52.2 67.0 40.9 57.2 All year 25.8 13.0 0.5 26.1 10.6 Monsoon 16.1 17.4 7.5 20.5 12.9 Post-monsoon 6.5 17.4 25.0 12.5 19.3 Wild boar None 61.3 68.1 94.0 80.7 83.8 All year 38.7 24.6 2.5 19.3 13.1 Monsoon 0.0 0.0 0.5 0.0 0.3 Post-monsoon 0.0 7.2 3.0 0.0 2.8 Porcupine None 96.8 92.8 99.0 92.0 96.1 All year 3.2 7.2 1.0 8.0 3.9 Birds None 100.0 98.6 100.0 96.6 99.0 All year 0.0 1.4 0.0 3.4 1.0 Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. local people, 51.3% reported that they were able to con- Mitigation of conflict between humans and elephants trol crop damage caused by different wildlife species, while Almost two-thirds of respondents reported that they used among immigrant populations, only 6.7% were able to active deterrent methods that involved scaring animals control crop damage (χ = 57.0, df = 1, p < 0.0001). with bright lights (i.e. setting fire to raw jute or tyres Among respondents (N = 169) who reported crop damage fixed at the end of bamboo sticks, locally known as caused by wild elephants, none stated that they could con- ‘bolems’, and torchlight) or loud noise (i.e. shouting, trol this damage. On the other hand, among respondents drumming and bamboo cracking), and in most cases, this (N = 157) who reported crop damage caused by other 2 varied statistically across the study areas (χ = 138.4, wildlife species, 17.2% claimed that they could control df = 21, p < 0.0001) (Table 3). In addition, villagers con- damage. structed thatched houses (huts) during the harvest season (especially of rice) from which to guard their fields at night. Such activities were always carried out in groups. Attitudes to crop damage Drumming, bamboo cracking and other means (e.g. use of Among all respondents, 25.8% (N = 388) were willing to buffalo to deter elephants, oil lamps and fencing made of tolerate damage caused by different wildlife species, but sharp thorns) were used less frequently, perhaps because this differed significantly between study sites (χ = 31.8, the more popular techniques are usually employed first. df = 3, p < 0.0001). Among local people, 32.9% were The majority of respondents agreed with the effectiveness tolerant of crop damage caused by different wildlife of traditional method(s) used to deter wildlife, especially species, while among immigrants very few respondents elephants, from crop fields, and this varied significantly (2.2%) reported that they would tolerate crop damage by 2 between study sites (χ = 55.3, df = 3, p < 0.0001) wildlife (χ = 33.9, df = 1, p < 0.0001). Moreover, of the (Table 3). We found that the traditional methods used to respondents who reported crop damage caused by wildlife deter elephants were time consuming, and this ultimately species such as birds, wild boar and porcupine (N = 157), increased the costs for farmers because, despite the deter- only a small proportion (8.3%) were tolerant, whereas rent, elephants still trampled crops until they moved on respondents (N = 169) who reported crop damages caused from a field. However, more than one-third of respondents only by wild elephants had absolutely no tolerance of reported that they were not using any active methods to such damages. Non-tolerance of crop damage was more deter elephants from crop fields. Among respondents who common in TGR (92.8%, N = 69), CWS (83.9%, N = 31) reported using no active deterrence methods, 0.5% lived and RF 2 (83.0%, N = 88) than in RF 1 (62.5%, N = 200) within 0.5 km of a park boundary, while 55.3% lived (χ = 31.8, df = 3, p < 0.0001). The level of tolerance 0.5 km to 1 km and 84.5% lived more than 1 km from of respondents increased significantly with the distance the park boundary (χ = 243.3, df = 14, N = 388, of their household from the PA boundary (χ = 273.8, p < 0.0001). df = 2, p < 0.0001). Respondents who lived within 0.5 km of the park boundary showed absolutely no tolerance People’s experience with conflict for crop damage caused by wildlife, whereas a small proportion of respondents living 0.5–1.0 km (14.0%) from Almost half of respondents claimed that conflict between a park boundary were tolerant of crop damage caused by humans and elephants had increased in their villages, wildlife other than elephant, and the proportion of tolerant and this differed significantly between the studied PAs respondents (86.0%) increased among those living beyond (χ = 54.8, df = 3, p < 0.0001) (Table 4). Human con- 1 km of the park boundary. flict with elephants was reported to be significantly higher International Journal of Biodiversity Science, Ecosystem Services & Management 169 Table 3. Percentages of people who responded concerning actions taken and their effectiveness for deterring wildlife from entering fields in each of the PAs included in the study. Deterrent techniques CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) None 6.5 11.6 60.0 11.4 36.1 Bolem 19.4 17.4 14.0 22.7 17.0 Torchlight 19.4 15.9 10.0 18.2 13.7 Shouting 25.8 17.4 2.5 22.7 11.6 Drumming 12.9 8.7 4.0 3.4 5.4 Bamboo cracking 0.0 2.9 3.0 4.5 3.1 Guarding from huts 6.5 1.4 1.5 6.8 3.1 Others 9.7 24.6 5.0 10.2 10.1 Effectiveness of action No 19.4 50.7 9.0 26.1 21.1 Yes 80.6 49.3 91.0 73.9 78.9 Notes: Use of buffalo, oil lamps and fencing with sharp thorns. CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. Table 4. Percentages of people who responded concerning whether the magnitude of human conflict with elephants is increasing and the reasons, and opinions of respondents for reducing such conflict in each of the PAs included in this study. Conflict increases CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) No 19.4 26.1 67.5 40.9 50.3 Yes 80.6 73.9 32.5 59.1 49.7 Reasons of conflict Habitat destruction 48.4 37.7 56.0 39.8 48.5 Increase in human 6.5 39.1 24.5 34.1 27.8 population and poverty Reserve size is too 38.7 13.0 10.0 17.0 14.4 small Crop raiding by 6.5 10.1 9.5 9.1 9.3 wildlife Options to reduce conflict Do not know 0.0 0.0 8.5 1.1 4.6 FD should respond in 45.2 27.5 35.5 29.5 33.5 time to deter elephants Need to introduce 41.9 60.9 27.5 59.1 41.8 compensation scheme Need to improve 3.2 5.8 18.5 8.0 12.6 elephant habitat Create buffer zone 9.7 5.8 10.0 2.3 7.5 through community involvement in a forestry programme Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. in southeast Bangladesh (i.e. CWS, TGR and RF 2) than reduce the conflict between humans and elephants, respon- in northern parts (i.e. RF 1). Respondents’ identifica- dents proposed several options that varied across study tion of reasons behind such conflict varied significantly sites (χ = 55.2, df = 12, p < 0.0001) (Table 4). between study sites (χ = 31.9, df = 9, p < 0.0001) Most respondents proposed introducing a compensation (Table 4). Almost half of respondents identified habi- scheme to protect against crop loss and demanded active tat destruction as a major cause of increasing human– support from the Forest Department to reduce problems elephant conflict, while more than one-quarter cited human associated with wild elephants. Very few respondents population growth and poverty as root causes. Very few emphasised improvement of elephant habitat to a sustain- respondents reported the size of the reserve or wildlife able level or adoption of social forestry as management crop raiding as a cause of human–elephant conflict. To strategies. 170 A.H.M. Raihan Sarker and E. Røskaft Predictors of interactions between humans and wildlife porcupine and birds, than respondents who lived further away from the PAs. Similarly, immigrants reported more Crop damage by elephants crop losses than local people. A step-wise linear regression analysis examining variations in perceptions of respondents towards whether elephants cause crop damage as dependent variable was tested with The ability to control crop damage caused by wildlife five independent variables (Table 5); only four of which proved to be significant contributors to the variation. The We tested the perception of respondents towards their abil- variable explaining most variation in this perception was ity to control crop damage caused by wildlife as a depen- distance of respondents’ homes from the park boundary. dent variable, with five independent variables (Table 5). The occupation of the respondent was the second most All independent variables explained 62.2% of the variation important predictor. Settlement status and land-holding on ability to control the crop damage caused by wildlife. status of respondents were other significant predictors The variable explaining most variation was the distance of accounting for most of the variation in this perception. respondents’ homes from the park boundary. Occupation The education status of respondents was not significant. and land-holding status were the second and third most All independent variables explained 51.0% of the varia- important predictors, respectively. Settlement status and tion in people’s perception of crop damage caused by wild education were not significant. Respondents who lived elephants. The analysis indicated that villagers who lived closer to the park reported less ability to control crop dam- closer to the park described higher levels of crop damage age caused by wildlife than those who lived further away caused by elephants than those who lived further away. from PAs. Farmers expressed least ability to control crop Farmers reported higher levels of crop damage by wild damage. Similarly, landless people, rather than landowners, elephants than non-farmers (e.g. daily labourers and other also reported least ability to control crop damage. occupations). Immigrants reported a higher level of crop loss than local people, and landless people reported even higher levels of crop loss than land-owners. Attitudes towards crop damage A step-wise linear regression with perception of crop dam- age caused by wildlife as dependent variable was used Crop damage by other wildlife species to test how much of the variation was explained by the The perception of whether other wildlife species (e.g. wild five independent variables (Table 5). Only two of the boar, porcupine and birds) cause crop damage at respon- tested variables proved to be significant contributors to dents’ farms or homestead gardens was tested with respect explain the variation. The most explanatory variable was to the five demographic variables. The step-wise linear the distance of respondents’ homes from the park bound- regression showed that the distance of respondents’ homes ary. Education was the second most important predictor. from the park boundary and the settlement status of respon- Settlement status, occupation and land-holding status were dents were the two most significant factors contributing to not significant. All independent variables explained 62.0% the variation in perception (Table 5). Education, occupation of the variation in tolerance of people towards crop damage and land-holding status of respondents were not signif- caused by wildlife. Respondents who lived further from icant. All independent variables explained 14.0% of the the reserve had a higher level of tolerance towards crop variation in people’s perception of crop damage caused by damage caused by wildlife than those who lived closer other wildlife. The analysis showed that respondents who to the park. Such attitudes were governed by the level of lived closer to the park boundary reported higher crop dam- education, since literate people showed a greater tolerance age caused by other wildlife species, such as wild boar, to crop damage than illiterate people. Table 5. Results of stepwise linear regression with perceptions as dependent variables in relation to various independent variables (R = rank, t = t-value). Crop raiding by Crop raiding by Ability to control Tolerance to crop elephant other wildlife crop damage damage Increasing HEC Independent variables Rt p-Value Rt p-Value Rt p-Value Rt p-Value Rt p-Value Distance of respondents’ 1 −12.8 0.0001 1 −5.2 0.0001 1 19.5 0.0001 1 23.9 0.0001 1 −15.6 0.0001 homes from park boundary Occupation 2 −3.9 0.0001 4 0.9 0.362 2 3.3 0.0001 4 −0.8 0.416 2 −5.3 0.0001 Settlement status 3 3.6 0.0001 2 3.1 0.002 4 −1.5 0.128 3 1.8 0.076 4 1.3 0.195 Land 4 −2.8 0.005 5 −0.3 0.776 3 3.3 0.0001 5 0.1 0.917 3 −3.1 0.002 Education 5 −1.4 0.173 3 −1.8 0.079 5 0.4 0.682 2 2.9 0.004 5 0.5 0.607 Constant 3 11.14 0.0001 5 4.08 0.0001 4 −9.72 0.0001 2 −14.47 0.0001 1 23.30 0.0001 R 0.510 0.0001 0.140 0.0001 0.622 0.0001 0.620 0.0001 0.528 0.0001 International Journal of Biodiversity Science, Ecosystem Services & Management 171 The perception of whether conflict is increasing local level (Dudley et al. 1992), their regional impact on agriculture is insignificant compared to other vertebrate We employed step-wise linear regression for the percep- and invertebrate pests (Naughton et al. 1999). Elephants tion of respondents on the magnitude of human conflict depredate at a high intensity mainly for just a few months with elephants in villages closer to PAs as dependent vari- of the year; the damage caused by wild boar and porcu- able to test how much of this variable could be explained pine is reported to be less, but it is continuous. The general by the five independent variables (Table 5). The variable impression given by farmers in the study areas is that ordi- explaining most variation was distance of respondents’ nary agricultural pests such as wild boar and porcupine can homes from the park boundary. Occupation was the second easily be managed with minimal effort and cost compared most important predictor. The third was land-holding sta- to wild elephants, which are more destructive and not easy tus, while settlement status and education of respondents to control. Therefore, some villagers residing in the study were not significant. All independent variables explained sites had tolerant attitudes towards crop damage caused by 52.8% of the variation in people’s opinions regarding the wildlife other than elephants. status of conflict between humans and elephants. Similar to There are seasonal fluctuations in crop raiding other cases studied, the respondents who lived closer to PAs that coincide with food availability and crop maturity reported a higher level of conflict with elephants than those (Marchais 2004; Mosojane 2004). In our study, we found who lived further away from the park boundary. Farmers that crop damage by wild elephants was reported more reported a higher level of conflict than non-farmers, and during the monsoon and post-monsoon season. Villagers landless people reported a higher level of conflict than reported that rice was the principal crop consumed by ele- landowners. phants, which is harvested during the monsoon and post- monsoon season. Since much of the cultivation around CWS and RF 2 occurs during the monsoon season, the Discussion results indicate that elephants are attracted to the crops Factors contributing to crop raiding during harvest time, and hence, there is a greater ten- dency for conflict between humans and elephants at this The study revealed that crop raiding by wildlife was a time than during the post-monsoon season. Reports of major cause of crop loss from farmers’ fields. Farms near crop damage in villages around the TGR were almost a forest or PA boundary, a water source or a migra- tion route suffer a disproportionate amount of damage the same in both seasons, since the TGR is a fundamen- (Naughton et al. 1999; Mosojane 2004). The attraction to tal habitat for Asian elephants in Bangladesh. The area crops increases due to the availability of water, which is surrounding RF 1 is a flood prone zone (Banerjee, unpub- relatively scarce in some seasons, particularly in the dry lished) where inhabitants are involved in illegal removal season. Artificially maintained water sources for expanding of sand and stones during the monsoon season because agriculture and rural livelihoods therefore attract elephants of flood prone nature of their crop fields and preference and other wildlife during times of drought (Sukumar 1990). for farming during the post-monsoon period. Hence, crop Naughton-Treves (1998) found that only 4.0–6.0% of crops damage is severe in this area during the post-monsoon were lost in fields cultivated within 500 m of Kibale season when new crops, leaves and fruits are common. National Park, and the area with the greatest loss was less Nchanji and Lawson (1998) suggested that crop raiding than 200 m from the forest boundary. Elephants are more occurs when food availability within the elephant habitat is low, but Sukumar (1989) noted that raiding in crop fields likely to raid along boundaries rather than going deep into was not an indicator of food availability or impending star- farming areas because the risk of detection is lowest in vation of elephants. Elephants with a surplus of natural areas that serve as a buffer between PAs and areas cleared food resources resorted to crop raiding and were probably for cultivation (Bandara and Tisdell 2002). This study also attracted by the greater palatability and food value (i.e. revealed that crop loss decreases with greater distance from nutrient content) of cultivated plants (Hart and O’Connell a PA boundary, while the ability of people to control crop 2000). Moreover, crops are more digestible than wild for- damage caused by wildlife decreases at shorter distances age (Rode et al. 2006). The determination of whether there from a park boundary. Therefore, many people residing is a definite link between degraded forest and crop-raiding closer to a park reported that they had transferred from incidents will require more detailed vegetation analysis. agricultural farming to non-farming activities due to their This would enable sound recommendations on species that inability to control crop damage by crop raiders (Sarker and Røskaft Forthcoming). might be re-planted to satisfy the nutritional requirements In our study, we found that there were no major of elephants. complaints about crop damage caused by other wildlife (wild boar, porcupine and birds) compared to elephants. Nevertheless, villagers who lived closer to PAs had lower Perceptions of crop damage levels of tolerance for crop damage by other wildlife than those living further from the park boundary, but in the Since the government has taken ownership of wildlife and case of wild elephant, villagers showed zero tolerance. prohibited extraction of natural resources, local human Although elephants may cause significant damage at the populations are no longer tolerant towards crop damage 172 A.H.M. Raihan Sarker and E. Røskaft caused by wildlife, especially elephants (Naughton-Treves between humans and wildlife intensifies (Yeager and Miller 1997). The study revealed that people had absolutely no 1986), and this is strongly related to the size of the PA. tolerance for damage caused by wild elephants. This exag- The study revealed that southeast Bangladesh (i.e. CWS, geration and distortion probably originates from feelings TGR and RF 2) is more susceptible to crop damage by of disenfranchisement and helplessness. The people’s per- wild elephants than the north of the country (i.e. RF 1), and ception of crop destruction by elephants is in fact much therefore, the conflict situation is more severe in the south- exaggerated compared to the extent of actual damage by east than in the north. Consequently, villagers in south- elephants. This exaggeration is possibly an attempt to draw east Bangladesh expressed a higher level of non-tolerance attention of conservation authorities to their ‘plight’ so that towards crop damage than those in north Bangladesh. resolution for poor rural economies can be found from the Therefore, before demarcation of the boundary of any PA loose implementation of rules regarding extraction of nat- through gazette notification, it is important to consider the ural resources or through some form of financial assistance home range of large mammals in relation to the current from the concerned authority. We found that immigrants population in the PA. are less tolerant of crop damage than local people because most of them are very poor and landless. They are often considered disadvantaged by conservation programmes, Conflict reduction measures and are exploited as a dissident voice by various political platforms. If people suffer too many losses due to wildlife, This study has revealed that in Bangladesh current mit- they will seek compensation, which could provide them igation measures are clearly inadequate to overcome the with financial aid, but when government funds are lacking, negative effects of conflict between humans and elephants such a provision is difficult to create. on rural communities and on the conservation of wild In Bangladesh, the extent of human–elephant conflicts elephants because farmers cannot control the damage. may be misperceived because this issue receives a great However, the mitigation techniques employed by farmers deal of political and media attention. Based on systemat- are similar to the short-term, active deterrent methods used ically gathered data, Dublin and Hoare (2004) concluded to control human–elephant conflicts throughout Africa and that there is a wide gap between the perceived and actual Asia (O’Connell-Rodwell et al. 2000; Sukumar 2004). The levels of conflict and damage caused by elephants. In fact, majority of farmers reported that they relied on ‘scaring it is often difficult to quantify the extent of the problem and chasing’ as control measures against elephants raiding because monitoring systems can be flawed. Methods for crops. Shouting and using fire to scare elephants away are assessing damage generally and partially depend on indi- traditional techniques that have been practised for centuries rect information supplied by villagers, some of whom may in Bangladesh and also used in African savannahs (Osborn have a vested interest in exaggerating the extent of the dam- and Hill 2005). The use of torches is the commonest and age because they see reporting damage as an opportunity most important deterrent method, and our field observa- to express frustration or anger about their helplessness in tions indicated that farmers probably use this mainly out dealing with conflicts with wildlife. There is an urgent need of tradition rather than efficiency. The respondents also to introduce well-designed and cost effective monitoring stated that guarding fields was especially beneficial in schemes that can produce quantitative data on the impact of reducing human conflict with elephants, as it provided an the various conflict mitigation programmes currently being early warning (Sitati et al. 2003). Huts or watchtowers are implemented in South Asia. constructed along boundaries of farms where elephants fre- quently enter fields. When elephants are spotted, farmers use a combination of loud noises, such as yelling, firecrack- ers, hitting metal objects and cracking whips. Despite the Factors contributing to human–elephant conflict use of lookout points (huts), we have found that human– This study revealed that habitat degradation and fragmen- elephant conflicts continue, indicating that the lookouts tation gradually increased the occurrence of human conflict may not be well positioned, and scientific studies are neces- with elephants, and this is therefore an important issue sary to determine the best place to construct lookout points. in Bangladesh. Extensive logging brings elephants closer We found that the presence of humans, use of loud noise to humans because elephants forage on the secondary and fire as well as other methods could deter elephants vegetation that grows after a disturbance (Lahm 1996). from crop fields, but these tactics are time consuming, Moreover, elephants have increased contact with humans allowing elephants to trample crops in the field for sev- due to changes in land use (Nelson et al. 2003). Isolated eral hours, ultimately causing greater financial losses to agricultural farms along park boundaries may expand the farmers. Farmers were unwilling to use fire guns, proba- human–elephant interface and create a land-use pattern bly because the elephant is a protected species, and thus that favours elephant foraging (Hoare and du Toit 1998). revealing any intent to destroy it may lead to prosecution. As people have continued to transform more land and Therefore, people living near PAs suggested that the Forest forest for settlement, shrinking habitats have compressed Department should assist in the deterrence of wild ele- wildlife populations to levels beyond their carrying capac- phants. Some studies have explored viable options to ity. When the carrying capacity is exceeded, the interaction deter crop raiding elephants (Sukumar 1989; Hoare 1995). International Journal of Biodiversity Science, Ecosystem Services & Management 173 There are many simple, low-cost, active deterrent meth- Natural Resource Management Schemes, CBNRMS ods available, such as burning ‘pepper dung’ or increased (O’Connell-Rodwell et al. 2000) have shown potential to guarding. ‘Pepper dung’ is elephant dung that has been decrease the human–elephant interface, and cultivation mixed with ground chillies and dried in the sun. These of palatable crops within or close to natural fences also cakes are then burned along field boundaries to create nox- achieves the same result (Nelson et al. 2003). Moreover, ious smoke (Osborn and Anstey 2002). Chillies (Capsicum the buffer zones can act as sustainable extractive reserves spp.) have been proven effective in discouraging African for people using forest resources, and this is one solution elephants (Loxodanta africana) from raiding crops (Sitati suggested by the respondents in this study for reducing the and Walpole 2006). conflict between humans and elephants. Another strategy is to make the cultivated areas less desirable for elephants (Chiyo et al. 2005). If it is economically feasible and if environmental conditions permit, farmers could plant a Management implications buffer zone of non-preferred crops like tea, tobacco and Traditional methods of deterring elephants have failed or coffee, which, ideally, can develop into near sub-optimal are disputed in many countries (Sukumar 1991). In many elephant habitats, increasing the barrier between the parts of the world, conservation of elephants by focus- cultivated crops and prime elephant habitats (Osborn and Parker 2003). ing on habitat integrity has proved successful. In this In 2004, the Nishorgo Support Project (NSP) initiated study, the villagers emphasised the improvement of ele- a co-management approach in five PAs (including TGR phant habitat as a way to ensure availability of food for wild and CWS) in Bangladesh on a pilot project funded by elephants. When an area is developed and managed for ele- the US Agency for International Development (USAID). phant conservation, it automatically serves for biodiversity The main challenge for the NSP is that it is hard to conservation, as the elephant is both a keystone and an convince villagers to accept this strategy (Sarker and umbrella species. Therefore, there is an urgent need for an Røskaft Forthcoming). Such a situation has frequently umbrella strategy for conflict mitigation as a conservation been found to arise, mainly due to human needs and tool for Asian elephants. This strategy could include the aspirations being ignored in the initial planning of PAs, following options: considering the home range of targeted often leading to conflict between indigenous people and elephant populations in order to develop landscape-scale conservation plans; vegetation analysis in elephant habi- managers (Corbett et al. 1998). Therefore, to mitigate tat to determine the nutritional value of plant species used human–elephant conflicts, the responsibility for control- by elephants; recording quantitative data on the impact of ling elephant damage could be transferred to farmers under the various conflict mitigation programmes in Asia; con- CBNRMS (Parker and Osborn 2006). In many African ducting scientific research to identify simple, low-cost, countries, CBNRMS have established systems whereby active deterrent methods at local level; and creating for- benefits from wildlife resources can be returned to rural est buffer zones. This can best be led by international communities (O’Connell-Rodwell et al. 2000). Moreover, bodies like the International Union for Conservation of governments and NGOs should introduce vocational train- Nature (IUCN) and the Asian Elephant Specialist Group ing and soft credit schemes for the subsistence of poor (AsESG) in collaboration with multilateral donors (e.g. people living around PAs to generate alternative livelihood Asian Development Bank, The World Bank, Norwegian activities. Agency for Development Cooperation, etc.). These mul- tilateral agencies could then help with mitigation options via investments in a range of countries. Conclusion State-sponsored and voluntary settlement programmes have been endorsed to encourage landless poor people to Wild elephants trample crop fields in seasonal raids take permanent residence around PAs to reduce resentment that coincide with crop maturity and food availability. or bitterness of villagers towards the conservation of PAs Southeast Bangladesh is more vulnerable to crop dam- and elephants (Sarker and Røskaft Forthcoming). With the age by wild elephants and other wildlife than the north current pressures of the human population, it is virtually of the country. Although a few people were able to con- impossible to relocate settlements, but this could prove trol damage caused by other wildlife, no respondents were worthwhile if the forest areas are brought under appropri- able to control damage caused by wild elephants; therefore, ate land-use management. Today, the human population of the affected people were hostile towards wild elephants. Bangladesh has increased to more than 162 million, reach- Habitat degradation and fragmentation, the most impor- ing a density of 1475 people per km (Bangladesh Bureau tant factors in wildlife conservation, have been increasing of Statistics [BBS] 2009). Of the total land area of 14.4 in Bangladesh, a country that has a large population and a million ha, only 2.44 million ha are under forest or tree high growth rate. These factors pose a real challenge not cover. This includes 0.3 million ha of homestead groves only for the conservation of elephants, but also for wildlife and 1.46 million ha of state-owned forest reserves and PAs in general that are simply impossible to control. Therefore, (Akhter and Sarker 1998). Thoughtful land-use strategies, it is important to search for a pragmatic solution to the for example forest buffer zones, under Community-Based elephant conservation problem. 174 A.H.M. Raihan Sarker and E. Røskaft We found that traditional methods for deterring ele- Medicine and Ecology Graduate Group; The University of California, Davis, CA, USA. p. 1–14. phants from crop fields were time consuming. However, Hoare R. 1995. Options for the control of elephants in conflict crop loss was found to decrease with increasing distance with people. Pachyderm. 19:54–63. from PA boundaries. Forest buffer zones close to PAs under Hoare R. 2000. African elephants and humans in conflict: the CBNRMS can ameliorate the conflict between humans outlook for co-existence. Oryx. 34:34–38. and elephants in Bangladesh by involving local resource Hoare RE, du Toit JT. 1998. Coexistence between people and elephants in African savannas. Conserv Biol. 13:633–639. users at every stage of conservation management to ensure Inskip C, Zimmerman A. 2009. Human-felid conflict: a review of a win–win situation in sustainable biodiversity conserva- patterns and priorities worldwide. Oryx. 43:18–34. tion. Therefore, there is a need to adopt a decentralised [IUCN] International Union for Conservation of Nature. 1994. decision-making process that involves all stakeholders at IUCN red list categories. Switzerland: IUCN Gland. a grassroots level to manage nature conservation. In addi- Kangwana KF. 1995. Human-elephant conflict: the challenge ahead. Pachyderm. 19:11–14. tion, we need to introduce vocational training and soft Khan MH, Khan SM, Biswas SR. 2004. Human–elephant credit schemes from local government and NGOs for the conflicts in Bangladesh and assessment of financial subsistence of poor people living around PAs. losses. IUCN-The World Conservation Union, Bangladesh Country Office, Dhaka. Conservation of Asian elephants in Bangladesh. Technical Report. Vol. 4, p. 1–25. Lahm SA. 1996. A nationwide survey of crop-raiding by ele- Acknowledgements phants and other species in Gabon. Pachyderm. 21:69–77. Løe J, Røskaft E. 2004. Large carnivores and human safety: a The current study was funded by NUFU (Project nr.2007/10108). review. AMBIO. 33:283–288. We are grateful to all the people who helped us to collect Marchais J. 2004. Research report on human–elephant conflict the data and to three anonymous referees and members of the in the Okavango Kopano Mokoro community trust villages. editorial board for valuable comments that have improved the Ngamiland, Botswana: Living With Elephants Foundation. paper. Special thanks to the Bangladesh Forest Department for NG 32. allowing us to use the map of Forest Zones in Bangladesh. Mishra C, Allen P, McCarthy T, Madhusudan MD, Bayarjargal Finally, we thank Richard Binns and an anonymous editorial A, Prins HHT. 2003. The role of incentive programs board member for their assistance in improving the English of our in conserving the snow leopard. Conserv Biol. 17: paper. 1512–1520. Mosojane S. 2004. Human–elephant conflict along the Okavango Panhandle in northern Botswana [MSc thesis]. [Pretoria, South Africa]: Department of Zoology and Entomology, References University of Pretoria. Akhter S, Sarker AHMR. 1998. Patronage, tenure, inclusive- Nath CD, Sukumar R. 1998. Elephant–human conflict in Kodagu, ness and local markets: ingredients of a successful strategy southern India: distribution patterns, people’s perceptions for social forestry at Betagi. For Trees People Newslett. and mitigation methods. Bangalore, India: Asian Elephant 35:36–41. Conservation Center, Centre for Ecological Sciences, Indian Bandara R, Tisdell C. 2002. Asian elephants as agricultural pests: Institute of Science. damages, economics of control and compensation in Sri Naughton L, Rose R, Treves A. 1999. The social dimensions of Lanka. Eco Ecol Environ. 65:1–39. human–elephant conflict in Africa: a literature review and [BBS] Bangladesh Bureau of Statistics. 2009. Statistical year case studies from Uganda and Cameroon. Madison (WI): book. Dhaka, Bangladesh: BBS, Government of People’s University of Wisconsin. Republic of Bangladesh. Naughton-Treves L. 1997. Farming the forest edge: vulnerable Chiyo PI, Cochrane EP, Naughton L, Basuta GI. 2005. Temporal places and people around Kibale National Park, Uganda. Geogr Rev. 87:27–46. patterns of crop raiding by elephants: a response to changes Naughton-Treves L. 1998. Predicting patterns of crop damage by in forage quality or crop availability? Afr J Ecol. 43:48–55. wildlife around Kibale National Park, Uganda. Conserv Biol. Corbett T, Lane M, Clifford C. 1998. Achieving indigenous 12:156–168. involvement in management of protected areas: lessons Nchanji AC, Lawson DP. 1998. A survey of elephant crop damage from recent Australian experience. Aboriginal Politics and around the Banyang – Mbo wildlife sanctuary, 1993–1996. Public Sector Management Research Paper no 5, Centre for Yaounde, Cameroon: Cameroon Biodiversity Project and Australian Public Sector Management. Griffith University. Wildlife Conservation Society. p. 1–29. Nelson A, Bidwell P, Sillero-Zubiri C. 2003. A review of Dublin HT, Hoare RE. 2004. Searching for solutions: the evolu- human–elephant conflict management strategies. Oxford tion of an integrated approach to understanding and mitigat- (UK): People and Wildlife Initiative, Wildlife Conservation ing human-elephant conflict in Africa. Hum Dimens Wildl. Research Unit, Oxford University. 9:271–278. O’Connell-Rodwell CE, Rodwell T, Rice M, Hart LA. 2000. Dudley JP, Mensah-Ntiamoah AY, Kpelle DG. 1992. Forest ele- phants in a rainforest fragment: preliminary findings from a Living with the modern conservation paradigm: can wildlife conservation project in southern Ghana. Afr J Ecol. agricultural communities co-exist with elephants? A five- 30:116–126. year case study in East Caprivi, Namibia. Biol Conserv. Gurung B, David Smith JL, McDougal C, Karki JB, Barlow 93:381–391. A. 2008. Factors associated with human-killing tigers in Ogra M, Badola R. 2008. Compensating human–wildlife con- Chitwan National Park, Nepal. Biol Conserv. 141:3069– flict in protected area communities: ground-level perspectives 3078. from Uttarakhanda, India. Hum Ecol. 36:717–729. Hart LA, O’Connell CE. 2000. Human conflict with African Osborn FV, Anstey S. 2002. Elephant–human conflict and and Asian elephants and associated conservation dilemmas. community development around the Niassa Reserve, Center for Animals in Society in the School of Veterinary Mozambique. Report of WWF/SARPO-MZEP. p. 1–29. International Journal of Biodiversity Science, Ecosystem Services & Management 175 Osborn FV, Hill C. 2005. Techniques to reduce crop loss: Sitati NW, Walpole MJ, Smith RJ, Leader-Williams N. 2003. human and technical dimensions in Africa. In: Woodroffe R, Predicting spatial aspects of human–elephant conflict. J Appl Thirgood S, Rabinowitz A, editors. People and wildlife; Ecol. 40:667–677. conflict and coexistence. Cambridge (UK): Cambridge Sukumar R. 1989. The Asian elephant – ecology and manage- University Press. p. 72–85. ment. Cambridge (UK): Cambridge University Press. Osborn FV, Parker GE. 2003. Towards an integrated approach for Sukumar R. 1990. Ecology of the Asian elephant in southern reducing the conflict between elephants and people: a review India II: feeding habits and crop raiding patterns. J Trop Ecol. of current research. Oryx. 37:80–84. 6:33–53. Packer C, Ikanda D, Kissui B, Kushnir H. 2005. Lion attacks on Sukumar R. 1991. The management of large mammals in rela- humans in Tanzania – understanding the timing and distri- tion to male strategies and conflict with people. Biol Conserv. bution of attacks on rural communities will help to prevent 55:93–102. them. Nature (Lond). 436:927–928. Sukumar R. 2004. The living elephants: evolutionary ecology, Parker GE, Osborn FV. 2006. Investigating the potential for behavior and conservation. New York: Oxford University chilli Capsicum spp. to reduce human–wildlife conflict in Press. Zimbabwe. Oryx. 40:343–346. Sukumar R. 2006. A brief review of the status, distribution Rode KD, Chiyo PI, Chapman CA, McDowell LR. 2006. and biology of wild Asian elephants. Int Zoo Yearbook. 40: Nutritional ecology of elephants in Kibale National Park, 1–8. Uganda, and its relationship with crop-raiding behaviour. J Thouless CR. 1994. Conflict between humans and elephants on Trop Ecol. 2:441–449. private land in northern Kenya. Oryx. 28:119–127. Røskaft E, Händel B, Bjerke T, Kaltenborn BP. 2007. Human Treves A, Karanth KU. 2003. Human–carnivore conflict and attitudes towards large carnivores in Norway. Wildl Biol. perspectives on carnivore management worldwide. Conserv 13:172–185. Biol. 17:1491–1499. Sarker AHMR, Røskaft E (Forthcoming). Human perceptions Wang SW, Macdonald D. 2005. Livestock predation by carni- towards conservation of protected areas: a case study from vores in Jigme Singye Wangchuck National Park, Bhutan. four protected areas in Bangladesh. Oryx. Biol Conserv. 129:558–565. Schwerdtner K, Bernd G. 2007. A conceptual framework Woodroffe R, Ginsberg JR. 1998. Edge effects and the extinc- for damage compensation scheme. Biol Conserv. 134: tion of populations inside protected areas. Science. 280: 354–360. 2126–2128. Sitati NW, Walpole MJ. 2006. Assessing farm-based measures Yeager R, Miller NN. 1986. Wildlife, wild death: land use and for mitigating human–elephant conflict in Transmara District, survival in eastern Africa. Albany (NY): State University of Kenya. Oryx. 40:279–286. New York Press. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Biodiversity Science, Ecosystem Services & Management Taylor & Francis

Human–wildlife conflicts and management options in Bangladesh, with special reference to Asian elephants (Elephas maximus)

Download PDF
12 pages

Loading next page...
 
/lp/taylor-francis/human-wildlife-conflicts-and-management-options-in-bangladesh-with-kkLOspx8y5

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Taylor & Francis
Copyright
Copyright Taylor & Francis Group, LLC
ISSN
2151-3732
eISSN
2151-3740
DOI
10.1080/21513732.2011.554867
Publisher site
See Article on Publisher Site

Abstract

International Journal of Biodiversity Science, Ecosystem Services & Management Vol. 6, Nos. 3–4, September–December 2010, 164–175 Human–wildlife conflicts and management options in Bangladesh, with special reference to Asian elephants (Elephas maximus) A.H.M. Raihan Sarker and Eivin Røskaft* Department of Biology, Norwegian University of Science and Technology (NTNU), Trondheim, Norway In this study, we explored the interaction between humans and Asian elephants (Elephas maximus). We analysed local mitiga- tion techniques and propose possible measures to reduce human–elephant conflicts. For four protected areas in Bangladesh, we investigated how people interact with elephants and other crop-damaging species. Crop-damaging species differed from site to site, but elephants were generally the most frequent offenders. Although some people were able to control crop dam- age caused by other wildlife, damage caused by wild elephants was mostly uncontrollable. Forest villagers were somewhat tolerant of damage caused by other wildlife but they were not tolerant of damage caused by wild elephants. Tolerance of crop damage caused by wildlife other than wild elephants tended to increase with distance of respondents from protected areas. Non-tolerance of crop damage caused by wildlife was more common in southeast Bangladesh than in the north. Habitat destruction, caused by high population growth and poverty, appeared to be a major cause for increasing human–elephant conflict. There is an urgent need for adoption of an umbrella strategy (e.g. estimate the size of elephant populations, develop landscape-scale conservation plans and create forest buffer zones under community-based natural resource management schemes) in the use of conflict mitigation as a conservation tool for Asian elephants. Keywords: human–elephant conflict; Bangladesh; attitudes; elephants; compensation 1991; Kangwana 1995; O’Connell-Rodwell et al. 2000). Introduction Elephants that traditionally lived in forests now frequently Human–wildlife conflict (HWC) is common where wildlife come out in groups or as scattered individuals and cause and human populations coexist and share limited resources financial and social losses (Sukumar 1989; Thouless 1994). (Schwerdtner and Bernd 2007), and has become a consider- The most frequently discussed losses are crop damage by able problem throughout the world (Wang and Macdonald elephants and injury or death of people (Hoare 2000). 2005). Conflicts are more intense where livestock holdings Elephants migrate seasonally to follow the accessibility of and agriculture are important parts of the rural livelihood water and favoured foods (Sukumar 1989). Their migra- (Ogra and Badola 2008). Crop raiding, property dam- tion patterns frequently expand not only beyond park age, livestock depredation and human casualties are the and reserve boundaries, but beyond national boundaries most common forms of conflict with wildlife (Inskip and as well (Hart and O’Connell 2000). This is true, for Zimmerman 2009). The most extreme form of human– example, for the Teknaf Game Reserve in Bangladesh wildlife conflict occurs when human lives are lost (Løe and the Alaungdaw Kathapa National Park in Myanmar, and Røskaft 2004; Packer et al. 2005; Gurung et al. 2008) South Asia (Personal communication, Mr. Nurul Kabir, and retaliation against wild animals follows (Treves and ACF, Bangladesh Forest Department). Therefore, the con- Karanth 2003). Retaliation, also for protection of livestock sequences of conflicts between humans and elephants are and agricultural crops, threatens the survival of wildlife not only a key concern for conservation, but also a major that comes into conflict with humans (Mishra et al. 2003). socio-economic and political issue because elephants move Large herbivores and carnivores are particularly affected great distances (Sukumar 1989). by this conflict and are either critically endangered or Elephant disturbances usually take place between dusk rapidly declining as a result (Woodroffe and Ginsberg and dawn and are strongly seasonal, corresponding with 1998). Human–elephant conflict is a key example of such crop harvesting periods (Sukumar 2004; Osborn and Hill an interaction. 2005). Farms adjacent to the park boundary are likely The human–elephant conflict arises from a range to be raided, especially during seasons (June–August and of direct and indirect negative interactions between October–December) when favoured foods (e.g. paddy) are humans and elephants, which can create negative atti- at the optimal stage of growth (Sukumar 1990). Conflicts tudes regarding the conservation of elephants (Røskaft are usually more severe close to protected areas (PAs), et al. 2007). Human pressures on elephants caused by which act as elephant refuges (Hoare 2000). Inevitably, the poaching and conflict over resources, as well as efforts to sharpest conflicts and greatest losses seem to occur where adjust the effects of elephants on vegetation and crops, villages are close to park borders (O’Connell-Rodwell et al. are widespread throughout Asia and Africa (Sukumar *Corresponding author. Email: roskaft@bio.ntnu.no ISSN 2151-3732 print/ISSN 2151-3740 online © 2010 Taylor & Francis DOI: 10.1080/21513732.2011.554867 http://www.infomaworld.com International Journal of Biodiversity Science, Ecosystem Services & Management 165 2000). Human–elephant conflicts arise in rural areas of Materials and methods Asia, posing significant problems to local communities, Study areas elephants and wildlife managers (Sitati et al. 2003). The This study was conducted in three PAs in south- conservation of the Asian elephant (Elephas maximus) east Bangladesh: Teknaf Game Reserve (TGR), Chunati is of both national and international significance because Wildlife Sanctuary (CWS), a reserve forest adjacent to it is considered to be a flagship species by international Dulahazara Safari Park (RF 2) and, in north Bangladesh, conservationist groups (e.g. the Asian Elephant Specialist a reserve forest in Sherpur district (RF 1) that lies between Group, AsESG) for the ecosystem it inhabits (Nath and the political border of Bangladesh and India (Figure 1). The Sukumar 1998). Therefore, resolution of human conflict TGR (IUCN category VI: a protected area managed mainly with elephants is a major concern and a high priority for for the sustainable use of natural ecosystems, IUCN 1994) conservation of elephants whose range includes countries was established in 1983 under the Cox’s Bazar (South) in Asia. ◦  ◦ Forest Division in southeast Bangladesh (21 04 N, 92 The elephant is a ‘Critically Endangered Species’ 09 E; total area 116.15 km ). TGR is part of a linear hill in Bangladesh, and it is included in Appendix 1 of range that gently slopes to rugged hills and cliffs; most the Convention on International Trade in Endangered of the streams are seasonal and dry up during the off- Species of Wild Fauna and Flora (CITES) and in the monsoon season. The CWS (IUCN category V: a protected Third Schedule of the Bangladesh Wildlife Preservation area managed mainly for landscape/seascape conserva- (Amendment) Act, 1974. Human settlement in forested tion and recreation, IUCN 1994) was established in 1986 areas, absence of tree camouflage inside the forest and and is managed by the Wildlife and Nature Conservation reductions in availability of food are considered to be Division. CWS is situated in southeast Bangladesh (21 the primary causes for the decrease in number of ele- ◦  2 58 N, 92 04 E, total area 77.64 km ). The sanctu- phants in the forests of Bangladesh (Khan et al. 2004). ary area is generally hilly with shallow to deep gul- Southeast Bangladesh is considered to be a primary lies and gentle to steep slopes, traversed by numerous conservation area for Asian elephants that provides a streams. RF 2 (IUCN category 1b: a protected area man- refuge for the existing population. An estimated 196– aged mainly for wilderness protection, IUCN 1994) is 227 elephants range over a fragmented landscape in surrounded by the Fasiakhali Range of the Cox’s Bazar Bangladesh (Sukumar 2006). However, the natural habi- ◦  ◦ (North) Forest Division (21 40 N, 92 04 E, total tat for the elephants in Bangladesh is rapidly diminish- area 27.1 km ) and was established in 1897. The topog- ing because more and more land is being cleared for raphy is undulating, and there are some natural peren- cultivation and development of infrastructure. Moreover, nial streams. RF 1 (IUCN category 1b) includes parts human encroachment on PAs in southeast and north ◦  ◦ of Jhenaigati (25 16 N, 90 01 E) and Nalitabari sub- Bangladesh has forced elephants to forage in non-protected ◦  ◦ district (25 13 N, 90 08 E) of Sherpur district and areas, thus increasing the disturbance caused to local was established in 2006. The total area of the RF 1 people. Therefore, with the dwindling populations of is 5.0 km , and it comprises a scattered secondary Sal Asian elephant, conflict between humans and elephants forest. has also become a progressively more serious issue in Bangladesh. Human–elephant conflict is a complex problem and Respondents a highly emotive issue that cannot be mitigated through reliance on a single mitigation technique (Sitati and In the early 1950s, the Bangladesh Forest Department Walpole 2006). Each field requires a specific deterrent at TGR, CWS and RF 2 allowed local landless and strategy. A range of mitigation techniques has proved poor people to live in these areas to protect the for- to be more effective in driving away raiders, but this est. We defined these inhabitants as local people because reportedly becomes less effective over time (Parker they were born in the areas and hold property rights and Osborn 2006). This is because, crop-raiding ele- from their ancestors. Similarly, indigenous tribal peo- phants soon learn to ignore deterrents and develop resis- ple who resided around RF 1 before the 1980s were tance to crop protection measures (Bandara and Tisdell also defined as local people. People who migrated to 2002). the study areas from the beginning of the 1980s to Ensuring the survival of Asian elephants in Bangladesh the present day and settled without government approval calls for an adequate and comprehensive understanding were defined as immigrants. Immigrants were reha- of their raiding patterns. This study aims to explore the bilitated around the study areas from other districts interactions between humans and wildlife in four PAs under the protection of various political and social in Bangladesh. An underlying assumption is that most groups and were often considered as strong potential human–wildlife interactions vary not only with season, but vote banks for political parties. However, in this study, also with distances of respondents’ homes from PAs. We we defined both local people and immigrants as ‘vil- also examined local mitigation techniques and their effec- lagers’. Sarker and Røskaft (Forthcoming) describe the tiveness, and propose possible remedies for reducing the demographic and socio-economic characteristics of the conflict between humans and elephants. respondents. 166 A.H.M. Raihan Sarker and E. Røskaft RF 1 RF 2 CWS TGR Source: Bangladesh Forest Department, 1999. Figure 1. Map showing the location of the four PAs studied in Bangladesh (RF 1 = Reserve forest in north Bangladesh, CWS = Chunati Wildlife Sanctuary, RF 2 = Reserve forest in southeast Bangladesh and TGR = Teknaf Game Reserve). Sampling and data collection sub-districts across the four study sites. We followed a structured questionnaire to reduce bias in the responses We prepared household maps in each PA using a GPS of respondents, while standardised questionnaires were up to a distance of 1 km from the park boundary. The used to ensure the reliability, generalisability and valid- demographic features of each village in the study area ity of the responses. A period of 45 days was spent in were collected from the local Union Council. We admin- each of the eight sub-districts, Teknaf (TGR), Lohagara istered a total of 388 standardised, structured and semi- and Banshkhali (CWS), Chokoria, Ukhia, and Ramu (RF structured questionnaires to 193 males and 195 females 2), Jhinaighati and Nalitabari (RF 1). The questionnaire over the age of 18 years. The participants were selected included socio-demographic variables and a set of closed randomly, and completion of questionnaires was facilitated (no/yes/don’t know) and open-ended questions related to through ‘face-to-face’ communication tools. The respon- perceptions of conflicts between humans and elephants dents were from 99 villages under 11 unions and eight International Journal of Biodiversity Science, Ecosystem Services & Management 167 (Box 1). Open-ended questions were asked at the end between perceptions and different independent variables of the structured portion to collect as much informa- were investigated using step-wise linear regression. tion as possible within the limited time (approximately 45 min per respondent) of the interview. Usually, house- hold heads (mostly males) were interviewed; however, in Results their absence, any member (mostly housewives) willing to Wildlife and crop losses participate was interviewed. The respondents were cho- Crop raiding by wild animals was an important issue in the sen for interview and categorised on the basis of their study areas, as almost two-thirds of respondents (64.1%) monthly income (e.g. <US$ 70 per month and >US$ 70 reported it as a significant cause of crop loss, while over per month), distance of home from a PA (e.g. <0.5 km, one-third (35.9%) reported natural disasters as most impor- 0.5–1.0 km and >1.0 km), land-holding status (e.g. 0 acre, tant (χ = 104.5, df = 3, N = 388, p < 0.0001). The 0.01–0.50 acre, 0.51–1.00 acre and >1.00 acre), occu- species that were responsible for crop damage differed sig- pation (e.g. farmer; other occupations such as business- nificantly from one site to another (χ = 129.0, df = 12, man, housewife, teacher, student, service provider, political p < 0.0001), but elephants were reported more frequently leader, etc. and daily labourer), education (e.g. illiterate or than other species (Table 1). Wild boar (Sus scrofa) ranked literate) and residential status (e.g. local or immigrant). second most likely species to be responsible for crop losses In this study, we defined both daily labourers and other around the PAs (Table 1). Southeast Bangladesh (i.e. CWS, occupations as ‘non-farmers’. TGR and RF 2) was more vulnerable to crop damage by wild elephants and other wildlife than north Bangladesh (i.e.RF1)(Table1). Box 1. Questions presented to respondents to assess their perceptions of human–elephant conflict Period of crop damage A. Which wild animals damage crops on your farm or homestead garden? Respondents claimed that wild elephants caused damage to B. Is crop-raiding by wild animals seasonal? crops mainly during the monsoon (June–September) and (Monsoon/post-monsoon/winter/all year) post-monsoon period (October–December), but this var- C. Are you able to control crop damage caused ied considerably between study sites (χ = 72.5, df = 9, by wildlife? (no/yes) p < 0.0001) (Table 2). Most of the crop damage occurred D. How do you look upon crop damage caused in north Bangladesh (i.e. RF 1) during the post-monsoon by wildlife? (non-tolerant/tolerant) period. In southeast Bangladesh (i.e. CWS, TGR and RF E. How do you deter wildlife from your field 2), crop damage caused by wild elephants was higher dur- and garden? ing the monsoon season and most frequent in and around RF 2. Only a few respondents claimed that wild boar F. Is this effective? (no/yes/do not know) (χ = 57.7, df = 9, p < 0.0001) and porcupine, Hystrix G. Do you think that human–elephant conflict indicus (χ = 10.5, df = 3, p < 0.015) raided crops at is increasing in your area? (no/yes) H. What is the main reason for the increase in any time of the year, but this varied considerably between conflict between humans and elephants? study areas (Table 2). There was no relationship between crop raiding birds and the period of crop damage across the PAs (χ = 7.4, df = 3, p < 0.060) (Table 2). Data analyses Ability to control crop damage Data were analysed using SPSS Version 16.0 (SPSS, Chicago, IL, USA). Perceptions were related to many Less than half of respondents (41.0%, N = 388) reported demographic variables. The significance level was set at that they were able to control crop damage caused by dif- p < 0.05. Differences between perceptions were tested ferent wildlife species, but this varied significantly between 2 2 using Pearson’s chi-square (χ ) test. The relationships study areas (χ = 87.8, df = 3, p < 0.0001). Among Table 1. Percentages of people who responded concerning crop damage caused by different wild animals in each of the PAs. Wildlife responsible for crop damage CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) None 6.5 10.1 60.0 11.4 35.8 Elephant 51.6 49.3 33.0 60.2 43.6 Wild boar 38.7 31.9 6.0 18.2 16.0 Porcupine 3.2 7.2 1.0 6.8 3.6 Birds 0.0 1.4 0.0 3.4 1.0 Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. 168 A.H.M. Raihan Sarker and E. Røskaft Table 2. Percentages of people who responded concerning the period of crop damage caused by different wild animals in each of the PAs. Wildlife responsible and the period of crop damage CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) Wild elephant None 51.6 52.2 67.0 40.9 57.2 All year 25.8 13.0 0.5 26.1 10.6 Monsoon 16.1 17.4 7.5 20.5 12.9 Post-monsoon 6.5 17.4 25.0 12.5 19.3 Wild boar None 61.3 68.1 94.0 80.7 83.8 All year 38.7 24.6 2.5 19.3 13.1 Monsoon 0.0 0.0 0.5 0.0 0.3 Post-monsoon 0.0 7.2 3.0 0.0 2.8 Porcupine None 96.8 92.8 99.0 92.0 96.1 All year 3.2 7.2 1.0 8.0 3.9 Birds None 100.0 98.6 100.0 96.6 99.0 All year 0.0 1.4 0.0 3.4 1.0 Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. local people, 51.3% reported that they were able to con- Mitigation of conflict between humans and elephants trol crop damage caused by different wildlife species, while Almost two-thirds of respondents reported that they used among immigrant populations, only 6.7% were able to active deterrent methods that involved scaring animals control crop damage (χ = 57.0, df = 1, p < 0.0001). with bright lights (i.e. setting fire to raw jute or tyres Among respondents (N = 169) who reported crop damage fixed at the end of bamboo sticks, locally known as caused by wild elephants, none stated that they could con- ‘bolems’, and torchlight) or loud noise (i.e. shouting, trol this damage. On the other hand, among respondents drumming and bamboo cracking), and in most cases, this (N = 157) who reported crop damage caused by other 2 varied statistically across the study areas (χ = 138.4, wildlife species, 17.2% claimed that they could control df = 21, p < 0.0001) (Table 3). In addition, villagers con- damage. structed thatched houses (huts) during the harvest season (especially of rice) from which to guard their fields at night. Such activities were always carried out in groups. Attitudes to crop damage Drumming, bamboo cracking and other means (e.g. use of Among all respondents, 25.8% (N = 388) were willing to buffalo to deter elephants, oil lamps and fencing made of tolerate damage caused by different wildlife species, but sharp thorns) were used less frequently, perhaps because this differed significantly between study sites (χ = 31.8, the more popular techniques are usually employed first. df = 3, p < 0.0001). Among local people, 32.9% were The majority of respondents agreed with the effectiveness tolerant of crop damage caused by different wildlife of traditional method(s) used to deter wildlife, especially species, while among immigrants very few respondents elephants, from crop fields, and this varied significantly (2.2%) reported that they would tolerate crop damage by 2 between study sites (χ = 55.3, df = 3, p < 0.0001) wildlife (χ = 33.9, df = 1, p < 0.0001). Moreover, of the (Table 3). We found that the traditional methods used to respondents who reported crop damage caused by wildlife deter elephants were time consuming, and this ultimately species such as birds, wild boar and porcupine (N = 157), increased the costs for farmers because, despite the deter- only a small proportion (8.3%) were tolerant, whereas rent, elephants still trampled crops until they moved on respondents (N = 169) who reported crop damages caused from a field. However, more than one-third of respondents only by wild elephants had absolutely no tolerance of reported that they were not using any active methods to such damages. Non-tolerance of crop damage was more deter elephants from crop fields. Among respondents who common in TGR (92.8%, N = 69), CWS (83.9%, N = 31) reported using no active deterrence methods, 0.5% lived and RF 2 (83.0%, N = 88) than in RF 1 (62.5%, N = 200) within 0.5 km of a park boundary, while 55.3% lived (χ = 31.8, df = 3, p < 0.0001). The level of tolerance 0.5 km to 1 km and 84.5% lived more than 1 km from of respondents increased significantly with the distance the park boundary (χ = 243.3, df = 14, N = 388, of their household from the PA boundary (χ = 273.8, p < 0.0001). df = 2, p < 0.0001). Respondents who lived within 0.5 km of the park boundary showed absolutely no tolerance People’s experience with conflict for crop damage caused by wildlife, whereas a small proportion of respondents living 0.5–1.0 km (14.0%) from Almost half of respondents claimed that conflict between a park boundary were tolerant of crop damage caused by humans and elephants had increased in their villages, wildlife other than elephant, and the proportion of tolerant and this differed significantly between the studied PAs respondents (86.0%) increased among those living beyond (χ = 54.8, df = 3, p < 0.0001) (Table 4). Human con- 1 km of the park boundary. flict with elephants was reported to be significantly higher International Journal of Biodiversity Science, Ecosystem Services & Management 169 Table 3. Percentages of people who responded concerning actions taken and their effectiveness for deterring wildlife from entering fields in each of the PAs included in the study. Deterrent techniques CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) None 6.5 11.6 60.0 11.4 36.1 Bolem 19.4 17.4 14.0 22.7 17.0 Torchlight 19.4 15.9 10.0 18.2 13.7 Shouting 25.8 17.4 2.5 22.7 11.6 Drumming 12.9 8.7 4.0 3.4 5.4 Bamboo cracking 0.0 2.9 3.0 4.5 3.1 Guarding from huts 6.5 1.4 1.5 6.8 3.1 Others 9.7 24.6 5.0 10.2 10.1 Effectiveness of action No 19.4 50.7 9.0 26.1 21.1 Yes 80.6 49.3 91.0 73.9 78.9 Notes: Use of buffalo, oil lamps and fencing with sharp thorns. CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. Table 4. Percentages of people who responded concerning whether the magnitude of human conflict with elephants is increasing and the reasons, and opinions of respondents for reducing such conflict in each of the PAs included in this study. Conflict increases CWS (N = 31) TGR (N = 69) RF 1 (N = 200) RF 2 (N = 88) Total (N = 388) No 19.4 26.1 67.5 40.9 50.3 Yes 80.6 73.9 32.5 59.1 49.7 Reasons of conflict Habitat destruction 48.4 37.7 56.0 39.8 48.5 Increase in human 6.5 39.1 24.5 34.1 27.8 population and poverty Reserve size is too 38.7 13.0 10.0 17.0 14.4 small Crop raiding by 6.5 10.1 9.5 9.1 9.3 wildlife Options to reduce conflict Do not know 0.0 0.0 8.5 1.1 4.6 FD should respond in 45.2 27.5 35.5 29.5 33.5 time to deter elephants Need to introduce 41.9 60.9 27.5 59.1 41.8 compensation scheme Need to improve 3.2 5.8 18.5 8.0 12.6 elephant habitat Create buffer zone 9.7 5.8 10.0 2.3 7.5 through community involvement in a forestry programme Note: CWS, Chunati Wildlife Sanctuary; TGR, Teknaf Game Reserve; RF 1, Reserve forest in north Bangladesh; RF 2, Reserve forest in southeast Bangladesh. in southeast Bangladesh (i.e. CWS, TGR and RF 2) than reduce the conflict between humans and elephants, respon- in northern parts (i.e. RF 1). Respondents’ identifica- dents proposed several options that varied across study tion of reasons behind such conflict varied significantly sites (χ = 55.2, df = 12, p < 0.0001) (Table 4). between study sites (χ = 31.9, df = 9, p < 0.0001) Most respondents proposed introducing a compensation (Table 4). Almost half of respondents identified habi- scheme to protect against crop loss and demanded active tat destruction as a major cause of increasing human– support from the Forest Department to reduce problems elephant conflict, while more than one-quarter cited human associated with wild elephants. Very few respondents population growth and poverty as root causes. Very few emphasised improvement of elephant habitat to a sustain- respondents reported the size of the reserve or wildlife able level or adoption of social forestry as management crop raiding as a cause of human–elephant conflict. To strategies. 170 A.H.M. Raihan Sarker and E. Røskaft Predictors of interactions between humans and wildlife porcupine and birds, than respondents who lived further away from the PAs. Similarly, immigrants reported more Crop damage by elephants crop losses than local people. A step-wise linear regression analysis examining variations in perceptions of respondents towards whether elephants cause crop damage as dependent variable was tested with The ability to control crop damage caused by wildlife five independent variables (Table 5); only four of which proved to be significant contributors to the variation. The We tested the perception of respondents towards their abil- variable explaining most variation in this perception was ity to control crop damage caused by wildlife as a depen- distance of respondents’ homes from the park boundary. dent variable, with five independent variables (Table 5). The occupation of the respondent was the second most All independent variables explained 62.2% of the variation important predictor. Settlement status and land-holding on ability to control the crop damage caused by wildlife. status of respondents were other significant predictors The variable explaining most variation was the distance of accounting for most of the variation in this perception. respondents’ homes from the park boundary. Occupation The education status of respondents was not significant. and land-holding status were the second and third most All independent variables explained 51.0% of the varia- important predictors, respectively. Settlement status and tion in people’s perception of crop damage caused by wild education were not significant. Respondents who lived elephants. The analysis indicated that villagers who lived closer to the park reported less ability to control crop dam- closer to the park described higher levels of crop damage age caused by wildlife than those who lived further away caused by elephants than those who lived further away. from PAs. Farmers expressed least ability to control crop Farmers reported higher levels of crop damage by wild damage. Similarly, landless people, rather than landowners, elephants than non-farmers (e.g. daily labourers and other also reported least ability to control crop damage. occupations). Immigrants reported a higher level of crop loss than local people, and landless people reported even higher levels of crop loss than land-owners. Attitudes towards crop damage A step-wise linear regression with perception of crop dam- age caused by wildlife as dependent variable was used Crop damage by other wildlife species to test how much of the variation was explained by the The perception of whether other wildlife species (e.g. wild five independent variables (Table 5). Only two of the boar, porcupine and birds) cause crop damage at respon- tested variables proved to be significant contributors to dents’ farms or homestead gardens was tested with respect explain the variation. The most explanatory variable was to the five demographic variables. The step-wise linear the distance of respondents’ homes from the park bound- regression showed that the distance of respondents’ homes ary. Education was the second most important predictor. from the park boundary and the settlement status of respon- Settlement status, occupation and land-holding status were dents were the two most significant factors contributing to not significant. All independent variables explained 62.0% the variation in perception (Table 5). Education, occupation of the variation in tolerance of people towards crop damage and land-holding status of respondents were not signif- caused by wildlife. Respondents who lived further from icant. All independent variables explained 14.0% of the the reserve had a higher level of tolerance towards crop variation in people’s perception of crop damage caused by damage caused by wildlife than those who lived closer other wildlife. The analysis showed that respondents who to the park. Such attitudes were governed by the level of lived closer to the park boundary reported higher crop dam- education, since literate people showed a greater tolerance age caused by other wildlife species, such as wild boar, to crop damage than illiterate people. Table 5. Results of stepwise linear regression with perceptions as dependent variables in relation to various independent variables (R = rank, t = t-value). Crop raiding by Crop raiding by Ability to control Tolerance to crop elephant other wildlife crop damage damage Increasing HEC Independent variables Rt p-Value Rt p-Value Rt p-Value Rt p-Value Rt p-Value Distance of respondents’ 1 −12.8 0.0001 1 −5.2 0.0001 1 19.5 0.0001 1 23.9 0.0001 1 −15.6 0.0001 homes from park boundary Occupation 2 −3.9 0.0001 4 0.9 0.362 2 3.3 0.0001 4 −0.8 0.416 2 −5.3 0.0001 Settlement status 3 3.6 0.0001 2 3.1 0.002 4 −1.5 0.128 3 1.8 0.076 4 1.3 0.195 Land 4 −2.8 0.005 5 −0.3 0.776 3 3.3 0.0001 5 0.1 0.917 3 −3.1 0.002 Education 5 −1.4 0.173 3 −1.8 0.079 5 0.4 0.682 2 2.9 0.004 5 0.5 0.607 Constant 3 11.14 0.0001 5 4.08 0.0001 4 −9.72 0.0001 2 −14.47 0.0001 1 23.30 0.0001 R 0.510 0.0001 0.140 0.0001 0.622 0.0001 0.620 0.0001 0.528 0.0001 International Journal of Biodiversity Science, Ecosystem Services & Management 171 The perception of whether conflict is increasing local level (Dudley et al. 1992), their regional impact on agriculture is insignificant compared to other vertebrate We employed step-wise linear regression for the percep- and invertebrate pests (Naughton et al. 1999). Elephants tion of respondents on the magnitude of human conflict depredate at a high intensity mainly for just a few months with elephants in villages closer to PAs as dependent vari- of the year; the damage caused by wild boar and porcu- able to test how much of this variable could be explained pine is reported to be less, but it is continuous. The general by the five independent variables (Table 5). The variable impression given by farmers in the study areas is that ordi- explaining most variation was distance of respondents’ nary agricultural pests such as wild boar and porcupine can homes from the park boundary. Occupation was the second easily be managed with minimal effort and cost compared most important predictor. The third was land-holding sta- to wild elephants, which are more destructive and not easy tus, while settlement status and education of respondents to control. Therefore, some villagers residing in the study were not significant. All independent variables explained sites had tolerant attitudes towards crop damage caused by 52.8% of the variation in people’s opinions regarding the wildlife other than elephants. status of conflict between humans and elephants. Similar to There are seasonal fluctuations in crop raiding other cases studied, the respondents who lived closer to PAs that coincide with food availability and crop maturity reported a higher level of conflict with elephants than those (Marchais 2004; Mosojane 2004). In our study, we found who lived further away from the park boundary. Farmers that crop damage by wild elephants was reported more reported a higher level of conflict than non-farmers, and during the monsoon and post-monsoon season. Villagers landless people reported a higher level of conflict than reported that rice was the principal crop consumed by ele- landowners. phants, which is harvested during the monsoon and post- monsoon season. Since much of the cultivation around CWS and RF 2 occurs during the monsoon season, the Discussion results indicate that elephants are attracted to the crops Factors contributing to crop raiding during harvest time, and hence, there is a greater ten- dency for conflict between humans and elephants at this The study revealed that crop raiding by wildlife was a time than during the post-monsoon season. Reports of major cause of crop loss from farmers’ fields. Farms near crop damage in villages around the TGR were almost a forest or PA boundary, a water source or a migra- tion route suffer a disproportionate amount of damage the same in both seasons, since the TGR is a fundamen- (Naughton et al. 1999; Mosojane 2004). The attraction to tal habitat for Asian elephants in Bangladesh. The area crops increases due to the availability of water, which is surrounding RF 1 is a flood prone zone (Banerjee, unpub- relatively scarce in some seasons, particularly in the dry lished) where inhabitants are involved in illegal removal season. Artificially maintained water sources for expanding of sand and stones during the monsoon season because agriculture and rural livelihoods therefore attract elephants of flood prone nature of their crop fields and preference and other wildlife during times of drought (Sukumar 1990). for farming during the post-monsoon period. Hence, crop Naughton-Treves (1998) found that only 4.0–6.0% of crops damage is severe in this area during the post-monsoon were lost in fields cultivated within 500 m of Kibale season when new crops, leaves and fruits are common. National Park, and the area with the greatest loss was less Nchanji and Lawson (1998) suggested that crop raiding than 200 m from the forest boundary. Elephants are more occurs when food availability within the elephant habitat is low, but Sukumar (1989) noted that raiding in crop fields likely to raid along boundaries rather than going deep into was not an indicator of food availability or impending star- farming areas because the risk of detection is lowest in vation of elephants. Elephants with a surplus of natural areas that serve as a buffer between PAs and areas cleared food resources resorted to crop raiding and were probably for cultivation (Bandara and Tisdell 2002). This study also attracted by the greater palatability and food value (i.e. revealed that crop loss decreases with greater distance from nutrient content) of cultivated plants (Hart and O’Connell a PA boundary, while the ability of people to control crop 2000). Moreover, crops are more digestible than wild for- damage caused by wildlife decreases at shorter distances age (Rode et al. 2006). The determination of whether there from a park boundary. Therefore, many people residing is a definite link between degraded forest and crop-raiding closer to a park reported that they had transferred from incidents will require more detailed vegetation analysis. agricultural farming to non-farming activities due to their This would enable sound recommendations on species that inability to control crop damage by crop raiders (Sarker and Røskaft Forthcoming). might be re-planted to satisfy the nutritional requirements In our study, we found that there were no major of elephants. complaints about crop damage caused by other wildlife (wild boar, porcupine and birds) compared to elephants. Nevertheless, villagers who lived closer to PAs had lower Perceptions of crop damage levels of tolerance for crop damage by other wildlife than those living further from the park boundary, but in the Since the government has taken ownership of wildlife and case of wild elephant, villagers showed zero tolerance. prohibited extraction of natural resources, local human Although elephants may cause significant damage at the populations are no longer tolerant towards crop damage 172 A.H.M. Raihan Sarker and E. Røskaft caused by wildlife, especially elephants (Naughton-Treves between humans and wildlife intensifies (Yeager and Miller 1997). The study revealed that people had absolutely no 1986), and this is strongly related to the size of the PA. tolerance for damage caused by wild elephants. This exag- The study revealed that southeast Bangladesh (i.e. CWS, geration and distortion probably originates from feelings TGR and RF 2) is more susceptible to crop damage by of disenfranchisement and helplessness. The people’s per- wild elephants than the north of the country (i.e. RF 1), and ception of crop destruction by elephants is in fact much therefore, the conflict situation is more severe in the south- exaggerated compared to the extent of actual damage by east than in the north. Consequently, villagers in south- elephants. This exaggeration is possibly an attempt to draw east Bangladesh expressed a higher level of non-tolerance attention of conservation authorities to their ‘plight’ so that towards crop damage than those in north Bangladesh. resolution for poor rural economies can be found from the Therefore, before demarcation of the boundary of any PA loose implementation of rules regarding extraction of nat- through gazette notification, it is important to consider the ural resources or through some form of financial assistance home range of large mammals in relation to the current from the concerned authority. We found that immigrants population in the PA. are less tolerant of crop damage than local people because most of them are very poor and landless. They are often considered disadvantaged by conservation programmes, Conflict reduction measures and are exploited as a dissident voice by various political platforms. If people suffer too many losses due to wildlife, This study has revealed that in Bangladesh current mit- they will seek compensation, which could provide them igation measures are clearly inadequate to overcome the with financial aid, but when government funds are lacking, negative effects of conflict between humans and elephants such a provision is difficult to create. on rural communities and on the conservation of wild In Bangladesh, the extent of human–elephant conflicts elephants because farmers cannot control the damage. may be misperceived because this issue receives a great However, the mitigation techniques employed by farmers deal of political and media attention. Based on systemat- are similar to the short-term, active deterrent methods used ically gathered data, Dublin and Hoare (2004) concluded to control human–elephant conflicts throughout Africa and that there is a wide gap between the perceived and actual Asia (O’Connell-Rodwell et al. 2000; Sukumar 2004). The levels of conflict and damage caused by elephants. In fact, majority of farmers reported that they relied on ‘scaring it is often difficult to quantify the extent of the problem and chasing’ as control measures against elephants raiding because monitoring systems can be flawed. Methods for crops. Shouting and using fire to scare elephants away are assessing damage generally and partially depend on indi- traditional techniques that have been practised for centuries rect information supplied by villagers, some of whom may in Bangladesh and also used in African savannahs (Osborn have a vested interest in exaggerating the extent of the dam- and Hill 2005). The use of torches is the commonest and age because they see reporting damage as an opportunity most important deterrent method, and our field observa- to express frustration or anger about their helplessness in tions indicated that farmers probably use this mainly out dealing with conflicts with wildlife. There is an urgent need of tradition rather than efficiency. The respondents also to introduce well-designed and cost effective monitoring stated that guarding fields was especially beneficial in schemes that can produce quantitative data on the impact of reducing human conflict with elephants, as it provided an the various conflict mitigation programmes currently being early warning (Sitati et al. 2003). Huts or watchtowers are implemented in South Asia. constructed along boundaries of farms where elephants fre- quently enter fields. When elephants are spotted, farmers use a combination of loud noises, such as yelling, firecrack- ers, hitting metal objects and cracking whips. Despite the Factors contributing to human–elephant conflict use of lookout points (huts), we have found that human– This study revealed that habitat degradation and fragmen- elephant conflicts continue, indicating that the lookouts tation gradually increased the occurrence of human conflict may not be well positioned, and scientific studies are neces- with elephants, and this is therefore an important issue sary to determine the best place to construct lookout points. in Bangladesh. Extensive logging brings elephants closer We found that the presence of humans, use of loud noise to humans because elephants forage on the secondary and fire as well as other methods could deter elephants vegetation that grows after a disturbance (Lahm 1996). from crop fields, but these tactics are time consuming, Moreover, elephants have increased contact with humans allowing elephants to trample crops in the field for sev- due to changes in land use (Nelson et al. 2003). Isolated eral hours, ultimately causing greater financial losses to agricultural farms along park boundaries may expand the farmers. Farmers were unwilling to use fire guns, proba- human–elephant interface and create a land-use pattern bly because the elephant is a protected species, and thus that favours elephant foraging (Hoare and du Toit 1998). revealing any intent to destroy it may lead to prosecution. As people have continued to transform more land and Therefore, people living near PAs suggested that the Forest forest for settlement, shrinking habitats have compressed Department should assist in the deterrence of wild ele- wildlife populations to levels beyond their carrying capac- phants. Some studies have explored viable options to ity. When the carrying capacity is exceeded, the interaction deter crop raiding elephants (Sukumar 1989; Hoare 1995). International Journal of Biodiversity Science, Ecosystem Services & Management 173 There are many simple, low-cost, active deterrent meth- Natural Resource Management Schemes, CBNRMS ods available, such as burning ‘pepper dung’ or increased (O’Connell-Rodwell et al. 2000) have shown potential to guarding. ‘Pepper dung’ is elephant dung that has been decrease the human–elephant interface, and cultivation mixed with ground chillies and dried in the sun. These of palatable crops within or close to natural fences also cakes are then burned along field boundaries to create nox- achieves the same result (Nelson et al. 2003). Moreover, ious smoke (Osborn and Anstey 2002). Chillies (Capsicum the buffer zones can act as sustainable extractive reserves spp.) have been proven effective in discouraging African for people using forest resources, and this is one solution elephants (Loxodanta africana) from raiding crops (Sitati suggested by the respondents in this study for reducing the and Walpole 2006). conflict between humans and elephants. Another strategy is to make the cultivated areas less desirable for elephants (Chiyo et al. 2005). If it is economically feasible and if environmental conditions permit, farmers could plant a Management implications buffer zone of non-preferred crops like tea, tobacco and Traditional methods of deterring elephants have failed or coffee, which, ideally, can develop into near sub-optimal are disputed in many countries (Sukumar 1991). In many elephant habitats, increasing the barrier between the parts of the world, conservation of elephants by focus- cultivated crops and prime elephant habitats (Osborn and Parker 2003). ing on habitat integrity has proved successful. In this In 2004, the Nishorgo Support Project (NSP) initiated study, the villagers emphasised the improvement of ele- a co-management approach in five PAs (including TGR phant habitat as a way to ensure availability of food for wild and CWS) in Bangladesh on a pilot project funded by elephants. When an area is developed and managed for ele- the US Agency for International Development (USAID). phant conservation, it automatically serves for biodiversity The main challenge for the NSP is that it is hard to conservation, as the elephant is both a keystone and an convince villagers to accept this strategy (Sarker and umbrella species. Therefore, there is an urgent need for an Røskaft Forthcoming). Such a situation has frequently umbrella strategy for conflict mitigation as a conservation been found to arise, mainly due to human needs and tool for Asian elephants. This strategy could include the aspirations being ignored in the initial planning of PAs, following options: considering the home range of targeted often leading to conflict between indigenous people and elephant populations in order to develop landscape-scale conservation plans; vegetation analysis in elephant habi- managers (Corbett et al. 1998). Therefore, to mitigate tat to determine the nutritional value of plant species used human–elephant conflicts, the responsibility for control- by elephants; recording quantitative data on the impact of ling elephant damage could be transferred to farmers under the various conflict mitigation programmes in Asia; con- CBNRMS (Parker and Osborn 2006). In many African ducting scientific research to identify simple, low-cost, countries, CBNRMS have established systems whereby active deterrent methods at local level; and creating for- benefits from wildlife resources can be returned to rural est buffer zones. This can best be led by international communities (O’Connell-Rodwell et al. 2000). Moreover, bodies like the International Union for Conservation of governments and NGOs should introduce vocational train- Nature (IUCN) and the Asian Elephant Specialist Group ing and soft credit schemes for the subsistence of poor (AsESG) in collaboration with multilateral donors (e.g. people living around PAs to generate alternative livelihood Asian Development Bank, The World Bank, Norwegian activities. Agency for Development Cooperation, etc.). These mul- tilateral agencies could then help with mitigation options via investments in a range of countries. Conclusion State-sponsored and voluntary settlement programmes have been endorsed to encourage landless poor people to Wild elephants trample crop fields in seasonal raids take permanent residence around PAs to reduce resentment that coincide with crop maturity and food availability. or bitterness of villagers towards the conservation of PAs Southeast Bangladesh is more vulnerable to crop dam- and elephants (Sarker and Røskaft Forthcoming). With the age by wild elephants and other wildlife than the north current pressures of the human population, it is virtually of the country. Although a few people were able to con- impossible to relocate settlements, but this could prove trol damage caused by other wildlife, no respondents were worthwhile if the forest areas are brought under appropri- able to control damage caused by wild elephants; therefore, ate land-use management. Today, the human population of the affected people were hostile towards wild elephants. Bangladesh has increased to more than 162 million, reach- Habitat degradation and fragmentation, the most impor- ing a density of 1475 people per km (Bangladesh Bureau tant factors in wildlife conservation, have been increasing of Statistics [BBS] 2009). Of the total land area of 14.4 in Bangladesh, a country that has a large population and a million ha, only 2.44 million ha are under forest or tree high growth rate. These factors pose a real challenge not cover. This includes 0.3 million ha of homestead groves only for the conservation of elephants, but also for wildlife and 1.46 million ha of state-owned forest reserves and PAs in general that are simply impossible to control. Therefore, (Akhter and Sarker 1998). Thoughtful land-use strategies, it is important to search for a pragmatic solution to the for example forest buffer zones, under Community-Based elephant conservation problem. 174 A.H.M. Raihan Sarker and E. Røskaft We found that traditional methods for deterring ele- Medicine and Ecology Graduate Group; The University of California, Davis, CA, USA. p. 1–14. phants from crop fields were time consuming. However, Hoare R. 1995. Options for the control of elephants in conflict crop loss was found to decrease with increasing distance with people. Pachyderm. 19:54–63. from PA boundaries. Forest buffer zones close to PAs under Hoare R. 2000. African elephants and humans in conflict: the CBNRMS can ameliorate the conflict between humans outlook for co-existence. Oryx. 34:34–38. and elephants in Bangladesh by involving local resource Hoare RE, du Toit JT. 1998. Coexistence between people and elephants in African savannas. Conserv Biol. 13:633–639. users at every stage of conservation management to ensure Inskip C, Zimmerman A. 2009. Human-felid conflict: a review of a win–win situation in sustainable biodiversity conserva- patterns and priorities worldwide. Oryx. 43:18–34. tion. Therefore, there is a need to adopt a decentralised [IUCN] International Union for Conservation of Nature. 1994. decision-making process that involves all stakeholders at IUCN red list categories. Switzerland: IUCN Gland. a grassroots level to manage nature conservation. In addi- Kangwana KF. 1995. Human-elephant conflict: the challenge ahead. Pachyderm. 19:11–14. tion, we need to introduce vocational training and soft Khan MH, Khan SM, Biswas SR. 2004. Human–elephant credit schemes from local government and NGOs for the conflicts in Bangladesh and assessment of financial subsistence of poor people living around PAs. losses. IUCN-The World Conservation Union, Bangladesh Country Office, Dhaka. Conservation of Asian elephants in Bangladesh. Technical Report. Vol. 4, p. 1–25. Lahm SA. 1996. A nationwide survey of crop-raiding by ele- Acknowledgements phants and other species in Gabon. Pachyderm. 21:69–77. Løe J, Røskaft E. 2004. Large carnivores and human safety: a The current study was funded by NUFU (Project nr.2007/10108). review. AMBIO. 33:283–288. We are grateful to all the people who helped us to collect Marchais J. 2004. Research report on human–elephant conflict the data and to three anonymous referees and members of the in the Okavango Kopano Mokoro community trust villages. editorial board for valuable comments that have improved the Ngamiland, Botswana: Living With Elephants Foundation. paper. Special thanks to the Bangladesh Forest Department for NG 32. allowing us to use the map of Forest Zones in Bangladesh. Mishra C, Allen P, McCarthy T, Madhusudan MD, Bayarjargal Finally, we thank Richard Binns and an anonymous editorial A, Prins HHT. 2003. The role of incentive programs board member for their assistance in improving the English of our in conserving the snow leopard. Conserv Biol. 17: paper. 1512–1520. Mosojane S. 2004. Human–elephant conflict along the Okavango Panhandle in northern Botswana [MSc thesis]. [Pretoria, South Africa]: Department of Zoology and Entomology, References University of Pretoria. Akhter S, Sarker AHMR. 1998. Patronage, tenure, inclusive- Nath CD, Sukumar R. 1998. Elephant–human conflict in Kodagu, ness and local markets: ingredients of a successful strategy southern India: distribution patterns, people’s perceptions for social forestry at Betagi. For Trees People Newslett. and mitigation methods. Bangalore, India: Asian Elephant 35:36–41. Conservation Center, Centre for Ecological Sciences, Indian Bandara R, Tisdell C. 2002. Asian elephants as agricultural pests: Institute of Science. damages, economics of control and compensation in Sri Naughton L, Rose R, Treves A. 1999. The social dimensions of Lanka. Eco Ecol Environ. 65:1–39. human–elephant conflict in Africa: a literature review and [BBS] Bangladesh Bureau of Statistics. 2009. Statistical year case studies from Uganda and Cameroon. Madison (WI): book. Dhaka, Bangladesh: BBS, Government of People’s University of Wisconsin. Republic of Bangladesh. Naughton-Treves L. 1997. Farming the forest edge: vulnerable Chiyo PI, Cochrane EP, Naughton L, Basuta GI. 2005. Temporal places and people around Kibale National Park, Uganda. Geogr Rev. 87:27–46. patterns of crop raiding by elephants: a response to changes Naughton-Treves L. 1998. Predicting patterns of crop damage by in forage quality or crop availability? Afr J Ecol. 43:48–55. wildlife around Kibale National Park, Uganda. Conserv Biol. Corbett T, Lane M, Clifford C. 1998. Achieving indigenous 12:156–168. involvement in management of protected areas: lessons Nchanji AC, Lawson DP. 1998. A survey of elephant crop damage from recent Australian experience. Aboriginal Politics and around the Banyang – Mbo wildlife sanctuary, 1993–1996. Public Sector Management Research Paper no 5, Centre for Yaounde, Cameroon: Cameroon Biodiversity Project and Australian Public Sector Management. Griffith University. Wildlife Conservation Society. p. 1–29. Nelson A, Bidwell P, Sillero-Zubiri C. 2003. A review of Dublin HT, Hoare RE. 2004. Searching for solutions: the evolu- human–elephant conflict management strategies. Oxford tion of an integrated approach to understanding and mitigat- (UK): People and Wildlife Initiative, Wildlife Conservation ing human-elephant conflict in Africa. Hum Dimens Wildl. Research Unit, Oxford University. 9:271–278. O’Connell-Rodwell CE, Rodwell T, Rice M, Hart LA. 2000. Dudley JP, Mensah-Ntiamoah AY, Kpelle DG. 1992. Forest ele- phants in a rainforest fragment: preliminary findings from a Living with the modern conservation paradigm: can wildlife conservation project in southern Ghana. Afr J Ecol. agricultural communities co-exist with elephants? A five- 30:116–126. year case study in East Caprivi, Namibia. Biol Conserv. Gurung B, David Smith JL, McDougal C, Karki JB, Barlow 93:381–391. A. 2008. Factors associated with human-killing tigers in Ogra M, Badola R. 2008. Compensating human–wildlife con- Chitwan National Park, Nepal. Biol Conserv. 141:3069– flict in protected area communities: ground-level perspectives 3078. from Uttarakhanda, India. Hum Ecol. 36:717–729. Hart LA, O’Connell CE. 2000. Human conflict with African Osborn FV, Anstey S. 2002. Elephant–human conflict and and Asian elephants and associated conservation dilemmas. community development around the Niassa Reserve, Center for Animals in Society in the School of Veterinary Mozambique. Report of WWF/SARPO-MZEP. p. 1–29. International Journal of Biodiversity Science, Ecosystem Services & Management 175 Osborn FV, Hill C. 2005. Techniques to reduce crop loss: Sitati NW, Walpole MJ, Smith RJ, Leader-Williams N. 2003. human and technical dimensions in Africa. In: Woodroffe R, Predicting spatial aspects of human–elephant conflict. J Appl Thirgood S, Rabinowitz A, editors. People and wildlife; Ecol. 40:667–677. conflict and coexistence. Cambridge (UK): Cambridge Sukumar R. 1989. The Asian elephant – ecology and manage- University Press. p. 72–85. ment. Cambridge (UK): Cambridge University Press. Osborn FV, Parker GE. 2003. Towards an integrated approach for Sukumar R. 1990. Ecology of the Asian elephant in southern reducing the conflict between elephants and people: a review India II: feeding habits and crop raiding patterns. J Trop Ecol. of current research. Oryx. 37:80–84. 6:33–53. Packer C, Ikanda D, Kissui B, Kushnir H. 2005. Lion attacks on Sukumar R. 1991. The management of large mammals in rela- humans in Tanzania – understanding the timing and distri- tion to male strategies and conflict with people. Biol Conserv. bution of attacks on rural communities will help to prevent 55:93–102. them. Nature (Lond). 436:927–928. Sukumar R. 2004. The living elephants: evolutionary ecology, Parker GE, Osborn FV. 2006. Investigating the potential for behavior and conservation. New York: Oxford University chilli Capsicum spp. to reduce human–wildlife conflict in Press. Zimbabwe. Oryx. 40:343–346. Sukumar R. 2006. A brief review of the status, distribution Rode KD, Chiyo PI, Chapman CA, McDowell LR. 2006. and biology of wild Asian elephants. Int Zoo Yearbook. 40: Nutritional ecology of elephants in Kibale National Park, 1–8. Uganda, and its relationship with crop-raiding behaviour. J Thouless CR. 1994. Conflict between humans and elephants on Trop Ecol. 2:441–449. private land in northern Kenya. Oryx. 28:119–127. Røskaft E, Händel B, Bjerke T, Kaltenborn BP. 2007. Human Treves A, Karanth KU. 2003. Human–carnivore conflict and attitudes towards large carnivores in Norway. Wildl Biol. perspectives on carnivore management worldwide. Conserv 13:172–185. Biol. 17:1491–1499. Sarker AHMR, Røskaft E (Forthcoming). Human perceptions Wang SW, Macdonald D. 2005. Livestock predation by carni- towards conservation of protected areas: a case study from vores in Jigme Singye Wangchuck National Park, Bhutan. four protected areas in Bangladesh. Oryx. Biol Conserv. 129:558–565. Schwerdtner K, Bernd G. 2007. A conceptual framework Woodroffe R, Ginsberg JR. 1998. Edge effects and the extinc- for damage compensation scheme. Biol Conserv. 134: tion of populations inside protected areas. Science. 280: 354–360. 2126–2128. Sitati NW, Walpole MJ. 2006. Assessing farm-based measures Yeager R, Miller NN. 1986. Wildlife, wild death: land use and for mitigating human–elephant conflict in Transmara District, survival in eastern Africa. Albany (NY): State University of Kenya. Oryx. 40:279–286. New York Press.

Journal

International Journal of Biodiversity Science, Ecosystem Services & Management – Taylor & Francis

Published: Dec 1, 2010

Keywords: human–elephant conflict; Bangladesh; attitudes; elephants; compensation

References

  • Human perceptions towards conservation of protected areas: a case study from four protected areas in Bangladesh