Role of social and ecological attributes in human-large mammal interaction: multi-species approach in Tamilnadu, India

Abstract

Escalating anthropogenic pressure and ensuing human-large mammal conflict (HLMC) incidents jeopardize the wildlife conservation goals as well as security of human life and livelihoods. It results in loss of human life and livestock, damage to crop and property, leading to negative perceptions among the local communities, and further retaliatory killing of conflict animals. Evidentially, large mammal populations are declining at a faster pace majorly due to anthropogenic pressures such as habitat loss and fragmentation, poaching, and retaliatory killing. The interface area, where the forest give way to the agricultural matrix and human settlements, is where the human-wildlife conflict (HWC) is most frequent, consequential and complex. These edge habitats act as a demographic sink, especially for wide-ranging species, having mortality rates and extinction risk within and beyond the forest reserve boundaries often higher than inside core PAs (Woodroffe and Ginsberg, 1998). It is crucial to ascertain the intensity of conflict and its ecological and anthropogenic determinants at these interface areas to develop appropriate mitigation measures to conserve the populations of large mammals, secure human life and promote co-existence in the shared landscapes. The major objectives of the study are to (a) assess socio-ecological characteristics of HLMC across the management gradients, (b) identify disparities in resource allocation between Protected Areas (PAs) and outside PAs and its effect on HWC, and (c) determine the effect of anthropogenic pressure on the distributional patterns of large mammals. My study species include tiger (Panthera tigris), leopard (Panthera pardus), dhole (Cuon alpinus), Asian elephant (Elephas maximus) and gaur (Bos gaurus), that contributes to a major share of HWC incidents in Tamil Nadu. A multispecies study in such large and contiguous landscape of conservation importance can bring out the actual pattern of HLMC in the State and help policy-makers to develop effective mitigation measures. Towards the first objective, I used a systematic grid-based (5x5 km) sampling framework (n=485), across an area of 14,200 km2 encompassing sixteen Forest Divisions (FD), covering PAs, Outside PAs and its Fringe Areas (FAs) of Western and Eastern Ghats part of Tamil Nadu WEGPTN), India, namely, Kanyakumari Wildlife Sanctuary (WLS), Kalakad-Mundanthurai Tiger Reserve (KMTR), Nellai WLS, Srivilliputhur Grizzled Squirrel Sanctuary, Megamalai WLS, Coimbatore FD, Kodaikanal WLS, Theni FD, Nilgiri FD, Gudalur FD, Mudumalai Tiger Reserve (MTR), Sathyamangalam Tiger Reserve (STR), Erode FD, Dharmapuri FD, Cauvery North WLS (CNWLS) and Hosur FD. The data was collected between November 2017 and March 2020 based on conflict incidents from tiger, leopard, dhole, elephant, and gaur for the past five years, through semi-structured interviews (n=1460) of local communities. I identified the local people’s perception towards coexistence and the socio-economic factors driving them. I observed that 90% (n=1248) of the surveyed population had a negative perception towards coexisting with wildlife, while 7% (n=103) expressed a positive perspective, and 3% (n=38) maintained an indifferent stance. Notably, 79% (n=1091) of the sampled population resided within Fringe Areas (FAs), 13% (n=178) in Protected Areas (PAs), and 9% (n=120) in Reserved Forests (RFs). Among them, 90% (n=160) of the population from PA had negative, 9% (n=16) had positive and 1% (n=2) had indifferent attitude towards coexisting with wildlife; In RFs, 73% (n=88) and 27% (n=32) had negative and positive view towards coexisting with wildlife, respectively. Whereas, in FAs, 92% (n=1000), 5% (n=55) and 3% (n=36) had negative, positive, and indifferent view towards coexisting with wildlife, respectively. Top multinomial regression models discerned the most influential variables determining the perception of local communities regarding coexistence with wildlife. These models elucidated that people’s perception of coexisting with wildlife, especially large mammals, was negatively influenced by decreased availability of basic amenities, lack of conductance of awareness program, an increased perception of the rise in HLMC incidents, increased prevalence of disease in livestock, increased negative view on Forest Department, increased fear of wildlife, decreased intensity of livestock depredation and increased crop depredation. Forty eight percent (n=694) of households reported having experienced human-carnivore conflict in the form of livestock depredations. Most of the livestock were killed by large carnivores in the mornings up until mid-day, between 6 am and 12 pm, coinciding with livestock grazing time inside the forest. Only a fewer tiger and leopard attacks occurred during the nights, followed by evening; the attacks were minimum during afternoon hours. Dhole attacks peaked in the early mid-evening hours and afternoon; no incidences of livestock predation by dholes occurred during night. Using the Generalized Linear Model (GLM) approach, the importance of potential predictor variables of HLMC was examined. Livestock predation risk by the tiger, leopard and dhole were found to be driven by the size of livestock species, dependency of local people on the forest, topography, proximity to the water body and forest boundary, precipitation, and forest cover. The probability of livestock depredation by tiger increased with a higher relative abundance of cow/buffalo, higher dependence of people on forests, greater proximity to water sources, shorter distance to forest boundaries, relatively 3 higher elevations and villages located within the forest areas, particularly in core areas of PAs; Livestock depredation risk probability from leopard increased with relatively higher abundance of goat/sheep, domestic dog, higher dependence on the forest by households and an increase in topographical slope. The probability of livestock depredation by dhole increased with higher forest cover loss, increase in the relative abundance of goat/sheep, greater dependence on the forest by households, and relatively lower precipitation. The risk of predation from leopard and dhole exhibited high spatial overlap, and predation by leopards was higher than dhole and tiger. Livestock predation by leopard and dhole was frequent in open areas of RF and buffer zones, while that from tiger occurred in densely forested core regions of PAs. Forty five percent (n=660) and ten percent (n=140) of households surveyed reported experiencing elephant and gaur conflict, respectively. Of these, both elephant and gaur are largely involved in crop damage, followed by property damage, human injury/death. Unlike large carnivores, most of the conflict incidents by elephant and gaur peaked in the night, followed by morning, evening and afternoon. Using GLM, I found that the major variables that influenced elephant and gaur conflict with man were: percentage of crop cover, distance to the water source, the number of major crops grown, the number of minor crops grown, mean annual precipitation and mean slope. The probability of human-elephant conflict increased with increased crop cover, increase in distance to the water source, number of major and minor crops grown, while the conflict decreased with an increase in mean annual precipitation and mean slope. The risk of humangaur conflict increased with lower percentage of crop cover, increase in distance to the water source, number of major crops grown, increase in mean annual precipitation and mean slope. I found that households inside RFs to report elephant conflict the most, followed by PAs and FAs. The reported gaur conflict was more in FAs, followed by RFs and PAs. I identified ca. 175 km2, 420km2, 430km2, 8900 km2, 625 km2 of the total sampled area, prone to tiger, leopard, dhole, elephant and gaur conflict with man, respectively. I conducted a semi-structured questionnaire survey with the Forest Range Officers (FROs) of all sixteen FDs across the WEGPTN, between November 2018 and March 2020, to address the second objective. I investigated the patterns of human-wildlife interactions, employed mitigation measures and their effectiveness, poaching incidents, and availability of manpower and resources. The study highlighted the disparity in resource allocation among PAs and outside PAs though the intensity of conservation issues was similar. Over 50% of the FROs reported an increase in human-wildlife conflict incidents during the past five years. Major species involved in the conflict in PAs and outside PAs were elephant, wild pig (Sus scrofa), 4 leopard, bonnet macaque (Macaca radiata), followed by sloth bear (Melursus ursinus), sambar (Rusa unicolor) (more in PAs), gaur (more in outside PAs), dhole, chital (Axis axis), and tiger. There were no variations in the mitigation measures employed in PAs and outside PAs. The endangered species, namely tiger (only in PAs), leopard, elephant, and pangolin (Manis crassicaudata) were amongst the fourteen wildlife species poached in the PAs and outside PAs, although wild pig, sambar, black-naped hare (Lepus nigricollis), and chital dominated the list. Prevalent methods for poaching were snares, trained dogs, guns (dominant in PAs), followed by hand-made bombs, poisoning, cage traps, electric traps (only outside PAs), and food baits. Fund availability, anti-poaching camps and anti-poaching watchers were insufficient in outside PAs. Over 50% of the FROs reported a decrease in poaching incidences in the past five years with improved wildlife management strategies. Based on the information gathered, I chose nine high-conflict zones, covering PAs and outside PAs across the WEGPTN (MTR, STR, CNWLS, Gudalur FD, Nilgiris FD, Coimbatore FD, Erode FD, Dharmapuri FD and Hosur FD) and conducted ecological sign survey and vegetation survey between May 2019 and March 2021 to ascertain the occupancy of large mammals and quantify degree of human disturbances inside the forests. Considering the average home range size of the dominant large mammal species, 10x10 km sampling grids were laid across the entire study area. I used a systematic grid-based Markovian model occupancy framework to identify the impact of ecological and anthropogenic variables (prey abundance, habitat features, direct human disturbance and habitat patchiness) on large mammal occupancy and mapped the Least Cost Pathway (LCP) and assess the Cost-Weighted Distance (CWD) analysis at a 100km2 scale. I found that tiger and dhole occupancy was driven by the abundance of large and medium-sized prey. However, leopards relied on the available small prey species and utilised perforated forest patches surrounding PAs. Broadly, large mammal occupancy was more in the contiguous forests with ample water sources and decreased in habitats with increased human disturbances and habitat patchiness. I found that tiger occupancy was highly confined to the contiguous forests of PAs while the rest of the large mammal species had scattered but concentrated occupancy in PAs. All the study species exhibited least habitat selection and permeability around highly fragmented landscapes, located farther from PAs. Accordingly, as the distance from PAs increased, the CWD also increased. I identified 87, 67, 78, 77 and 81 actively used LCPs respectively for tiger, leopard, dhole, elephant and gaur. There were 17, 41, 23, 26 and 23 optimized LCPs for tiger, leopard, dhole, elephant, and gaur, respectively. Among the 17 optimized LCP for tiger, Euclidean distance value ranged from 22 to 86 km with CWD values between 25 and 150 km. From 41 optimized LCPs for leopard, the Euclidean (5) distance values ranged between 14 and 139 km, whereas, CWD values stretched between 34 and 1,928 km. The Euclidean distance values of dhole were between 22 and 106 km with a CWD value ranging from 24 and 512 km in the generated 23 optimized LCPs. For elephant, there were 26 optimized LCPs, with Euclidean distance extending between 22 and 86 km; however, the CWD values ranged between 24 and 134 km. Among the 23 optimized LCPs for gaur, Euclidean distance value ranged from 22 and 78 km but the CWD value extended between 24 and 116 km. The LCP maps identified several pinch points for large mammal movement particularly outside PAs, including fragmented parts of Nilgiris, Gudalur, Coimbatore, and Dharmapuri FDs, highlighting critical regions for restoring core habitat connectivity. The substantial proportion of large mammals inhabiting outside PAs demands strong efforts for habitat restoration and minimising further habitat degradation, which can also reduce the risk of HLMC. HLMC are inevitable when natural landscapes undergo modifications. Bringing down the HLMC incidents to a tolerable limit through effective mitigation measures is the viable solution to ensure harmonious existence of people and wildlife in a forest ecosystem. Findings underscore the importance of identifying factors affecting people’s perception towards coexistence, priority conflict hotspots, large mammal movement pathways, ecological and anthropogenic determinants of HLMC and large mammal occupancy, and the effectiveness of employed conflict mitigation measures. The conflict hotspot and corridor maps provide a powerful visual tool to understand patterns of HLMC and movement pattern of large mammals, respectively, at a landscape level to the stakeholders. The identified HLMC zones and the high resistance areas, regardless of PAs/outside PAs, should be provided with sufficient resources to mitigate conflict and landscape impermeability. Improved livestock husbandry practices, constructing predator-proof enclosures, and using shock-collars in high livestock depredation areas can be effective in decreasing the frequency of human-large carnivore conflict. Changing the cropping patterns, using a combination of elephant proof trench and electric fence, using early warning systems through Artificial Intelligence (AI) sensors, mobile technology and drones in the high conflict areas can reduce elephant and gaur conflict with humans. Providing timely compensation to the affected household is also an important mitigation measure to reduce the negative attitude of local communities. The study emphasises the need for forest habitat restoration by abating biotic pressure within the forest boundary, thereby enhancing prey abundance, availability of water resources, improved habitat connectivity to increase large mammal occupancy, reduce large mammal movement resistance within the forests, reduce HLMC and promote coexistence.