Multiscale distribution models for conserving widespread species: the case of sloth bear Melursus ursinus in India

Abstract

Aim Information on patterns and determinants of spatial distributions remains poorly available for many widespread species of conservation importance. The sloth bear Melursus ursinus in the Indian subcontinent exemplifies this requirement. We aimed at assessing (1) distribution patterns of sloth bears at two spatial scales, (2) ecological and anthropogenic factors that determine bear occupancy. Location We estimated sloth bear habitat occupancy at a nationwide scale across India and at the landscape scale (38, 540 km2) in the Western Ghats of Karnataka, India. Methods We used a grid‐based occupancy approach to determine sloth bear distribution patterns. At the nationwide scale, we used data from questionnaire surveys of field experts (grid‐cell size ~2818 km2; 1326 cells). At the landscape scale, we conducted field surveys of bear signs (grid‐cell size = 188 km2; 205 cells). Detection/non‐detection data from both surveys were analyzed using occupancy modelling methods that account for imperfect detection. We examined the influence of scale‐specific ecological and social covariates on patterns of occupancy. Results Nationwide, sloth bears occupied an estimated 67% of plausible bear habitat in contrast to 46% derived from methods that disregard detectability. Bear distribution was positively influenced by deciduous forests, scrub and barren areas, regions with high human densities and cultural tolerance. At the landscape scale, bears occupied 61% of the area versus 54% estimated from methods ignoring detectability. Occupancy probabilities increased with forest cover and topographic heterogeneity, whereas annual precipitation and human disturbance showed negative effects. Main conclusions Our study underlines the need to integrate human‐modified areas with existing conservation landscapes. Given its widespread nature, functional role, conservation status and relatively benign interactions with humans, we propose recognizing sloth bear as an umbrella species for securing unprotected habitats in India. Protection of widespread species like the sloth bear in other landscapes may complement current conservation strategies for large mammalian communities.

Substituting space for time: Empirical evaluation of spatial replication as a surrogate for temporal replication in occupancy modelling

“…In several field surveys of species’ occupancy, the focus has increasingly shifted towards collection of data from non‐invasive methods such as indirect sign surveys (Karanth et al., ; Thorn, Green, Bateman, Waite, & Scott, ), camera trap surveys (Burton, Sam, Balangtaa, & Brashares, ; Linkie, Dinata, Nugroho, & Haidir, ), interview surveys of experts or local informants (Pillay, Johnsingh, Raghunath, & Madhusudan, ) and field sampling for analysis of faecal DNA (Long, Donovan, MacKay, Zielinski, & Buzas, ). We chose camera trap surveys and indirect sign surveys for our study because both methods involved sampling along forest roads/trails, which are known to be used extensively by sloth bears (Puri, Srivathsa, Karanth, Kumar, & Karanth, ). For our surveys, we defined sites as 13‐km 2 grid cells, large enough to encompass sloth bear home ranges.…”

“…We chose camera trap surveys and indirect sign surveys for our study because both methods involved sampling along forest roads/ F I G U R E 1 Location of Bhadra Tiger Reserve and surrounding land-cover matrix in the Western Ghats landscape, India. White areas represent open agriculture or urban/semi-urban areas trails, which are known to be used extensively by sloth bears (Puri, Srivathsa, Karanth, Kumar, & Karanth, 2015). For our surveys, we defined sites as 13-km 2 grid cells, large enough to encompass sloth bear home ranges.…”

“…Terrain heterogeneity was calculated as the index of ruggedness described by Riley, DeGloria, and Elliot (1999), using the Terrain Analysis tool in QGIS F I G U R E 2 Schematic of the sampling design, showing camera trap locations, sign survey routes and the grid cells (sites) where the two surveys overlap (version 2.2.0). We used the coefficient of variation in ruggedness across pixels in each site, as this measure better represented the heterogeneity in terrain (Puri et al, 2015). We used the mean value of Normalized Difference Vegetation Index (NDVI) in each site as an index of primary productivity of vegetation; lower values of mean NDVI indicating drier areas with lower canopy cover, which would positively influence sloth bear occurrence.…”

Human footprint differentially impacts genetic connectivity of four wide-ranging mammals in a fragmented landscape

“…Empirical evidence across taxa suggests that organisms with specific habitat requirements (and hence habitat association) are usually characterized by low gene flow resulting in strong genetic differentiation (Stuart-Fox et al, 2001; Zayed et al, 2006; Khimoun et al, 2016; Kierepka et al, 2016; Harvey et al, 2017). Among our study species, tigers and sloth bears show stronger habitat associations (Kitchener & Dugmore, 2000; Ramesh et al, 2012; Joshi et al, 2013; Das et al, 2014; Yumnam et al, 2014; Puri et al, 2015) and hence we expect them to have higher genetic differentiation than jungle cats and leopards (Athreya et al, 2013; Gray et al, 2016; Stein et al, 2016).…”

Staying cool or staying safe in a human-dominated landscape: which is more relevant for brown bears?

“…Humans cause most mortality in virtually all large carnivore populations (Woodroffe and Ginsberg 1998 ; Treves 2009 ), including bears (e.g., Sánchez-Mercado et al 2008 ; Bischof et al 2009 ). Human activities cause habitat fragmentation and habitat loss for different bear species (Liu et al 1999 ; Naves et al 2003 ; Escobar et al 2015 ; Puri et al 2015 ; Andersen and Aars 2016 ). Bears perceive changes in the level of risk posed by human activities, which can trigger behavioral responses (e.g., Stillfried et al 2015 ) and/or stress responses (Støen et al 2015 ; Ditmer et al 2015 ).…”