|dc.description.abstract||The snow leopard (Panthera uncia) is an elusive large carnivore distributed in the mountains of Central Asia. The highly camouflaged fur and harsh environment of the cat’s habitat have made ecological studies challenging. To date, most snow leopard research have been conducted in small study areas, or “hotspots”, and less than 2% of the global snow leopard distribution range has been sampled systematically. Basic ecological knowledge of the species across large landscapes is essential in order to develop effective conservation and management plans. The main aim of this thesis was therefore to provide data on some key ecological aspects of snow leopard ecology on a large spatial scale. The study was primarily based on field surveys of prey density, collection of snow leopard scats for studying diet and estimating density, and semi-structured interviews for assessing conflicts with humans. Most of the data were collected within 26 grid cells of 5x5 km2 each, totaling 650 km2 in a 4.393 km2 area within two protected areas (Annapurna and Manaslu) in the central Himalayas, Nepal.
Individual snow leopards were identified from fecal DNA, and abundance and density were estimated using spatially explicit capture recapture models (SECR). The average density estimate of snow leopards for the entire projected model area (15,322 km2) was 0.95 (SE 0.19) animals per 100 km2 (95% CL = 0.66 - 1.41). The predicted density within the sampling area of ca. 4393 km2 was 1.07/100 km2. An extrapolation based on this estimate rendered an abundance of 48 individuals within our 4393 km2 sampling area, and a country-wide abundance of 140 animals within 12815 km2 of potential snow leopard habitat. This estimate is less than half of the current, official estimate made for the whole country (DNPWC 2017). The best SECR model described the variation in density as a quadratic function of altitude, and detection of scats as a linear function of topography.
The diet analyses of snow leopards and wolves, based on genotyped scats (snow leopards, n=182 scats; wolves n=57 scats), revealed that both species preferred wild prey. Snow leopards typically preferred cliff-dwelling wild ungulates (mainly bharal), whereas wolves preferred plain-dwellers (Tibetan gazelle, kiang and Tibetan argali). Livestock content in the diet of both predators was low compared to their availability (snow leopard = 27%; wolf = 24%) and snow leopards significantly avoided livestock. A seasonal difference in diet was noted in wolves, but not in snow leopards, and I believe this was mainly due to seasonal differences in the availability of important prey animals, such as marmots. Snow leopards avoided yaks, probably due to their large size, and preferred horses as they are often left unattended in the pastures, and also preferred goats as they are of optimal body size. Male snow leopards consumed more livestock than females.
My investigation on human-carnivore conflicts revealed that the overall losses to various carnivores was only around 1% (N=428) of the total livestock holdings. More than half of the livestock depredation events (62%) was caused by snow leopards. Small stock (goats and sheep) were killed more often by snow leopards and wolves than larger domestic animals (yak, horse, cattle). Losses occurred in all seasons, however, a significantly larger proportion of livestock depredation occurred during summer than in winter. More kills were observed in summer than in winter by wolves and by other predators. Among farmers owning large proportions of large stock (yak, horse and cow), the probability of loss increased with herd size; such a trend was not evident among farmers owning mainly small stock. The relationship between wild prey density and livestock loss was complex; the lowest loss probability occurred in areas of high wild prey density and low livestock density, but it was highest in areas where both livestock and wild prey were abundant.
This dissertation has demonstrated the importance of spatial scale in research, monitoring and protection of snow leopards. I detected marked spatial variations in snow leopard diet and density, which entails that small-scale studies may produce highly variable results due to contrasting ecological conditions. Moreover, the overall livestock loss to carnivores was minimal, but with large local variations. Ongoing conflict resolution strategies are being implemented on an ad hoc basis, and there is a need to develop new plans that take into account local variations in order minimize risk of loss, as well as to improve systems for recording losses. Furthermore, there is a need to increase monitoring efforts outside the protected areas to ensure the survival of snow leopards and other predators in the Himalayas.||nb_NO