Abstract: 1. Conservation policy for the high mountains of Asia increasingly recognises the need to encompass large multi-use landscapes beyond the protected area network. Due to limited long-term research in this region, our understanding of even fundamental processes, such as factors regulating large mammal populations is poor.
2. Understanding the factors that regulate animal populations, especially those generating cyclicity, is a long-standing problem in ecology. Long-term research across multiple taxa (mainly from Europe and North America) has focussed on the relative roles of food and predation in generating cyclicity in population dynamics. It remains unclear how trophic interactions that are influenced by anthropogenic stressors can affect population dynamics in human-modified landscapes.
3. We present a 10-year study to compare the effects of livestock grazing on density-dependent dynamics in two populations of bharal, Pseudois nayaur, in the Himalayas. We combine this with a mechanistic understanding of whether density dependence in these two sites acts predominantly by affecting adult survival or recruitment. We compared and quantified density dependence in the bharal population by fitting Bayesian Gompertz state-space models.
4. We found evidence for negative density dependence which indicates possible cyclic dynamics in the bharal population of the site (Tabo) with low livestock density. The population dynamics of this site were driven by recruited offspring—with a 2-year density-dependent lag effect—rather than adult survival. In the site with high livestock density (Kibber), this density dependence was not detected. We postulate the potential role of excessive grazing by livestock in affecting offspring recruitment, thereby affecting the bharal population in Kibber.
5. Synthesis and applications: Our results suggest that conservation action to facilitate wild herbivore population recovery, such as the development of protected areas and village reserves, needs to account for density-dependent regulation. Sites with trophy hunting require continuous monitoring to understand the effects of density dependence so that appropriate hunting quotas can be formulated.

Abstract: Habitat selection of Snow Leopard ( Unica unica) in Beita Mountain of the Altay Mountain system in northeast Xinjiang was conducted from September to October 2004. Six habitat features of 59 sites used by Snow Leopard and 30 random plots were measured by locating 15 transects surveys in the study area . Vanderploge and Scaviaps selectivity index was used to assess Snow Leopardps selection for the different habitat parameters. Principal Component Analysis was used as the primary factor . The results indicated that Snow Leopard preferred the altitude between 2000 – 2 200 m and avoided 2 600 – 3 000 m ; selected cliff base , ridgeline and avoided hillside and valley bottom ; utilized the shrub and rejected the forest ; selected the nongrazing area and avoided the slightly broken region ; preferred north orientation and rejected the south orientation. The results show that grazing status , vegetation type , topography and the ruggedness are the primary factors for the habitat selection of Snow Leopard.