Predator-prey interactions in subtropical forest and ecology and conservation of swamp deer or barasingha (Rucervus duvaucelii duvaucelii) : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Conservation Biology at Massey University, Auckland, New Zealand

The tiger, an apex predator, is an indicator of a healthy ecosystem. Therefore, the conservation of this immensely important umbrella species necessitates ecological knowledge on its interaction with co-predators and status of significant prey species sharing the habitat. Royal Bengal tiger (Panthera tigris tigris) and common leopard (Panthera pardus fusca) are two sympatric carnivores in suitable habitats of the Indian subcontinent where sympatric ungulates cervids are prevailing as major prey species. This thesis investigated the diet overlap of these sympatric apex predators and the population status of prey species in the subtropical lowland landscape popularly known as Terai Arc Landscape (TAL), Nepal. This thesis further explored ranging behaviour, habitat preference, movements, conservation genetics and food habits of the swamp deer or barasingha (Rucervus duvaucelii duvaucelii), the largest cervid prey species of tiger in the western portion of TAL, Nepal. The line transect survey revealed high density and biomass of major cervid prey species compared to other tiger bearing protected areas, with large-sized swamp deer, medium-sized spotted deer (Axis axis) and hog deer (Axis porcinus) being main prey species. Scat analysis revealed that wild ungulates were the main food of both predators. Large-sized wild preys (mainly swamp deer) were found more frequently in tiger scats, and domestic cattle were found more frequently in leopard scats. Both predators consumed the medium-sized spotted deer in high proportions resulting in a high diet overlap. Swamp deer played a critical role in diet partition between the two cats. Dominant tigers displaced leopards to the degraded fringe habitat where the latter subsisted on the domestic livestock, a major cause of human-wildlife conflict. The monitoring of nine radio-collared swamp deer showed a high 95% Minimum Convex Polygon (MCP) and 95% Fixed Kernel (FK) home ranges compared to other sympatric ungulates such as spotted deer, hog deer, and barking deer (Muntiacus muntjac). There was no significant variation in 95% FK annual home ranges between sexes and across three seasons. Overall, the grassland was the most preferred habitat of swamp deer in all the seasons, and the dense Sal (Shorea robusta) forest was the least preferred. In the monsoon season, the moderately dense Sal forest followed the grassland. Within grassland, swamp deer preferred Imperata cylindrica assemblage followed by Imperata cylindrica – Narenga porphyrocoma and Narenga porphyrocoma assemblage. A range shift from the grassland to the Sal (Shorea robusta) forest, only during the monsoon season, indicated that the grassland had reached the carrying capacity during this season. The radio-collared swamp deer crossed the international border. During the rutting season, they consistently used contiguous habitat patches of Lagga-Bagga area of Pilibhit Tiger Reserve, India, thus emphasising the need for transboundary cooperation to conserve this species. The genetic variability, population structure and effective population size of Shuklaphanta National Park (SNP) and Bardia National Park (BNP) populations of swamp deer were examined using the mitochondrial (mt) DNA and microsatellite markers. The analysis revealed moderate to high genetic diversity compared to other swamp deer populations in India. Neutrality tests, which are used to evaluate demographic effects, did not support population expansion. The multimodal pattern of mismatch distribution indicated that both swamp deer populations are under demographic equilibrium. Furthermore, population bottleneck analysis indicated no signature of a bottleneck for both populations. Bayesian cluster analysis and population differentiation test revealed two population clusters with low population differentiation. The effective population size in BNP was below 50, which is often regarded as a threshold below which inbreeding depression is likely to occur. It is recommended to design and implement an effective conservation strategy to enhance the genetic diversity and increase the population size of swamp deer in BNP through an in-situ conservation program and translocation of some breeding individuals from SNP to BNP. Faecal microhistological analysis of swamp deer, sympatric hog deer, and spotted deer from both grassland and the Sal forest habitats revealed that graminoids constituted the majority of the diet of these species in both habitats. However, the proportion of woody plants in diets of spotted deer was significantly higher than the other two. Apart from the graminoids, woody plant Shorea robusta and herb Phoenix humilis were major plant species consumed in the Sal forest. Among graminoid species, early successional tall grasses, especially Saccharum spontaneum, were the dominant food of all three deer species in both habitats. The importance of early successional tall grasses in their diet emphasised the key role of the threatened alluvial floodplains in conserving threatened mammal species in South Asia. Swamp deer foraged more in late succession tall grasses (Saccharum narenga and Themeda spp.) and short grasses (Imperata cylindrica, Cyperus spp., Chrysopogon zizanioides, Cymbopogon spp.) than hog deer and spotted deer. Despite the similarity of their diet, the three ungulates coexisted through differential consumption of plants species and seasonal habitat partitioning