DNA metabarcoding reveals dietary niche partitioning among threatened Himalayan Ungulates based on chloroplast rbcL...

Diet is the most important axis of niche differentiation in ungulates and mediates coexistence, habitat requirements. Himalaya amplify dietary dynamics because seasonal snow cover and plant phenology decreases resource availability. In this study fecal DNA metabarcoding was used to quantify diet and the dietary niche separation among five sympatric wild ungulates Hangul (Cervus hanglu hanglu), Kashmir markhor (Capra falconeri cashmiriensis), Kashmir musk deer (Moschus cupreus), Himalayan goral (Naemorhedus goral), and Himalayan ibex (Capra sibirica). Fresh fecal samples were collected throughout the Kashmir Valley, western Himalaya across the elevational and habitat gradients. We amplified two plant barcodes (chloroplast rbcL and nuclear ITS2) and sequenced the amplicons using Oxford Nanopore. Diets were analyzed by relative read abundance, functional traits, dietary diversity, niche breadth, and interspecific overlap. Across the seasonal diet profiles, we detected 208 plant species from 174,871 assigned reads. Hangul summer diet exhibits the broadest diet among species (richness = 115 taxa, H' = 4.48, Levins' B = 64.7) with a diffuse core of many co-dominant taxa, indicating generalist mixed feeding during peak plant productivity. Musk deer during autumn showed the narrowest diet (richness = 50, B = 14.3), indicating strong seasonal concentration on a smaller set of plants. Pairwise dietary overlap (Pianka's O) ranged from 0.136 to 0.985, showing both convergence and partitioning among the ungulate species. The results are consistent with a landscape-level coexistence model in which sympatric Kashmir ungulates share a broad forage base but segregate through differences in functional feeding (grazing vs. browsing tendencies), dominance structure, and seasonal shift. The study used composite species-season templates, the results are interpreted as landscape level diet profiles rather than individual or population level variance estimates.