Abstract
Diet composition is among the most critical dimensions of animal ecology, yet seasonal dietary diversity has rarely been investigated in sympatric herbivores. This study used DNA metabarcoding to conduct an analysis of seasonal variations in diet composition and trophic niches for sympatric sika deer, Reeves' muntjac, and Chinese hare in Taohongling National Nature Reserve (TNNR). The results showed that Smilax china (11.79%) was the leading food eaten by sika deer in summer, whereas dominated by Rubus spp. (36.42%) and Loropetalum chinense (25.48%) in winter; Rubus spp. accounted for the majority of Reeves' muntjac's diet throughout the year. In comparison, the Chinese hare primarily consumed Smilax china from winter to spring but changed to Poa annua (10.81%) and Setaria viridis (23.05%) in summer and fall. Compared to other seasons, significant differences (Shannon index, p < 0.05) occurred in spring and summer, showing higher diversity of food items across the three herbivorous. Nonmetric multidimensional scaling (NMDS) analysis suggested significant partitioning in the food items of Chinese hares compared to the two ruminants. Both sika deer and Reeves' muntjac occupied a wider niche breadth and dietary diversity in summer, reflecting generalised feeding habits (Sd: Ba = 0.06; Rm: Ba = 0.04) and lower in fall (Ba = 0.01) with stronger selectivity and specialization, which was consistent with the optimal foraging theory. Notably, no significant difference was indicated in seasonal niche breadth for Chinese hare (p > 0.05). The niche overlap indices were 0.989 (fall) and 0.831 (winter) between sika deer and Reeves' muntjac, indicating a higher dietary similarity and overlap. However, differences in foraging plant taxa and abundance ratios may facilitate dietary niche partitioning. The diet of herbivores reflected plant-herbivore interactions and seasonal diet differences were correlated with feeding strategies, which facilitate coexistence and reduce competition of co-occurring species in the food dimension.