Tibetan red deer (<i>Cervus canadensis wallichi</i>) diet composition patterns and associations during lean seasons in Tibet, China

Tibetan red deer (<i>Cervus canadensis wallichi</i>) in the high-altitude environment of the Qinghai-Tibet Plateau could face seasonal challenges from food shortages and nutritional deficiencies but the nutritional requirements are complex. Analyzing diet composition pattern(s) is the first step to disentangle this complexity. From a systematic perspective, we hypothesize that: (A) diet composition pattern or patterns exist within the population and (B) a portion of the diet beyond characterized diet combinations will consist of random combinations. In this study, we investigated diet composition patterns of a Tibetan red deer population distributed in the Sangri Red Deer Reserve, Tibet Autonomous Region, during the harsh lean season. In March 2021 and 2022, we searched for Tibetan red deer in the reserve and collected freshly defecated samples. Diet composition at the individual level was determined using micro-histological analysis, followed by k-means clustering and co-occurrence network analysis to reveal population level diet composition patterns. Diet composition of Tibetan red deer included 14 and 19 plant species (or genera) in 2021 and 2022, respectively. K-means clustering indicated two distinct diet patterns within the population across both sampling periods. In 2021, diet composition of both clusters was dominated by <i>Salix</i> spp. (58.49% and 33.67%). In 2022, <i>R. macrophylla</i> had the highest ranking and occupied 34.83% of diet composition in the first cluster while <i>Salix</i> spp. (39.39%) was the most consumed food in the second cluster. Results of co-occurrence networks showed positively associated food combinations of less dominant food items, with a staple food occurring in all food item pairs in both years. However, randomness accounted for 95.83% and 93% of all food item pairs in 2021 and 2022, respectively, which implies a stable dietary complex system. The 2022 co-occurrence network displayed complex associations, while the 2021 network exhibited limited and simple associations. Our results suggest that Tibetan red deer fulfill their nutritional requirements by consuming high quantities of several food items or a balanced combination of foods with complex co-occurrence associations to overcome potential food shortages, but multilayer networks containing nutritional values and food availabilities are necessary to entangle the complexity of the dietary system.