Abstract:Animal items consumed are important energy and nutrition resources for the Great Bustard (Otis tarda) during the breeding period. However, due to limitations of traditional diet analysis methods, the animal derived diet is still unclear, and the diet difference between different breeding sites is not well explored. The high-throughput sequencing (HTS) has offered major advantages over traditional methods in diet analysis, such as low work cost, big dataset and taxa assignment at a high fine scale level. Based on fecal sampling, we applied the HTS approach to identify animal items consumed by the Great Bustard in two core areas of Tumuji Nature Reserve (Kaoshan and Maanshan) during the breeding season, and compared the spatial variation in diet diversity. The diet diversity was estimated by Simpson and Shannon index, and the diet difference was compared between Kaoshan and Maanshan using ANOSIM test and nonmetric multi-dimensional scaling analysis (NMDS). The minimum sampling intensity (n = 11) is sufficient to make the detection limit of MOTUs reach the plateau (Fig. 1). Of 24 fecal samples, 29 unique prey DNA (Table 1) were discovered, corresponding to invertebrate species, with the prey group of the Coleoptera (the Arthropoda) accounting for the highest proportion (44.83%). Below the order level, the family Scarabaeidae accounted for the highest (24.14%), followed by Acridiidae (13.79%), Meloidae (10.34%) and Thripidae (6.89%). The consuming frequency of a given dietary species and the number of prey per individual in Maanshan were both significantly higher than that in Kaoshan, as well as the diet diversity index (Fig. 2). The diet of Great Bustard was different between two key breeding sites (Fig. 3). This study facilitates investigating the relationship between diet and habitat selection, as well as understanding of foraging strategies during the breeding period. By providing dietary reference data, it could help the conservation managers to effectively protect and restore the habitats of Great Bustard. Meanwhile, the basic framework could be applied in studying the feeding ecology of other species using the HTS method.