3.2 Dietary composition and seasonal variation in animal food items
We examined seasonal variations in the diet composition of A. squamipes , especially during times of resource limitation (e.g., in winter). Based on the full year, our results showed that although some small insects (ants, spiders, crickets, and beetles) were consumed, the Chinese mole shrew is primarily an earthworms-eating shrew with a semi-fossorial foraging mode. Using order-level taxonomy only, species of Haplotaxida, Stylommatophora, Hymenoptera, Orthoptera, and Moniligastrida dominated the diet with species of Haplotaxida representing the highest FO (100%) and highest taxonomic richness (>74%) of consumption (Supplementary Table 3). Notably, the consumption of Haplotaxida significant decreased (spring vs. winter: 83% vs. 45%, P=0.002; summer vs. winter: 81% vs. 45%, P=0.03; autumn vs. winter: 90% vs. 45%, P=0.004) during winter (Supplementary Table 3). Thus, earthworms were considered as the major food item in the diet of A. squamipes. In addition, as the common prey of shrews, arthropods (such as Orthoptera, Coleoptera, Dermaptera, Diptera, and Lepidoptera) were also detected but at low frequencies and relative abundances in A. squamipes diet (Supplementary Table 3).
At the species-level, the dominant (top five) animal species in terms of both FO and relative abundance were Metaphire californica ,Amynthas morrisi , Amynthas corticis , Deroceras laeve , and Camponotus thadeus (Figure 4A, 5 and Table 2). Among the total animal food items, 12 different species of earthworms belonging to four families (Megascolecidae, Enchytraeidae, Moniligastridae, and Lumbricidae) accounted for 70%−80% of the animal-derived diet (Table 2, Figure 5 and Supplementary Table 4), indicating that these soil invertebrates are extremely abundant and diverse in the studied region. Among them, Metaphire californicawas most frequently detected in all samples, contributing 19.8%−60% of the relative abundance of overall prey consumption and peaking at 60% in autumn (Table 2 and Figure 5). The second-richest prey (Amynthas morrisi ) were eaten more frequently and made up a larger proportion (>34%) of the available prey in spring and summer than in autumn and winter (<5%; Manne−Whitney U test: P=0.008). In addition, we found a trend in the consumption of earthworms that shifted from higher numbers of earthworms during spring and summer to lower levels during autumn and the least in winter (Figure 1B and Supplementary Table 4). Thus, the proportions of earthworms consumed by Chinese mole shrew during winter were significantly decreased (P < 0.01) (Figure 1B). Meanwhile, the proportions of the all earthworm species significantly decreased (P<0.01) during winter (Figure 1B), because their availability of was reduced. Our analysis indicated the animal-derived diets of A. squamipescontain a high prevalence and diversity of earthworms. However, during winter, Chinese mole shrew predominantly preyed on Camponotus thadeus and Deroceras leave with a high FO (50%) and in higher proportions compared to other seasons (Table 2, Figure 4A and 5). Therefore, our study revealed that Chinese mole shrews have a broad diet comprising many different invertebrates of various sizes (dominantly earthworms) based on COΙ metabarcoding approaches.