Tracking Ungulate Diet: Comparing Observational and DNA Metabarcoding Tools

Large mammalian herbivores (LMH) are abundant in grazing ecosystems and play a pivotal role in shaping vegetation characteristics. However, accurately determining their diets through traditional methods, such as direct observations, remains challenging, particularly in natural communities and mixed-species grazing systems. Recent studies have shown that DNA metabarcoding can effectively identify the plant composition in LMH diets as well as the plant-dwelling arthropods (PDA) incidentally ingested by LMH while grazing. Given the high specificity of herbivorous insects to their host plant, we hypothesize that DNA metabarcoding of arthropods ingested by LMH could offer valuable insights into their feeding preferences. The goal of this study is to evaluate the accuracy of plant and arthropod DNA metabarcoding methods in identifying the diets of sheep and cattle and to compare their performance with direct observations and known dietary patterns from the literature. To test this, we collected fecal samples from sheep and cattle grazing in the northeast Asian grasslands. We amplified arthropod DNA using COI mitochondrial markers and plant DNA using ITS1 markers, followed by Illumina sequencing. Additionally, we conducted field observations to identify plants grazed by sheep and cattle. The DNA metabarcoding methods provided a comprehensive view of the LMH diet. Both DNA metabarcoding methods successfully detected dietary differences between sheep and cattle, with sheep primarily consuming nutrient-rich forbs and cattle predominantly grazing on Poaceae, consistent with known foraging behaviors. While the constant presence of arthropods across multiple samples suggests that DNA of ingested arthropods could provide complementary information regarding LMH foraging behavior, we found such to be rather limited. However, our findings confirm that plant DNA metabarcoding is a reliable and accurate method for identifying LMH diets.