Prey partitioning between sympatric wild carnivores revealed by DNA metabarcoding: a case study on wolf (Canis lupus) and coyote (Canis latrans) in northeastern Washington

• Published in: Conservation genetics Vol. 22; no. 2; pp. 293 - 305
• Main Authors: Shi, Yue, Hoareau, Yves, Reese, Ellen M., Wasser, Samuel K.
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.04.2021; Springer Nature B.V
• Subjects: Alces alces; Animal Genetics and Genomics; Biodiversity; Biomedical and Life Sciences; Canis latrans; Canis lupus; Carnivores; Composition; Conservation Biology/Ecology; Coyotes; Deer; Deoxyribonucleic acid; Diet; DNA; Domestic animals; Ecology; Ecosystem management; Environmental changes; Environmental impact; Evolutionary Biology; Feces; Lepus americanus; Life Sciences; Livestock; Lupus; Moose; Niche overlap; Odocoileus; Partitioning; Plant Genetics and Genomics; Prey; Research Article; Spatial variations; Species; Sympatric populations; Technology; Temporal variations; Wolves; Scats; Carnivore-livestock conflicts; Prey partitioning; Wolf recovery; Diet; DNA metabarcoding
• ISSN: 1566-0621; 1572-9737
• DOI: 10.1007/s10592-021-01337-2

Accurate dietary inference of wild carnivores is essential to understand their impacts on the ecosystem and manage carnivore-livestock conflicts. Diet analysis with fecal DNA metabarcoding technology can help deliver valuable insights with fine-grained resolution. The recovery of wolves in Washington, USA offers an excellent opportunity to study the diet of this iconic carnivore species and explore prey partitioning between sympatric wolves and coyotes. We characterized the diet composition and spatiotemporal dietary variations in each species using fecal DNA metabarcoding technology on 202 fecal samples from wolves (N = 99) and coyotes (N = 103) collected across three wolf pack ranges and two seasons in northeastern Washington. We also quantified the diet niche overlap between these two canid species. In total, 19 different prey items were detected, with most assigned at the species level. Frequency of occurrence (FOO) data showed that wolves primarily preyed upon deer ( Odocoileus sp . ; 47.5%) and moose ( Alces alces ; 42.4%). Coyotes also consumed moose (30.1%) and deer (21.4%), but snowshoe hares ( Lepus americanus ) were the most common prey (61.2%) in the coyote diet. Multiple samples were found to contain DNA from domestic animals, including pig, rabbit, and cow. Results on diet composition using FOO and relative read abundance were qualitatively similar, indicating a strong biological pattern. We found significant spatial variations in the wolf diet composition ( p  = 0.001) and significant spatio temporal variations in the coyote diet (pack range: p  = 0.003; season: p  = 0.023). Dietary overlap between these two canid species varied with pack ranges and seasons ( O = 0.08–0.74). Our study demonstrates that fecal DNA metabarcoding is an efficient non-invasive tool to characterize high-resolution diet profiles of carnivores and monitor their dietary changes over space and time. Limitations of fecal DNA metabarcoding are also discussed.