DNA metabarcoding identifies diet composition of introduced rodents on Oʻahu, Hawaiʻi

Sara M. Gabrielson, Biological Sciences, Northern Arizona University, Flagstaff, AZ, Rebecca L. Mau and Jeff T. Foster, Pathogen and Microbiome Institute, Northern Arizona University, Flagstaff, AZ, Donald R. Drake, Botany Department, University of Hawai'i at Manoa, Honolulu, HI, Jinelle H. Sperry, Engineer Research and Development Center, Champaign, IL, Jinelle H. Sperry, Natural Resources and Environmental Sciences, University of Illinois at Urbana-Champaign, Champaign, IL

Background/Question/Methods
Rodents are some of the most ubiquitous and successful invasive species globally, and are particularly destructive on islands. Three rodent species, black rat, Rattus rattus; Pacific rat, R. exulans; and house mouse Mus musculus, are abundant throughout forested ecosystems on Oʻahu, Hawaiʻi. They eat a wide variety of plants and animals due to their omnivorous diets. Detailed diet analyses can provide a more complete understanding of the roles of invasive rodents in ecosystems and can identify potential indirect effects of rodent control efforts. Invasive rodent diets have been studied through captive-feeding trials, visual stomach content identification, and stable isotope analysis; however, none of these methods have been able to identify natural diet diversity at a fine taxonomic scale (e.g. species). Using DNA metabarcoding analyses we provide a detailed description of invasive rodent diets at fine taxonomic levels and compare natural diets of these rodent species across multiple sites. Rodents were trapped at seven forest communities on Oʻahu from November 2014 to November 2016, and 531 fecal samples were collected. DNA from samples were extracted and sequenced, using metabarcoding targeting the ITS2 region and COI gene, to identify plant and arthropod species in the samples, respectively.
Results/Conclusions
A broad range of native and non-native plants and arthropods were identified, with significant diet differences among some comparisons of rodent species and sites. Preliminary results indicate that house mice ate a more diverse plant diet than black rats, plant species richness was higher in house mouse diets compared to black rats (H = 4.2, p = 0.041), while no richness difference was found in Pacific rat. There were no strong plant species richness differences among diets at different sites. Black rat plant diet community composition was different compared to house mice (pseudo-F 6.9 p = 0.001) and Pacific rat (pseudo-F 2.8 p = 0.004), indicating that black rats ate a distinct suite of plants. At the site level, rodent plant diet composition at each of the seven sites were all significantly different (pseudo-F 7.7, p = 0.001). Arthropod diet analyses are ongoing but indicate that rodents consume both native and non-native taxa particularly spiders, flies, moths, and beetles, and these varied by rodent species and site. Clear site differences in rodent diets suggest that the omnivorous nature of invasive rodents allows for broad and varying diet choices and therefore indirect rodent control impacts will be localized.