Skin microbiomes of frogs vary among body regions, revealing differences that reflect known patterns of chytrid infection

Sonia L. Ghose^*Jonathan A. Eisen

• University of California, Davis, Davis, California, United States

The amphibian skin microbiome is an important line of defense against pathogens including the deadly chytrid fungus, Batrachochytrium dendrobatidis (Bd). Bd is known to preferentially infect ventral skin surfaces and feet of host amphibians, often leaving dorsal surfaces like the back absent of infection. Within-individual variation in infection distribution across the skin, therefore, may relate to differences in microbiomes among skin regions. However, characterization of microbiome variation within amphibian individuals is needed. We utilized 16S rRNA gene amplicon sequencing to compare microbiomes of ten body regions from nine captive Rana sierrae individuals and their tank environments. These individuals were naive to Bd, allowing us to assess whether their microbiomes differed among body regions prior to any impacts that may be caused by infection. We compared frog and tank environment microbiomes, finding that they harbored distinct microbial communities. Notably, the bacterial families Burkholderiaceae (phylum Proteobacteria) and Rubritaleaceae (phylum Verrucomicrobia) were dominant on frog skin, driven in large part by the relative abundances of undescribed members of these families that were significantly higher on frogs than in their environment. Within individuals, we detected differences between microbiomes of body regions where Bd infection would be expected compared to regions that infrequently experience infection. Putative Bd-inhibitory relative abundance was significantly higher on body regions where Bd infection is often localized. These results suggest that microbiomes in certain skin regions may be predisposed for interactions with Bd. Further, these findings highlight the importance of considering intraindividual heterogeneities, which could provide insights relevant to predicting localized interactions with pathogens.