Pathogens and mutualists: Temporal dynamics in the skin microbiome of the Sierra Nevada yellow-legged frog.

Background/Question/Methods

The Sierra Nevada yellow-legged frog (Rana sierrae), which has been extirpated from over 90% of its historic range, is one of the most threatened amphibian species in California.  In order to develop conservation techniques to promote the recovery of this once abundant species, we closely monitored a population of R. sierrae in the Desolation Wilderness, CA, that continues to persist in spite of infection with the amphibian chytrid fungus, Batrachochytrium dendrobatidis (Bd).  We marked all individuals using passive integrated transponder (PIT) tags, and we monitored individual infection levels throughout the summer active season of 2013 using qPCR.  In addition, we collected bacterial cultures from the skin of 60 individuals, and we conducted in vitro Bd-inhibition trials in order to identify strains of bacteria with anti-fungal properties, which could be used as probiotic treatments in future conservation interventions.  Finally, we collected skin microbiome samples from individually marked R. sierrae at four intervals throughout the summer active season, and we used next generation sequencing to analyze changes in species composition and community structure in R. sierrae’s resident skin microbes through time.

Results/Conclusions

From a sample of 725 skin swabs, we detected an overall prevalence of 63.7% Bd infection, with an average infection intensity of 591.1 ZE.  Of 57 strains of bacteria isolated from the skin of wild-caught R. sierrae, approximately one quarter inhibited the growth of Bd in vitro, with some strains impeding Bd’s growth far more than others.  Finally, we used an Illumina Mi-Seq to track changes in the skin microbiome of 10 adult R. sierrae, 4 juveniles, 4 tadpoles, and 6 environmental samples, at 4 intervals throughout the summer active season.  We found that the skin microbiome of R. sierrae was distinct from bacterial communities in the environment, and that host life stage is an important predictor of bacterial community structure.  We also describe variation in species composition and community structure through time.  This longitudinal analysis will contribute to the successful development of future bioaugmentation interventions, since a basic understanding of microbial community dynamics will allow us to better predict the ability of a probiotic treatment to colonize a frog’s skin and persist in the long term.