OOS 45-2
Structure-function relationships in the amphibian skin microbiome

Thursday, August 14, 2014: 1:50 PM
307, Sacramento Convention Center
Lisa K. Belden, Biological Sciences, Virginia Tech, Blacksburg, VA
Myra C. Hughey, Department of Biological Sciences, Virginia Tech, Blacksburg, VA
Eria A. Rebollar, Department of Biology, James Madison University
Thomas P. Umile, Department of Chemistry, Villanova University
Stephen C. Loftus, Department of Statistics, Virginia Tech, Blacksburg
Elizabeth A. Burzynski, Department of Chemistry, Villanova University
Kevin P. C. Minbiole, Department of Chemistry, Villanova University, Villanova, PA
Leanna L. House, Department of Statistics, Virginia Tech, Blacksburg
Roderick V. Jensen, Department of Biological Sciences, Virginia Tech, Blacksburg, VA
Matthew H. Becker, Department of Biological Sciences, Virginia Tech, Blacksburg, VA
Jenifer B. Walke, Biological Sciences, Virginia Tech, Blacksburg
Daniel Medina, Department of Biological Sciences, Virginia Tech, Blacksburg, VA
Roberto Ibáñez, Smithsonian Tropical Research Institute, Panama
Reid N. Harris, Department of Biology, James Madison University and Amphibian Survival Alliance, Harrisonburg, VA
Background/Question/Methods

Reduced species diversity leads to decreased ecosystem function in at least some systems. Vertebrates, including amphibians, host symbiotic microbes that may contribute to host disease resistance. The chytrid fungus Batrachochytrium dendrobatidis (Bd) causes a lethal amphibian skin disease, and may be a strong selective agent on the diversity and function of the amphibian skin microbiome. In Panama, many amphibians are threatened by Bd, and the spread of Bd has been closely tracked. In 2012, we completed a field survey in Panama to examine structure-function relationships in the skin microbiota in the context of chytrid infection. We focused on 3 amphibian host species, and sampled them across 4 sites that varied in Bd history (total N=138).  We collected two skin swab samples from each individual; one swab to assess Bd infection status and bacterial community structure and one to assess metabolite diversity, as the production of anti-fungal metabolites is an important disease resistance function of these microbes. We used ordinations to examine variation in bacterial community structure and metabolite diversity across species and sites, and based on Bd infection status.  In addition, we developed a model to identify bacterial OTUs and metabolites that were important predictors of Bd infection status.  

Results/Conclusions

Overall, across all 4 sites, only 36 individuals were infected with Bd.  Ordinations indicated clear clustering of the bacterial community structure based on both species and site, but there was no clustering based on Bd infection status. There was no clear clustering of metabolite diversity based on species, site, or Bd infection, although there was evidence that dispersion differed for the species and sites. With our model, we identified 9 bacterial OTUs and 5 metabolites that were significant predictors of Bd infection status. Results from our ordinations suggest that there may not be a strong link at the community level between specific community structure and function, as we observed greater variation in metabolite diversity despite distinct clustering of bacterial community structure among species and sites.  However, there could be important key bacterial taxa and metabolites within these communities that predict disease outcome or that are altered by Bd infection. Experimental manipulations of Bd status will be required to sort out these potential correlations.