ABSTRACT: The amphibian chytrid fungus, Batrachochytrium dendrobatidis (Bd), is a globally emerging pathogen that has caused widespread amphibian population declines, extirpations, and extinctions. However, Bd does not occur in all apparently suitable amphibian populations, even within regions where it is widespread, and it is often unclear why Bd occurs in some habitats but not others. In this study, we rigorously surveyed the amphibian and invertebrate biodiversity of 29 ponds in Missouri, screened resident amphibian larvae (Rana (Lithobates) sp.) for Bd infection, and characterized the aquatic physiochemical environment of each pond (temperature pH, conductivity, nitrogen, phosphorus, and chlorophyll-a). Our goal was to generate hypotheses toward answering the question, "Why does Bd not occur in all apparently suitable habitats?" Bd occurred in assayed amphibians in 11 of the 29 ponds in our study area (38% of ponds). We found no significant relationship between any single biotic or abiotic variable and presence of Bd. However, multivariate analyses (nonmetric multidimensional scaling and permutational tests of dispersion) revealed that ponds in which Bd occurred were a restricted subset of all ponds in terms of amphibian community structure, macroinvertebrate community structure, and pond physiochemistry. In other words, Bd ponds from 6 different conservation areas were more similar to each other than would be expected based on chance. The results of a structural equation model suggest that patterns in the occurrence of Bd among ponds are primarily attributable to variation in macroinvertebrate community structure. When combined with recent results showing that Bd can infect invertebrates as well as amphibians, we suggest that additional research should focus on the role played by non-amphibian biota in determining the presence, prevalence, and pathogenicity of Bd in amphibian populations.
