Invasibility of a North American soil ecosystem to amphibian-killing fungal pathogens

North American salamanders are threatened by the intercontinental spread of chytridiomycosis, a deadly disease caused by the fungal pathogen Batrachochytrium salamandrivorans (Bsal). To predict the potential dispersal of Bsal spores to salamander habitats, we evaluated the capacity of soil microbial communities to resist invasion. We determined the degree of habitat invasibility using soils from five locations throughout the Great Smoky Mountains National Park, a region with a high abundance of susceptible hosts. Our experimental design consisted of replicate soil microcosms exposed to different propagule pressures of the non-native pathogen, Bsal, and an introduced but endemic pathogen, B. dendrobatidis (Bd). To compare growth and competitive interactions, we used quantitative PCR, live/dead cell viability assays, and 16S rRNA amplicon sequencing. We found that soil microcosms with intact bacterial communities inhibited both Bsal and Bd growth, but inhibitory capacity diminished with increased propagule pressure. Bsal showed greater persistence than Bd. Linear discriminant analysis (LDA) identified the family Burkolderiaceae as increasing in relative abundance with the decline of both pathogens. Although our findings provide evidence of environmental filtering in soils, such barriers weakened in response to pathogen type and propagule pressure, showing that habitats vary their invasibility based on the properties of their local microbial communities.

Methods
General methods include soil collection in the field, experimental microcosm setup with microbially active (non-autoclaved) and inactive (autoclaved) soil samples, inoculation with two different Batrachochytrid pathogens, and incubating for two weeks while analyzing samples at different time points (see manuscript for detailed methods). Samples were processed for live/dead cell counts to obtain a proxy of microbial activity, and pathogen load via Quantitative PCR to determine pathogen response, and sequenced using 16S rRNA methods to understand bacterial community structure. Each of these results was then analyzed using computational methods to gain insight into the process of microbial invasion for this system.