Background/Question/Methods Communities of hosts can change through space and time, altering the transmission potential of parasites. We investigated how changes in host composition, total host abundance, and environmental conditions affect a ranavirus in larval amphibian communities. Ranavirus is a genus of generalist parasites of amphibians and are spread via contact and environmental transmission. We developed a dynamical model for ranavirus in a 2-species community, and we manipulated this model in terms of community composition, host abundance, and virus half-life. Next, we measured community competence, diversity (richness and evenness), and overall composition (principal component analysis) for each community. Community competence is the propensity for a community of hosts to support a parasite and was calculated as the sum of the products of each species’ competence (estimated by mean viral load) and their relative abundance in a community. Community competence serves as a site-level metric of community composition that is relevant to parasite transmission. All data were previously collected and include observations from 20 wetlands over 6 months for 17 species of larval amphibians. After calculating these metrics, we used data visualization and correlation tests to detect patterns between our variables of interest (composition, abundance, and environment) and transmission potential of ranavirus. Results/Conclusions Our transmission model showed that manipulations of community composition, host abundance, and viral half-life can all enhance a community’s ability to support a parasite, and that when combined, the effects of these variables resulted in communities highly susceptible to ranavirus. Analysis of real host communities confirmed that similar host compositions (according to PCA ordination) had similar community competence, but dissimilar compositions could still result in similar community competence. In our diversity analysis, we did not detect a traditional dilution effect pattern but we did find effects of composition across an evenness gradient. At low evenness, communities have either high or low levels of community competence, suggesting that these communities were dominated by either low or high competent species. Further analysis showed that communities with high community competence were often the most abundant and occurred in months with lower temperatures (longer viral half-life). Community competence, host abundance, and water temperature were all significantly correlated with infection prevalence in host communities. Our results suggest that changes in community composition, host abundance, and environmental conditions can result in highly susceptible host communities. We also show that richness alone may not be sufficient in detecting the influence of community composition on parasite transmission.
