1. Heterogeneities in infections among host populations may arise through differences in environmental conditions through two mechanisms. First, environmental conditions may alter host exposure to pathogens via effects on survival.Second, environmental conditions may alter host susceptibility, making infection more or less likely if contact between a host and pathogen occurs. Further, host susceptibility might be altered through acquired resistance, which hosts can develop, in some systems, through exposure to dead or decaying pathogens and their metabolites. Environmental conditions may alter the rates of pathogen decomposition, influencing the likelihood of hosts developing acquired resistance.2. The present study primarily tests how environmental context influences the relative contributions of pathogen survival and per capita transmission on host infection prevalence using the amphibian chytrid fungus (Batrachochytrium dendrobatidis; Bd) as a model system. Secondarily, we evaluate how environmental context influences the decomposition of Bd because previous studies have shown that dead Bd and its metabolites can illicit acquired resistance in hosts. We conducted Bd survival and infection experiments and then fit models to discern how Bd mortality, decomposition and per capita transmission rates vary among water sources [e.g. artificial spring water (ASW) or water from three ponds].3. We found that infection prevalence differed among water sources, which was driven by differences in mortality rates of Bd, rather than differences in per capita transmission rates. Bd mortality rates varied among pond water treatments and were lower in ASW compared to pond water. 4. These results suggest that variation in Bd infection dynamics could be a function of environmental factors in waterbodies that result in differences in exposure of hosts to live Bd. In contrast to the persistence of live Bd, we found that the rates of decomposition of dead Bd did not vary among water sources, which may suggest that exposure of hosts to dead Bd or its metabolites might not commonly | 171

Understanding local-scale variability in disease dynamics can be important for informing strategies for surveillance and management. For example, the amphibian chytrid fungus (Batrachochytrium dendrobatidis; Bd), which is implicated in population declines and species extinctions of amphibians, causes spatially variable epizootics and extirpations of its hosts. Outbreak heterogeneity could be driven by differential survival of zoospores, the free-living infectious life stage of Bd, or the persistence of dead zoospores and/or its metabolites in water, which could induce resistance among hosts. To gain a mechanistic understanding of the potential for variation in local transmission dynamics of Bd, we conducted Bd survival and infection experiments and then fit models to discern how Bd mortality, decomposition, and per-capita transmission rate vary among water sources. We found that infection prevalence differed among water sources, which was driven by differences in mortality rates of Bd zoospores, rather than differences in per-capita transmission rates. Specifically, zoospore mortality rates varied significantly among pond water treatments and were lower in artificial spring water compared to pond water sources. These results suggest that variation in Bd infection dynamics could be a function of differences in exposure of hosts to live Bd. In contrast to the persistence of live zoospores, we found that rates of decomposition of dead zoospores did not vary among water sources. These results may suggest that exposure of hosts to dead Bd or its metabolites, which have been shown to induce acquired resistance, might not commonly vary among nearby sites. Ultimately, a mechanistic understanding of the drivers of variable epizootics of Bd could lead to increases in the effectiveness of surveillance and management strategies.