Multiple genera of fungi. In the first study to show inoculation of bats with Pd causes WNS, Lorch and colleagues noted that their captive study was not long enough to result in mortality despite histological evidence of infection among inoculated bats. Density-dependent growth of Pd may explain why mortality did not occur within the time period of their study, which inoculated bats with 500 000 conidia. Our mortality data are only suggestive of self-inhibition in Pd, however, and research documenting germination at varying concentrations of conidia is needed to directly address this hypothesis. Such research, along with studies documenting natural exposure dynamics among freeranging bats, are needed to better TWS119 inform captive studies of WNS, which typically inoculate bats with 500 000 conidia. Inoculations resulting in mortality patterns that differ from wild populations may produce misleading insights into WNS. It is notable that we were often unable to detect Pd DNA on swabs from bats inoculated with 500 conidia. This demonstrates that the number of Pd conidia did not exponentially increase on bats in this treatment. We hypothesize that bats have some ability to control the fungal infection at this level of exposure. Although the mechanism of control is uncertain, the increased frequency of periodic arousals observed in these treatments likely plays some role. Arousals provide opportunities for euthermic rest, grooming, and immune upregulation, although the brevity of periodic arousals in bats compared to other hibernating mammals likely limits potential immune responses to Pd. As previously discussed, however, the number of arousals bats can energetically sustain are limited, and the frequent arousals in bats inoculated with 500 conidia resulted in high mortality despite an ability to control the fungus. Furthermore, Pd always remained on some bats within the 500 conidia treatment groups, serving as vectors for continued Pd exposure within this hibernation chamber. Mortality in the remaining inoculation treatments was not significantly greater than controls in our model. This lack of difference was driven by the low mortality observed in the remaining inoculation treatments hibernated at 4uC and relatively high mortality in the 10uC control group. The mortalities in the 10uC control group are well explained by the logistic regression model. Mortalities in this group were primarily males with body condition indices at the onset of hibernation that were below the median body condition. It is well documented that lower temperatures are more energetically favorable for hibernating bats, a conclusion supported by our own data. Thus, it is not surprising that we observed high mortality among male bats, which aroused more frequently from hibernation, with low fat reserves when placed in an energetically unfavorable environment. Mortality among bats with low body condition in both control groups may also result from placing bats in environmental conditions that differ from their native hibernacula. The 4u and 10uC environmental chambers represented temperatures that are colder and warmer, respectively, than both of the hibernacula we sampled in Illinois and Michigan.