FIGURE 2.
(a) Change in overwinter prevalence of Pseudogymnoascus destructans in the year that P. destructans was first detected at a site. Lines show fits of a GLMM with a binomial distribution with timing of first detection interacting with species and an additive effect of date with site as a random effect. (b) Twice as many sites had the first detection of P. destructans in late hibernation as in early hibernation