Table 2.

Candidate models used to evaluate hypotheses describing presence-absence or abundance of Pseudogymnoascus destructans within hibernacula (sites).

Model	Covariate	Description
Environment
Spatial	X	Distance from the entrance
	Z	Height from the floor
	Type	Sediment or wall-swab
Temporal	WNStime	Time since first confirmation of WNS in bats at the site. Each unit coincides with a sampling interval (approximately every four months).
	Time of year	Time of year sample was collected; early hibernation, late hibernation, summer
Climate	Tdmean	Daily mean temperature
	Tdrange	Daily temperature range
Bats
Spatial	Regiona	Sampling region within the site (near entrance, middle, deep)
	Zb	Height from the floor
Temporal	WNSWintersc	Winters since first confirmation of WNS in bats at a site
Climate	Tdmean	Daily mean temperature within 2 m of roost location
	Tdrange	Daily temperature range within 2 m of roost location
Host	Species	Myotis lucifugus, Perimyotis subflavus, or Myotis sodalis
	Sex	male, female
	BMI	Body mass index [mass (g) / forearm (mm)]
Clinical signs	Fungus	Extent of visible fungus on the wing
	UV	Extent of UV fluorescence on the wing

^aIncluded in models using full dataset.

^bIncluded in models using subset dataset with paired temperature information because we included temperature at roost location as a covariate. X was not included due to collinearity with other covariates in models.

^cOmitted in models using subset dataset due to a singularity in response by site, which was included as a random effect.