COS 148-3 - CANCELLED - Predicting hibernating bat roost establishment in spatiotemporally complex hibernacula

Background/Question/Methods

The realized niche can be difficult to define for a given population due to the stochasticity and complexity inherent in natural systems. Subterranean ecosystems are relatively isolated and stable, allowing us to more easily define the relevant factors defining realized niche space. We examine the niche space used by little brown bats (Myotis lucifugus) to hibernate in a spatiotemporally variable hibernaculum infected with the fungus causing white nose syndrome (WNS). Importantly, bat populations hibernating in this hibernaculum exhibit minimal WNS mortality despite regional population declines. By understanding the environmental conditions allowing WNS survival in this population, we can develop environmentally-dependent population risk assessment. To do this, we gathered detailed temperature and humidity data and used this to predict roost occurrence probability based on midwinter survey observations of hibernating M. lucifugus populations. We developed a Bayesian hierarchical model that predicts underground temperature as a function of above-ground conditions, which informs a logistic regression predicting M. lucifugus roost presence at a given microsite within the hibernaculum.

Results/Conclusions

We found that M. lucifugus hibernation roost locations are established as a combination of relatively warm temperatures, stable microclimates, and sociality. Predictions of roost presence were significantly improved by hierarchically estimating underground conditions as a function of above-ground climate, which allows assessment of underground dynamics when microclimate measurements are unavailable. We demonstrate this by using our hierarchical model to predict roosting microsites under climate change thirty years in the future. By defining hibernation niche space, we can better understand how bats interact with available microclimates to survive hibernation with a pathogen that has caused mass mortality in other locations. This allows us to improve risk assessment of hibernating bat populations threatened by rapidly changing environmental conditions due to disease and climate change.