Large-scale climate variation modifies the winter grouping behavior of endangered Indiana bats

Power laws describe the functional relationship between 2 quantities, such as the frequency of a group as the multiplicative power of group size. We examined whether the annual size of well-surveyed wintering populations of endangered Indiana bats (Myotis sodalis) followed a power law, and then leveraged this relationship to predict whether the aggregation of Indiana bats in winter was influenced by global climate processes. We determined that Indiana bat wintering populations were distributed according to a power law (mean scaling coefficient α = −0.44 [95% confidence interval {95% CI} = −0.61, −0.28). The antilog of these annual scaling coefficients ranged between 0.67 and 0.81, coincident with the three-fourths power found in many other biological phenomena. We associated temporal patterns in the annual (1983–2011) scaling coefficient with the North Atlantic Oscillation (NAO) index in August (βNAOAugust = −0.017 [90% CI = −0.032, −0.002]), when Indiana bats are deciding when and where to hibernate. After accounting for the strong effect of philopatry to habitual wintering locations, Indiana bats aggregated in larger wintering populations during periods of severe winter and in smaller populations in milder winters. The association with August values of the NAO indicates that bats anticipate future winter weather conditions when deciding where to roost, a heretofore unrecognized role for prehibernation swarming behavior. Future research is needed to understand whether the three-fourths–scaling patterns we observed are related to scaling in metabolism.