Thermal diversity of North American ant communities: Cold tolerance but not heat tolerance tracks ecosystem temperature

In ectotherms, gradients of environmental temperature can regulate metabolism, development and ultimately fitness. The thermal adaptation hypothesis assumes that thermoregulation is costly and predicts that more thermally variable environments favour organisms with wider thermal ranges and thermal limits (i.e., critical thermal minima and maxima, CT_min and CT_max) which track environmental temperatures. We test the thermal adaptation hypothesis at two biological levels of organization, the community and species level. Location: Continental USA. Time period: May–August 2016 and May–August 2017. Major taxa studied: Ants (Hymenoptera:Formicidae). Methods: We used ramping assays to measure CT_max and CT_min for 132 species of North American ants across 31 communities spanning 15.7° of latitude. Results: Ants were cold tolerant in cooler environments particularly at the community level where CT_min was positively correlated with the maximum monthly temperature (CT_min = 0.24T_max− 0.4; R² =.39, p <.001). In contrast, most ant communities included some highly thermophilic species, with the result that CT_max did not covary with environmental temperature means or extremes. Consequently, we found no evidence that thermally variable environments supported ant communities with broader thermal ranges. We found a strong phylogenetic signal in CT_max but not CT_min. Species level responses paralleled community data, where maximum monthly temperatures positively correlated with species CT_min but not CT_max, which was significantly lower in subterranean species. Main conclusions: Our results suggest a large fraction of continental trait diversity in CT_max and CT_min can be found in a given ant community, with species with high CT_max widely distributed regardless of environmental temperature. Species level analyses found the importance of local microclimate and seasonality in explaining thermal tolerances. Frequent invariance in CT_max of insects at a large scale might be caused by (a) local adaptations to a site's microclimates and (b) species acclimation potential, both of which cannot be accounted for with mean annual temperatures.