When does temperature limit bee flight? Identifying the missing pieces of the puzzle

Bees are the most important insect pollinators, and nearly all bee-mediated pollination depends on flight. In this review, we identify five key gaps that limit our ability to predict how abiotic factors shape bee flight performance and ecology. First, while clade strongly influences endothermy and flight temperature limits, we still know little about how phylogeny, body size, and ecological traits shape thermal biology across the full diversity of bee species - most data come from temperate, northern hemisphere  Apinae. Second, because the mechanisms of thermal balance during flight have only been studied in a handful of species, and these vary, we clearly lack a predictive understanding of when bees will be physiologically stressed by temperature during flight. Third, although some studies suggest desiccation may limit flight more than overheating, the links between thermal balance and water regulation remain poorly understood. Fourth, we know very little about bees’ capacity to respond to thermal variation through acclimation, developmental plasticity, or evolutionary adaptation. Finally, we need advances in biophysical modeling to better simulate heat and water exchange in bees, including the role of cuticular structures and internal heat transfer among body regions. Filling these gaps is essential for building predictive, mechanistic models of how climate change will affect bee physiology and the pollination services bees provide.

Methods
The table-1 presents clade-level variation in the thermal biology of bees, which was created by searching scholar.google.com and Web of Science using relevant keywords such as: bees, temperature, thermal, flight, activity, foraging, CTmax, and thermal limit. Many of the references were not found through web searches, but rather by investigating the bibliographies of in-hand papers. We are sure that additional relevant papers exist, but feel that it is very likely that the papers found and used will provide representative numbers for each parameter. Clades were based on Henriquez-Piskulich et al. 2024. To indicate the approximate number of species per clade, we relied on Danforth et al. 2019. To indicate the number of species in the clade whose flight thermal biology has been studied, we summed the number of species whose data contributed to Table 1. The endothermic index (thorax temperature – operative temperature) was taken from papers that directly measured minimal air and thorax temperatures for flight or from the largest thorax and air temperature differential for lab studies. We did not include field data for the thorax and air differential, as bees in cool conditions routinely raise their body temperatures by solar basking, so such data do not provide a good index of endothermy (Herrera et al. 2023). Thermal stability was calculated as (1 minus the slope of thorax temperature on air temperature), including for bees measured in either the field or laboratory, as surprisingly, thermal stability appeared similar for bees measured in both conditions, and there were some clades for which we lacked field data (Supplementary Table 1). Many studies reported these values directly. Herrera et al. 2023 and Herrera 2024 did not usually report clade-level slopes of thorax on air temperature, so we calculated these using linear regressions in Excel and the provided online data for these studies. The lowest and highest air temperatures and thorax temperatures at which bees could fly were calculated by pooling studies, which explicitly attempted to measure these in lab conditions and field studies that reported that they attempted to assess the full breadth of temperatures at which their target bees. If the study did not report these values explicitly, they were estimated from the presented figures. There were many more studies assessing minimal air temperatures for flight than maximal. For Herrera et al. 2023 and Herrera 2024, which provided raw data online, we assessed minimal and maximal air and thorax temperatures for each clade studied. CTmax data were mostly taken from Johnson et al. 2023, who recently compiled such data for pollinators, with the inclusion of a few new relevant studies. For each parameter and clade, we reported the median and range in Table 1.

We used the data in Table 1 to test for correlations across bee clades between degree of endothermy and thermal stability during flight, the lowest air and thorax temperature for flight, the highest air and thorax temperature for flight, and CTmax. We used Prism 10.5 for macOS for statistical analyses. All variables passed the Kolmogorov-Smirnov normality test, so we used Pearson's r to test for significant correlations.