Morphological and Ecological Traits Explain Variation in Desiccation Tolerance Among Bees

Rising temperatures associated with global warming are often accompanied by shifts in precipitation patterns, increasing the risk of desiccation for insects. However, our understanding of how functional traits shape desiccation tolerance remains limited, particularly among pollinators. We assessed the desiccation tolerance of 22 bee species (13 genera, five families) from the Greek island of Lesbos, covering a range of body sizes, hairiness, nesting habitats, and social structures. We measured three desiccation tolerance metrics under standardized acute desiccating conditions (27°C and 0% RH): survival time, water loss rate (WLR), and water content at mortality (critical water content, CWC). In addition, we evaluated the phylogenetic signal of these three desiccation-related traits and examined the influence of morphology (body mass and hair length), nesting (nest guild and nest location), and sociality on each. Survival time, WLR, and CWC varied widely among species. Larger-bodied species exhibited lower WLR and survived longer under desiccation stress. Survival time decreased with increasing WLR and CWC. Hair length was positively associated with CWC but did not affect survival time or WLR. Ecological traits did not predict survival time, but nesting guild influenced both WLR and CWC, and sociality influenced CWC. Eusocial species nesting in large cavities exhibited higher WLR and CWC, whereas solitary species generally showed lower CWC. Phylogenetic signal was moderate (WLR) to strong (survival time and CWC) but was not statistically significant. Our results reveal significant interspecific variation in bee desiccation tolerance and demonstrate that morphological and ecological traits shape physiological responses to water stress. These results highlight trait-based mechanisms underlying desiccation tolerance and suggest that functional trait variation may influence pollinator vulnerability to increasing drought and climatic variability.