Does the Bergmann's rule apply in wild bees?

A phenotypic response has been widely reported following a climate cline: the decrease in mean body size in warmer conditions and conversely the increase of mean body size in colder conditions. This rule was first highlighted by Bergmann (1847). Initially, Bergmann's intention was to describe this rule based on endothermic vertebrate data and within species. However this rule is now widely tested among non-vertebrate species and between species assemblages. Even if this pattern is well established in endotherms vertebrates, it is not generalizable in ectotherms.
Initially the rule was suggested to be due to an adaptive response through mechanisms of thermoregulation in which smaller surface area to volume ratio improves the heat conservation. Alternative adaptive (e.g. costs and benefits of life history traits and natural selection) and non-adaptive (e.g. effects of temperature on biochemical processes) mechanisms have also been proposed as explanation. Though bees are of particular ecological and economic importance since they are efficient pollinators of both wild flowers and crops, very few studies assessed the Bergmann's rule among them and none of them at the European scale.
We used data set including 615 bee species distributed in 21 genera and five families which is the most exhaustive bees 'study about the relationship between body size (with intertegulae distance as proxy) and latitude. Our aim is to assess relationships between size and latitude/bioclimatic factors.
We found different responses in the relationship of average body-size / climate throughout the different bee clades. All bees taken together, mean body size of bee assemblages followed the Bergmann's rule, i.e. mean body size significantly increased with increasing latitudes. Similarly, the half of bee clades followed the Bergmann's rule (i.e. Andrenidae and Apidae families and Andrena, Dasypoda, Lasioglossum, Panurginus and Panurgus genera), unlike Colletidae and Melittidae families and Bombus and Melitta genera which followed the converse Bergmann's rule. No significant relationship were found for Halictidae family and Ceratina and Halictus genera. Previously, only one study analyses the variation of bees body size at a family level and at the continental scale, in the United States (Hawkins, 1995). In that study, Andrenidae was the only family to follow the Bergmann's rule and Melittidae the only family to follow the converse Bergmann's rule, similarly to our results. Potential drivers of those contrasting patterns are developed.