Highlighting the article by da Silva, Stevens and Schwarz in Insectes Sociaux
Written by Insectes Sociaux Associate Editor Miriam Richards
My favourite social bee is generally the one that I am writing a paper on, which at the moment is the Eastern Carpenter Bee, Xylocopa virginica. The more we study this bee, the more complex and interesting its social behaviour gets, which is mildly ironic, because recently a reviewer suggested that despite complicated behavioural interactions among co-nesting females, these carpenter bees are not “truly social”. The still rather widespread presumption that caste-based eusociality is the only “true” form of sociality certainly influences how we think about social evolution in insects.  The recent publication in Insectes Sociaux of a paper entitled “Casteless sociality in an allodapine bee and evolutionary losses of social hierarchies” by da Silva, Stevens and Schwarz (2015), provides a timely reminder of the enormous breadth of social behaviour in bees, which includes both caste-based and casteless forms.
Allodapine bees are the best studied of four tribes in the carpenter bee clade (Xylocopinae), thanks to years of research by Mike Schwarz and his colleagues at Flinders University in Australia. Allodapine bees exhibit several distinctive characteristics, including progressive provisioning of their offspring, which are raised in chambers excavated in plant stalks and branches. They also exhibit a wide range of colony types, from mostly subsocial species to eusocial, including a highly eusocial species whose morphologically distinct castes represent an independent transition to this form of obligate eusociality (Dew et al., 2012). Da Silva and colleagues studied Braunsapis puangensis nesting at several locations in Fiji (!), collecting nests and examining their contents. These bees live in small groups of up to five females, with the modal group size being one, suggesting that females can nest either solitarily or socially. By comparing the number of brood in a nest to the number of adult females, da Silva et al. estimated the per capita rate of brood production in nests with differing numbers of females. Two main trends emerged. First, social nests almost invariably contain brood whereas solitary nests often do not. This is consistent with a general pattern in allodapine bees in which sociality protects against total brood loss. Second, in social nests with brood, the main predictor of per capita brood productivity was the length of the nest cavity, with per capita brood productivity declining as group size increased. This is a surprising conclusion because in previous studies of allodapines, Schwarz’s group has found a positive relationship between group size and per capita productivity. The study thus provides “a clear benefit to social nesting in terms of preventing brood loss”, but no advantage (indeed, a detriment) when group size increases above two females.
Perhaps the most interesting finding in this paper is that referred to in the title: multi-female groups comprise casteless societies. In most social insects, the familiar queen and worker castes exhibit a negative association between reproduction and work. An idiosyncracy of social carpenter bees is that the reproductive castes can be based on a positive association between reproduction and work (Hogendoorn and Velthuis, 1999). However, in Braunsapis puangensis, there is no evidence of either type of caste division – co-nesting females are no more different in wing wear (which measures foraging activity) and ovarian development (which measures reproductive potential) than randomly selected pairs of solitary females. Although Schwarz’s group prefer the term “casteless sociality”, the more traditional term “communal” would also appear to be an apt descriptor of colony social organization in B. puangensis. Phylogenetically, B. puangensis seems to represent an evolutionary transition from a hierarchical, caste-based form of sociality to an egalitarian, casteless form of sociality. Although it was once predicted that communal and eusocial behaviour were mutually exclusive and not to be found in the same clade, but it would appear that B. puangensis provides us with a rare piece of evidence refuting this hypothesis.
How rare is communal behaviour in bees, and is the almost complete absence of communal behaviour in eusocial lineages real? It’s difficult to tell, because current approaches emphasize caste-biased sociality, relegating other forms to the “not truly social” category. For instance, in a recent phylogenetic study investigating evolutionary transitions between solitary and social behaviour in halictine bees, communal species were classified as solitary while species with caste-based societies were classified as social (Gibbs et al., 2012).  In another recent contribution to Insectes Sociaux, Dew, Tierney and Schwarz (2015) emphasize that our ability to identify egalitarian, casteless societies in bees depends on how we define communal, egalitarian behaviour and the observational methods used to distinguish it from hierarchical forms of colony social organization. I would not be surprised to find out that in bees, at least, casteless and other types of non-eusocial societies, are a lot more common than we currently think.
 Of course, I disagreed with the reviewer and included the word “social” in the title – not once, but twice! (Richards, 2011)
 Some of the authors of this paper have been forced to endure considerable haranguing from me about behavioural differences between solitary and communal bees, which they did with great good humour.
da Silva, C.R.B., Stevens, M.I., and Schwarz, M.P. 2015. Casteless sociality in an allodapine bee and evolutionary losses of social hierarchies. Insectes Soc. http://link.springer.com/10.1007/s00040-015-0436-0
Dew, R.M., Rehan, S.M., Tierney, S.M., Chenoweth, L.B., and Schwarz, M.P. 2012. A single origin of large colony size in allodapine bees suggests a threshold event among 50 million years of evolutionary tinkering. Insectes Soc. 59: 207–214
Dew, R.M., Tierney, S.M., and Schwarz, M.P. 2015. Social evolution and casteless societies: needs for new terminology and a new evolutionary focus. Insectes Soc.
Gibbs, J., Brady, S.G., Kanda, K., and Danforth, B.N. 2012. Phylogeny of halictine bees supports a shared origin of eusociality for Halictus and Lasioglossum (Apoidea: Anthophila: Halictidae). Mol. Phylogenet. Evol. 65: 926–939
Hogendoorn, K., and Velthuis, H.H.W. 1999. Task allocation and reproductive skew in social mass provisioning carpenter bees in relation to age and size. Insectes Soc. 46: 198–207
Richards, M.H. 2011. Colony social organisation and alternative social strategies in the eastern carpenter bee, Xylocopa virginica. J. Insect Behav. 24: 399–411
da Silva, C.R.B., Stevens, M.I., and Schwarz, M.P. 2015. Casteless sociality in an allodapine bee and evolutionary losses of social hierarchies. Insectes Soc.