Understanding eusociality requires both proximate and ultimate thinking and due consideration of individual and colony‐level interests

Abstract

In his recent book, ‘The evolution of social wasps’, Hunt(2007) advocates a purely proximate approach whenstudying the evolutionary origins of eusociality, and heproposes a research agenda for paper wasps that focusesexclusively on how sociality evolved for the good of thecolony. At the same time, he dismisses ultimate levels ofexplanation and the consideration of individual-level inter-ests as a ‘‘blindered view of social evolution focusedexclusively on [...] inclusive fitness’’ (p. 164), and hedevotes two chapters to analysing the supposed flaws of kinselection theory and its applications. Hunt denies anyexplanatory power to ultimate thinking, that is attempts tounderstand why sociality evolved, which he sees as beingmerely ‘interpretations’ (p. 193)!We agree with Hunt that understanding the evolutionand maintenance of sociality requires proximate-levelstudies of the underlying genetic, physiological, develop-mental and ecological mechanisms. But we most stronglydisagree with his subordination of ultimate to proximatethinking, and his sharp criticism of kin selection theory.Rejection of ultimate thinking and individual level selectionleads at best to an incomplete understanding of eusociality,and at worst to erroneous interpretations of social beha-viour. The vast majority of students of social evolutionacknowledge the need for a balanced approach that takesinto account Tinbergen’s four questions (Krebs and Davies1997), that is combine studies on the selective value, theproximate-level mechanisms, the ontogenic developmentand the evolutionary history of social behaviours (Helantera¨2007, Korb and Heinze 2008). Understanding socialitythus requires combining, rather than opposing, proximateand ultimate levels, and it also requires considering bothindividual and colony-level interests.Hunt overlooks the importance of integrating proximatemechanisms and ultimate causation. His exclusively prox-imate scenario for the evolution of eusociality in Polistespaper wasps merely proposes that the solitary Polistesancestor had two generations per year (bivoltinism). Thefirst generation would have reproduced immediately afteremergence. The second generation would have entereddiapause to survive the unfavourable season (dry or cold intropical and temperate regions, respectively), and it wouldhave reproduced the following season. This ancestralbivoltinism would have given rise to the production oftwo broods in modern Polistes, a first brood of workersfollowed by a second brood of foundresses (Hunt 2007).This proximate level analysis explains some differencesbetween first and second brood females, such as fat storageand ability to overwinter (this fat storage mechanism alsooccurs in the sweat bee Halictus ligatus (Richards and Packer1994), and it may be common in facultative eusocialsystems). However, it does not explain why most first broodfemales behave as workers rather than attempt to reproducein the first place, and why this proportion varies acrossspecies or even populations within the same species. Kinselection theory does just that.Hunt also fails to recognize that individual selection caninterfere with colony level selection for eusocial organiza-tion. He considers only selection at the level of the group,but West-Eberhard (1975) and others have pointed outthat members of a colony are not only ‘soma’ of a super-organism, but individuals that may pursue selfish repro-ductive strategies despite being related and having parallel(but not identical) collective interests. He also ignoresevidence that these social conflicts can significantly reducecolony productivity in Polistes, where workers sometimescompete over reproduction by mutual egg eating, a costlybehaviour that cannot be explained by colony level thinking(Liebig et al. 2005, numerous similar examples of socialconflicts in insect societies are known e.g. Ratnieks et al.2006, Rankin et al. 2007, Beekman and Oldroyd 2008).