Climate has long been suggested to affect population genetic structures of eusocial insect societies. For instance, Hamilton [Journal of Theoretical Biology 7 (1964) 17] discusses whether temperate and tropical eusocial insects may show differences in population-level genetic structure and viscosity, and how this might relate to differences in the degree of synchrony in their life cycles or modes of nest founding. Despite the importance of Hamilton's 1964 papers, this specific idea has not been tested in actual populations of wasps, probably due to the paucity of studies on tropical species. Here, we compare colony and population genetic structures in two species of primitively eusocial paper wasps with contrasting ecologies: the tropical species Polistes canadensis and the temperate species P. dominulus. Our results provide important clarifications of Hamilton's discussion. Specifically, we show that the genetic structures of the temperate and tropical species were very similar, indicating that seasonality does not greatly affect population viscosity or inbreeding. For both species, the high genetic differentiation between nests suggests strong selection at the nest level to live with relatives, whereas low population viscosity and low genetic differentiation between nest aggregations might reflect balancing selection to disperse, avoiding competition with relatives. Overall, our study suggests no prevalence of seasonal constraints of the life cycle in affecting the population genetic structure of eusocial paper wasps. These conclusions are likely to apply also to other primitively eusocial insects, such as halictine bees. They also highlight how selection for a kin structure that promotes altruism can override potential effects of ecology in eusocial insects.

Stable social groups usually consist of families. However, recent studies have revealed higher level social structure, with interactions between family groups across different levels of social organization in multiple species. The explanations for why this apparently paradoxical behaviour arises appear to be varied and remain untested. Here, we use automated radio-tagging data from over 1000 wasps from 93 nests and social network analyses of over 30 000 nest visitation records to describe and explain interactions across levels of social organization in the eusocial paper wasp Polistes canadensis . We detected three levels of social organization (nest, aggregation and community) which exchange ‘drifter’ individuals within and between levels. The highest level (community) may be influenced by the patchiness of high-quality nesting habitats in which these insects exist. Networks of drifter movements were explained by the distance between nests, the group size of donor nests and the worker-to-brood ratios on donor and recipient nests. These findings provide some explanation for the multi-level social interactions, which may otherwise seem paradoxical. Fitness benefits across multiple levels of social organization should be considered when trying to understand animal societies.