Does the sociality of pollinators shape the organisation of pollination networks?

• Published in: Oikos 2019-05, Vol.128 (5), p.741-752
• Main Authors: Maia, Kate P., Rasmussen, Claus, Olesen, Jens M., Guimarães, Paulo R.
• Format: Article
• Language: English
• Subjects: Bees; betweenness; Biodiversity; Biological evolution; Castes; centralisation; closeness; Colonies; Foraging; Insects; Interactions; mutualism; Networks; Niche overlap; Niches; Plant reproduction; Pollination; Pollinators; Shape; Social organization; Species; weighted modularity; weighted nestedness
• ISSN: 0030-1299; 1600-0706
• DOI: 10.1111/oik.05387

A striking structural pattern of pollination networks is the presence of a few highly connected species which has implications for ecological and evolutionary processes that create and maintain diversity. To understand the structure and dynamics of pollination networks we need to know which mechanisms allow the emergence of highly connected species. We investigate whether social pollinator species are highly connected in pollination networks, and whether network structure is affected by the presence of high proportions of social pollinator species. Social insects are abundant, with long activity periods and, at the highest level of social organisation, specialised foraging castes. These three attributes are likely to increase the number of interactions of social species and, consequently, their role in pollination networks. We find that social species have, on average, more prominent network roles than solitary species, a possible mechanism being the individual‐rich colonies of social insects. However, when accounting for the shared evolutionary history of pollinators, sociality is only associated with highly interactive roles in Apidae. For apid bees, our structural equation analysis shows that the effect of sociality on species network roles is an indirect result of their high levels of interaction frequency. Despite the relative importance of sociality at a species‐level, an increasing proportion of social species in pollination networks did not affect overall network structure. Our results suggest that behavioural traits may shape patterns of interaction of individual species but not the network‐level organisation of species interactions. Instead, network structure appears to be determined by more general aspects of ecological systems such as interaction intimacy, patterns of niche overlap, and species abundance distributions.