Reply from G. Bernasconi and J.E. Strassmann

Abstract

It is unfortunate that small, initially closed, newly founded ant colonies are so difficult to study in the field. If this were not the case, we might have much more complete information on the advantages of multiple-founding queens. As it is, we rely principally on laboratory studies1. Advantages of group nesting in the face of intraspecific brood raiding during colony founding were first reported in Myrmecocystus mimicus2, where brood raiding also occurs between mature colonies after ritualized territorial fights2. In My. mimicus, Messor pergandei and Acromyrmex versicolor, evidence that workers of neighbouring incipient nests reciprocally steal brood comes from the laboratory1,3,4. To date, brood raiding has been investigated and observed in the field only in the red imported fire ant, Solenopsis invicta3 (with one excellent field study of Me. pergandei not observing any occurrences5). From these field studies6,7, we know that the intensity of brood raiding and colony survival depend on queen density, as well as on the density and local spacing of new nests. Consistent with density-dependent benefits, queens are more likely to form groups when crowded, with local queen densities within one or a few metres probably being relevant for joining decisions8. Density also increases the frequency of usurpation attempts by raided queens6. Density-dependent advantages accruing to multiple-queen colonies are, therefore, potentially less important in the species’ native range (as suggested by Brown4 and previous studies6), where mound densities are lower than in the USA (Ref. 9) (but similar at selected sites9); thus, alate densities might also be lower.