A multispecies aphid–ant association: density dependence and species-specific effects

Abstract

In the central Rocky Mountains of Colorado 4 species of aphids feed on fireweed, and 10 species or ants have been observed tending 3 of these species of aphids. Local populations of aphids and their associated ants were counted nondestructively at weekly intervals for up to 11 weeks. The effects of ants upon the persistence of recently initiated populations and the growth of populations during 1-week intervals were assessed by multidimensional contingency table analysis. Macrosiphum valerianae, which is not tended by ants, was affected negatively by ants. Aphis varians and A. helianthi were affected by tending, but A. salicariae was not affected. The responses of both A. varians and A. helianthi to tending were density dependent; i.e., small populations that were untended were more likely to decrease than tended populations, but large populations that were tended were either more likely to decrease than untended populations or equivalent to them. Possible reasons for this density-dependent effect are discussed, and its implications for the stability of the mutualism between aphids and ants are considered. Different species of ants had different effects. Populations of A. varians that were tended by Formica neorufibarbis or Tapinoma sessile performed no differently than untended populations, but low-density populations that were tended by F. cinerea or F. fusca were less likely to decline than untended populations.