Moths, ants, and pitcher-plants: Small and large-scale biogeography of a tri-trophic interaction

Background/Question/Methods

The pitcher plant Sarracenia purpurea is a model system for ecological studies ranging from plant ecophysiology and demography to analysis of aquatic food web that inhabits its water-filled pitchers. Virtually all of these studies consider the plant and its associate food web – the "Sarracenia microecosystem" – as a closed system; their relationship with the surrounding fauna is poorly known. Two important "external" taxa that affect the Sarracenia microecosystem are ants and moths. Ants are the primary prey captured by pitcher plants and also feed at the plant's extrafloral nectaries. Specialist noctuid (Papaipema appassionata, Exyra fax) and tortricid (Endothenia daeckeana) moths selectively feed on Sarracenia rhizomes, leaves, and seeds, respectively. We hypothesized that within bogs, ants would nest preferentially near to large plants that produce large amounts of nectar, but that these plants would also be attacked more commonly by moths. We further hypothesized that because ant abundance in bogs declines dramatically with latitude that spatial associations between moths, ants, and pitcher-plants would diminish with increasing latitude. To test these hypotheses, we assessed plant size distributions and sampled joint occurrences of ant nests and pitcher-plant moths in a single 81-m² plot in each of 77 bogs from southwestern Rhode Island to northern Maine.

Results/Conclusions

Bog ant assemblages were numerically dominated by four species: Formica subaenescens, Tapinoma sessile, Myrmica lobifrons, and Dolichoderus pustulatus. Ant abundance declined significantly with latitude and elevation, but moth abundance was uncorrelated with biogeographic variables. Across all sites, moths were most commonly found feeding on leaves, rhizomes, and seeds of large plants. However, in southern New England where ants were abundant, the spatial association of ant nests with moth-attacked plants was negative. This result suggests that plants attacked by moths reduce nectar production and that ants had relocated their nests to be in closer proximity to unattacked plants. In northern New England, where ants were less abundant, locations of ant nests and moth-attacked plants were not spatially associated. Spatial patterns of ants, moths and plants depend not only on local interactions between species (including herbivory, nectivory, and nest-site selection), but also on geographical gradients of climatic variables that control regional species diversity and abundance of ants.