An evaluation of methods for sampling ants (Hymenoptera: Formicidae) in British Columbia, Canada

Abstract

AbstractAnts (Hymenoptera: Formicidae) are ubiquitous and of considerable ecological interest, yet poorly studied in Canada. Given their natural history attributes and relatively low density and diversity in cool boreal forests, there is a need to evaluate the applicability of commonly used approaches to sampling and analysis. We examined the relative utility of two pitfall trap designs, mini-Winkler litter extractions, and colony versus forager sampling for ecological studies. First, we found that Laurent (conventional) and Nordlander (modified to exclude larger nontarget organisms) pitfall traps were equally efficacious for estimating total species richness. Second, pitfall trapping yielded marginally higher total species richness than mini-Winkler litter sampling, by the incidence-based estimator (Chao2). Third, two studies considering the relationship between captures of individual ants in pitfall traps and identified ant colonies argued for caution in how pitfall captures are interpreted. In the first study, Nordlander traps placed in a grid surrounding nests ofFormica obscuripesForel revealed unique and highly patchy captures of individuals, with no patterns relating to proximity to the nest. In the second study, abundance estimates to compare ant assemblage structure in a simple grassland ecosystem by pitfall trapping (relative) and area-based hand sampling (absolute) for colonies, revealed that relative sampling does not reflect the absolute ant assemblage structure. Our results support, equivocally, the use of pitfall traps (Nordlander) over mini-Winklers in the cool moist forests of west-central British Columbia, but individual specimen counts should not be used when analysing the data.