Sampling and Analysis Methods for Ant Diversity Assessment

The current demand for studies on ants in general and their communities in particular has exponentially increased during the last decades. Much has already been said about ant sampling techniques, but we attempt to present a critical approach to the different available methods for sampling ants, their limitations, and their complementarity. We focus on sampling methods for ant inventories; entire ant colony sampling for a range of purposes, such as behavioral, cytogenetic, or population studies; and ant community studies. Methods presented here are valid for both tropical and temperate regions taking into account that ants are essentially thermophilous and found in lower richness and abundance in cold regions or during cold seasons. Sampling depends on the stratum of interest. Thus, different methods, or a combination of them, may be selected for soil-, litter-, or vegetation-associated ants. Regardless of the method used, some considerations must always be taken into account. The first of these is compatibility of data between the sampling methods. Be they pitfall traps, Winkler sack samples, or whatever, such sampling units must be treated independently from each another; otherwise the inferential statistics used may not be valid or used with caution as violation of independence may occur in two non-mutually exclusive general forms: pseudoreplication and spatial (or temporal) autocorrelation, for example, which are common mistakes in ant studies. We provide recommendations for statistical approaches to the data and different suggestions of analyses that can be used for the different kinds of data taken with ants.