Ants are largely used for monitoring and inventory programs, resulting in great amounts of material sampled with several methodologies, financial limitations and lack of experts to do the taxonomy of many taxa. To contribute for the elaboration of alternative protocols, we analyzed the complementarity of two sampling methods in the detection of species richness. In 6 plots of 1 km² distributed in an area of 33,000 ha of continuous primary forest, in the Brazilian state of Pará, the diversity and the frequency of species of genera Crematogaster,Gnamptogenys and Pachycondyla were investigated, as part of the "Ant methodology Protocol" of the Project TEAM. Using the Winkler extractor and pitfall traps, 10 sub-samples were collected per transect. In each plot, 4 transects of 100 m were installed. We registered 36 species, and 15 are new records for the state, and 29 for the study site. Seventeen species were common to both sampling methods, 11 were captured with pitfall and 8 with Winkler. The proportions of singletons were high for both methods (>0,4). Nineteen species were captured by only one method and three were represented by only one individual in both methods. Ordination analysis (SSH, Semi-Strong Hybrid Multi-Dimensional Scaling) indicated differences in community structure between both sampling methods. There was a complementarity's effect between these two methods in order to obtain a more complete inventory of the species richness, indicating that both must be used in the protocol. Other resources like rarefaction and/or sample dilution must be evaluated in future investigations.
Ants; Winkler extractor; Pitfall traps; Amazon
