Abstract
In this study we assess the bias in estimates of bird richness in a tropical wetland. We used 18 months of data from point counts to estimate the species richness in a coastal lagoon in Oaxaca and computed eight species richness estimators. We evaluated estimate inaccuracy using the sum of total richness that we obtained with combined sampling techniques between 2006 and 2010. We selected the richness estimator of minimum inaccuracy and used randomized species accumulation curves to investigate differences in species densities obtained with 10 and 20-minute counts. To ascertain if habitat type influenced richness estimates, we compared the bird diversity associated with contrasting mangrove and estuarine environments. The total richness for the lagoon was 185 species and the jackknife 2 richness estimator showed the minimum inaccuracy. The randomized accumulation curves of 10 and 20 minutes displayed partially overlapping confidence intervals and an increasing time lag between the count durations at equivalent densities. The rarefaction curves from contrasting habitat types showed significantly higher richness and species density in the estuary while the jackknife 2 estimator indicated similar sample coverage in both habitats. We identify important consequences of the sampling methodology that we used and conclude that point count sampling resulted in incomplete species detection but facilitated comparative analyses.