There exists a gap in the knowledge of the effect of the bird survey methods on the accuracy of estimation of bird abundance, richness, and diversity indices. To test for the interaction between survey methods and habitat, we carried out a mist net and point count monitoring study over one year in Central México in a fragment of tropical deciduous forest adjacent to a cropland and in a fragment of tropical deciduous forest area converted to park adjacent to the former fragment. We estimated bird abundance, richness, inverse of the Simpson index, the exponential of the Shannon index, and the Hill evenness index. The goal of using both methods was obvious only in the case of richness estimation: species detected by both methods could be considered for pooled analysis. Point count survey detected higher evenness in bird species set pooled for resident, endemic, migratory categories and in resident birds, and lower evenness in endemic and in migratory birds than mist nets. For most bird categories there was no significant relationship between bird abundances assessed with both methods. The analysis of the method-habitat interaction allowed us to infer that for pooled bird categories, and for endemic and migratory species the evenness was larger in the park, and in resident birds it was larger in the forest. Although it was impossible to estimate the true value of different indices or the true shape of different relationships, the combination of methods allowed us to conclude if a given index was either larger or smaller in given habitat. Here, we propose the use of at least two survey methods not to sum or to average the indices estimated (excepting richness) but rather to infer the direction of the change of a given diversity estimator in line with the change of the habitat type.

There exists a gap in the knowledge of the effect of the bird survey methods on the accuracy of estimation of bird abundance, richness, and diversity indices. To test for the interaction between survey methods and habitat, we carried out a mist net and point count monitoring study over one year in Central México in a fragment of tropical deciduous forest adjacent to a cropland and in a fragment of tropical deciduous forest area converted to park adjacent to the former fragment. We estimated bird abundance, richness, inverse of the Simpson index, the exponential of the Shannon index, and the Hill evenness index. The goal of using both methods was obvious only in the case of richness estimation: species detected by both methods could be considered for pooled analysis. Point count survey detected higher evenness in bird species set pooled for resident, endemic, migratory categories and in resident birds, and lower evenness in endemic and in migratory birds than mist nets. For most bird categories there was no significant relationship between bird abundances assessed with both methods. The analysis of the method-habitat interaction allowed us to infer that for pooled bird categories, and for endemic and migratory species the evenness was larger in the park, and in resident birds it was larger in the forest. Although it was impossible to estimate the true value of different indices or the true shape of different relationships, the combination of methods allowed us to conclude if a given index was either larger or smaller in given habitat. Here, we propose the use of at least two survey methods not to sum or to average the indices estimated (excepting richness) but rather to infer the direction of the change of a given diversity estimator in line with the change of the habitat type.