Advocating better habitat use models in birds: how to choose the right one in the light of a diversity of approaches?

Studies on habitat use represent a basic aspect of bird ecology, due to its importance in natural history, distribution, response to environmental changes, management and conservation. Basically, a statistical model that identifies environmental variables linked to a species presence is searched for. In this sense, there is a great diversity of analytical methods which identify important explanatory variables within a model, with higher explanatory and predictive power than classical parametric approaches. In particular, these allow dealing with non-normal distributions, heterogeneous variances and non-independent errors. However, most of these powerful models are not widespread and remain underused in ornithological studies, partly because of their complex theory, and in some cases, difficulties on their implementation and interpretation. I describe and compare some of these models applied to common Neotropical birds, namely generalized linear (mixed) models, generalized additive (mixed) models, occupancy models, and classification and regression trees. Their use in bird studies is encouraged, given their huge potential as statistical tools.