Understanding mechanisms of avian flight by integrating observations with tests of competing hypotheses

A long-standing problem in the study of avian flight is determining how biomechanics and physiology are associated with behaviour, ecological interactions and evolution. In some avian clades, flight mechanisms are strongly linked to ecology. Hummingbirds, for example, exhibit traits that support both hovering flight and nectar foraging. In most avian clades, however, features such as wing shape are highly variable among taxa without clear relationships to biomechanics, energetics or ecology. In this Commentary, we discuss challenges to understanding associations between phenotype and performance in avian flight. A potential pitfall in studies that attempt to link trait specialization with performance is that the most relevant traits and environments are not being considered. Additionally, a large number of studies of the mechanisms of avian flight are highly phenomenological. Although observations are essential for hypothesis development, we argue that for our discipline to make progress, we will need much more integration of the observational phase with developing crucial tests of competing hypotheses. Direct comparison of alternative hypotheses can be accomplished through analytical frameworks as well as through experimentation.