Limitations of ecomorphological analysis in explaining macrohabitat segregation in a songbird guild1

It is not uncommon for different species within a guild to be non-randomly segregated within a landscape through the occupation of different preferred macrohabitats. The explanation most amenable with ecological theory is that different guild members are adapted for exploiting the different opportunities offered by such macrohabitats. Morphological characters are possible adaptations reflecting habitat preferences. We investigated morphological traits in a guild of small foliage-gleaning, insectivorous birds breeding in central Ontario, Canada (4 families; 23 species). We measured 27 skeletal features and compared 2 types of differences: those between conspecific sexes (that by necessity occupy the same macrohabitat) and those between different guild members (that tend to occupy different macrohabitats). We predicted that if macrohabitat differences are reflected in different morphologies, the differences between conspecific sexes would be less than the differences between species, at least after correcting for size. We used (a) principal components analysis (PCA), (b) distance matrices derived from PCA scores, and (c) Mantel tests. Although conspecific male and female morphologies were correlated, nearest neighbours in morphological space were frequently non-conspecifics. Accordingly, because morphological differences between similar species that tend to occupy different macrohabitats are often smaller than morphological differences between conspecific sexes, our findings indicate that skeletal morphology provides no basis for explaining patterns of within-guild macrohabitat segregation.