HEAD-SCRATCHING AMONG NORTH AMERICAN WOOD-WARBLERS (PARULIDAE)

A survey of 40 species of 14 genera of North American wood-warblers (Parulidae) reveals that the head-scratching method employed is surprisingly stable within species. The experiments by Nice & Schantz (1959a, b) induced some normally overwing head-scratchers to scratch the head under the wing. It is suggested that this was because their leg rings became caught in secondaries, making normal overwing head-scratching impossible. A few exceptional head-scratching patterns under different conditions invariably involved normally overwing head-scratchers employing the underwing method, and a few species head-scratch under the wing as nestlings but change to overwing before fledging. In all, 31 species appear to be normally overwing head-scratchers, seven are underwing head-scratchers, one species uses both methods and one remains uncertain. There is no evidence for lateral preference in either head-scratching method, and head-scratching is only loosely linked sequentially to preening. The functional significance of head-scratching may be related to blockage of the eustachian tube, or to cleaning and oiling the feathers. Avian head-scratching is more difficult to homologize with mammalian head-scratching than others have considered it to be, but all evidence suggests that within birds underwing head-scratching is phylogenetically primitive. In wood-warblers, the head-scratching method does not correlate with taxonomy. However, ground-dwelling wood-warblers tend to scratch the head under the wing and arboreal wood-warblers over the wing. This correlation provides the first strong clue to the functional significance of the difference between methods in passerine birds.