PLUMAGE AND SIZE VARIATION IN THE MOURNING WARBLER

The Mourning Warbler (Oporornis philadelphia) is highly variable throughout its breeding range (Hall 1979, Pitocchelli 1990) but a thorough study of geographic variation in this species is lacking. Therefore, the nature of this variation is poorly understood. I studied macrogeographic variation in plumage and skeletal characters of 364 specimens from most of the breeding range in North America. There was no clear pattern of variation in external or skeletal measurements. Univariate analyses of external characters and skeletal characters revealed little variation between localities for most characters. Multivariate analyses also yielded little structure or subdivision among populations. Regression analyses failed to find significant north-south clines in several size measurements (P > 0.05). The Mourning Warbler did not conform to predictions based on Bergmann's rule. Spatial Autocorrelation of plumage. external and size measurements only found one character-TAIL length-which conformed to the isolation by distance model. A significant Canonical Correlation (P < 0.05) was found between external measurements and climatic data but only 36% of the external variation could be explained by the environmental variables. There appears to be little geographic variation among populations from the breeding range of the Mourning Warbler. There was a weak relationship between some morphological measurements and climatic variables. Environmental selection does not appear to play a role in shaping the pattern of variation on the breeding grounds in North America.