Dorsal color variation among subspecies of the "Oregon" Dark-eyed Junco complex

Initial descriptions of avian subspecies were based on expert opinions of geographic variation in phenotypes and are inherently subjective. Although best practices for subspecies delimitation continue to be debated, reassessing subspecies limits with current, quantitative methods is important toward refining and improving taxonomic treatments. Plumage coloration is the basis of many subspecies diagnoses, but is potentially problematic because of the historical lack of quantitative methods to quantify color. Recently developed methods, such as colorimetry, provide repeatable measurements of color variation that can be used to reassess subspecies limits. In this study, we reassess color variation among subspecies of the Oregon Dark-eyed Junco (Junco hyemalis [oreganus Group]) complex, in which differences in back and hood color were established as diagnostic characters. We measured back and hood brightness and hue in 206 museum specimens among five Oregon Dark-eyed Junco subspecies using a colorimeter. We then compared mean measurements among subspecies and conducted a discriminant function analysis to assess how well dorsal color predicted subspecies. Our data correctly classified only 67.9% of males and 82.5% of females to their presumed subspecies. Furthermore, no adjacent subspecies pairs passed the "75% rule" due to extensive overlap in plumage characters. Thus, back color alone is not as effective in diagnosing Oregon Dark-eyed Junco subspecies as originally described, suggesting a possible taxonomic revision. Specifically, similarity in phenotypic and genetic data suggests that some combination of thurberi, montanus, and shufeldti may be lumped to recognize broad, clinal variation in dorsal color alongside clinal variation in other phenotypes and extensive gene flow.