Figure 3. As the difference in plumage between males and females increases, the likelihood of mating (observed number of pairs – expected number of pairs due to random chance) decreases. Plumage was quantified by scoring 11 plumage patches on males and females of each phenotypic class (see Supporting Information for plumage scoring criteria). Below the x -axis are representative illustrations of the different pair types (top row: males, bottom row: females), depicting the male-female plumage divergence with increasing values of x.
Appendix 1. A plumage scoring criterion. We modified the method of Toews et al. (2016) to rank Vermivora spp. plumage with the exception of three edits (denoted using asterisks within table). Intermediate throat pigmentation (“Thr.”) between a Golden-winged Warbler (Thr. = 0) and Blue-winged Warbler (Thr = 2) should resemble a Brewster’s Warbler (i.e ., white in color) rather than that of a second-year male Golden-winged Warbler as in Teows et al. (2016). Additionally, breast plumage color (“Breast”) and belly plumage color (“Belly”) for adult male Blue-winged Warblers are bright yellow (rather than yellow-green) so those plumage categories were extended (from 4- to 5). All other plumage scoring criteria were identical to those used by Toews et al. (2016): wing bar width (“W. Bar Wd.”), wing bar color (“W. Bar. Col.”), nape color (“Nape”), back color (“Back”), rump color (“Rump”), auricular pigmentation pattern (“Auric.”), supercillium color (“Super.”), and malar color (“Malar”). A male Blue-winged Warbler with typical plumage would score 40 using this protocol whereas a Golden-winged Warbler would score 0.