-We used discriminant functions analysis to identify specimens of wood pewees (Contopus) from the Great Plains. Size and shape differences between C. virens and C. sordidulus are slight, but we found significant differences in multivariate space. Specimens from places where both species occur are not phenetically intermediate; hence we have no evidence of hybridization. Singing birds were morphologically like those of the species whose song they sang; hence we have no evidence of song switching. Our analyses show that C. virens occurs in summer as far west as Crockett Co., Texas; Morton Co., Kansas; and Phillips Co., Montana. Contopus sordidulus occurs in summer east to Phillips Co., Montana; it is a fairly common and regular migrant through extreme western Kansas. The Eastern and Western wood pewees (Contopus virens and C. sordidulus) are difficult to separate both in the field and in the hand (Phillips et al. 1966:169-170, Browning 1977). The most conspicuous difference between the two-their distinctive primary songs-is of restricted use: the females do not sing; spring migrants sing rather infrequently and fall migrants probably do not sing at all; in collections, song type is rarely indicated specimen labels. Additionally, the song may be partially or wholly learned and thus be an imperfect indicator of kind especially in sympatry (where individuals may have learned either both songs or the wrong one). In practice, non-singing pewees are often identified on geographical grounds, and indeed the largely allopatric distributions of the forms leaves this an expedient and acceptable solution in most cases. However, their ranges do overlap both where they winter in Central and South America (A.O.U. 1957) and in Mexico and the western Great Plains where they occur in passage. In the Great Plains they occur, sympatrically in some places, as uncommon breeders (Fig. 1; Short 1961, Barlow and Rising 1965, Godfrey 1966, Rising 1974). The apparent intermediacy of several specimens from western Kansas and Nebraska suggests that hybridization between the forms may occur in the Great Plains (Short 1961, Barlow and Rising 1965, Schueler and Rising 1976). To date, however, the identification of these specimens from the Great Plains has been imprecise. For two reasons, we became interested in determining whether or not these pewees hybridize. First, the answer would help us to understand the significance of bright display plumages of birds. Sibley (1957) argued that such plumages have evolved because they facilitate accurate species recognition and thus reduce hybridization. Rohwer et al. (1980), however, pointed out hat hybridization seems to occur most commonly among individuals of both the most dimorphic and the brightest species. On the basis of feather color intermediacy, for example, it has been argued that buntings (Passerina), grosbeaks (Pheucticus) and orioles (Icterus) all hybridize rather commonly in the Great Plains (West 1962, Rising 1970, Emlen et al. 1975). But do these hybridize more commonly than less brightlycolored birds, or do they merely seem to hybridize more commonly because hybrids (intermediates) are more easily defined in species with distinctive plumages? Perhaps hybridization occurs in species that are similar in appearance such as chickadees (Parus), wood pewees, and crested flycatchers (Myiarchus), and we just do not recognize the intermediates. Second, Mayr and Short (1970:60) suggested that these two pewees are conspecific; information pertaining to mate selection in sympatry will clarify their genetic relationship. Our purposes are (1) to establish rigorous criteria for identification of specimens, and using these criteria (2) to ide tify, as accurately as possible, our pewee specimens from the Great Plains as well s to reassess earlier identifications (e.g., those in Barlow and Rising 1965, Rising 1974).
Read full abstract
Jun 1, 1987
Ron J Cleevely 
Open Access
