KNOWN SECONDARY CONTACT AND RAPID GENE FLOW AMONG SUBSPECIES AND DIALECTS IN THE BROWN-HEADED COWBIRD

We analyze geographic variation in morphology for Sierra Nevadan populations of Brown-headed Cowbirds (Molothrus ater) in relation to two levels of geographic structure of populations: 1) subspecies and 2) vocal dialects. Two morphologically distinct subspecies occupy opposite slopes of the Sierra Nevada: M. a artemisiae on the east slope is larger than M. a. obscurus on the west slope, and its juveniles have predominantly white versus yellow rictal (bill) flanges. Populations of obscurus moved into California from the lower Colorado River around 1900 and invaded the western Sierra during the 1930's. A relatively constant type of flight whistle occurs in obscurus populations up to 700 km apart, but east-slope artemisiae are divided into many distinct dialects. The means of seven morphological measurements and of principal component 1 (i.e., body size) for 2,287 individuals show similar clinal patterns for adult males, yearling males, and females over a 60-km north-south transect along the eastern Sierra Nevada: large sizes occur at the southernmost study site at Bishop, decrease clinally to the north to Mammoth Lakes, and then increase clinally to the north from Mammoth Lakes to the northernmost site at Lee Vining. This reversal in cline slope is reiterated for variation in the frequency of white-flanged juveniles. Our data do not support adaptive explanations for the clines. Rather, the most parsimonious explanation is that there has been flow of obscurus genes into artemisiae from the west slope across the crest of the Sierra Nevada into the Mammoth Lakes area. Concordantly, the Sierran crest at Mammoth Lakes is considerably lower than any other point within about 80 km to the north or 180 km to the south. Two other findings strongly support this interpretation. First, wing lengths of our longest-winged samples (at Bishop and Lee Vining) are significantly shorter than those of cowbirds collected in the same region between 1912 and 1922. Second, the wing lengths of Mammoth Lakes adult males and females during 1978-1981 are significantly larger than those of the 1982-1985 period. This putative hybridization must have been rapid, as there have been fewer than 50 years since secondary contact could have occurred. Calculation of the gene-flow parameter Ι of Endler (1977) indicates that the gene-flow rate reported here is higher than for nearly all of the species he surveyed. This gene flow occurs between two subspecies with different flight whistles and across several cowbird flight-whistle dialects within artemisiae. Therefore, vocal differences among these dialects do not appear to be a strong deterrent to gene flow.