Mitochondrial-DNA Variation in the Polytypic Alaskan Song Sparrow

Abstract

--Previous studies of mitochondrial-DNA (mtDNA) variation in the polytypic Song Sparrow (Melospiza melodia) have not found geographic structure of haplotype distributions. We used restriction-fragment analysis of mtDNA to analyze geographic variation of Song Sparrows in Alaska, where phenotypic differentiation between populations is greatest. We examined 42 individuals, representing five subspecies, from island and mainland localities. In addition, we examined two specimens each from Washington and California. An average of 55 sites per individual distinguished 12 mtDNA haplotypes, which were all closely related (d = 0.001-0.008). Geographic structure was evident in the distribution of haplotype frequencies, but not in their phylogenetic affinities. Received 21 January 1991, accepted 1 August 1991. THE STUDY of morphological variation between allopatric populations of birds has led to inferences about microevolutionary processes based on the rarely tested assumption that the measured traits have a large additive genetic variance and, therefore, indicate at some level the genetic structure of a population (Mayr 1942, 1963). Consequently, multiple subspecies have been named for many North American bird taxa based on geographic variation in morphology (AOU 1957). Attempts to describe the genetic structure of avian populations more directly using allozymes have been stymied by the relative lack of population differentiation in birds compared with other vertebrates, presumably because of their potential for great vagility via flight (Barrowclough 1983). In genetic terms, population differentiation will occur only if the effective population size is small relative to the level of gene flow. Analysis of mitochondrial-DNA (mtDNA) variation has proven to be a sensitive molecular assay of intraspecific genetic variation (Wilson et al. 1985, Avise et al. 1987, Shields and Wilson 1987b) in part because the effective population number for mtDNA genes is reduced (Birky et al. 1983). Mitochondrial DNA is maternally inherited without recombination (Giles et al. 1980). Because of this uniparental inheritance, the evolutionary effective population size for mitochondrial genes is expected to be equal to the number of females under the neutral-mutation •Present address: Department of Genetics, University of Georgia, Athens, Georgia 30602, USA. theory. Assuming a sex ratio of 1:1, the mean time to fixation or loss of mitochondrial mutations is one-half that for nuclear genes (Birky et al. 1983). The difference in effective-population number between mitochondrial and nuclear genes means that mitochondrial genes can be subdivided in a population with migration rates that result in panmixis for nuclear genes. In addition, sequence evolution in mtDNA is rapid compared to single-copy nuclear loci so that sequence differences will accumulate in mtDNA over shorter divergence times (Brown et al. 1979). Application of mtDNA analysis to birds at the intraspecific level has revealed genetic structure with a variety of geographic patterns and degrees of divergence (Avise and Ball 1991). Two of the more extensive surveys provide a useful comparison. Red-winged Blackbirds (Agelaius phoeniceus) from across the continent possessed many closely related clones of mtDNA (the maximum genetic distance p = 0.008) with limited geographic structure (Ball et al. 1988). In contrast, the Seaside Sparrow (Ammodramus maritimus) had relatively large genetic distances (maximum p = 0.013) between Atlantic and Gulf Coast populations, but within each of those populations, p averaged only 0.002 and clonal diversity was low (Avise and Nelson 1989). This discontinuous pattern of mtDNA variation presumably results from high levels of historical contact among populations within each region (similar to that in Red-winged Blackbirds), but not between regional populations due to separation by a long-term zoogeographic barrier (Avise et al. 1987, Avise and Nelson 1989). Both species are polytypic as