PHYLOGEOGRAPHY OF THE MALLARD (ANAS PLATYRHYNCHOS): HYBRIDIZATION, DISPERSAL, AND LINEAGE SORTING CONTRIBUTE TO COMPLEX GEOGRAPHIC STRUCTURE

Auk, The, Jul 2005 by Kulikova, Irina V, Drovetski, Sergei V, Gibson, Daniel D, Harrigan, Ryan J, Et al

Demographic analysis of mtDNA data showed signs of exponential population growth. The observed distributions of the pairwise nucleotide differences among group A and B haplotypes did not differ from the distribution expected under Rogers' (1995) model of sudden population expansion (P-values > 0.07; Fig. 4). Tajima's (1989) D also differed from zero for both the group A and B haplotype clades (D^sub A^ = -2.14, P 9.00, df = 1, P 9.00, df=1, P

POPULATION DIFFERENTIATION

Overall, 36.3% of observed genetic variation was partitioned among populations. Within clade A, 3.6% of the variation was accounted for by variation among populations, and 44.0% by variation among populations in the B clade. Thus, variation was concentrated primarily within populations, but much more so in clade A than in clade B.

Western Russia differed from the other regions in its lack of group B haplotypes. By contrast, Alaska had the highest frequency of B haplotypes (22 of 27). In North Asia and the Aleutian Islands, B haplotypes occurred infrequently. Group A and B haplotypes were not distributed randomly among the four different sampling regions (G = 36.24, P

Pairwise OST values calculated for all haplotypes ranged from 0.0018 (North Asia vs. Western Russia) to 0.4956 (Western Russia vs. Alaska), with the highest values observed in pairwise comparisons that included mainland Alaska (Φ^sub ST^ > 0.4112; Table 3). All other pairwise Φ^sub ST^ values were relatively low, indicating low levels of genetic divergence among Asian populations. The Φ^sub ST^ values for group A haplotypes varied from 0.0015 (Alaska vs. Western Russia) to 0.1461 (Aleutian Islands vs. Alaska; Table 3). In this series of pairwise comparisons, only the Aleutian Islands differed from other localities, but not for the Aleutian Islands vs. Alaska with a Bonferroni-corrected α = 0.0083 for six pairwise comparisons (Φ^sub ST^ = 0.0728-0.1461; Table 3).

DISCUSSION

PHYLOGEOGRAPIIIC STRUCTURE

The mtDNA phylogeny included two divergent haplotype groups that correspond to Avise et al.'s (1990) group A and B and the type 1 and 2 haplotypes identified by Johnson and Sorenson (1999). Group A haplotypes were most common and were found in Western Russia, North Asia, the Aleutian Islands, and mainland Alaska (Fig. 5). By contrast, most group B haplotypes occurred only in Alaska or the Aleutian Islands, though three group B haplotypes occurred in the Primorye region of the Russian Far East. Those three haplotypes, however, formed a separate subclade nested within the group B clade, differing by at least five mutation steps from all other group B haplotypes (Kulikova et al. 2004).

The paraphyly of Mallard mtDNA is well documented. Mallard mtDNA haplotypes are paraphyletic with respect to the Hawaiian Duck, American Black Duck, Mexican Duck, Mottled Duck, and Eastern Spot-billed Duck (Avise et al. 1990, McCracken et al. 2001, Rhymer 2001, Kulikova et al. 2004). Group A and B mtDNA haplotypes were previously found in North American and Asian Mallards, whereas only group B haplotypes occurred in the abovementioned monochromatic relatives of the Mallard from North America and the Hawaiian Islands. Eastern Spot-billed Duck, which is sympatric and hybridizes with Mallards in the Russian Far East and northeastern China, had both types of haplotypes, but B haplotypes in that species (hereafter "group SB") were different from North American B haplotypes found in American Black Duck, Mottled Duck, and Mexican Duck (Kulikova et al. 2004).

The distribution of group A and B haplotypes in Eurasian and North American Mallards can be explained by at least two hypotheses. The "Asian invasion" hypothesis, first articulated by Palmer (1976), albeit not in genetic terms, implies "out of Asia" colonization and subsequent introgressive hybridization between the Mallard and its close relatives as the cause of mtDNA paraphyly (Johnson and Sorenson 1999, McCracken et al. 2001, Kulikova et al. 2004). In the context of the mtDNA analyzed here, that corresponds to Mallards bearing group A haplotypes colonizing and introgressing with closely related species characterized by group B haplotypes. The conspicuous absence of group B haplotypes from Western Russia is consistent with hybridization as the source of group A and B haplotype mixing in sampled populations. In Western Russia and Europe, the Mallard distribution does not overlap with any other closely related species. The same is true of Mallards breeding in North Asia, though those populations probably hybridize with Eastern Spotbilled Ducks where they co-occur in the Russian Far East (e.g. Primorye), Korea, northern Japan, and northeastern China (Kulikova et al. 2004). The three group B haplotypes we identified in Mallards from Primorye (SB6, SB7, SB9) are group SB Eastern Spot-billed Duck haplotypes (Kulikova et al. 2004), and their presence in the Mallard gcno pool supports hybridization as the source of shared group A and B haplotypes in Mallards and Eastern Spot-billed Ducks from Primorye.