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

Haplotype diversity (H) and nucleotide diversity (π) were calculated using ARLEQUIN, version 2.0 (Schneider et al. 2000). Analysis of molecular variance (AMOVA; Excoffier et al. 1992), pairwise mismatch distributions, and Rogers' (1995) model of sudden population expansion for species and haplotype lineages were also calculated using ARLEQUIN, version 2.0. Pairwise ST values and P-values were calculated (1) among all populations including all haplotypes and (2) among all populations excluding group B haplotypes. Bonferroni correction factors (a = 0.05; Sokal and Rohlf 1981) were used for those tests. We used a 4 × 2 G-test of independence (Sokal and Rohlf 1981) to test the null hypothesis that group A and B haplotypes (Avise et al. 1990) were distributed randomly in the four regions.

We used coalescent analysis as implemented in FLUCTUATE, version 1.3 (Kuhner et al. 1998) to estimate the neutral parameter theta (θ = N^sub e^μ, where N^sub e^. is the effective female population size and μ is the mutation rate) and the growth rate parameter (#) from posterior distributions of mtDNA gene trees. Transition-transversion ratios for the A and B clades were set to 29.35 and 17.55, respectively, using values obtained from MODELTEST, version 3.06 (Posada and Crandall 1998), and Markov-chain Monte Carlo analyses were repeated with different numbers of short and long chains until parameter estimates stabilized. Final parameter estimates were obtained with 10 short chains of 1,000 steps and three long chains of 15,000 steps, with posterior sampling increments every 20 generations. We used hierarchical likelihood ratio tests to compare the model with g set to zero (i.e. a stable population) and allowed to vary. Significance was determined using the chi-square distribution with one degree of freedom.

RESULTS

MITOCHONDRIAL DNA CONTROL REGION SEQUENCE VARIATION

We identified 103 unique haplotypes comprising 666-667 nucleotides among 152 control region sequences. Of the 667 control-region nucleotide positions, 87 (13.0%) were variable, and 60 (9.0%) were parsimony informative. All but four variable positions occurred within the first 351 positions of the 5' end of the control region; the other substitutions occurred at positions 372, 521, 639, and 661. Transitions occurred at 82 positions, and transversions occurred at six positions. Two positions possessed both transitions and transversions. Three one-base deletions were present in single individuals at positions 165, 206, and 210. A fourth indel was present at position 211. A gap at that position was observed in Mallards from North Asia (7), the Aleutian Islands (2), and Alaska (24), but not in Mallards from Western Russia. That indel generally discriminates the two divergent haplotype groups identified in previous studies (Avise et al. 1990, McCracken et al. 2001, Kulikova et al. 2004, M. D. Sorenson and R. J. Harrigan unpubl. data). In our data set, however, three Primorye Mallards with group A haplotypes in addition to all group B haplotypes had a gap at position 211 instead of thymine or cytosine. Average nucleotide base composition was 33.3% C, 25.9% T, 25.3% A, and 15.5% G and similar to that of other birds (Baker and Marshall 1997). No evidence of coamplified nuclear mtDNA (Sorenson and Fleischer 1996, Sorenson and Quinn 1998) was observed in our data.