Skip to main content
NCBI home page
Proceedings of the Royal Society B: Biological Sciences logoLink to Proceedings of the Royal Society B: Biological Sciences
. 2004 Dec 6;272(1558):71–78. doi: 10.1098/rspb.2004.2908

Natural selection on mitochondrial DNA in Parus and its relevance for phylogeographic studies

Robert M Zink 1
PMCID: PMC1634939  PMID: 15875572

Abstract

I examined mitochondrial ND2 sequences in species of tits (genus Parus) to determine whether substitutions were neutral or under selection. Haplotype trees in these species are shallow and geographically unstructured, which could be a signature of recent demographic expansion or purifying selection. McDonald–Kreitman tests revealed a pattern of excess replacement polymorphisms for closely related taxa. Replacement substitutions tended to be less common in the basal parts of haplotype trees, suggesting that they are selected against over evolutionary time. Thus, the pattern of selection is consistent with a model of mildly deleterious haplotypes. To investigate the significance of this selection, the ND2 gene was subdivided into surface and transmembrane portions, and synonymous–non-synonymous sites, and FST-values (the amount of variance distributed among populations) computed for each. Among sampling localities in P. major, both partitions gave similar estimates of FST, which is consistent with demographic expansion as the cause of the shallow trees. In P. montanus, two localities contained high percentages of individuals that had a single replacement substitution in the transmembrane portion of the gene, which inflated FST-values relative to the surface sites. Thus, the interpretation of the genetic differentiation of these two sites could be either geographical isolation or selection. Given that the sole substitution causing the high FST-values was a replacement one, selection is implicated. However, this substitution occurred in only 2 of 12 localities, revealing that overall the phylogeographic pattern was not biased by selection, once the cause of the two high pairwise FST-comparisons was recognized. Investigators should consider the source of significant genetic heterogeneity within species, to determine if the variation is likely caused by demographic isolation or selection. However, it seems unlikely that most phylogeographic analyses are invalidated by the effects of natural selection.

Keywords: natural selection, ND2, mitochondrial DNA, selective neutrality, McDonald–Kreitman test

1. Introduction

A plethora of phylogeographic studies (Avise 2000) exists based on variation in the DNA found in the mitochondrion (mtDNA). A basic premise is that the variation observed is selectively neutral and that genetic patterns reflect the historical demography of populations and not the effects of natural selection. Testing for natural selection has been sporadic or superficial, despite the importance of the neutrality assumption. Several studies have found evidence for selection on mtDNA (Messier & Stewart 1997; Wise et al. 1998; Creevey & McInerney 2002; Eyre-Walker 2002; Jansa et al. 2003). This has led some researchers (e.g. Ballard & Whitlock 2004) to challenge the evolutionary interpretation of mtDNA variation in some contexts. Thus, it is timely to investigate the nature of selection and its significance for phylogeographic inquiry, which has become a major area of study (Avise 2000).

A large number of mtDNA phylogeographic studies have been conducted on birds (summaries in Avise & Walker 1998; Zink 2004). A typical result of many studies, especially those of passerine birds (songbirds), is a shallow (sequences differ by less than 1%) and geographically unstructured haplotype tree. This result is interpreted to be a signature of population increase and range expansion, and that populations are connected either currently or in the very recent past by gene flow (Zink et al. 2003). It is also possible that shallow haplotype trees are due to purifying selection. In the latter case, our understanding of recent population history could be biased because the tree structure does not represent demographic history.

Wise et al. (1998) studied variation in ND2 in Homo sapiens and Pan troglodytes and found significant differences in the F-value of Fu (1997) in both species, indicating either non-neutral sequence evolution or recent demographic expansion. The authors subdivided the ND2 gene into surface and transmembrane portions. Using the McDonald–Kreitman (1991) test, Wise et al. (1998) found evidence for selection in humans, in the form of an excess of replacement polymorphisms within the transmembrane portion of the gene only. This would not be expected if demographic expansion were the cause of the significant Fu's F-values because each region should be affected concordantly. This type of result calls into question the use of mtDNA data for phylogeographic inference because divergence within and among species might not be due solely to drift.

In this study, I explore variation in newly acquired sequences of the ND2 gene in several species of tits (genus Parus) from Eurasia. Two of the species, P. major and P. montanus have been subjected to multiple phylogeographic analyses, based on sequence variation in the non-coding mitochondrial control region (Kvist et al. 2001, 2003; Pavlova 2004). In both species (sensu stricto), a shallow haplotype tree was discovered, with no significant phylogeographic divisions. This type of tree could result from recent demographic expansion, expected for species after amelioration of Ice Age climates. Alternatively, purifying selection might be responsible for the lack of genetic heterogeneity over the ranges of each species. My goals were to test for deviations from selective neutrality and to determine whether any observed departures influenced phylogeographic interpretations.

2. Material and methods

Three focal taxa included were great tit (Parus major, n = 87), Japanese tit (P. minor, n = 29), and willow tit (P. montanus, n = 139); details of sampling are given in Pavlova (2004). Four individuals of the Turkestan tit (Parus bokharensis), the sister species to P. major (Kvist et al. 2003), were used. A single individual of the marsh tit (P. palustris) was used as an outgroup to P. montanus. GenBank accession numbers are AY732496-733007, AY734248-734251. Basic sequence statistics were calculated for each species using DnaSP (Rozas et al. 2003), including: nucleotide diversity (π), theta (θ), D (Tajima 1989), D (Fu & Li 1993), F (Fu & Li 1993), and F (Fu 1997). Using a model (Persson & Argos 1994) of ND2, the sequences were divided into surface (372 bp) and transmembrane (666 bp) sections (the stop codon was ignored). McDonald–Kreitman (1991) (M–K) tests implemented in DnaSP were used to compare all base pairs, surface sites, and transmembrane sites. M–K tests compare the ratio of fixed (invariable) synonymous (SI) to non-synonymous (RI) substitutions to the ratio of polymorphic (variable) synonymous (SV) to non-synonymous variable (RV) substitutions. The SI:RI ratio is taken as the expectation for neutral evolution of sequences. Only unique haplotypes were used in M–K tests.

Because M–K tests are pairwise, one cannot infer whether departures from neutrality were in one or both taxa being compared. To explore the phylogenetic deployment of putative selective effects, I used the relative ratio test of Creevey & McInerney (2002). In this method, a rooted phylogenetic tree is computed (and assumed correct). The software constructs a neighbour-joining tree using distances from synonymous sites, non-synonymous sites, or their ratio; results presented below were independent of the distance measure used for neighbour-joining trees. Hypothetical ancestral sequences are constructed for each node, and in the event of ambiguities, resolved using matrices of zero-, two- and fourfold degenerate positions. All substitutions are identified on the tree and classified as either non-synonymous or synonymous. For each internal branch of the tree, substitutions are counted from the branch to the descendant tips, resulting in four values: RI, RV, SI and SV. A G-test or Fisher's exact test (Sokal & Rohlf 1981) is used to compare the ratio of SV:RV to SI:RI (the expected value under selective neutrality). If significant differences occur, one can infer whether the positive selection is directional (high RI) or non-directional (high RV). Computations are implemented in software Crann written by C. Creevey (available at http://bioinf.may.ie/crann/).

Creevey & McInerney (2002) also describe a method to detect negative or purifying selection. Here, the ratio of replacement to silent sites is computed for each lineage in the tree, and the number of replacement and silent changes counted on each branch. The expectation is that the ratio of replacement:silent changes on each branch will be predicted by the overall ratio of the two types of site on the entire tree. It is possible to observe both positive and negative selection on the same branch if overall there is strong purifying (negative) selection but a few highly favoured replacement substitutions.

Creevey & McInerney (2002) and others (Maynard Smith 1994; Hughes 1999) note that if species compared are too distant, saturation will obscure the pattern of replacements. In my analyses, I found that P. montanus and P. major differed by over 10% uncorrected sequence divergence. Hence, analyses were done independently on two major groups, P. major, P. minor and P. bokharensis, and P. palustris and P. montanus (which differ by ca. 7% uncorrected sequence divergence; R. M. Zink, personal observation).

I used a structural model of ND2 (Persson & Argos 1994) to divide the gene into surface (codons 1–3, 67–81, 142–183, 244–264, 325–345, 406–426, 487–561, 637–711, 796–816, 901–963, 1024–1038) and transmembrane (remaining codons) sections. To measure the pattern and extent of population subdivision, I used DnaSP to compute FST for all bases, surface and transmembrane sections; individuals were grouped by geographical locality. Because of the apparent pattern of selection in P. montanus, FST was computed for synonymous sites only and a mismatch distribution (Rogers & Harpending 1992) was produced. The mismatch distribution was compared with the theoretical expectation for an expanding population implemented in DnaSP (Rozas et al. 2003).

3. Results

Nucleotide diversity (table 1) was similar for P. major (0.0012) and P. minor (0.0012), whereas P. montanus was more variable (0.0021). For all three species, Fu's F was significant, as were Tajima's D, Fu and Li's D*, and Fu and Li's F-values (table 1). Transmembrane sites exhibited higher nucleotide diversity than surface sites in P. minor but not in P. major or P. montanus (table 1). In M–K tests of P. major and P. minor, an excess of RV was observed for all sites, transmembrane and surface regions (table 2), although the test for surface sites was not significant. For example, comparing all base positions, there were no fixed replacement substitutions (out of 23 substitutions), whereas within species 10 out of 46 polymorphisms involved replacement substitutions. All M–K tests involving P. bokharensis were not significant (small sample size is a likely issue), although again an apparent excess of polymorphisms within species involved replacement substitutions. The M–K test for P. palustris versus P. montanus was not significant despite an apparent excess of RV substitutions (SI = 43, RI = 7, SV = 53, RV = 12; Fisher's exact test p-value 0.62). Values shown in table 2 for P. montanus versus P. major, P. bokharensis and P. minor are for comparison, but because of the likelihood of saturation (at least a 40% underestimate based on maximum-likelihood distance correction), they are not interpreted.

Table 1.

Basic sequence statistics for species of Parus at ND2.

(Values are not presented for P. bokharensis because only four individuals were sequenced.)

taxon/region S π θ Tajima's D Fu and Li's D Fu and Li's F Fu's F
P. major (n = 87)
all sites 34 0.0012 0.0065 −2.57** −5.30* −5.10* −26.17**
surface 10 0.0014 0.0054 −1.94* −2.74* −2.92* −10.23**
transmembrane 24 0.0010 0.0072 −2.58* −5.35* −5.15* −24.07*
Parus minor (n = 29)
all sites 13 0.0012 0.0034 −2.22** −3.01* −3.24* −13.14**
surface 2 0.0005 0.0014 −1.25 −0.73 −1.01 −1.63
transmembrane 11 0.0015 0.0046 −2.20* −3.20* −3.40* −12.70**
Parus montanus (n = 139)
all sites 63 0.0021 0.0114 −2.55** −5.43** −5.00* —
surface 21 0.0024 0.0112 −2.25** −3.80** −3.84* −28.40**
transmembrane 42 0.0020 0.0115 −2.51** −5.16* −4.87* −34.48**
Open in a new tab

*p < 0.05

**p < 0.01.

Table 2.

(a) Synonymous and non-synonymous mutations in ND2 between three species of Parus. (b) As in (a) but for transmembrane sites only.

(Top and bottom halves of the matrix contain comparisons for surface and all sites, respectively. Groups of four numbers off diagonal refer to numbers of synonymous (top row) and non-synonymous (bottom row) mutations fixed between species (left column) and polymorphic within species (right column). For example, comparing P. major and P. minor, the ratio of synonymous:non-synonymous mutations fixed between species was 23:0 and the ratio for polymorphic sites was 36:10 (p = 0.014). p-values are from Fisher's exact test.)

P. major P. minor P. montanus P. bokharensis
(a)
P. major 6 10 23 30 2 9
0 2 13 3 1 1
p = 0.53 p = 0.008 p = 0.94
P. minor 23 36 17 21 22 9
0 10 12 3 13 1
p = 0.014 p = 0.039 p = 0.19
P. montanus 70 77 68 63 22 21
35 19 34 15 14 2
p = 0.038 p = 0.05 p = 0.018
P. bokharensis 6 29 30 12 72 54
2 7 2 3 38 12
p = 0.89 p = 0.38 p = 0.028
(b)
P. minor 17 26
0 8
p = 0.040
P. montanus 50 47 51 42
22 16 22 12
p = 0.57 p = 0.43
P. bokharensis 4 20 21 11 50 33
1 6 1 2 24 10
p = 0.67 p = 0.64 p = 0.40
Open in a new tab

Relative ratio tests for branches in the haplotype tree of P. major and P. bokharensis (rooted with all P. minor haplotypes) were not significant, indicating no overall trend for directional (positive) selection. However, the pattern of substitutions (figure 1) suggested that non-synonymous substitutions were underrepresented in the basal parts of the haplotype tree. That is, as one proceeds from the tips to the base of the tree, silent substitutions increase but replacement substitutions increase at a much lower rate (figure 2). The same pattern was observed with surface and transmembrane sites (results not shown), irrespective of whether the neighbour-joining tree was constructed with distances calculated from synonymous substitutions non-synonymous substitutions or the ratio of the two.

Figure 1.

Figure 1

Open in a new tab

Neighbour-joining tree based on values of D for P. major, P. bokharensis and P. minor (root). Stars refer to nodes with significant deficiencies of replacement substitutions. The two boxes show results of M–K tests (1991) that were not significant but which show the nature of substitutions at the bases of major clades.

Figure 2.

Figure 2

Open in a new tab

Numbers of synonymous (filled squares) and replacement (filled diamonds) substitutions at nodes on tree in figure 1 (values for nodes 34–37 not plotted). The number of synonymous changes increases with distance from the tip of the tree faster than replacement substitutions.

The haplotype tree for P. montanus (rooted with P. palustris) showed one significant relative ratio difference due to an excess of RV substitutions, indicating positive non-directional selection (node 56, SI = 71, RI = 9, SV = 71, RV = 29). As in P. major, the haplotype tree (figure 3) showed a relatively large number of basal nodes in which replacement changes were underrepresented. Additionally, one terminal branch of the tree (connecting haplotypes 44, 9, 34, 41, 4 and 57) showed a preponderance of branches with underrepresented replacement substitutions. The number of synonymous substitutions increases from the tips of the tree to the base whereas replacement substitutions levels off (not shown), as in figure 2. Thus, non-synonymous substitutions are younger on average.

Figure 3.

Figure 3

Open in a new tab

Neighbour-joining tree for P. montanus rooted with P. palustris. Stars refer to nodes with significant deficiencies of replacement substitutions. At node 56 there was a significant M–K test.

Multiple sampling localities are represented in the P. major and P. montanus datasets. The plot of FST-values obtained from surface and transmembrane sites among samples of P. major (figure 4) show that neither gene region provides qualitatively different estimates of geographical structuring. At a higher level of divergence, the same result was obtained for the comparison of P. major, P. bokharensis and P. minor (figure 5).

Figure 4.

Figure 4

Open in a new tab

Plot of FST-values derived from transmembrane portion of ND2 versus surface portion for samples of P. major. Although there appears to be scatter, all individual FST-values are low and not significant (all values of χ2 > 0.50).

Figure 5.

Figure 5

Open in a new tab

Plot of FST-values derived from transmembrane portion of ND2 versus surface portion for samples of P. major, P. minor and P. bokharensis.

For P. montanus, FST-values based on the surface and transmembrane portions (not shown) differed, because in several instances, low values of the former were paired with high values of the latter. Inspection of the results showed that the high transmembrane FST-values involved comparisons of samples from Anadyr and Kamchatka with other localities. To explore this result further, FST-values based on all base pairs and synonymous sites only were compared (figure 6), with the same result—pairwise comparisons in which one of the sites was either Anadyr or Kamchatka were much higher for all base pairs. Lastly, mismatch distributions for P. montanus show close correspondence between observed and expected values for all sites (figure 7) and synonymous sites (figure 8) only.

Figure 6.

Figure 6

Open in a new tab

Plot of FST-values derived from silent sites versus all base positions in ND2 for samples of P. montanus. Filled squares represent comparison of the sample from Anadyr and open square comparison of samples to Kamchatka. The dash indicates comparison of Anadyr and Kamchatka. Filled diamonds represent all other comparisons.

Figure 7.

Figure 7

Open in a new tab

Mismatch distribution for 139 individual P. montanus using all of ND2 (observed: histogram; expected: line). Average number of pairwise differences: 2.18. The fit of observed and expected distributions (Kolmogorov–Smirnov test, χ2 = 0.72, p > 0.9) implies a recent and extensive increase in population size.

Figure 8.

Figure 8

Open in a new tab

Mismatch distribution for 139 individual P. montanus using only synonymous substitutions in ND2 (observed: histogram; expected: line). Average number of pairwise differences: 1.70. The fit of observed and expected distributions (Kolmogorov–Smirnov test, χ2 = 0.18, p > 0.9) implies a recent and extensive increase in population size.

4. Discussion

The role of natural selection on DNA sequences in phylogenetics and phylogeography is unclear. Barton & Etheridge (2004, p. 1115) remarked that their analysis supported previous simulations (e.g. Neuhauser & Krone 1997) which showed that ‘…selection has surprisingly little effect on genealogies’. Whether natural selection also has little effect on phylogeographic studies is unknown, although Ballard & Whitlock (2004) suggest that selection on mtDNA could compromise evolutionary inferences. If true, hundreds of analyses could be biased. Divergence thought to be a result of isolation and genetic drift might instead reflect geographical patterns of selection. For P. major, P. bokharensis and P. minor, the M–K tests (table 2) suggest an excess of replacement substitutions segregating within species. However, these replacement substitutions tend to be at the tips of the haplotype tree rather than towards the base (figure 1). The significance of the relative basal decrease in replacement substitutions can be inflated by testing the tree at multiple nested levels. However, it is clear that replacement sites did not increase at the same rate as synonymous sites (figure 2), which themselves might be underestimated at deeper nodes (owing to saturation). Thus, the sequences appear to be evolving in a manner consistent with a model of slightly deleterious alleles. Consequently, the age of replacement substitutions is less than silent ones, the former of which are removed more quickly by selection. A similar phylogenetic distribution of replacement substitutions was noted by Nielsen & Weinreich (1999). The consequence of this relatively weak selection for phylogeographic inference appears minimal because FST-values from surface positions showed the same pattern as those from the transmembrane portion of the gene (figures 4 and 5). Thus, although replacement substitutions appear to be mildly deleterious, phylogeographic conclusions based on the sequence data appear unaffected for P. major. Furthermore, there were no fixed replacement substitutions between P. major and P. minor, and only one between P. bokharensis and P. major, suggesting that mtDNA divergence accompanying the early stages of lineage divergence and speciation is selectively neutral.

The situation in P. montanus is somewhat different. Here again there is a disproportionately large number of polymorphic replacement sites (table 2), and a strong trend for replacement substitutions to be relatively rare toward the base of the haplotype tree (figure 3). As with P. major, this is a signature of mildly deleterious alleles. However, when the pattern of FST is examined for surface versus transmembrane (not shown) and synonymous versus all sites (figure 6), there are clear discrepancies. Inspection of figure 6 revealed that the mismatched FST estimates occurred in comparisons of two northeastern sample localities, Anadyr and Kamchatka. The large values of FST occurred because of a replacement substitution at base position 452 (a transmembrane site) in a haplotype (#12 in figure 3) that is fixed in Anadyr (10/10) and common (4/5) in Kamchatka. There are no fixed synonymous differences for these localities. Hence, although it is possible by chance for a single fixed difference to be a replacement one, it is also possible that this haplotype is under positive selection, against a background of mildly deleterious alleles (Creavey & McInerney 2002). Thus, one might interpret the significant FST-values for Anadyr and Kamchatka as evidence of incipient divergence caused by geographical isolation, or selection for a new favourable haplotype. These two localities are geographically proximate, which could mean similar selective pressures are at work, or that gene flow occurs between them (or both). A definitive decision seems unlikely given that only a single substitution is involved. Nonetheless, considering the broader picture, only 2 out of 12 localities appear to be affected. Apart from being cautious about interpreting the evolutionary significance of the high FST-values for comparisons involving Anadyr and Kamchatka, overall phylogeographic inferences are not altered by the existence of mildly deleterious haplotypes.

Besides assessing population differentiation, mtDNA data are also used for assessing recent population history such as population growth. A typical procedure is to compute a mismatch distribution (Rogers & Harpending 1992) which, if unimodal, is consistent with recent large expansion in population (and likely range) size. A pattern of purifying selection could mimic population expansion by limiting mtDNA variation to a single common haplotype and a series of closely related (but neutral) ones, leading to a unimodal mismatch distribution. Analyses presented here suggest a deficit of replacement changes, which could clearly affect mismatch distributions. However, in P. montanus, there are only 12 non-synonymous changes, and the mismatch distributions based only on synonymous changes (n = 53) do not differ substantially from that based on all base pairs (figures 7 and 8). Thus, it appears reasonable to interpret the unimodal mismatch distributions as evidence for recent population expansions, which makes sense for species living in environments that have expanded greatly after the last Ice Ages.

It seems premature to be overly critical of phylogeographic inferences derived from mtDNA out of concern for natural selection. There simply have not been enough studies of the effects of selection on phylogeographic inferences based on mtDNA gene sequences. Although it would be useful to have estimates of population parameters from multiple loci, in general nuclear loci (including microsatellites) coalesce too slowly to provide tests of incipient patterns that are discernable in mtDNA gene trees. It will be critical to assess the extent of selection in other species to determine the generality of the results presented here.

Acknowledgments

The US National Science Foundation funded many of the studies summarized here. S. Rohwer, A. Pavlova and S. Drovetski contributed markedly to this work. I thank A. Pavlova for permission to analyse unpublished ND2 sequences. R. Blackwell-Rago and M. Westberg performed much of the laboratory work. B. Barber, C. Creevey, S. Drovetski, A. Jones, A. Kessen, A. Pavlova and two anonymous reviewers made comments on the manuscript.

References

  1. Avise J.C. Phylogeography. Harvard University Press; Cambridge, MA: 2000. [Google Scholar]
  2. Avise J.C., Walker D. Pleistocene phylogeographic effects on avian populations and the speciation process. Proc. R. Soc. Lond. B. 1998;265:457–463. doi: 10.1098/rspb.1998.0317. doi:10.1098/rspb1998.0317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ballard J.W.O., Whitlock M.C. The incomplete natural history of mitochondria. Mol. Ecol. 2004;13:729–744. doi: 10.1046/j.1365-294x.2003.02063.x. [DOI] [PubMed] [Google Scholar]
  4. Barton N.H., Etheridge A.M. The effect of selection on genealogies. Genetics. 2004;166:1115–1131. doi: 10.1093/genetics/166.2.1115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Creevey C., McInerney J.O. An algorithm for detecting directional and non-directional positive selection, neutrality and negative selection in protein coding DNA sequences. Gene. 2002;300:43–51. doi: 10.1016/s0378-1119(02)01039-9. [DOI] [PubMed] [Google Scholar]
  6. Eyre-Walker A. Changing effective population size and the McDonald–Kreitman test. Genetics. 2002;162:2017–2024. doi: 10.1093/genetics/162.4.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fu Y.-X. Statistical tests of neutrality against population growth, hitchhiking and background selection. Genetics. 1997;147:915–925. doi: 10.1093/genetics/147.2.915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fu Y.-X., Li W.-H. Statistical tests of neutrality of mutations. Genetics. 1993;133:693–709. doi: 10.1093/genetics/133.3.693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jansa S.A., Lundrigan B.L., Tucker P.K. Tests for positive selection on immune and reproductive genes in closely related species of the murine genus. Mus. J. Mol. Evol. 2003;56:294–307. doi: 10.1007/s00239-002-2401-6. [DOI] [PubMed] [Google Scholar]
  10. Hughes A.L. Adaptive evolution of genes and genomes. Oxford University Press; 1999. [Google Scholar]
  11. Kvist L., Martens J., Ahola A., Orell M. Phylogeography of a Palearctic sedentary passerine, the willow tit (Parus montanus) J. Evol. Biol. 2001;14:930–941. [Google Scholar]
  12. Kvist L., Martens J., Higuchi H., Nazarenko A.A., Valchuk O.P., Orell M. Evolution and genetic structure of the great tit (Parus major) complex. Proc. R. Soc. Lond. B. 2003;270:1447–1454. doi: 10.1098/rspb.2002.2321. doi:10.1098/rspb 2002.2321 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. McDonald J.H., Kreitman M. Adaptive protein evolution at the Adh locus in Drosophila. Nature. 1991;351:652–654. doi: 10.1038/351652a0. [DOI] [PubMed] [Google Scholar]
  14. Maynard Smith J. Estimating selection by comparing synonymous and substitutional changes. J. Mol. Evol. 1994;39:123–128. doi: 10.1007/BF00163800. [DOI] [PubMed] [Google Scholar]
  15. Messier W., Stewart C.-B. Episodic adaptive evolution of primate lysozymes. Nature. 1997;385:151–154. doi: 10.1038/385151a0. [DOI] [PubMed] [Google Scholar]
  16. Neuhauser C., Krone S.M. The genealogy of samples in models with selection. Genetics. 1997;145:519–534. doi: 10.1093/genetics/145.2.519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Nielsen R., Weinreich D.M. The age of nonsynonymous and synonymous mutations in animal mtDNA and implications for the mildly deleterious theory. Genetics. 1999;153:497–506. doi: 10.1093/genetics/153.1.497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Pavlova A. Ph.D. dissertation. University of Minnesota; 2004. Comparative phylogeographies of Eurasian birds. [Google Scholar]
  19. Persson N., Argos P. Prediction of transmembrane segments in proteins utilising multiple sequence alignments. J. Mol. Biol. 1994;237:182–192. doi: 10.1006/jmbi.1994.1220. [DOI] [PubMed] [Google Scholar]
  20. Rogers A.R., Harpending H. Population growth makes waves in the distribution of pairwise genetic differences. Mol. Biol. Evol. 1992;9:552–569. doi: 10.1093/oxfordjournals.molbev.a040727. [DOI] [PubMed] [Google Scholar]
  21. Rozas J., Sanchez-DelBarrio J.C., Messequer X., Rozas R. DnaSP, DNA polymorphism analysis by the coalescent and other methods. Bioinformatics. 2003;19:2496–2497. doi: 10.1093/bioinformatics/btg359. [DOI] [PubMed] [Google Scholar]
  22. Sokal R.R., Rohlf F.J. Biometry. Freeman; San Fransisco, CA: 1981. [Google Scholar]
  23. Tajima F. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics. 1989;123:585–595. doi: 10.1093/genetics/123.3.585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Wise C.A., Sraml M., Easteal S. Departure from neutrality at the mitochondrial NADH dehydrogenase subunit 2 gene in humans, but not in chimpanzees. Genetics. 1998;148:409–421. doi: 10.1093/genetics/148.1.409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Zink R.M. The role of subspecies in obscuring avian biological diversity and obscuring conservation policy. Proc. R. Soc. Lond. B. 2004;271:561–564. doi: 10.1098/rspb.2003.2617. doi:10.1098/rspb.2003.2617 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Zink R.M., Drovetski S.V., Questiau S., Fadeev I.V., Nesterov E.V., Rohwer S. Recent evolutionary history of the bluethroat across Eurasia. Mol. Ecol. 2003;12:3069–3075. doi: 10.1046/j.1365-294x.2003.01981.x. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society

RESOURCES