GEOGRAPHIC VARIATION IN THE MITOCHONDRIAL CONTROL REGION OF BLACK-THROATED BLUE WARBLERS (DENDROICA CAERULESCENS) WENDY E. GRUS,1,3 GARY R. GRAVES,2,4 AND TRAVIS C. GLENN1,5 1Savannah River Ecology Laboratory, University of Georgia, Drawer E, Aiken, South Carolina 29802, USA; and 2Department of Vertebrate Zoology, MRC-116, National Museum of Natural History, Smithsonian Institution, P.O. Box 37012, Washington, D.C. 20013, USA Abstract.?We investigated the genetic population structure of the Black-throated Blue Warbler (Dendroica caerulescens), a Nearctic?Neotropic migrant passerine that breeds in cool mixed deciduous?coniferous forests in eastern North America. A cline in plumage color in breeding populations in the central Appalachian Mountains suggests either a contact zone between two formerly allopatric populations or the presence of a strong contemporary selection gradient. Analysis of ??? base pairs of the mitochondrial control region from ??? individuals sampled from ?? populations revealed relatively high haplotype diversity, low nucleotide diversity, and limited but signi?cant phylogeographic structure across the breeding range (analysis of molecular variance [AMOVA], variation among populations  ?.?%; P  ?.??) and between northern and southern population groups (AMOVA, variation among groups  ?.?%; P  ?.??). Genetic di?erentiation among populations did not conform to an isolation-by-distance model. Nucleotide diversity was generally highest in the central Appalachians and lower in geographically peripheral populations. Populations from the northwestern periphery of the breeding range in Michigan had the lowest haplotype diversity and were genetically distinct from populations in the southern Appalachians. The star-shaped haplotype network, extensive sharing of common haplotypes among populations, and the haphazard distribution of rare haplotypes are most likely attributable to the combined e?ects of postglacial expansion from a single refugium (??,??????,??? years ago) and long-distance dispersal events. The existence of a cline in plumage color, in the face of inferred recent gene ?ow, suggests that a strong selection gradient is operating, perhaps related to the migratory divide postulated from stable- isotope data. Received ?? September ????, accepted ? October ????. Key words: Appalachian Mountains, Black-throated Blue Warbler, Dendroica caerulescens, genetic structure, glaciation, mitochondrial DNA, plumage color, stable isotopes. Variaci?n Geogr??ca en la Regi?n Control Mitocondrial de Dendroica caerulescens Resumen.?Investigamos la estructura gen?tica poblacional de Dendroica caerulescens, un paseriforme que migra entre el Ne?rtico y el Neotr?pico y que se reproduce en ?reas frescas de bosques mixtos de con?feras en el este de Norte Am?rica. Una clina en la coloraci?n del plumaje en las poblaciones reproductivas de las monta?as Apalaches sugiere que existe una zona de contacto entre dos poblaciones que fueron alop?tricas, o la presencia de un fuerte gradiente de selecci?n contempor?nea. El an?lisis de ??? pares de bases de la regi?n control mitocondrial de ??? individuos muestreados en ?? poblaciones, revel? una diversidad de haplotipos relativamente alta, una baja diversidad de nucle?tidos y una estructura ?logeogr??ca escasa pero signi?cativa en el ?rea de distribuci?n reproductiva (an?lisis de varianza molecular [AMOVA], variaci?n entre poblaciones  ?.?%; P  ?.??) y entre los grupos formados por las poblaciones del norte y del sur (AMOVA, variaci?n entre grupos  ?.?%; P  ?.??). La diferenciaci?n gen?tica entre poblaciones no se ajust? al modelo de aislamiento por distancia. La diversidad de nucle?tidos fue generalmente m?s alta en el centro de las Apalaches y menor en las poblaciones perif?ricas. Las poblaciones de la periferia en el noroeste del ?rea de distribuci?n reproductiva en Michigan tuvieron la menor diversidad de haplotipos y fueron gen?ticamente diferentes a las de las monta?as Apalaches del sur. La red de haplotipos con forma de estrella, la gran cantidad de haplotipos comunes compartidos entre poblaciones y la distribuci?n al azar de haplotipos raros, son atribuibles probablemente al efecto combinado de la expansi?n postglacial a partir de un ?nico refugio (??,??????,??? a?os atr?s) y eventos de dispersi?n de gran distancia. La existencia de una clina en la coloraci?n del plumaje a pesar del ?ujo gen?tico reciente inferido, sugiere que est? operando un fuerte gradiente de selecci?n, quiz?s relacionado a la divisi?n migratoria postulada con datos de is?topos estables. ? 198 ? The Auk 126(1):198?210, 2009 ? The American Ornithologists? Union, 2009. Printed in USA. 3Present address: Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N, C3-168, P.O. Box 19024, Seattle Washington 98109, USA. 4Address correspondence to this author. E-mail: gravesg@si.edu. Grus and Graves contributed equally. 5Present address: Department of Environmental Health Science, University of Georgia, Athens, Georgia 30602, USA. The Auk, Vol. ???, Number ?, pages ???????. ISSN ????-????, electronic ISSN ????-????. ? ???? by The American Ornithologists? Union. All rights reserved. Please direct all requests for permission to photocopy or reproduce article content through the University of California Press?s Rights and Permissions website, http://www.ucpressjournals. com/reprintInfo.asp. DOI: ??.????/auk.????.????? JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 199 Pleistocene glacial cycles were important drivers of phyletic diversi?cation, genetic reorganization, and lineage extinction in plants and animals. Contemporary populations of Holarctic ver- tebrates often exhibit low haplotype diversity in formerly glaciated regions (Sage and Wol? ????) or clines of decreasing haplotype diversity along transects from unglaciated to recently glaciated landscapes (Hayes and Harrison ????, Hewitt ????, Meril? et al. ????, Mil? et al. ????). Such spatial patterns can be caused by the rapid expansion of bottlenecked populations from glacial refugia (Rogers and Harpending ????, Rogers ????) or by range-wide se- lective sweeps in which one or more favored haplotype spreads across a species? geographic range (Maruyama and Birky ????). Weak phylogeographic structure observed in formerly glaciated regions may result from the consequences of postglacial range expansion or from ongoing demographic processes. In birds, for example, migratory behavior and high natal dispersal may act to spread mitochondrial haplotypes among distant populations (Zink ????). Discriminating between the ancient e?ects of post- glacial population expansion and current demographic processes in mobile organisms is, therefore, a major challenge for phylogeo- graphic hypothesis-testing. Here, we examine variation in the mitochondrial DNA (mtDNA) control region of Black-throated Blue Warbler (Dendro- ica caerulescens; hereafter ?warbler?), a Nearctic?Neotropic mi- gratory songbird whose breeding range is centered in cool mixed deciduous?coniferous forests of eastern North America (Holmes ????). The highest breeding densities occur in mesic forests (???? ?,??? m above sea level) in the Appalachian Mountains (Wilcove ????, Graves ????b, Haney et al. ????). More than two dozen avian species (representing ?? families) share a similar breeding distribution in the Appalachians, occurring southward at higher altitudes to at least the Great Smoky Mountains (~??nN latitude). Taxonomists have applied trinomials to many Appalachian popu- lations (American Ornithologists? Union ????), but morphologi- cal di?erentiation is subtly clinal in most taxa (Zink and Remsen ????). The warbler is unique among Appalachian species in ex- hibiting geographic variation in plumage color that is distinctive enough to be quanti?ed in the ?eld (Graves ????a). Breeding pop- ulations in the southern Appalachians (Dendroica caerulescens cairnsi) have signi?cantly darker plumage than those breeding on glaciated landscapes from Pennsylvania northward (D. c. cae- rulescens). Rudiments of this geographic pattern were recognized more than a century ago (Coues ????, Ridgway ????). Recent sys- tematic collections (G. R. Graves unpubl. data) have revealed the existence of a cline in plumage color between southern Virginia (??nN) and the Susquehanna River (??nN), which lies near the southernmost limit of the Wisconsinan glaciation (??,??? years ago) in the Appalachians (Williams et al. ????). This phenotypic pattern could represent a contact zone between two formerly allopatric populations or signal the presence of a strong selection gradient in the Appalachians (Mayr ????). Stable-isotope analysis of feather keratins provides inferen- tial support for a north?south subdivision of warbler populations (Chamberlain et al. ????, Rubenstein et al. ????). Hydrogen isotope signatures suggest that ?blue-backed? populations from the glaci- ated portion of the species? breeding range winter in the western Caribbean, whereas ?black-backed? populations breeding in the central and southern Appalachians appear to winter in the eastern Caribbean (Rubenstein et al. ????). Such migratory divides in pas- serine birds are often associated with partial barriers to gene ?ow and population di?erentiation (Bensch et al. ????, Chamberlain et al. ????, Ruegg and Smith ????, Bearhop et al. ????). Other lines of evidence, however, predict that gene ?ow among breeding populations of warblers may be substantial. The settlement patterns of yearlings and older males at ?ne spatial scales appear to be driven by heterogeneity in habitat quality and by the despotism of older philopatric males (Holmes ????, Hol- mes et al. ????). A survey of ?? breeding populations distributed throughout the core range revealed that the proportion of year- ling males in local populations increased as total relative abun- dance decreased northward and westward from the Appalachian Mountains (Graves ????b). This implies that yearlings from high- density source areas in the Appalachians may disperse to low-quality habitats that support lower population densities near the northern and western margins of the breeding range (Graves ????b). These factors may result in the spread of haplotypes and weak or imper- ceptible phylogeographic structure among contemporary breeding populations. The hypothesis that male natal dispersal is prevalent at local and perhaps regional spatial scales is supported by carbon isotope data from an altitudinal gradient in the Appalachians that indicated that yearling males rarely returned to breed in the altitu- dinal zone where they were hatched (Graves et al. ????). The phylogeography of malarial parasites provides further indirect evidence of gene ?ow among breeding populations of warblers (Fallon et al. ????). Analyses of malarial infections in ?? populations of warblers sampled across the species? breeding range showed that lineages of Plasmodium and Haemoproteus are geographically widespread and did not provide site-speci?c infor- mation. The wide distribution of malarial parasites is thought to re?ect natal dispersal of their hosts as well as mixing of breeding populations on the wintering grounds. Finally, Davis et al. (????) investigated geographic variation of the mtDNA control region and microsatellite markers in war- blers from four breeding populations, three located north and one south of the Last Glacial Maximum. These analyses revealed signi?cant genetic variation within populations but weak phylo- geographic structure overall, and little di?erentiation between southern and northern populations. The strength of Davis et al.?s (????) conclusion about di?erentiation between northern and southern populations, however, is limited by the number of popu- lations sampled, particularly in the southern Appalachians, where phylogeographic structure may be more pronounced. We sought to build on the work of Davis et al. (????) with a more extensive analysis of geographic variation in the mtDNA control region of the warbler based on ??? specimens from ?? breeding localities. We addressed two fundamental questions: (?) Do haplotypes exhibit geographic variation among breeding pop- ulations? And (?) did glaciation history a?ect phylogeography? METHODS Sample collection.?Because nucleotide and haplotype diversity may exhibit clinal variation from unglaciated to glaciated land- scapes, we sampled breeding populations at strategic geographic intervals both north and south of the Last Glacial Maximum (Fig. ?): Cooper Creek, Georgia (??.??nN, ??.??nW; coordinates converted 200 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 to decimals; n  ?? individuals); Santeetlah Creek, North Carolina (??.??nN, ??.??nW; n ??); Unaka Mountains, Tennessee (??.??nN, ??.??nW; n  ??); Blackwater Falls, West Virginia (??.??nN, ??.??nW; n  ?); Ricketts Glen, Pennsylvania (??.??nN, ??.??nW; n  ??); Catskill Mountains, New York (??.??nN, ??.??nW; n  ??); Adirondack Mountains, New York (??.??nN, ??.??nW; n  ??); Round Lake, Ontario (??.??nN, ??.??nW; n  ??); Saint John River, New Brunswick (??.??nN, ??.??nW; n  ??); and Paint Lake, Mich- igan (??.??nN, ??.??nW; n  ??) (Graves ????b, Fallon et al. ????). Population samples included both yearling (?rst breeding season) and older males (second or later breeding season), which can be distinguished by several plumage characters (Graves ????a). To ensure that neither migrating nor wandering postbreeding birds were collected, males were collected during the peak of the breed- ing season (????? June, ?????????; Graves ????). Specimens were packaged and frozen whole in liquid nitrogen immediately after collecting. Voucher specimens were deposited in the research col- lections of the National Museum of Natural History, Smithsonian Institution, Washington, D.C. We combined our data with an overlapping section of mtDNA control-region sequence previously published for breed- ing populations of the warbler (Davis et al. ????): Coweta, North Carolina (??.??nN, ??.??nW; ?? males, ? females); Hubbard Brook, New Hampshire (??.??nN, ??.??nW; ?? males, ?? females); Sara- nac Lake, New York (??.??nN, ??.??nW; ? males, ? female); and Hi- awatha National Forest, Michigan (??.??nN, ??.??nW; ?? males, ?? females). Davis et al. (????) did not collect voucher specimens or specify collection dates for blood sampled from the North Carolina, Michigan, and New York populations. DNA sequencing.?Detailed laboratory protocols are avail- able online (see Acknowledgments). Brie?y, genomic DNA was isolated from striated muscle samples by digesting ?.? g of muscle in ??? ?L of standard proteinase K digestion bu?er (Sambrook et al. ????). For the populations from the Adirondack and Catskill mountains, Ricketts Glen, and Santeetlah Creek, digestion was followed by a guanidine thiocyanate with diatomaceous-earth extraction protocol (Carter and Milton ????). For the other six populations, the digestion was followed by a phenol chloroform ex- traction (Sambrook et al. ????). The genomic DNA was quanti?ed on ?% agarose gels and diluted with TLE (??mM Tris pH ?, ?.? mM EDTA). A ???-base pair (bp) section of domain ? of the mtDNA control region was ampli?ed with polymerase chain reaction (PCR) primers L? (?`TTCTTGCTTTAAGGGTATGT) and H? (?`TCAATAGATAACCATGTCCT) (Milot et al. ????) for ??? in- dividuals. Ampli?cation was done in ??-?L volumes with ?nal re- action concentrations of ?.? mM MgC??, ??? ?M of each dNTP, ?.?? ?M of each primer, ? unit of Taq polymerase, and ?? ng of DNA. The PCR amplicons were generated in Eppendorf gradient Mastercyclers using a touchdown protocol (Don et al. ????) with an initial annealing temperature of ??nC and an ending anneal- ing temperature of ??nC. Ampli?cations were con?rmed on ?.?% agarose gels containing ethidium bromide. Sequences from both DNA strands were determined directly from PCR products using Big-Dye terminator chemistry and an ABI ???-?? automated sequencer (Applied Biosystems, Foster City, California). All cycle-sequencing reactions were carried out in ??- ?L volumes with ?.?? ?M primer, ?? ng PCR product (?.???.? ?L of unpuri?ed PCR product), and Big-Dye Terminators using ABI speci?cations except that the terminator mix was diluted ?:? with ?r automated sequencing dilution bu?er (?? mM MgC??, ??? mM Tris-HCl pH ?.?). Control-region sequences have been deposited in GenBank (accession numbers EU???????EU??????). Data analysis.?Sequences were edited using SEQUENCHER, version ?.?.? or ?.?.? (Gene Codes, Ann Arbor, Michigan). To con- ?rm the mitochondrial origin of PCR products, we compared Black- throated Blue Warbler sequence with those from Yellow Warbler (D. petechia; Milot et al. ????) in GenBank and computed transi- tion:transversion (Ti:Tv) ratios to compare with those of other bird species. The Ti:Tv ratio for Black-throated Blue Warbler (?.? for our data alone, ?.? for the combined data) is similar to the average ra- tio for neutral mtDNA sites in birds (Belle et al. ????). We also note the absence of heterozygotes, which would be expected in nuclear copies of mitochondrial genes. We examined the likely relationships between haplotypes with an unrooted minimum spanning network constructed in TCS?.?? (Clement et al. ????). To investigate possible phylogeographic structure among the populations, we used ARLE- QUIN, version ?.?? (see Acknowledgments), to calculate an analysis of molecular variance (AMOVA; Exco?er et al. ????), which deter- mines the amount of haplotype variation within populations, among populations in a group, and between groups of populations. We con- ducted AMOVA on our data and those of Davis et al. (????) to en- sure that the two data sets did not di?er systematically. We pooled the data in subsequent analyses because they appeared to be comple- mentary. Di?erent groupings of populations were then investigated to determine whether genetic structure existed to match phenotypic and migratory di?erences between the populations. We categorized populations, a priori, into three geographic pools: northern (south to the Adirondack Mountains), central (Catskill Mountains to Ricketts Glen), and southern (Blackwater Falls southward). Four di?erent FIG. 1. Distribution of sampling localities within the core breeding range (shaded gray) of Black-throated Blue Warbler in eastern North America (Fallon et al. 2006). The thick black line indicates the glacial front at the Wisconsinan maximum, ~18,000 years ago. T7 Z5 Saint John RiveF "^ RoundLake SaranLI \t * \_t Lak=VHubbard( Adirondack Mtns *^-Catskill Mtns Ricketts Glen Cooper Creek JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 201 AMOVA tests were performed: (?) no geographic pools; (?) three geographic pools?northern, central, and southern; (?) two geo- graphic pools?central combined with northern; and (?) two geo- graphic pools?central combined with southern. We conducted additional spatial analyses of molecular vari- ance (SAMOVA), using both geographic and genetic distances among populations to maximally di?erentiate groups of breeding populations (Dupanloup et al. ????). Spatial analyses of molecular variance require the number of groups to be prede?ned, but group membership is determined through the analysis. We ran separate SAMOVAs for two and three groups. The relationship between ge- netic and geographic distances among populations was assessed with Mantel tests (Bohonak ????). All P values are two-tailed. To investigate Holocene demographic history, we used ARLEQUIN to calculate haplotype (h) and nucleotide diversity (?) for each population, to perform Tajima?s test (Tajima ????) and Fu and Li?s test (Fu and Li ????) for neutrality, and to calculate a mis- match distribution of pairwise di?erences between haplotypes. We also calculated FS and R? (Ramos-Onsins and Rozas ????) in DnaSP (Rozas et al. ????) to evaluate recent demographic processes. We estimated the time since the most recent common an- cestor (Tmrca) with two di?erent methods. First, we estimated the number of generations since Tmrca using the quick calculation of ?  ? ?t, where t is the number of years since coalescence (gen- eration time is one year) and ? is the mutation rate (Rogers and Harpending ????, Schneider and Exco?er ????), which we mod- eled at ?.???.? substitutions site?? MA?? (Avise and Walker ????). We then used ARLEQUIN to estimate ? and ??% con?dence inter- vals (CI) of ? under a model of pure demographic expansion. We also used BEAST, version ?.?.? (Drummond and Rambaut ????), to estimate Tmrca using parameters similar to those used by Davis et al. (????), except that we used a general time reversible (GTR) model with gamma rate distribution and invariable sites, a total chain length of ?? million sampling every ??,??? to achieve an ef- fective sample size of ??? for Tmrca, and only one substitution rate in the simulations (?.? substitutions site?? MA??; complete param- eters can be obtained from T. C. Glenn). We used Chao?s (????) equation to obtain a nonparametric estimate of the true number of haplotypes (H`) present in con- temporary populations of the warbler based on the number of rare haplotypes present in the breeding sample, `  ? ?? ? ?? H H a bobserved 2 2 where Hobserved represents the observed number of haplotypes in a sample, a is the number of observed haplotypes that are repre- sented by only a single individual in the sample (i.e., singletons), and b is the number of observed haplotypes represented by exactly two individuals in that sample (i.e., doubletons). A ??% CI for the true number of haplotypes in a sample was calculated with Chao?s (????) variance equation: var / / / `  ? ?? ? ??  ? ?? ? ?? ? ? ? ? ? ? H b a b a b a b 4 4 4 3 2 ? ? RESULTS Haplotype diversity and distribution.?For the combined data set, the ???-bp section of the mtDNA control region exhibited ?? variable sites (??%) and ?? unique haplotypes among the ??? typed individuals (Fig. ?). Most haplotypes di?ered by only a few base-pair changes (Fig. ?). The K??uf + I + G model (Hasegawa et al. ????, Yang ????, Gu et al. ????) was selected by the hierarchi- cal likelihood ratio test of MODELTEST, version ?.?? (Posada and Crandall ????), as the best ?t for the control-region data. Param- eters estimated for this model were as follows: transition rate  ??.?; transversion rate ? (AlT, ClG)  ?.?; transversion rate ? (AlC, GlT)  ?.? (for an average transition?transversion [Ti:Tv] ratio  ?.?); gamma shape parameter  ?.??; base frequencies A  ?.??, C  ?.??, G  ?.??, and T  ?.??; and proportion of invariable sites  ?.??. The minimum spanning network of haplotypes was highly re- ticulated and organized around two principal and two subsidiary foci (Fig. ?). Most haplotypes were represented by singletons (?? of ??), ?? haplotypes were shared by ??? individuals, ? haplotypes were shared by ? individuals, and ? common haplotypes were shared by ????? individuals (Fig. ?). The four most common hap- lotypes, which accounted for ??% of all individuals, were widely distributed geographically (Fig. ?). Most of the rarer haplotypes di?ered from common haplotypes by one or two nucleotides; re- lated haplotypes often occurred in widely separated populations (Fig. ?). Pairwise nucleotide diversity (?  ?.???????.?????) was highest in Pennsylvania near the center of the breeding range and lowest in northern populations on glaciated landscapes and at the southern extreme of the breeding range in Georgia (Table ? and Figs. ? and ?A). The number of haplotypes detected per popula- tion was signi?cantly correlated with sample size (r?  ?.??; P  ?.????). Haplotype diversity (overall h  ?.??) varied from ?.?? (Paint Lake) to ?.?? (Blackwater Falls and Catskill Mountains) (Table ? and Fig. ?B). Nucleotide and haplotype diversity were uncorrelated (P  ?.??), and neither diversity index was correlated with latitude (P  ?.??). Population history.?Warblers exhibited signi?cant among- population variation (AMOVA, ?.??%, P  ?.???) that was incon- sistent with the predictions of an isolation-by-distance model (Mantel test, r  ?.??, P  ?.??; Fig ?C). Among-population vari- ation was also signi?cant when populations were pooled into two regional groups (highest support for the central and south- ern populations lumped; AMOVA, ?.??%, P  ?.??) or three re- gional groups (northern, central, and southern; AMOVA, ?.??%, P  ?.??). The two-group SAMOVA clustered the central and southern populations, except that Georgia was grouped with the northern populations (?.??% variation among groups, P  ?.???; ?.??% variation among populations within groups; ??.??% varia- tion within populations). When individual populations were compared in a pairwise fashion, ?? combinations exhibited signi?cant Fst values (P  ?.??). After P values were adjusted for the number of simultaneous obser- vations (?.??/??  ?.?????), six pairwise combinations of popula- tions showed signi?cant Fst values. The Hiawatha population was signi?cantly di?erent from the Catskill Mountains population and three others from the southern Appalachians?Santeetlah Creek, Coweta, and the Unaka Mountains. The Paint River population 202 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 FI G . 2 . V ar ia bl e ba se p os iti on s in th e m tD N A c on tr ol re gi on a nd fr eq ue nc y of h ap lo ty pe s in B la ck -t hr oa te d B lu e W ar bl er (s ee te xt fo r d et ai ls o f b as e po si tio ns a nd p op ul at io n la be ls ). Fi gu re 2 is c on tin ue d on th e ne xt p ag e. oocoooooooooooo oooooooo o o o o o o ^^?^?^?^?^?^?^?^?^?^?IJ JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 203 FI G . 2 . C on tin ue d. V ar ia bl e ba se p os iti on s in th e m tD N A c on tr ol re gi on a nd fr eq ue nc y of h ap lo ty pe s in B la ck -t hr oa te d B lu e W ar bl er (s ee te xt fo r d et ai ls o f b as e po si tio ns a nd po pu la tio n la be ls ). (D (D (D (D (D (D (DO ? ? ? O ? O OOOO (JO ? O O OOOOO ? ? O O ? \- ? ? O ? ? OO ? ? ? O O 204 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 from Michigan also di?ered from populations in the Catskill and Unaka mountains. Tests for evidence of population expansion gave con?icting results. The subtly bimodal distribution of pairwise nu- cleotide di?erences (Fig. ?) suggests that warbler populations are in equilibrium and that a sudden-population-expansion model can be rejected (raggedness index  ?.??, P ?.??; sudden expansion sum of squared deviation  ?.???, P  ?.???; Rogers and Harpending ????). By contrast, Rozas et al.?s (????) statistics reject the constant-popu- lation-size model (FS  ???.??, P  ? r ????; R?  ?.???, P  ?.??). The rejection of neutrality hypotheses by Fu and Li?s (????) D* statistic (D*  ??.??, P  ?.??) and Tajima?s (????) D (D  ??.??, P  ?.??) provides additional evidence for population expansion. Although the strong rejection of neutrality could have other explanations be- sides population expansion, those explanations, such as selective sweeps (Maruyama and Birky ????, Moyer et al. ????) due to the linked coding regions of mtDNA, are rarely observed in mtDNA of animal populations (Gerber et al. ????). The star-shaped topol- ogy of the haplotype tree is consistent with expectations of a recent range expansion (Avise and Walker ????). Under the model of pure demographic expansion, we esti- mated a mean ?  ?.?? (??% CI: ?.????.??), yielding the number of years since Tmrca (?  ? ?t) of ??,??? (??% CI: ?,??????,??? years ago), TABLE 1. Genetic variation in breeding populations of Black-throated Blue Warblers. Population Sample size Number of haplotypes Nucleotide diversity (P) Haplotype diversity (h) Cooper Creek, Georgia 12 7 5.081 0.8788 Coweta, North Carolina 21 12 6.359 0.9095 Santeetlah Creek, North Carolina 29 15 6.846 0.9236 Unaka Mountains, Tennessee 16 11 7.427 0.9083 Blackwater Falls, West Virginia 8 7 5.611 0.9643 Ricketts Glen, Pennsylvania 10 7 8.124 0.8667 Catskill Mountains, New York 21 15 7.884 0.9619 Adirondack Mountains, New York 19 9 5.230 0.8830 Hubbard Brook, New Hampshire 57 29 7.153 0.8929 Saranac Lake, New York 10 6 4.968 0.8889 Round Lake, Ontario 15 10 5.710 0.9451 Saint John River, New Brunswick 14 11 6.272 0.9429 Hiawatha NF, Michigan 37 12 4.237 0.8093 Paint Lake, Michigan 18 8 5.416 0.7908 FIG. 3. Minimum spanning network for haplotypes of Black-throated Blue Warbler. Each oval represents a haplotype listed in Figure 2. Oval size is proportional to sample size. Missing haplotypes are indicated by small circles, and each branch indicates one nucleotide difference. D38 D42 h36 D41 D20 h17 JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 205 FI G . 4 . M in im um s pa nn in g ne tw or k fo r ea ch p op ul at io n of B la ck -t hr oa te d B lu e W ar bl er . T he n et w or k fr om F ig ur e 3 is s ho w n fo r ea ch p op ul at io n, h ig hl ig ht in g th e ha pl ot yp es fo un d in e ac h po pu la tio n. F ig ur e 4 is c on tin ue d on th e ne xt p ag e. > d O 206 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 FI G . 4 . C on tin ue d. M in im um s pa nn in g ne tw or k fo r ea ch p op ul at io n of B la ck -t hr oa te d B lu e W ar bl er . T he n et w or k fr om F ig ur e 3 is s ho w n fo r ea ch p op ul at io n, h ig hl ig ht in g th e ha pl ot yp es fo un d in e ac h po pu la tio n. u B I u -*?> O u JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 207 which suggests that the warbler was restricted to a single refu- gium near the end of the Pleistocene. The more complex coales- cence simulations using BEAST yielded a median Tmrca of ??,??? years ago (??% CI: ??,??????,??? years ago). If a slower mutation rate of ?.??? mutations per site per million years is assumed (cf. Davis et al. ????), the Tmrca is estimated to be about four times older. DISCUSSION Breeding populations of Black-throated Blue Warblers exhibit rel- atively high haplotype diversity, low nucleotide diversity (cf. Mil? et al. ????, Milot et al. ????), and subtle phylogeographic struc- ture. We observed ?? haplotypes (?? singletons and ?? double- tons) in ??? individuals (?.?? haplotypes individual??) sampled from ?? populations. A nonparametric estimate (Chao ????, ????) of the lower bound of the true number of haplotypes for the pooled sample was estimated as ??? (??% CI: ??????? haplotypes). The lower bound of the true number of haplotypes estimated from the four populations sampled by Davis et al. (????) was ??? (??% CI: ?????? haplotypes). The higher estimate of haplotype diversity from the combined data set indicated that analyses based on fewer geographic localities substantially underestimated the true num- ber of haplotypes present in contemporary populations. The com- bined power obtained from sampling both more individuals and more localities was also important in revealing variation among population groups. It is clear that most variation (??%) in the mtDNA control region in warblers occurs within breeding populations. Limited but signi?cant di?erentiation was observed between southern ?black-backed? populations (D. caerulescens cairnsi) and the northern ?blue-backed? populations (D. c. caerulescens). The two westernmost populations in Michigan were genetically distinct from populations in the Catskill Mountains and in the south- ern Appalachians of Tennessee and North Carolina, which indi- cates that populations from the southern and western edges of the breeding range may have diverged signi?cantly from one another. On the other hand, the haphazard distribution of rare haplotypes (e.g., doubletons) among widely separated locations (Fig. ?) and the distribution of malarial parasite lineages among warbler pop- ulations (Fallon et al. ????) suggests that long-distance dispersal events occur with some regularity. Because ??% of the sampled individuals were male, and because natal dispersal in songbirds is invariably female-biased (Greenwood ????, Clarke et al. ????), our analyses more likely represent an optimistic than a conservative estimate of genetic di?erentiation among warbler populations. Nucleotide diversity was highest in populations near the cen- ter of the breeding range and tended to be lower in northern pop- ulations on glaciated landscapes and in Georgia at the southern terminus of the species? breeding range. Haplotype diversity was lowest in Michigan near the western periphery of the breeding range. Collectively, these data suggest that breeding populations from the central part of the breeding range, from West Virginia to New York, are more genetically diverse. Lower genetic diversity in peripheral populations is, presumably, a consequence of geographic isolation and smaller e?ective population sizes. Alternatively, FIG. 5. Relationship of (A) nucleotide diversity and (B) haplotype diver- sity with latitude and (C) pairwise Fst values with geographic distance. FIG. 6. Mismatch distribution for pairwise differences in Black-throated Blue Warbler haplotypes. The broken line represents the pairwise mis- match distribution. The expected distribution under the sudden-expansion model (Ti:Tv  6.4) is represented by the solid line. 36 40 44 Latitude (N) 36 40 44 Latitude (N) 500 1000 1500 Distance (km) 48 48 2000 12000 10000- %* 8000 6000 4000- 2000 15 20 25 30 Number of pairwise differences 208 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 higher genetic diversity in the central portion of the breeding range may represent a contact zone between two formerly allo- patric populations that originated in separate Pleistocene re- fugia. However, coalescence analysis suggests that warbler populations were restricted to a single refugium between ??,??? and ??,??? years ago (?  ?.?), an interpretation consistent with the star-shaped topology of the haplotype tree. This implies that geographic variation in plumage color and migratory behavior ob- served in contemporary breeding populations originated no ear- lier than the last glacial period. Davis et al. (????) drew a similar conclusion but estimated that Tmrca ranged from ??,??????,??? years ago (?  ?.?) to ??,???????,??? years ago (?  ?.???). ?Blue-backed? populations, which occupy the whole of the glaciated breeding range, are separated from the ?black-backed? populations of the unglaciated southern Appalachians by a cline in plumage melanism extending from southern Virginia (??nN) to the Susquehanna River in Pennsylvania (??nN) (G. R. Graves unpubl. data). This phenotypic variation, which is believed to be genetically controlled, is exhibited in yearling and older males. Although it is possible that geographic variation in plumage color existed in refugial populations, we suspect that the ob- served pattern is a relatively recent phenomenon associated with sexual selection and a strong selective factor, perhaps related to the migratory divide postulated from stable-isotope data (Ru- benstein et al. ????). Although the genetic di?erences are small and relatively recent in origin, our data suggest that warblers are more appropriately managed as two conservation units rather than a single unit. Although there is no direct evidence that breeding popula- tions segregate on the wintering grounds, stable-isotope studies suggest that populations breeding on glaciated landscapes win- ter primarily in Cuba and Jamaica, whereas populations from the southern Appalachians winter largely in Hispaniola and Puerto Rico (Rubenstein et al. ????). Migratory behaviors of songbirds can have a strong genetic component that may evolve rapidly (Berthold and Querner ????; Helbig ????, ????; Berthold ????; Bearhop et al. ????), and genetic di?erentiation is often associated with migratory divides in the breeding range (Bear- hop et al. ????). The extent to which the hypothesized migra- tory divide in warbler breeding populations coincides with the position of the cline in plumage color is unknown. However, se- lection for wintering destination is a plausible mechanism for the maintenance of geographic variation in plumage color in the presence of recent (postglacial) or, possibly, ongoing gene ?ow. Under such selection, o?spring from hybridization between individuals that winter on di?erent islands in the Greater An- tilles may be genetically programmed to end their migrations over open water in the Caribbean. In any event, genetic factors that encode plumage color and migratory behavior in warblers are not likely to be re?ected in the selectively neutral mtDNA control region or the coding regions to which it is linked. New, massively parallel DNA sequencing technologies will facilitate powerful investigations of the genetic basis of phenotypic varia- tion in birds. Future research into the mechanisms that gen- erated and maintain the phenotypic variation among warbler populations will make use of the large number of individuals and populations that have already been sampled (Graves ????, Fallon et al. ????). ACKNOWLEDGMENTS We thank P. Angle, J. Dean, C. Dove, C. Gebhard, C. Milensky, J. Ososky, A. Ross, and B. Schmidt for preparing specimens and several anonymous reviewers for comments on the manuscript. Scienti?c licenses to sample populations were issued by the Cana- dian Wildlife Service (Atlantic and Ontario Region), U.S. Fish and Wildlife Service, U.S. Department of Agriculture Forest Service , Georgia Department of Natural Resources, Michigan Department of Natural Resources, New York Department of Environmental Conservation, North Carolina Wildlife Resources Commission, Pennsylvania Game Commission, Tennessee Wildlife Resources Agency, West Virginia Department of Natural Resources, and Wisconsin Department of Natural Resources. M. Schable and C. Outz helped in laboratory analyses. Funding was provided by the Alexander Wetmore Fund, the Research Opportunities Fund, the Biodiversity Surveys and Inventory Program of the National Museum of Natural History, the Scholarly Studies Program, the Smithsonian Migratory Bird Center (all Smithsonian Institution), National Science Foundation Award ??????? to the Savannah River Ecology Laboratory Research Experience for Undergradu- ates program, and U.S. Department of Energy (contract numbers DE-FC-??-??SR????? and DE-FC??-??SR?????) to the Univer- sity of Georgia?s Savannah River Ecology Laboratory. For detailed laboratory protocols, see http://www.uga.edu/srel/DNA_Lab/ protocols.htm. ARLEQUIN is available at lgb.unige.ch/arlequin/. LITERATURE CITED American Ornithologists? Union. ????. Check-list of North American birds, ?th ed. American Ornithologists? Union, Balti- more, Maryland. Avise, J. C., and D. Walker. ????. Pleistocene phlylogeographic e?ects on avian populations and the speciation process. Proceed- ings of the Royal Society of London, Series B ???:???????. Bearhop, S., W. Fiedler, R. W. Furness, S. C. Votier, S. Wal- dron, J. Newton, G. J. Bowen, P. Berthold, and K. Farn- sworth. ????. Assortative mating as a mechanism for rapid evolution of a migratory divide. Science ???:???????. Belle, E. M. S., G. Piganeau, M. Gardner, and A. Eyre-Walker. ????. An investigation of the variation in the transition bias among various animal mitochondrial DNA. Gene ???:?????. Bensch, S., T. Andersson, and S. ?kesson. ????. Morphologi- cal and molecular variation across a migratory divide in Willow Warblers, Phylloscopus trochilus. Evolution ??:?????????. Berthold, P. ????. Genetic basis and evolutionary aspects of bird migration. Advances in the Study of Behavior ??:???????. Berthold, P., and U. Querner. ????. Genetic basis of migratory behavior in European warblers. Science ???:?????. Bohonak, A. J. ????. IBD (isolation by distance): A program for analyses of isolation by distance. Journal of Heredity ??:???????. Carter, M. J., and I. D. Milton. ????. An inexpensive and simple method for DNA puri?cations on silica particles. Nucleic Acids Research ??:????. Chamberlain, C. P., S. Bensch, X. Feng, S. ?kesson, and T. Andersson. ????. Stable isotopes examined across a migratory divide in Scandinavian willow warblers (Phylloscopus trochilus trochilus and Phylloscopus trochilus acredula) re?ect their African JANUARY 2009 ? MTDNA VARIATION AMONG BLACK-THROATED BLUE WARBLERS ? 209 winter quarters. Proceedings of the Royal Society of London, Series B ???:?????. Chamberlain, C. P., J. D. Blum, R. T. Holmes, X. Feng, T. W. Sherry, and G. R. Graves. ????. The use of isotope tracers for identifying populations of migratory birds. Oecologia ???: ???????. Chao, A. ????. Non-parametric estimation of the number of classes in a population. Scandinavian Journal of Statistics ??:???????. Chao, A. ????. Estimating the population size for capture?recapture data with unequal catchability. Biometrics ??:???????. Clarke, A. L., B.-E. S?ther, and E. R?skaft. ????. Sex biases in avian dispersal: A reappraisal. Oikos ??:???????. Clement, M., D. Posada, and K. A. Crandall. ????. TCS: A computer program to estimate gene genealogies. Molecular Ecol- ogy ?:?????????. Coues, E. ????. Characters of Dendroica c?rulescens cairnsi. Auk ??:?????. Davis, L. A., E. H. Roalson, K. L. Cornell, K. McClanahan, and M. S. Webster. ????. Genetic divergence and migration patterns in a North American passerine bird: Implications for evolution and conservation. Molecular Ecology ??:?????????. Don, R. H., P. T. Cox, B. J. Wainwright, K. Baker, and J. S. Mat- tick. ????. ?Touchdown? PCR to circumvent spurious priming during gene ampli?cation. Nucleic Acids Research ??:????. Drummond, A. J., and A. Rambaut. ????. BEAST: Bayesian evo- lutionary analysis by sampling trees. [Online.] BMC Evolution- ary Biology ?:???, doi:??.????/????-????-?-???. Dupanloup, I., S. Schneider, and L. Excoffier. ????. A simu- lated annealing approach to de?ne the genetic structure of popu- lations. Molecular Ecology ??:?????????. Excoffier, L., P. E. Smouse, and J. M. Quattro. ????. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restric- tion data. Genetics ???:???????. Fallon, S. M., R. C. Fleischer, and G. R. Graves. ????. Malarial parasites as geographic markers in migratory birds? Biology Let- ters ?:???????. Fu, Y.-X., and W.-H. Li. ????. Statistical tests of neutrality of muta- tions. Genetics ???:???????. Gerber, A. S., R. Loggins, S. Kumar, and T. E. Dowling. ????. Does nonneutral evolution shape observed patterns of DNA variation in animal mitochondrial genomes? Annual Review of Genetics ??:???????. Graves, G. R. ????a. Age determination of free-living male Black- throated Blue Warblers during the breeding season. Journal of Field Ornithology ??:???????. Graves, G. R. ????b. Geographic clines of age ratios of Black- throated Blue Warblers (Dendroica caerulescens). Ecology ??: ?????????. Graves, G. R. ????. Testicular volume and asymmetry are age- dependent in Black-throated Blue Warblers (Dendroica caerule- scens). Auk ???:???????. Graves, G. R., C. S. Romanek, and A. Rodriguez Navarro. ????. Stable isotope signature of philopatry and dispersal in a migratory songbird. Proceedings of the National Academy of Sci- ences USA ??:?????????. Greenwood, P. J. ????. Mating systems, philopatry and dispersal in birds and mammals. Animal Behaviour ??:?????????. Gu, X., Y.-X. Fu, and W.-H. Li. ????. Maximum likelihood estima- tion of the heterogeneity of substitution rate among nucleotide sites. Molecular Biology and Evolution ??:???????. Haney, J. C., D. S. Lee, and M. Wilbert. ????. A half-century com- parison of breeding birds in the southern Appalachians. Condor ???:???????. Hasegawa, M., H. Kishino, and T. Yano. ????. Dating of the human?ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution ??:???????. Hayes, J. P., and R. G. Harrison. ????. Variation in mitochondrial DNA and the biogeographic history of woodrats (Neotoma) of the eastern United States. Systematic Biology ??:???????. Helbig, A. J. ????. SE- and SW-migrating Blackcap (Sylvia atrica- pilla) populations in central Europe: Orientation of birds in the contact zone. Journal of Evolutionary Biology ?:???????. Helbig, A. J. ????. Genetic basis, mode of inheritance and evolution- ary changes of migratory directions in Palearctic warblers (Aves: Sylviidae). Journal of Experimental Biology ???:?????. Hewitt, G. M. ????. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society ??:???????. Holmes, R. T. ????. Black-throated Blue Warbler (Dendroica caer- ulescens). In The Birds of North America, no. ?? (A. Poole and F. Gill, Eds.). Academy of Natural Sciences, Philadelphia, and Amer- ican Ornithologists? Union, Washington, D.C. Holmes, R. T., P. P. Marra, and T. W. Sherry. ????. Habitat- speci?c demography of breeding Black-throated Blue Warblers (Dendroica caerulescens): Implications for population dynamics. Journal of Animal Ecology ??:???????. Maruyama, T., and C. W. Birky, Jr. ????. E?ects of periodic selec- tion on gene diversity in organelle genomes and other systems without recombination. Genetics ???:???????. Mayr, E. ????. Animal Species and Evolution. Belknap Press of Har- vard University Press, Cambridge, Massachusetts. Meril?, J., M. Bj?rklund, and A. J. Baker. ????. Historical demography and present day population structure of the green- ?nch, Carduelis chloris?An analysis of mtDNA control-region sequences. Evolution ??:???????. Mil?, B., D. J. Girman, M. Kimura, and T. B. Smith. ????. Genetic evidence for the e?ect of a postglacial population expan- sion on the phylogeography of a North American songbird. Pro- ceedings of the Royal Society of London, Series B ???:?????????. Milot, E., H. L. Gibbs, and K. A. Hobson. ????. Phylogeography and genetic structure of northern populations of the Yellow War- bler (Dendroica petechia). Molecular Ecology ?:???????. Moyer, G. R., K. O. Winemiller, M. V. McPhee, and T. F. Turner. ????. Historical demography, selection, and coales- cence of mitochondrial and nuclear genes in Prochilodus species of northern South America. Evolution ??:???????. Posada, D., and K. A. Crandall. ????. MODELTEST: Testing the model of DNA substitution. Bioinformatics ??:???????. Ramos-Onsins, S. E., and J. Rozas. ????. Statistical properties of new neutrality tests against population growth. Molecular Biol- ogy and Evolution ??:?????????. Ridgway, R. ????. The birds of North and Middle America, part ?. Bulletin of the U.S. National Museum, no. ??. Rogers, A. R. ????. Genetic evidence for a Pleistocene population explosion. Evolution ??:???????. 210 ? GRUS, GRAVES, AND GLENN ? AUK, VOL. 126 Rogers, A. R., and H. Harpending. ????. Population growth makes waves in the distribution of pairwise genetic di?erences. Molecular Biology and Evolution ?:???????. Rozas, J., J. C. S?nchez-DelBarrio, X. Messeguer, and R. Rozas. ????. DnaSP, DNA polymorphism analyses by the coales- cent and other methods. Bioinformatics ??:?????????. Rubenstein, D. R., C. P. Chamberlain, R. T. Holmes, M. P. Ayres, J. R. Waldbauer, G. R. Graves, and N. C. Tuross. ????. Linking breeding and wintering ranges of a migratory songbird using stable isotopes. Science ???:?????????. Ruegg, K. C., and T. B. Smith. ????. Not as the crow ?ies: A his- torical explanation for circuitous migration in Swainson?s Thrush (Catharus ustulatus). Proceedings of the Royal Society of Lon- don, Series B ???:?????????. Sage, R. D., and J. O. Wolff. ????. Pleistocene glaciations, ?uctu- ating ranges, and low genetic variability in a large mammal (Ovis dalli). Evolution ??:?????????. Sambrook, J., E. F. Fritsch, and T. Maniatis. ????. Molecular Cloning: A Laboratory Manual, ?nd ed. Cold Spring Harbor Lab- oratory Press, Cold Spring Harbor, New York. Schneider, S., and L. Excoffier. ????. Estimation of past demo- graphic parameters from the distribution of pairwise di?erences when the mutation rates vary among sites: Application to human mitochondrial DNA. Genetics ???:?????????. Tajima, F. ????. Statistical methods for testing the neutral muta- tion hypothesis by DNA polymorphism. Genetics ???:???? ???. Wilcove, D. S. ????. Changes in the avifauna of the Great Smoky Mountains: ?????????. Wilson Bulletin ???:???????. Williams, J. W., T. Webb III, P. H. Richard, and P. Newby. ????. Late Quaternary biomes of Canada and the eastern United States. Journal of Biogeography ??:???????. Yang, Z. ????. Maximum-likelihood estimation of phylogeny from DNA sequences when substitution rates di?er over sites. Molecu- lar Biology and Evolution ??:?????????. Zink, R. M. ????. Comparative phylogeography in North American birds. Evolution ??:???????. Zink, R. M., and J. V. Remsen, Jr. ????. Evolutionary processes and patterns of geographic variation in birds. Pages ???? in Current Ornithology, vol. ? (R. F. Johnston, Ed.). Plenum Press, New York. Associate Editor: J. Klicka