Trans-Beringia Comparisons of Mitochondrial Dna Differentiation in Birds

The Condor 97:63%649 0 The Cooper Ornithological society 1995

TRANS-BERINGIA COMPARISONS OF MITOCHONDRIAL DNA DIFFERENTIATION IN BIRDS

ROBERT M. ZINK Bell Museum, Universityof Minnesota, St. Paul, MN 55108

SIEVERT ROHWER Burke Museum BD- IO and Department of Zoology, Universityof Washington,Seattle, WA 98195

ALEXANDER v. fiNDREEV Institutefor BioIogicalProblems of the North, Magadan, Russia

DONNA L. D~MANN Museum of Natural Science,Louisiana State University,Baton Rouge,LA 70803

Abstract. We comparedmitochondrial DNA (mtDNA) restrictionfragment profiles from samplesof 13 bird speciesthat occuron both sidesof Beringia.All but two species,Lapland Longspur(Calcarius lapponicus)and Green-wingedTeal (Anas crecca),exhibited evidence of geneticdifferentiation, albeit at varying degrees.Several speciesexhibited mtDNA differentiation consistentwith species-leveldistinctness: Marbled Murrelet (Brachyramphus marmoratus),Three-toed Woodpecker(Picoides tridactylus), Whimbrel (Numeniusphaeo- pus), Mew Gull (tints canus), Black-billed Magpie (Pica pica), American Pipit (Anthus rubescens),and Rosy Finch (Leucostictearctoa). Other speciesexhibited levels of mtDNA differentiationintermediate betweenpopulations and species:Barn Swallow (Hirundo rustica), Common Tern (Sterna hirundo), Common Snipe (Gallinago galhnago), and Pelagic Cormorant (Phalacrocoraxpelagicus).Because of smallsample sizes, we do not recommend formal taxonomicchanges, although our data could be combinedwith otherdata to raise severaltaxa to specieslevel. Our data do not indicatea consistentlevel of mtDNA differ- entiationbetween putatively conspecific populations on differentsides of Beringia,suggesting differenttimes of colonizationor cessationof geneexchange. Most comparisonsof birds within continentsexhibit lessmtDNA differentiationthan our tram-Beringiacomparisons, suggestinglimited geneflow betweencontinents. Key words: Geographicvariation; mitochondrial DNA; geneticdtfherentiation; taxonomy; Siberia; North America; evolution,

INTRODUCTION pede geographicdifferentiation (Edwards 1993). Analyses of mitochondrial DNA (mtDNA) dif- In this study, we compared populations of 13 ferentiation within avian specieshave provided species sampled on opposite sides of Beringia. information on levels of genetic variation, his- Although some speciesmigrate between the con- torical patterns of population fragmentation, hy- tinents, our comparisons are some of the first to bridization, and population structure (Ball and test for inter-continental differences. Our Sibe- Avise 1992; Bermingham et al. 1992; Degnan rian samples came from four general regions in and Moritz 1992; Moore et al. 1991; Zink and northeastern Russia, Cherskiy and Anadyr in the Dittmann 1993a, 1993b; Zink 1994). These arctic, the region around Magadan on and near mtDNA analysescompared populations from the the Okhotsky Sea, and the southern third of the same continent. Although large geographicdis- Kamchatka Peninsula; the North American sam- tances often separatedpopulations compared for ples came from a variety of geographicsites (Ap- mtDNA variation (Ball et al. 1988; Ball and Av- pendix 1). Our goal was not an assessmentof the ise 1992; Zink and Dittmann 1993a, 1993b), the effects of Beringia per se; rather, we estimated potential exists for long distance dispersal to im- the level of mtDNA distinctiveness for 13 species that are distributed acrossBeringia on different continental land masses.Our data thus provide another perspective on the growing body of information on the geographyof mtDNA variation I Received4 August1994. Accepted 9 February1995. within bird species. 640 ROBERT M. ZINK ET AL.

TABLE 1. Composite haplotypesfor comparisonswith Old World and New World specimens.Letters indicate fragment profiles for the following restriction endonucleases:Avu I, Ava II, BumH I, &z~?nI, Bun II, Bgl I, Bg/ II, EcoR I, Hint II, Hind III, Ncl I, and Stu I. A “.” indicates the “A” pattern. RUS = Russia, and USA = United States.

Haplotype Location n Composite pattern Phulucrocoruxpelagicus 0.0016 0.0076 1 RUS 1 AAAAAAAAAAAA 2 RUS 1 ...... B . . . - 3 RUS . . B B.. - 4 USA : . . . i:i:.. . BB Anus creccu 0.023’ NC? 1 RUS 1 AAAAAAAAAAAA 2 RUS, USA 1, 3 BBBB.BB.BBBB 3 RUS 1 C...... BC.. GulIinugogullinugo 0.002 1 0.006 1 RUS 2 AAAAAAAAAAAA 2 RUS 1 ...... 3 USA 1 . . . . B...... : 4 USA 1 . . . . B ...... Numeniusphueopus 0.0037 0.047 1 RUS 2 AAAAAAAAAAAA 2 RUS 1 . B 3 USA 2 .CBBB:B:: :BB 4 USA 1 .CBBB.B...BC Lurus cunus 0 0.020 1 RUS 3 AAAAAAAAAAAA 2 USA 1 BBBBBBB...B. Sterna hirundo 0.0055 0.0075 1 RUS 2 AAAAAAAAAAA- RUS 1 . B , . B B. .B- : RUS 1 .C..B::.. B- 4 USA 1 .B..B.B..BBB USA . . . . B . B B 2 USA : .D..B.B:::B: BruchyrumphusmurmorutuS 0.0045 0.060 RUS 1 AAAAAAAAAAAA : RUS 2 B.. B 3 USA 1 BB: BBC..B1CB 4 USA 1 BC. BBC. .B.CC 5 USA 1 BD. BBC. .B.CC Picoidestriductylus 0.0 0.055 RUS 5 AAAAAAAAAAAA : USA 4 BBBBBBB.BBBB Hirundo rusticu 0.0056 0.015 1 RUS 1 AAAAAAAAAAAA RUS 1 . . . . B . . . : USA 1 , BB. C B : B : 4 USA 1 .BB. D::: B . B . Pica pica 0.0034 0.039 RUS 2 AAAAAAAAAAAA : RUS 1 . . B 3 USA 1 .BBbb:BBBBBb 4 USA 2 .BBBB.BBBBBC Anthus spinolettu 0.0017 0.029 1 RUS 3 AAAAAAAAAAAA INTER-CONTINENTAL MTDNA DIFFERENCES IN BIRDS 641

TABLE 1. Continued.

Haplotype Location n Compositepattern

RUS 1 ...... B . : RUS B:. 4 USA : : B:BBBBC.B:B Calcarius lapponicus 0.0042 0.0044 1 RUS AAAAAAAAAAAA RUS : .BB.....B.B. 3’ USA 1 .c B . C . 4 USA 2 .CB:::::B.C USA 1 . B .B C: : RUS 1 .BB::::. B : C . Leucostictearctoa 0.0019 0.0190 1 RUS 2 AAAAAAAAAAAA RUS 1 B...... : RUS 1 . . . . . B : 4 RUS 1 i:“: 5 US‘4 1 cli:: BCBB: .ti; ’ Averagedacross all 3 haplotypes. 2 Not computed;see text.

METHODS differentiation among subspeciesin a qualitative We surveyed mtDNA restriction fragment length manner, based on comparisons of study skins at polymorphisms from 13 species (Table 1). We the Burke Museum, personal observations, and first purified mtDNA from frozen tissuesfollow- the literature. Becauserange disjunctions also play ing standard protocols (Lansman et al. 199 1, a role in interpreting mtDNA patterns, we de- Dowling et al. 1990). Each individuals’ mtDNA termined the nature of discontinuities in range was digested with 12 restriction endonucleases, between Siberian and North American samples end-labeled with Yj, and electrophoresedthrough in addition to Beringia itself. For consistencywe 0.8-1.2% agarose gels. Autoradiography re- follow the specificnomenclature of the 1983 edi- vealed fragment profiles for each restriction en- tion of the AOU Check-List, but we augment donuclease, and different profiles were given a this with other citations for controversial species. different letter. The 12 letters for each individual Subspecificnames are from Vaurie (1959, 1965) constitute its composite haplotype. We scored and the 1957 edition of the AOU Check-List; the presenceand absenceof each restriction frag- the best range maps for Russian subspeciesare ment, and used the method of Nei and Li (1979) found in Dement’ev and Gladkov (1966-1970). to estimate the percentagenucleotide divergence (p). Although the pattern of restriction sites is RESULTS preferred for estimating phylogenetic relation- Little intraspecificsequence divergence was found ships, for estimating p we have found that re- within continents (Table 1). Differentiation be- striction fragments and sites yield close corre- tween continents was observed in both terrestrial spondence (Avise 1994). Also, most haplotypes and aquatic birds, migratory and nonmigratory were closely related (i.e., separated by one or a species,and those with and without phenotypic few restriction sites)which further eliminates bi- differentiation (see below). The degree of popu- ases in using restriction fragments (Dowling et lation differentiation between haplotypes from al. 1990). We used the matrix ofp-values to con- North American and Siberian samples ranged struct a UPGMA phenogram that portrays the from zero (Calcarius lapponicus) to 6.0% (Bra- pattern of genetic similarity among the haplo- chyramphus marmoratus). For all but Anus crec- types; however, our sample sizes are too small ca and Calcarius lapponicus, mtDNA compari- to pursue haplotype phylogenies between taxa. sons revealed apparently diagnostic differences We summarized the degree of morphological between continents. At one extreme, the samples 642 ROBERT M. ZINK m AL.

of Picoides tridactylus differed at 11 of 12 re- where phylogeographicbreaks (Avise et al. 1987) striction endonucleases.At the other extreme, occurand whether they are associatedwith subtle samples of Sterna hirundo differed at only one. discontinuities in morphology, which would in- In Anas crecca, very divergent haplotypes were dicate evolutionary divisions of perhaps species discovered (e.g., haplotypes 1 and 3 vs. 2) but status. haplotype 2 was found both in Siberia and North Samples of Anas creccarepresented two trans- America. continentally distributed subspeciescrecca (Eur- asia) and Caroline&s (North America), which differ in nuptial coloration of males. Apart from DISCUSSION the beringia water barrier, there are no major TAXONOMIC COMMENTS range disjunctions. We found three haplotypes, Limits of intra- and inter-specific taxa are some- two of which were highly divergent; haplotypes times difficult to judge with molecular methods. 1 vs. 2 gave a p-value of 0.036. This level of Some pairs of avian speciesappear to have ex- haplotype differentiation is consistentwith spe- tremely similar mtDNA fragment profiles (p = ties level distinctness.However, haplotype 2 was O.OOl), such as Zonotrichia leucophrysand Z. found both in Russia and U.S., implying gene atricapilla (Zink et al. 199 l), whereas some sib- exchangebetween the continents. This is consis- ling species, such as Limnodromus griseus and tent with wintering birds of the “wrong” race L. scolopaceus,are highly differentiated (p = regularly being reported from Europe and from 0.082; Avise and Zink 1988). There is no par- both coastsof North America. As in some other titular level of differentiation that is equatedwith waterfowl, mixing of divergent mtDNA haplo- a given taxonomic rank (Johnsonand Zink 1983); type lineages among populations has occurred rather, the pattern of variation setsthe limits of (Avise et al. 1990,1992). Further study is needed taxa (Cracrafi 1983). Furthermore, mtDNA data to determine specieslimits in A. crecca. represent a single gene tree that is embedded in Our samples of Numenius phaeopusincluded the organismal phylogeny (Avise 1994). Thus, the subspeciesvariegatus (Russia) and hudsoni- we do not advocate speciesstatus unless “fixed” cus (U.S.), which differ in morphology. Numen- mtDNA differences are associated with other ius p. variegatusis smaller than hudsonicusand character differences such as plumage patterns. also differs in coloration. Numeniusp. variegatus Of course, well-differentiated taxa for which no has a whitish rump, whereas that of hudsonicus fixed molecular differences are apparent are is uniformly dark, contrasting little in appear- sometimes considered distinct evolutionary en- ante with its back; variegatusalso lacks the bully tities (Barrowcloughand Gutitrrez 1990). In rec- backgroundcolor in its venter that characterizes ognizing the limitations of our sampling (both in hudsonicus.Their differencesin backgroundcol- terms of numbersof individuals and named taxa), oration gives variegatus a strikingly whiter or we do not recommend any formal taxonomic grayer overall appearance.A large hiatus in cen- changesbelow. tral Siberia apparently separates the breeding Our Phalacrocorax pelagicus sample repre- rangeof variegatusin the Far East from nominate sented the subspeciespelagicus (Russia) and res- phaeopus of northern Europe (Dementev’ and plendens(U.S.), which are morphologically sim- Gladkov 1966-1970). The isolation of the Far ilar (resplendensis somewhat smaller; Palmer East form, its morphological distinction, and the 1962) and essentially continuously distributed striking mtDNA differences that distinguish it (pelagicusis transberingian in its distribution in from hudsonicusof North America (P = 0.047) the north Pacific). Phalacrocorax p. pelagicusis suggeststrongly that two speciesare involved. found in the Russian Far East and Alaska, where- Comparison of these two forms with the Euro- as resplendensis found on the west coastof North pean race would be worthwhile. At least three America from British Columbia southward. Our speciesmay be representedbecause the nominate resplendensspecimens came from Puget Sound race, phaeopus, of Europe is equally as distinct in Washington. MtDNA revealed 4 of 12 restric- morphologically as the two forms we have com- tion endonucleasesas distinct, but the overall pared genetically (although the variation be- level of differentiation, 0.0076, is relatively low tween phaeopus and eastern Siberian variegatus in comparison to other speciespairs (Avise and might be clinal; Cramp and Simmons 1983). Nu- Zink 1988). Further study is neededto determine men& p. hudsonicus also has two disjunct INTER-CONTINENTAL, MTDNA DIFFERENCES IN BIRDS 643 breeding populations in North America, one in billed birds of Europe and breed more or less the Hudson Bay region and the other in the continuously from the Atlantic coast westward northwestern arctic (Godfrey 1986), but the two to the Rocky Mountains. There is, thus, a very populations have not been described racially. large gap separating the breeding ranges of the Our samples of Gallinago gallinago include two forms we sampled. The mtDNA data suggest the transcontinental subspeciesdelicata (North very limited differentiation, with only a single America) and gallinago (Eurasia); there are no fixed restriction site, and p-value of 0.0075. Fur- major range disjunctions on either continent. ther sampling is required to determine if the Despite being rather similar morphologically, black-billed race of the Far East can be charac- these two forms are sometimes considered dif- terized by a qualitative break in mtDNA, which ferent species(Sibley and Monroe 1990). They would indicate two phylogenetic species.Sibley differ primarily in number of rectrices( 14 in gal- and Monroe (1990) indicate no species-levelcon- linago, 16 in delicata) and width of the outer troversy. rectrices. Our data revealed a single mtDNA re- Brachyramphus marmoratus includes two striction site difference, and a very low p-value subspecies,marmoratus (U.S.) and perdix (Rus- betweenU.S. and Russianhaplotypes (P = 0.006). sia). The presumedbreeding ranges of theseforms Further sampling is required to discern whether (few eggshave been found in western Alaska or there is a discre?emtDNA boundary that would in Russia) are separated by a distance of about correspondgeographically to the break in mor- 500 km at the straits lying between the Com- phology. mander Islands in Russia and the outer-most Our Lants canus sample representedthe sub- group of the Aleutians in the United States(Seal- specieskamtschatschensis (Russia) and brachyr- ey et al. 1982). There is no overlap in size,perdix hynchus(U.S.), which are separatedonly by the being considerably larger, with a disproportion- Bering Sea. Larus c. kamtschatschensisis a larger ately longer and more slender bill, white eye-ring and heavier-billed form than North American in alternate plumage, and lacking the rich reddish brachyrhynchus,and slightly darker dorsally, es- browns in its breeding plumage that characterize pecially in juvenal plumage; tail patterns differ marmoratus. The mtDNA genomes of the two in first basic plumages (Cramp and Simmons taxa are highly differentiated (p = 0.060), more 1983). The mtDNA data suggestthat two species so than most avian species-leveltaxa (Avise and are represented,a possibility also raised by Sibley Zink 1988); we suggestthat they are specifically and Monroe (1990). Previous molecular studies distinct (as did Ridgway [ 19 191).Despite at least revealed little differentiation over a large area in 10 fall and winter records of perdix from inland a gull species (e.g., Bell 1992). Although gulls areas throughout North America (Sealey et al. tend to wander, there are no North American 199 l), philopatry and biogeographic barriers records of L. c. kamtschatschensis.The species- must have been in place for a considerable pe- level mtDNA differentiation of p = 0.02 was riod. consistentwith the morphologicaldifferentiation Our samples of Picoides tridactylus included of the east Siberian form. the subspeciestridactylus and albidior (Russia) Our Sterna hirundo sample represented the and faciatus and dorsalis(U.S.). Populations are subspecieslongipennis (Russian Far East) and continuously distributed acrossthe northern co- hirundo (U.S.); hirundo is also the race found in niferous forests of Eurasia and North America. Europe, and the two subspeciesappear to inter- Alaskan specimens at the Burke Museum are grade (Cramp 1985). Sterna h. longipennis is rather similar in coloration to a series of speci- slightly larger and darker than hirundo, but dif- mens from the Magadan region. Yet our Asian fers principally during the breeding season in and North American specimensare unambigu- having the bill black, and feet dark brown, red ously differentiated genetically, both in number or blackish(for other differencessee Cramp 1985). of restriction endonucleaseswith different pat- In Eurasia, the transition from hirundo to lon- terns (11 of 12) and inp (0.055). In other mtDNA gipennis occurs in a broad region of central Si- surveysof continental variation in woodpeckers, beria where populations exhibit much variation little or no differentiation was found (Ball and in bill and foot color (Vaurie 1965, Dementev’ Avise 1992, Moore et al. 1991). Hence, we sug- and Gladkov 1966-l 970). North American pop- gest that two speciesare representedon the two ulations of Common Terns resemble the red- continents, a view that seemspreviously to have 644 ROBERT M. ZINK ET AL. been unrecognized (Sibley and Monroe 1990). (Vaurie 1959); differences apparently also exist On both continents, birds from southern popu- in calls (Goodwin 1976). Our Kamchatkan spec- lations (many of which are isolated) are much imens also differ from North American birds in darker than those in the north (Dement’ev and having shorter tails and heavier bills. North Gladkov 1966-l 970, Vaurie 1965). This within- American specimens are more similar morpho- continent differentiation between northern and logically to the European populations. Vaurie southern forms is much more impressive than (1959) observed that four geographicallyisolated the more subtle differences between birds from sets of populations occur in the Old World, one either side of the Bering Sea. Given the sedentary ranging from Europe to central Asia, another in nature of thesewoodpeckers, more extreme mor- northwest Africa, the one we sampled in north- phological differentiation might be expected east Asia, and another in southeastAsia. Thus, across Beringia. Perhaps their morphological there are land barriers in addition to the Bering similarity is due to similarities in their northern Sea water barrier. We suggestthat multiple spe- environments; their extreme genetic divergence cies exist in P. pica, an opinion consistent with suggestsa long history of isolation. that of Goodwin (1976), but not mentioned by Our samples of Hirundo rustica included the Sibley and Monroe (1990). Given the propensity subspeciessaturnata (Russia) and erythrogaster for individuals of thesetaxa to wander (Goodwin (U.S.); we follow Vaurie’s (1959) taxonomy. 1976), one might have predicted little differen- These two races are similar enough that De- tiation. Apparently barriers to dispersal were in ment’ev and Gladkov (1966-l 970) consider the place for a considerable period, and mtDNA northeastern-most populations of Asia (that we evolution may have proceeded faster than that sampled) also to be erythrogaster.Hirundo r. sa- of plumage. turnata is variable in ventral coloration, with The specific and subspecific status of Anthus individuals varying from almost pure white to spinolettafrom Asia and North America are con- russet-ochrein their venters. Our two specimens, troversial. The Water Pipit complex of Europe taken near Magadan, are almost white ventrally, and central Asia is often treated as specifically thus being strikingly different from the American distinct (Anthus spinoletta) from the American subspecieserythrogaster, with its reddish belly. Pipit of Siberia and North America (Anthus ru- These taxa are separatedprimarily by the Bering bescens;see AOU 1989). Water and American Sea. Our samples differed at 5 of 12 restriction Pipits reportedly are sympatric without inter- endonucleases (p-value 0.015). Hirundo r. sa- breeding in the region of Lake Baikal (see Sibley turnata closely resembles the European taxon, and Monroe 1990). Ventrally American Pipits rustica. In the intervening region, geographic are heavily streaked whereas Water Pipits are variation in coloration is extreme. For example almost unstreaked; they also differ greatly in vo- in the region of Lake Baikal (other Burke Mu- calizations (Alstrom and Mild 1987). What is seum specimens), north to Yukutsk on the Lena striking about our sample is that, although the River, Barn Swallows have uniformly dark, Bering Sea appears to represent a major barrier chestnut-colored underparts (Dement’ev and to gene flow between Asia and North America, Gladkov 1966-l 970). Zones of contact between it has not resulted in the development of any these very different forms should be examined. significant trans-Beringian morphological differ- We suggestthat two or more speciesexist in H. entiation. Vaurie (1954) reports being unable to rustica. separatebreeding seasonspecimens of American Our samples of Pica pica included the sub- (“Water”) Pipits from the Commander Islands species camtschatica (Russia) and hudsonia and from the mouth of the Lena River from (U.S.). The isolated race found in Kamchatka breeding specimenstaken in North America. Our (the source of our specimens) and the Anadyr own comparisons of Russian and American region of Russia is situated in closestgeographic specimensat the Burke Museum corroborate his proximity to North America. Nonetheless, our findings. Yet, eight of 12 restriction endonucleas- Russian and U.S. samples are highly differenti- es separatedthe Russian and U.S. samples, with ated in mtDNA (p = 0.039). Morphologically a p-value of 0.029. If morphological or other camtschaticais among the most divergent of the genetic differencesbetween thesepopulations are eurasianraces, having much more extensivewhite found, we think they should be treated as differ- in the primaries and the greenest gloss of all ent species.Anthus rubescenswinters in southern INTER-CONTINENTAL MTDNA DIFFERENCES IN BIRDS 645 regions of both continents, suggestingthat east suffice for estimating degree of mtDNA differ- Asian and North American populations were entiation. For example, if one were to draw at isolated long enough to have differentiated in random two or three Song Sparrow (Melospiza migratory habits (Dement’ev and Gladkov 1966- melodia) haplotypes from each of two geograph- 1970). Several other trans-Beringian colonists ically distant populations, which are not genet- make biannual migrations acrossthe Bering Sea ically differentiated, it is very unlikely that they to winter only in one continent or the other, de- would suggestgeographic differentiation (Zink spite breeding in both. and Dittmann 1993a). Although our presentstudy Our sample of Calcarius lapponicus repre- requires amplification with more data, several sented different subspecies (alascenis in U.S. inferences about inter-continental genetic differ- and lapponicusin Russia). Although there is little entiation can be drawn. morphological differentiation between them, Trans-Beringia comparisons of 13 speciesin- most of our Russian male specimens are more dicated a variety of degrees of differentiation. heavily marked with black than North American Some “species,” such as Picoides tridactylus, breeding specimens from Alaska. A phenogram Brachyramphus marmoratus, Anthus spinoletta, of haplotypes (not shown) did not reveal sepa- Pica pica, Larus canus, and Numenius phaeopus ration into two continental groups (and would are highly differentiated relative to other avian not even if “rooted”). MtDNA data reveal as species (Avise and Zink 1988). These six taxa much differentiation between haplotypes from have clearly had relatively long independent his- the same continent as between haplotypes from tories on either side of Beringia. At the other different continents. Either the two taxa were re- extreme, some speciesappear to have been sep- cently isolated, such that mtDNA lineages are arated relatively recently (Gallinago gallinago, paraphyletic with respect to subspeciesbound- Phalacrocorax pelagicus, Sterna hirundo). Only aries (Avise 1994) or there is current gene ex- Calcarius lapponicusand Anus creccaappear to change. The absence of mtDNA differentiation have current gene exchange across Beringia, as does not disprove the existence of subspecies,as evidenced by the occurrence of the same hap- morphological change is thought to evolve faster lotypes on both continents. Thus, unlike many than mtDNA monophyly in some birds (Zink studies of North American birds (e.g., Ball et al. and Dittmann 1993a), but not others (e.g., Pi- 1988; Zink and Dittmann 1993a, 1993b; Ball coidestridactylus, Pica pica; this study). and Avise 1992) the majority of trans-Beringian Specieslimits in the genus Leucosticteare con- comparisons exhibited intraspecific isolation, at troversial (AOU 1983); recently, the AOU (1993) least to some degree. The variation in degree of elevated the three main North American taxa to mtDNA differentiation could be attributed to dif- specieslevel, rather than as subspeciesof L. arc- ferent dispersal rates, rates of evolution, or his- toa. Our single U.S. specimen is referable to L. torical factors. Unfortunately, reliable estimates tephrocotislittoralis, and our Siberian specimens of dispersaldistances are unavailable, which pre- to L. arctoa pustulata. These taxa are morpho- vents determining if genetically differentiated logically differentiated, and the range is frag- species are relatively more sedentary. Without mented by land barriers in addition to the Bering outgroup comparisons, it is not possible to de- Sea itself. The North American specimen dif- termine if rates of mtDNA differentiation vary fered at eight of 12 restriction fragment profiles acrossthe speciescompared (Avise 1994). from the Russian ones, ap-value of 0.0 19, a level Our comparisons suggestthat the avifauna of commensurate with several avian species(Zink each continent consistsof specieswith a variety et al. 1991). It seems likely that two speciesare of past histories, some with long periods of iso- represented in our comparison, which is consis- lation, othersrelatively recently isolated, and still tent with the opinion of Sibley and Monroe others currently in contact. This suggestsa his- (1990). tory of community membership that is at odds General comments.Our study provides insight with the principal assumption of vicariance bio- into the general nature of mtDNA differentiation geography,which presumesthat ancestral biotas within species-leveltaxa distributed on two con- were stable and widespread, allowing taxa to dif- tinents. Rather than study one species inten- ferentiate in concert (Neigel and Avise 1986, Av- sively, we compared multiple specieson conti- ise 1992). Many paleoecologists (e.g., Bennett nents. Relatively small samples of individuals 1990, Pielou 199 1) suggestthat communities un- 646 ROBERT M. ZINK ET AL. dergo marked changes in species membership gists’ Union Check-list of North American birds. over relatively short geologicaltime periods. Our Auk 106:532-538. AMERtcAN0 RNITHOLOGISTSUNION.’ 1993. Thirty- study suggests that species colonized North ninth Supplementto the American Ornithologists’ America and Siberia at different times, judging Union Check-list of North American birds. Auk from the varying degreesof genetic differentia- 110:675-682. tion, resulting in a mosaic of different histories AVISE,J. C. 1992. Molecular population structureand the biogeographic history of a regional fauna: a being represented within a single extant com- casehistory with lessonsfor conservationbiology. munity. Oikos 63:62-76. If each species-pair was placed in a phyloge- ARISE,J. C. 1994. Molecular markers,nautral history, netic context, it might be possible to determine and evolution. Chapman and Hall, New York. the direction of invasions of each continental ARISE,J. C., ANDR. M. ZINK. 1988. Molecular genetic divergencebetween avian siblingspecies: King and region. For example, Hirundo rwstica has differ- Clapper Rails, Long-billed and Short-billed Dow- ent close relatives on both continents; thus, we it&e&, Boat-tailed&d Great-tailed Grackles,and might ask whether this swallow differentiated in Tufted and Black-crestedTitmice. Auk 105:516- Asia and later invaded North America or vice 528. AVISE,J. C., C. D. ANKNEY,AND W. S. NEI~ON. 1990. versa. Assuming that the Siberian and North Mitochondrial gene trees and the evolutionary re- American populations of H. rustica are sistertaxa, lationship of Mallard and Black Ducks. Evolution this resolves into the question of where the clos- 44:1109-1119. est relative of this sisterpair resides.Ifthe nearest AVISE,J. C., R. T. ALISAUSKAS,W. S. NELSON,AND C. phylogenetic relative of this sister pair is an Old D. ANKNEY. 1992. Matriarchal population genetic structure in an avian species with female World swallow, one might reasonably infer an natal philopatry. Evolution 46:1084-1096. invasion of the New World from Asia. Suchanal- AVISE,J. C., J. ARNOLD,R. M. BALL,E. BERMINGHAM, yses could reveal general rules about the nature T. LAMB, J. E. NEIGEL,C. A. REED,AND N. C. of fauna1 exchange between Asia and North SAUNDERS.1987. Intraspecific phylogeography: the mitochondrial DNA bridge between popula- America. tion geneticsand systematics.Ann. Rev. Ecol. Syst. 18:489-522. ACKNOWLEDGMENTS Bw, R. M., JR., AND J. C. AVISE. 1982. MtDNA We are deeplygrateful to GarrettEddy for havingthe phylogeographicdifferentiation among avian pop- vision to encouragethe BurkeMuseum ’s expeditions ulations, and the evolutionary significanceof sub- to Russia,for his supportof thoseexpeditions and the species.Auk 109:626-636. lab workfor theseanalyses and, especially,for hiswill- BALL, R. M., JR., S. FREEMAN,F. C. Jm, E. BER- MINGHAM,AND J. C. AVISE. 1988. Phylogeo- ingnessto let our larger scientificobjectives drive every araphic ponulation structureof Red-winged Black- aspect of our work in Russia, even when it took us birds assessedby mitochondrial DNA. Proc. Natl. farther afield than originally planned. For their extraor- Acad. Sci. U.S.A. 85:1558-1562. dinary help and hospitality in Russia we thank K. Re- BARROWCLOUGH,G. F., AND R. J. GUTIBRREZ. gel, N. Dokuchaev, D. Banin, V. Masterov, and S. 1990. Genetic variation and differentiation in the Spot- Drovetskiy. Thanks also to S. W. Cardiff and C. S. ted Owl (Strix occidentalis).Auk 1071737-744. Wood for collecting specimens, and to C. Wood for BELL,D. A. 1992. Hybridization and sympatry in the logistical efforts that literally made the Burke Mu- Western Gull/Glaucous-winged Gull complex. seum’s Russianexpeditions work. We thank S. W. Car- Ph.D. diss. Univ. of California, Berkeley, CA. diff, J. Klicka, A. J. Fry, F. H. Sheldon and J. V. Rem- BENNE~~.K. D. 1990. Milankovitch cvclesand their sen, Jr., for comments on the manuscript. This study effects on speciesin ecological and evolutionary is basedentirely upon the collectionsof the University time. Paleobiology 16: l-2 1. of Washington Burke Museum and the Louisiana State 1 BERMMGHAM,E., S. ROH~ER, S. FREEMAN,AND C. University Museum of Natural Science. WOOD. 1992. Vicariance biogeography in the LITERATURE CITED Pleistocene and speciation in North American wood warblers: a test of Mengel’s model. Proc. ALSTROM, P., ANDK. MILD. 1987. Some notes on the Natl. Acad. Sci. U.S.A. 89:6624-6628. taxonomy of the Water Pipit complex. Proc. 4th CRAW, J. 1983. Speciesconcepts and speciation Int. Identification Meeting (Eilat) :4748. analysis. Current Ornithology 1:159-187. AMERICANORNITHOLOGISTS UNION.’ 1957. Check- &, S. [ED.]. 1985. Handbook of the birds of Eu- list of North American birds, 5th ed. American rope, the Middle East, and North Africa: the birds Ornithologists’ Union, Washington, D.C. of the western Palearctic. Vol. IV. Oxford Univ. AMERICANORNITHOLOGISTS UNION.’ 1983. Check- Press, Oxford, U.K. list of North American birds, 6th ed. American CRAMP,S., ANDK.E.L. SIMMONS.[EDS.]. 1983. Hand- Ornithologists’ Union, Washington, D.C. book of the birds of Eurone. the Middle East. and AME~CAN ORNITHO~~GIS~~UNION. 1989. Thirty- North Africa: the birds of the western Palearctic. seventh Supplement to the American Omitholo- Vol. III. Oxford Univ. Press, Oxford, U.K. INTER-CONTINENTAL MTDNA DIFFERENCES IN BIRDS 641

DEGNAN,S. M., AND C. MORTIZ. 1992. Phylogeog- life to glaciatedNorth America. The Univ. of Chi- raphy of mitochondrial DNA in two species of cago Press, Chicago. white-eyes in Australia. Auk 109:80&8 11. R~DGWAY,R. 19 19. The birds of North and Middle DEMENT’EV,G. P.. ANDN. A. GLADKOV.IEDS.~. 1966- America, Part VIII. Government Printing Office, 1970. [Birds of the Soviet Union]. Israel Program Washington, D.C. for Scientific Translations. Jerusalem. SEALEY, S. G., H. R. CARTER, AND D. ALISON. DOWLING, T. E., C. MORITZ, AND J. PALMER. 1990. 1982. Occurrencesof the Asiatic Marbled Murre- Nucleic acids II. Restriction site analysis,p. 250- let [Bruchyramphusmarmorutus perdix (Pallas)] 3 19. In D. M. Hillis and C. Moritz [eds.], Molec- in North America. Auk 99: 778-78 1. ular Systematics.Sinauer Assoc., Sunderland, MA. SEALEY,S. G., H. R. CARTER,W. D. SHUFCIRD,K. D. EDWARDS,S. V. 1993. Long-distancegene flow in a POWERS,AND C. A. CHASE,III. 1991. Long-dis- cooperative breeder detectedin genealogiesof mi- tance vagrancyof the Asiatic Marbled Murrelet in tochondrial DNA sequences.Proc. R. Sot. Lond. North America, 1979-l 989. West. Birds 22: 145- B 252:177-185. 156. GODFREY,W. E. 1986. The birds of Canada, revised SIBLFY,C. G., AND B. L. MONROE,JR. 1990. Distri- edition. National Museums of Canada, Ottawa. bution and taxonomy of birds of the world. Yale GOODWIN,D. 1976. Crows of the world. Cornell Univ. Univ. Press, New Haven, CT. Press, Ithaca, NY. VAURIE,C. 1954. Systematicnotes on Palearcticbirds. JOHNSON,N. K., AND R. M. ZONK. 1983. Speciation No. 7. Alaudidae and Motocillidae (genusAnthus). in sapsuckers(Sphyrupicus): I. Genetic differen- Am. Mus. Novit. no. 1672:l-l 3. New York. tiation. Auk 100:871-884. VAURIE, C. 1959. The birds of the Palearctic fauna. LANSMAN,R. A., R. 0. SHADE,J. F. SHAPIRA,AND J. Passeriformes.Witherby, London. C. AVISE. 1981. The use of restriction endonu- VAURIE. C. 1965. The birds of the Palearctic fauna. cleasesto measure mitochondrial DNA sequence Non-Passeriformes. Witherby, London. relatednessin natural populations III. Techniques ZINK, R. M. 1994. The geographyof mitochondrial and potential annlications. J. Mol. Evol. 17:214- DNA variation, population structure, hybridiza- 226.. -- tion, and specieslimits in the Fox Sparrow (Pas- MOORE,W. S., J. H. Gm, ANDJ. T. PRICE. 1991. serelfuiliucu). Evolution 48:96-l 11. Mitochondrial DNA variation in the Northern ZINK, R. M., ANDD. L. D~~~MANN. 1993a. Gene flow, Flicker (Coluptesaurutus, Aves). Mol. Biol. Evol. refugia, and evolution of geographicvariation in 8:327-344. the Song Sparrow (Melospizumelodiu). Evolution NEI, M., AND W. H. LI. 1979. Mathematical model 47~717-129. for studying genetic variation in terms of restric- ZONK,R. M., AND D. L. DITTMANN. 1993b. Popula- tion endonucleases.Proc. Natl. Acad. Sci. U.S.A. tion structureand gene flow in the Chipping Spar- 76~5269-5273. row (Spizellapussen nu),’ and a hypothesisfor evo- NEIGEL,J. E., AND J. C. AVISE. 1986. Phylogenetic lution in the genusSuizellu. Wilson Bulletin 105: relationships of mitochondrial DNA under vari- 399-413. - - ous demographic models of speciation, p. 515- ZINK, R. M., D. L. DI-~~MANN,AND W. L. ROOTES. 534. In S. Karlin and E. Nevo [eds.], Evolutionary 1991. Mitochondrial DNA variation and the phy- processesand theory. Academic Press,New York. logeny of Zonotrichiu. Auk 108:578-584. PIELOU,E. C. 1991. After the Ice Age: the return of