Intraspecific Molecular Phylogeny in the Yellow Warbler (Dendroica Petechia), and Implications for Avian Biogeography in the West Indies

Evolution, 48(6), 1994, pp, 1914-1932

INTRASPECIFIC MOLECULAR PHYLOGENY IN THE YELLOW WARBLER (DENDROICA PETECHIA), AND IMPLICATIONS FOR AVIAN BIOGEOGRAPHY IN THE WEST INDIES

NEDRA K. KLEIN'AND WESLEY M. BROWN Museum ojZoology and Department ofBiology, University ojMichigan, Ann Arbor, Michigan 48109 E-mail: Wesley.Brownenon.cc.umich.edu

Abstract.-A phylogenetic analysis of mitochondrial DNA (mtDNA) restriction sites was used to examine the evolutionary history of populations of yellow warbler (Dendroica petechia) sampled from North America, Central America, South America, and the West Indies. Thirty-seven hap lotypes were identified, and only one was found in more than one of these regions. Estimated sequence divergence among haplotypes ranged from 0.14 to 3.17%, and mtDNAs from North American migratory populations clearly were differentiated from those ofmost tropical sedentary populations. Parsimony analysis ofhaplotypes suggested multiple colonizations ofthe West Indies archipelago and of individual Caribbean islands. The inference of multiple colonizations has important implications for studies ofavian ecology and evolution in this region.

Key words.-Bird, Caribbean region, Dendroica petechia, island biogeography, mtDNA, phylogeny, West Indies.

Received September 28, 1992. Accepted March 18, 1994.

Studies of island biogeography have contrib er taxa, and most likely have arrived at their uted greatly to our understanding of ecological present distributions via dispersal from sur andevolutionary processes (MacArthurand Wil rounding continental areas and from other is son 1963, 1967; Lack 1976; Berry 1983; Wil lands (Bond 1963, 1978; Seutin et al. 1993). liams 1989). Such inferences about process can Nevertheless, testable predictions of sources become even more powerful, however, when they of colonists can be made based on distances to are made within an historical framework pro the nearest land masses. In a series ofpapers on vided by estimates of phylogenetic relationship WestIndian birds, Bond (1948, 1961, 1963, 1978, among island populations and species (Rosen 1979, 1993) documented distributions and dis 1975; Eldredge and Cracraft 1980; Wiley 1981; cussed taxonomic relationships of West Indian Guyer and Savage 1986; Kluge 1988; Miller and birds. Basing his hypotheses on current faunal Miller 1989). distributions, phenotypic similarities, and in Caribbean islands have acquired their fauna ferred affinities to mainland groups, he proposed eitherfrom overwaterdispersal (Darlington 1938; two main sources of the Caribbean avifauna: Bond 1963, 1978; Pregill198l; Koopman 1989), northern South America for many ofthe Lesser or as relicts through fragmentation ofa formerly Antillean species, and Central and North Amer widespread distribution (Rosen 1975; Kluge ica for Greater Antillean and some Lesser An 1988). Vicariance explanations of distribution tillean species. He proposed multiple coloniza and evolutionary diversification have been pro tions from these sources for island populations posed for several faunal groups on Caribbean of widespread species. islands (Rosen 1975; Kluge 1988; Miller and Whether single or multiple colonizations have Miller 1989). However, many avian species are occurred in the West Indies is an important dis too recent in origin (Brodkorb 1971; Feduccia tinction for inferences about evolution within and Martin 1976; Feduccia 1977; Olson 1985) this region. The implication ofa widespread spe for the vicariance explanations proposed for oth- cies having colonized the archipelago multiple times from different sources is that island pop ulations would not constitute a monophyletic group and would thus not be each others' closest I Present address: Department ofOrnithology, Amer relatives. Monophyly ofisland populations is an ican Museum of Natural History, Central Park West at 79th, New York, NY 10024-5192, E-mail: implicit assumption of many evolutionary and [email protected]. ecological studies of species on archipelagoes 1914

(MacArthur and Wilson 1967; Ricklefs and Cox We present here an estimate of the phyloge 1972, 1978). netic relationships among populations ofyellow Another assumption of many island biogeo warbler (Dendroica petechia), a widespread, pol graphic studies ofcolonizing species is that pop ytypic avian species. The phylogenetic recon ulations on adjacent islands are each others' clos struction is based on a restriction endonuclease est relatives; that is, their distributions are the analysis of mitochondrial DNA (mtDNA). We result ofa stepping-stone process ofcolonization use the mtDNA phylogeny to examine the evo (Darlington 1938; MacArthur and Wilson 1967; lutionary history of resident populations in the Ricklefs and Cox 1972, 1978). An alternative is West Indies and tropical continental areas. We a colonization pattern that is random with re test the hypothesis that the West Indies has been spect to geography. Under the stepping-stone colonized multiple times by different source pop model, islands arranged linearly (e.g., the Lesser ulations ofyellow warbler, as suggested by Bond Antilles) are colonized sequentially, beginning (1963, 1978) for widespread Caribbean birds. If with that closest to the mainland source (Wil multiple colonizations have occurred, the Carib liams 1989). Individual island populations would bean samples should not form a monophyletic be derived from those on other islands or the group relative to othersamples. The analysis also adjacent mainland, and heritable differences is used to infer colonization sources, as has been among them should thus be due to selection or done for other groups ofisland organisms (Ash random differentiation processes rather than to ley and Wills 1987; Phillips et al. 1989), and to different phylogenetic histories. If, however, a test the critical prediction of the stepping-stone widespread species has colonized the archipelago colonization model that populations on adjacent multipletimesfrom different source populations, islands are closely related to each other. there may be areas where populations on adja cent islands are descendants ofdifferent ancestral METHODS colonizers. Differences among such populations Specimens were collected by shotgun or mist may thus be due to their different genealogies. net from 1986-1990. A total of 194 yellow war A phylogenetic inference of relationships blers was used in this study. General localities among island and mainland populations of a and sample sizes were: Michigan (17), North widespread taxon is useful both because it allows Carolina (9), Washington (6), Baja California (1), testing the models of single and multiple colo Florida Keys (1), Jamaica (9), Dominican Re nizations of archipelagoes (different patterns of public (10), Puerto Rico (17), U.S. Virgin Islands relationship are predicted under each) and be (24), Montserrat (8), Guadeloupe (5), Dominica cause potential interpretive problems associated (6), Martinique (4), St. Lucia (4), Venezuela (40), with missing taxa (islands that were never col Panama (13), Costa Rica (20). Populations from onized or whose populations have been extir Michigan, North Carolina, and Washington are pated) can be eliminated. Under the model of a wholly migratory. The rest are wholly sedentary single colonization ofan archipelago, monophyly tropical residents. of island populations is predicted. Polyphyly is Although a few estimates of historical rela predicted under the multiple-colonization mod tionships among paruline warblers exist (Mengel el, provided founders came from different source 1964; Barrowclough and Corbin 1978; Raikow populations. It is difficult to test models ofisland 1978; Bledsoe 1988; Sibley and Ahlquist 1990), colonization pattern within archipelagoes, how none of the studies include Dendroica petechia. ever, because (1) any tree topology is consistent Preliminary allozyme evidence (Avise et al. 1980) with a random colonization model (i.e., the ran suggests that the yellow warbler may fall outside dom model is not falsifiable), and (2) although a ofa group consisting ofother Dendroica. In the pectinate tree is predicted under the stepping absence ofan explicit hypothesis ofrelationship stone model, if colonization has occurred rap among D. petechia and other warblers, choice of idly, or if there has been any postcolonization outgroup taxa was done with the goal ofencom movement among islands, the comblike pattern passinga wide range ofinferred divergence levels ofhistorical relationship will not be recoverable. relative to the ingroup (Maddison et al. 1984). Nevertheless, one can test the prediction of the Outgroup species (with sample sizes and locali stepping-stone model that populations on adja ties) were: D. pensylvanica (2, Michigan), D. dis cent islands should be each others' closest rela color (2, Florida), Setophaga ruticilla (1, Mich tives. igan), Parula americana (1, North Carolina), 1916 N. K. KLEIN AND W. M. BROWN

Geothlypis triehas (2, Michigan), and Basileute passing the geographic range ofsamples) digested rus eulicivorus (I, Venezuela). All outgroups ex with that enzyme were run together on one set ceptB. eulicivorus are migratory NorthAmerican of agarose and polyacrylamide gels to facilitate species. Basileuterus is a wholly sedentary, Neo comparison of fragment sizes among different tropical genus that is the probable sister taxon haplotypes. Other gel sets consisted of. the re to the genera of largely migratory North Amer maining individuals. DNA was digested to com ican warblers (Raikow 1978; Sibley and Ahlquist pletion (2-14 h) with an excess of the following 1990). enzymes: Aval, AvaIl, BamHI, Bell, BglII, BstEII, Within 5 to 15 min after death, whole birds Clal, HindIII, Kpnl, Ndel, Nhel, Pvull, and SpeI. were frozen on dry ice or tissues were removed Digestions were performed underconditions rec and frozen in liquid nitrogen. Tissues removed ommended by the suppliers (Boehringer-Mann in the field were stored in liquid nitrogen until heim, New England Biolabs, and Toyobo). Frag they were transported back to the Laboratory of ments were end-labeled with p32 (Brown 1980; Molecular Systematics at the University of Wright et al. 1983; Dowling et al. 1990), run in Michigan Museum ofZoology, where they were 1x TBE buffer on both agarose (0.8-1.2%) and stored at -70°C until processing. Birds originally polyacrylamide (3.5-5%) vertical gels, and vi frozen whole on dry ice were also stored at -70°C sualized by autoradiography (Brown 1980). A until laboratory analysis, when the tissues were size standard of lambda phage DNA digested removed and processed before thawing. with HindIII mixed with rf>x 174 phage DNA Voucher specimens (museum study skins, digested with Haelll was included on each gel. skeletons, or skins and partial skeletons) were Fragment sizes were estimated from calibration prepared for most individuals from which curves plotted from 10glO fragment size versus mtDNA was purified. These specimens are de distance migrated of size-standard fragments. posited in collections at the University ofMich Mean size estimates ofthe mtDNA molecule for igan Museum ofZoology, Ann Arbor; American each species were calculated from the sizes es Museum ofNatural History, New York; Florida timated from fragments generated by each en State University, Tallahassee; North Carolina zyme for the mapped individuals. State Natural History Museum, Raleigh; Burke Because of potential errors in assuming ho Museum, Seattle; National Museum of Natural mology ofcleavage sites that produce fragments History, Washington, D.C.; Museo Nacional de of similar size, direct mapping of relative posi Costa Rica, San Jose; Museo de Historia Natural tions of cleavage sites is recommended for phy Fundacion La Salle, Caracas, Venezuela; Museo logenetic analyses (Templeton 1983b; Moritz et de Zoologia Facultad de Ciencias, Universidad al. 1987; Dowling et al. 1990; Swofford and Ol Autonoma de Mexico, Mexico City; and the De sen 1990). Even in closely related organisms, partmentofNatural Resources, San Juan, Puerto mtDNA fragments of similar size can be pro Rico. duced by cleavage at nonhomologous sites (Pumo All birds used in the analysis were local resi et al. 1988; Klein pers. obs.). Cleavage sites for dents as inferred from the presence ofnests, ter each outgroup species and for two yellow warbler ritorial behavior, or relative development ofgo mtDNA lineages (WI-A from the Dominican nads. Republic and NA-E from Michigan; estimated Isolation, purification, and restriction endo sequence divergence of 2.3%) were therefore nuclease digestion of mitochondrial DNA fol mapped independently, using double and triple lowed methods outlined in Lansman et al. (198 1) digests (Brown and Vinograd 1974; Dowling et and Dowling et al. (1990). Purified mtDNA was al. 1990) (table 1). Cleavage-site losses relative isolated from liver, heart, or pectoral muscle. to the mapped sites and site homology were in The amount of tissue used ranged from 0.03 g ferred from fragment pattern comparisons with to 0.3 g, and the homogenization buffer consisted those for mapped haplotypes (Vawter and Brown ofone part 0.5 M sucrose in TE to five parts 200 1986; Dowling and Brown 1989). The positions mM EDTA, 10 mM NaCl, 10 mM Tris. ofcleavage sites that were gained relative to the Restriction enzymes producing variable frag mapped haplotypes were determined with ad ment patterns among three samples (from Mich ditional double digests. Additional double-di igan, Costa Rica, and Puerto Rico) were used to gests were also used to verify positions in un digest the remaining mtDNAs. Foreach enzyme, mapped individuals of synapomorphic restric 15 samples (1 from each of 15 localities encom- tion sites that uniquely identified clades. TABLE I. Restriction-site map positions (in kb, to nearest 0.05 kb) relative to ClaI A, and presence/absence (I/O) of site. ?, presence/absence of site not determined. Outgroup species are OG-A, Dendroica pensylvanica; OG-B, Dendroica discolor; OG-C, Parula americana; OG-D, Geothlypis trichas; OG-E, Setophaga ruticilla; OG- F, Basileuterus culicivorus. Enzyme abbreviatons as follows: AI: AvaI, All: AvaIl, Ba: BamHI, Be: Ben, B: BgnI, Bs: EstEll, C: ClaI, H: HindIII, K: KpnI, Nd: NdeI, N: NheI, P: PvuIl, Sa: SadI, S: SpeI.

Haplotypes West Indies Central America Venezuela North America Outgroups WI- CA- VE- NA- OG- Position Site ACEFGBHIJK ABCDEFGHIJ AFBCDE ACDEFGIJLMN ABCDEF -0- C 1111111111 1111111111 111111 11111111111 111111 0.1* Nd 0000000000 0000000000 000000 00000000001 000000 0.6 S 0000000000 0000000000 000000 00000000000 000100 0.75 H 1111001111 0000010100 001111 00000000000 000000 1.1 H 1111111111 1111111111 111111 11111111111 111111 1.2 All 1111111111 1111111111 111111 11111110111 111111 1.2 H 0000000000 0000000000 000000 00000000000 000010 1.4 Sat 1???1???1? ?????????? ?????? ???l??????? 111111 1.5 S 1111111111 1111111111 111111 11111111111 111111 1.7 All 1111111111 1111111111 111111 00000000000 001000 1.8 BC 1111111111 1111111111 111111 11111111111 110101 1.85 All 1111111111 1111111111 111111 11111111111 111111 2.05 N 1111111111 1111111111 111111 11111111111 111111 2.25 All 0000000000 0000000000 000000 00000000000 000101 2.3 S 1111111111 1111111111 111111 11111111111 110111 2.4 N 1111111111 1111111111 111111 11111111111 111110 2.5 Be 0000000000 0000000000 010000 00000000000 001000 2.55 H 0000000000 0000000000 000000 00000000000 010000 2.6 All 0000000000 0000000000 000000 00000000000 000011 2.75§ All 0000000000 0000000000 000000 00000000000 000001 2.85 S 0000000010 0000000000 000000 00000000000 000100 3.0 H 0000000000 0000000000 000000 00000000000 111111 3.0 AI 1111111111 1111111111 111111 11111111111 000000 3.1 Bs 0000000000 0000000000 000000 00000000000 000100 3.6 K 0000000000 0000000000 000000 00000000000 010001 3.6 P 1111111111 1111111111 111111 11111111111 110111 4.3§ All 0000000000 0000000000 000000 00000000000 000101 4.4 Nd 0000000000 0000000000 000000 00000000101 010000 4.6 All 1111111111 1111111111 111111 11111111111 111111 4.7 C 0000000000 0000000000 000000 00000000000 001000 4.9 All 1111110001 1111111111 111111 11111111111 101111 4.9 K 1111101111 1111111111 111111 11111111111 111111 5.05 All 0000000000 0000000000 000000 00000000000 001000 5.1 S 1111111111 1111111111 111111 11111111111 110110 5.2 H 1111111111 1111111111 111111 11111111110 011111 5.3 AI 0000110000 0000000000 110100 00000000000 110000 5.5 Bs 1111111111 0011110101 111111 11111111111 100011 5.5 Be 1111111111 1111111111 111111 11111111111 111100 5.7 AI 1111111111 1111111111 111111 11111111111 111111 5.8 Bs 1111111111 1111111111 110100 11111111111 100000 5.8 P 1111111111 1111111111 111111 11111111111 111111 5.9 P 0000000000 0000000000 000000 00000000000 000001 6.2 P 0000110010 0000000000 110100 11111111111 111100 6.25 S 1111111111 1111111111 111111 11111111111 101101 6.5 Nd 1111111111 1111111111 111101 11111111111 111101 6.5 B 0001000000 0000000000 000000 00000000000 000000 6.8 S 0000000000 0000000000 000000 00000000000 001000 7.0 Bs 1111001111 0000010100 001011 00000000000 000000 7.0 AI 0000000000 0000000000 000000 00000000000 010000 7.1 B 1111111111 1111111111 111111 11111111111 010000 7.2 S 0000110000 0000000000 110000 00000000000 000000 7.3 B 0000111111 1111111111 111111 11111111111 110000 7.4 All 0000000000 0000000000 000000 00000000000 000100 7.4 H 0000000000 0000000000 000000 00000000000 000001 TABLE 1. Continued.

West Indies Central America Venezuela North America Outgroups WI- CA- VE- NA- OG- Position Site ACEFGBHIJK ABCDEFGHIJ AFBCDE ACDEFGIJLMN ABCDEF 7.5 H 1101111111 1111111111 111111 10111111111 111111 7.6 Be 1111000000 0000000000 000000 00000000000 000000 7.8 N 0000000000 0000000000 000000 00000000009 000100 7.9 K 0000000000 0000000000 000000 00000000000 000001 8.1 H 1111111111 1111111111 111111 11111111111 111110 8.5 AI 1111111110 0011011100 111111 11111111111 000100 8.6 Bs 0000000000 0000000000 000000 00000000000 011000 8.6 AI 0000110000 0000000000 110100 00000000000 000000 8.9 S 1111111111 1111111111 111111 11111111111 111111 9.1 H 0000000000 0000000000 000000 00000000000 001010 9.3 Ba 1111111111 1111111111 111111 11111111111 110000 9.4 H 1111111111 1011111111 111111 11111111111 111011 9.45 S 0000000000 0000000000 000000 00000000000 000001 9.6 Nd 0000000000 0000000000 000000 11111111111 011111 9.7 All 1111111111 1111111111 111111 11111111111 100001 9.8 S 1111111111 1111111111 111111 11111111111 101001 9.8 Ba 1111111111 1111111111 111111 11111111111 111010 10.0 C 0000000000 0000000000 000000 00000000000 000001 10.3 Be 0000000000 0000000000 000000 00000000000 000001 10.4 Bs 0000000000 0000000000 000000 00000000000 100101 10.5 N 0000000000 0000000000 000000 00000000000 001000 10.5 Ba 1111111111 1111111111 111111 11111111111 111111 10.6 Be 1111111111 1111111111 111111 11111111111 111011 10.7 S 0000000000 0000000000 000000 00000000000 010110 10.8 N 1111111111 1111111111 111111 11111111111 010000 11.0 All 0000000000 0000000000 000000 00000000000 001000 11.1:j: Ba 0000000110 0000000000 000000 00000000000 000000 11.1 K 1111111111 1111111111 111111 00000000000 000000 11.4 All 1111111111 1111111111 110101 11111111111 111111 11.5 AI 0000000000 0000000000 000000 00000000000 000100 11.7 All 1111001111 0000010100 001011 00000000000 000000 11.9 P 1111111111 1111111111 111111 11111111111 011110 12.1 p 1111111111 1111111111 111111 00000111111 010010 12.15 All 1111111111 1111111111 111111 11111111111 101010 12.5 H 0000000000 0000000000 000000 00000000000 000110 12.7 S 1111111111 1111111111 111111 11111111111 111101 12.7 All 0000000000 0000000000 000000 00000000000 000010 12.8 N 0000000000 0000000000 000000 00000000000 000010 12.8 Bs 1111111111 1111111111 111111 00000111111 110010 13.0 H 0000000000 0000000000 000000 00000000000 000010 13.1 N 0000000000 0000000000 000000 00100000000 000000 13.4 Ba 1111111111 1111111111 111111 11111111111 111101 13.5 Nd 1111111111 1111111111 111111 11111111111 111111 13.75 S 1111111111 1111111011 111111 11111101010 101110 13.8 Nd 1111111111 0001100001 110100 11111011000 000000 14.1 S 0000000000 0000000000 000000 00000000000 000100 14.1 Nd 0000000000 0000000000 000000 00000100000 000000 14.2 N 1111111111 1111111111 111111 00000000000 110010 14.3 Bs 0000000000 0000000000 000000 00000000000 000111 14.4 P 0000000000 0000000000 000000 00000000000 010000 14.5 AI 1111111111 1111111111 111111 11111111111 111111 14.9 Ba 1111111001 1111111111 001111 00001000000 111101 14.9 All 0000000110 0000000000 110000 11110111111 000001 15.2 B 1111111111 1111111111 111111 11111111111 111111 15.35 H 0000000000 0000000000 000000 00000000000 000101 15.4 All 0000000000 0000000000 000000 00000000000 000011 15.45 K 1111111111 1111111111 111111 11111111111 111111 15.6 All 0000000000 0000000000 000000 00000000000 000001 15.8 All 0100000000 0000000000 000000 00000000000 000000 YELLOW WARBLER MTDNA PHYLOGENY 1919

TABLE I. Continued.

West Indies Central America Venezuela North America Outgroups WI- CA- VE- NA- OG- Position Site ACEFGBHIJK ABCDEFGHIJ AFBCDE ACDEFGIJLMN ABCDEF 16.0 Nd 1111111111 1111111111 l1lll1 11111111111 111111 16.5 AI 1111111111 1111111111 l1lll1 u n n n n i lll1l1 16.8 Sat 1???1???l? ?????????? ?????? ???l??????? lll1l1 16.8 P 0000000000 0000000000 000000 111O1111ll1 000000 17.0 All l11ll11111 1111111111 111111 11111111111 111111 17.0 AI 0000000000 0000000000 000000 00000000000 000100 17.1 All 0000000000 0000100010 000000 00000000000 000000 17.1 C • Other possible location is at 3.2 kb. t Most individuals were not screened for SadI because these two sites are conserved in vertebrates. SadI sites not included in phylogenetic analysis. :\: Other possible location is at 12.9 kb. § Position not certain because fragment lacks internal sites ofother mapped enzymes.

The mapping strategy employed was double Sequence divergence between pairs ofmtDNA digests of all enzymes with BglII and with Clal haplotypes was estimated from cleavage map to produce a crude map. All warbler mtDNAs comparisons using equations (21) and (28) ofNei containeda single Clal site and two or three BglII and Tajima (1983). Estimates ofdivergence were sites. The Clal site appears to be conserved in calculated separately for each type ofenzyme (6 all warblers examined in this study (see below). base, 6-base degenerate, or 5-base degenerate) Additional double and triple digests with other with the final estimate a weighted average ofthese enzyme combinations were then used to deter values (weighted in accordance with the total mine more precisely the relative site positions. number ofbase pairs recognized by each enzyme Maps of yellow warbler and outgroup mt type). DNAs were aligned with the single Clal site. This The matrix of cleavage-site presence/absence site appears to be conserved in warblers, in other was used to generate hypotheses of relationship passerines examined in this lab (including Coer among yellow warbler mtDNAs. Individual hap ebidae: Coereba, and Estrildidae: Vidua, Pytilia, lotypes were the operational taxonomic units Hypargos, and Lagonosticta; Klein pers. obs.), (OTUs). Because vertebrate mtDNA probably is and in the chicken, Gallus gallus (Desjardins and transmitted across generations without recom Morais 1990). This Clal site is found also in bination (Brown 1983, 1985; Moritz et al. 1987), higher primates, rodents, ungulates, and bats there is a direct ancestor-descendent relation (Bibb et al. 1981; Anderson et al. 1982; Carr et ship. Treating haplotypes as OTUs means that al. 1986; Hixson and Brown 1986; Phillips et al. a given population may consist ofmore than one 1991). Homology of the conserved Clal site in OTU, and one OTU can span several popula yellow warblers and all outgroup taxa was con tions. firmed by its constant position relative to two Evolutionary trees were generated by parsi Sadl sites that mark a 1.72-kb fragment and that mony analysis using PAUP version 3.0s (Swof are highly conserved among vertebrate mtDNAs ford 1991). Ten heuristic searches were com (Brown 1985; Carr et al. 1987; Moritz et al. 1987). pleted, using the random addition sequence and One Sadl site is located within the 12S ribo tree bisection-reconnection branch swapping op somal RNA (rRNA) gene, the other in the 16S tions. All six outgroup species were included in rRNAgene (Hixson and Brown 1986; Desjardins the analysis without specifying relationships and Morais 1990). Clal cleaved this 1.72-kb among them. Results presented are those gen fragment into a 1.4-kb and a 0.32-kb fragment erated when haplotype CA-B was excluded from in the mtDNA of all warblers examined, and a the phylogenetic analyses (fig. 3), and those gen Clal site is found in the published chicken erated when haplotypes CA-B and CA-J were mtDNA sequence, between the Sadl sites, 332 excluded (fig. 4). These haplotypes were excluded bp (within the range ofmeasurement error ofthe because each differed from another Central 320 bp fragment found in the warbler mtDNAs) American haplotype by single site losses and their from the Sadl site in the 12S rRNA gene. inclusion resulted in a large increase in the num- 1920 N. K. KLEIN AND W. M. BROWN

TABLE 2. Haplotypes and fragment patterns (indicated by letters) and locality where sampled. Locality abbre viations as in figures I and 2. Enzyme order: AvaI, AvaIl, Bamlii, BelI, BglII, BstEII, ClaI, HindIII, KpnI, NdeI, NheI, Pvull, SpeI.

Haplotype Fragment patterns N Localities (N) West Indies WI-A AAAAAAAAAAAAA 52 DR (6), PR (10), ST (5), SJ (8), SC (8), MO (8), GU (2), DO (I), SL (4) WI-C AGAAAAAAAAAAA 5 DR (1), PR (3), ST (I) CA-F* AAABBAAAABAAA 2 DR (2) WI-E AAAAAAADAAAAA I DR (1) WI-F AAAACAAAAAAAA 6 PR (4), SJ (I), SC (1) WI-G BBABBCABAAADB 11 GU (2), DO (5), MA (4) WI-B BBABBCABCAADB I GU (1) WI-H ADABBAAAAAAAA 2 JA (2) WI-I ACCBBAAAAAAAA 5 JA (5) WI-J ACCBBAAAAAADC 2 JA (2) WI-K CAABBAAAAAAAA I FL (1) Central America CA-A CBABBBABABAAA 11 CR (11) CA-B CBABBBAEABAAA 2 CR (2) CA-C ABABBCABABAAA 8 CR (5), CPa (2), PPa (1) CA-D ABABBCABAAAAA 2 PPa (2) CA-E CHABBCABAAAAA I PPa (I) CA-F* AAABBAAAABAAA 5 CPa (5) CA-G ABABBBABABAAA I CR (1) CA-H AAABBAAAABAAD I CPa (1) CA-I CHABBBABABAAA 1 CR (1) CA-J CBABBCABAAAAA 1 CPa (I) Venezuela VE-A BJBBBCABAAADB 19 Moe (7), Mor (12) VE-B AIABBEAAABAAA 11 Par (6), Anc (5) VE-C BBABBCAAAAADA 2 Par (I), Anc (I) VE-D AIABBEAAADAAA 5 Anc (5) VE-E AAABBEAAABAAA 2 Anc (2) VE-F BJBGBCABAAADB I Moe (1) North America NA-A AFBBBDABBCBFA 22 MI (13), NC (9) NA-C AFBBBDACBCBFA 1 MI (I) NA-D AFBBBDABBCCFA 1 MI (I) NA-E AFBBBDAABCBBA I MI (1) NA-F AEABBDABBCBFA 1 MI (1) NA-G AFBBBCABBIBHA 1 Baja (I) NA-I AFBBBCABBCBHD 2 WA(2) NA-J AKBBBCABBCBHA I WA (1) NA-L AFBBBCABBKBHD I WA (I) NA-M AFBBBCABBLBHA 1 WA(I) NA-N AFBBBCAFBJBHD 1 WA(I) * Ca-F found in West Indies and Central America.

ber ofequally parsimonious trees. Inclusion/ex phyly ofdifferent groups ofhaplotypes. Because clusion ofCA-B and CA-J resulted in essentially using only D. pensylvanica as the outgroup re the same topologies; that is, they were placed in sulted in the same number ofmost parsimonious an unresolved group of Central American hap trees and ingroup topologies as did use ofall six lotypes and branching patterns in the rest ofthe outgroup species together, the tests of'rnonophy tree were unaffected by their presence/absence. ly of different haplotype groups were done with To reduce computer run time, subsequent only D. pensylvanica as the outgroup. searches were therefore conducted without CA-B The statistical significance of differences be and CA-J. These searches were those done under tween the shortest tree and those trees generated various topological constraints to test the mono- under topological constraints was evaluated with YELLOW WARBLER MTDNA PHYLOGENY 1921

TABLE 3. Mean and standard deviation of size esti mates (in kb) ofmitochondrial DNA molecules ofwar bier species.

Species Mean SD Dendroica petechia 17.1 0.13 D. pensylvanica 17.0 0.20 D. discolor 17.0 0.15 Setophaga ruticilla 17.1 0.16 Parula americana 17.0 0.09 Geothlypis trichas 17.0 0.25 Basileuterus culicivorus 17.2 0.24 the nonparametric method ofTempleton (l983b). Because this test cannot be done on consensus FIG. 1. Distribution of North American (NA-) hap lotypes. Numbers refer to these general localities: 1, trees, tree number seven from each constrained Washington (WA); 2, Michigan (MI); 3, North Caro analysis was arbitrarily selected for comparison lina (NC); 4, Baja California (Baja). with tree number seven from the unconstrained analysis. To make inferences about the evolution ofmi Genetic discontinuity on a large geographic gratory behavior within yellow warblers, an ad scale in the yellow warbler is suggested by the ditional parsimony analysis was done, adding regionally limited distributions of most haplo migratory behavior as a character and including types. Although some of the West Indian and all outgroup species. Of the latter, only B. culi Venezuelan haplotypes (WI-G, WI-B, VE-A, VE civorus is sedentary. Because the topology ofthe F) shared more cleavage sites with each other trees summarized in a consensus was essentially than they did with other haplotypes in their re the same as that generated by analysis of DNA spective regions, only one haplotype (CA-F) was data alone, only simplified versions of the trees found in more than one of the four geographic are shown (fig. 5). regions studied (North America, Central Amer ica, Venezuela, and West Indies) (table 2). CA-F REsULTS was found in the West Indies (Dominican Re Haplotype Distributions. - Thirty-seven yel public) and in Central America (on islands off low warbler haplotypes were identified after di the Caribbean coast ofPanamaat Bocas del Toro). gestion with 13 restriction endonucleases (table Haplotypes also often had limited geographic 2). Outgroups were intraspecifically monomor distributions within a region (figs. 1,2), but dif phic with respect to mtDNA haplotypes, but ferences between the most common haplotype sample sizes were small and each species was and others within a region were usually only one represented by single-locality samples only. One or two restriction-site gains/losses. hundred and eighteen restriction sites were sur Phylogenetic Analyses. - When CA-J was in veyed (51-62 per individual), 41 ofwhich were cluded and CA-B excluded from the analysis, in restricted to outgroup species. Twenty-nine sites excess of 3000 shortest trees were generated on were phylogenetically informative within yellow the first replication in PAUP. A subsequent run warblers. The mapped positions ofsites and their was then undertaken in which no more than 100 presence/absence in each haplotype are listed in trees of minimum length were saved from each table 1. replication, resulting in 500 most parsimonious Size estimates ofthe mtDNA molecules were trees, 165 steps each (fig. 3). When CA-J was also similar for all species (table 3), similar to those excluded, 105 trees of 164 steps each were re previously reported for warblers (Kessler and tained after ten replications (fig. 4), savinga max Avise 1985), and within the range reported for imum of 1000 trees per replication. other birds (Shields and Helm-Bychowski 1988). A major split divides the North American Estimated pairwise sequence divergence among (NA-) haplotypes (including NA-G from a trop yellow warbler mtDNAs ranged from 0.14-3.17% ical sedentary population in Baja California) from (mean: 1.5%), and that between mtDNAs from the rest, which are all from tropical sedentary yellow warblers and the other warbler species populations in Venezuela, Central America, and used as outgroups ranged from 2.87 to 8.53%. the West Indies. This North American clade is 1922 N. K. KLEIN AND W. M. BROWN

11 K'.".·...• o 300

~ ~o 0 19 ·~o 0 A,B,G '-12 o / H,I,J 18/ ~I>'" 20 ...... <, A 9-- A,G " <, WI- ~ -, 21...... 22 " -, G (/- A -, ------o \ ...... I" \ \ CA- \ \ \ \ I I / ~.\ I/ VE- \ ,I \ J" I 1 '.

FiG. 2. Distribution ofhaplotypes from tropical" populations. Numbers refer to these general localities: 5, Costa Rica (CR); 6a, Pacific coast; Panama (PPa); 6b, Caribbean coast, Panama (CPa); 7, Venezuela, Mochima (Moe); 8, Venezuela, Morrocoy (Mor); 9, Venezuela, Paraguana (Par); 10, Venezuela, Ancon de Iturre (Anc); 11, Florida Keys (FL); 12, Jamaica (JA); 13, Dominican Republic (DR); 14, Puerto Rico (PR); 15, St. Croix (SC); 16, St. Thomas (ST); 17, St. John (SJ); 18, Montserrat (MO); 19, Guadeloupe (GU); 20, Dominica (DO); 21, Martinique (MA); 22, St. Lucia (SL).

supported by one restriction-site gain and two consisted of two Lesser Antillean haplotypes site losses. Little resolution exists within North (WI-G and WI-B) and three Venezuelan haplo America, except that haplotypes from Washing types (VE-A, VE-F, and VE-C) (figs. 3, 4). An ton state and Baja California form a group sup other, identified in 57% ofthe shortest trees, con ported by two restriction-site gains (fig. 3). sisted of the rest of the West Indian haplotypes Within the tropics, haplotypes from within except CA-F (fig. 4). Within this clade, WI-A regions (Central America, West Indies, Vene (which was the most common and widespread zuela) did not form monophyletic groups. Al of all West Indian haplotypes) formed a group though sequence divergence among Central with three other haplotypes probably derived American haplotypes was low (mean pairwise from it. Haplotypes from Jamaica formed a group divergence of0.56%, range: 0.16-1.14%), two of within this larger West Indian assemblage. the haplotypes (CA-F and CA-H) were united Monophyly of West Indian haplotypes requires with some West Indian and Venezuelan haplo seven extra steps, and the difference between the types to the exclusion ofthe others from Central most parsimonious tree and the constrained tree America (figs. 3, 4). Three extra steps are re is statistically significant (table 4). quired for monophyly ofCentral American hap Venezuelan haplotypes (VE-) were also divid lotypes, but the difference between the most par ed among two clades, the one mentioned above simonious tree and the constrained tree is not (with Lesser Antillean haplotypes), and one with statistically significant (table 4). some Central American haplotypes (identified in West Indian haplotypes (WI-) were divided 71% ofthe trees) (figs. 3,4). Monophyly of'Ven among two clades. One, identified in all trees, ezuelan haplotypes requires seven extra steps and YELLOW WARBLER MTDNA PHYLOGENY 1923

WlA rIl-lHIf--r--:~IWest Indies WIC WIF WIE I------CAF DR,Panama WIF 57 ,..- WIH :' I[arnaica WQ WII 1------WIK Ftcnde Keys WIJ VEO L....- WIK VED Iwest Venez '----VEE ..... CAF '------CAH Panama I------CAA YEB 71 r-----lf------CAC YED I------CAD L....- VEE I------CAE Cent Am '-- CAH I------CAG I------CAI '------CAj 57 --"-J-----::~ I Lesser Antilles VEA --.,.~_ YEF I east Vcnez '------VEC westVenez I------~~ IMich,NC /------NAE r==== NAG Baja ~~~ IWash NAN '-----NAM

I NAF Mich l- --'======D.pm '------other outgroup species

FIG. 3. Strict consensus of 500 most parsimonious trees. Length = 165 steps. Includes haplotype CA-J. CI = 0.48, HI = 0.52 (excluding uninformative char acters), RI = 0.79, RC = 0.48. +/-, gain/loss ofre striction site. White bar, unique synapomorphy; solid black bar, homoplastic synapomorphy; checkered bar, unique synapomorphy relative to ingroup only. 1------D. pen. '------other outgroup species the difference between the constrained tree and FIG. 4. Majority-rule consensus of 105 trees. Length = 164 steps. Excludes haplotype CA-J. Numbers, % of the most parsimonious tree is statistically sig trees with that node. No numbers, node appeared in nificant. 100% oftrees. Cl = 0.48, HI = 0.52 (excluding unin formative characters), RI = 0.79, RC = 0.49. DISCUSSION Intraspecific Genetic Divergence and Hierar biers (0.14-3.17%) also suggests a temporal com chical Structure. - As indicated by the trees ob ponent to intraspecific diversity. Molecular tained from parsimonyanalysis and summarized (DNA) divergence may not be completely metric in the consensus topology (figs. 3, 4), some hap (DeSalle and Templeton 1987; Templeton 1987; lotypes (and populations) within Dendroica pe Vawter and Brown 1986), yet a positive corre techia, share a much more recent common an lation has been demonstrated between genetic cestor than do others. Some other avian species divergence and temporal divergence (time since show such phylogenetic structure (Mack et al. splitting) (Farris 1981; Brown 1983; Wilson et 1986; Shields and Wilson 1987; Avise and Nel al. 1985; Ayala 1986; Moritz et al. 1987). Al son 1989; Van Wagner and Baker 1990; Zink though migratory populations that nest in North 1991; Gill and Slikas 1992; Seutin et al. 1993), America have at times been considered a sepa but many do not (Ball et al. 1988; Shields and rate species (D. aestiva) (Hellmayr 1935) and Helm-Bychowski 1988; Fleischer et al. 1991; should perhaps still be so classified (Ridgely Moore et al. 1991; Ovendon et al. 1991; Zink et 1976), substantial variation in intraspecific di al. 1991; Hare and Shields 1992). vergence is found even among haplotypes sam The range of estimated sequence divergence pled only from sedentary tropical populations values between pairs ofhaplotypes in yellow war- (0.14-2.22%). 1924 N. K. KLEIN AND W. M. BROWN

TABLE 4. Wilcoxon matched-pairs signed ranks test from tropical sedentary populations (including (two-tailed) of a most-parsimonious tree with no to NA-G) form a monophyletic group distinct from pological constraints (67 steps) compared with other hypotheses of relationship (with topological con a monophyletic group ofhaplotypes from North straints). Tree A, Central American haplotypes mono American migratory populations requires six ex phyletic; tree B, West Indian haplotypes monophyletic; tra steps (table 4). Given the conservative nature tree C, Venezuelan haplotypes monophyletic; tree D, ofthe two-tailed test, the difference between the Tropical haplotypes (including NA-G) monophyletic constrained and the most parsimonious tree (P and North American haplotypes (migratory popula tions) monophyletic; tree E, all regions monophyletic. < 0.06) is probably significant. For all analyses, multiple shortest trees ofequal length It is also surprising that NA-G is placed in the were generated. Therefore, one tree from each analysis middle of the North American group, making (arbitrarily chosen to be tree 7) was used for the com the migratory populations paraphyletic. How parisons. All trees compared were those generated us ing only Dendroica pensylvanica as the outgroup. For ever, because monophyly ofthe haplotypes from the two topologies being compared, "No. of sites" is migratory populations (with NA-G as the sister the number ofrestriction sites that undergo the greater taxon to them) requires only one extra step, sup (+) or lesser (-) number ofchanges listed in the cor port for this paraphyletic arrangement is not responding score column; "score" is the greater (+) or lesser (-) number of changes each site undergoes on strong. the alternative tree relative to the most parsimonious Within North America, some hierarchical tree; and rank is the relative rank ofthe scores for each structure exists among haplotypes from migra of those sites. T is the test statistic, the sum of the tory populations of yellow warbler. Haplotypes negative ranks. The absence ofnegative scores or ranks from Washington state (NA-I, J, L, M, N) form for all comparisons suggests the most parsimonious tree is equally or more compatible with the data sets a group (with NA-G) within the larger North than are any ofthe alternative trees, and is significantly American assemblage. North Carolina birds all more compatible for all comparisons except that with shared the same haplotype, which was also the trees A and D. most common one in Michigan (NA-A). Genetic differentiation ofWashington state birds relative No. of Tree Length Sites Score Rank T P to those from Michigan and North Carolina is consistent with their placement in currently rec A 70 3 +1 2 0 > 0.10 B 74 5 +1 3 0 < 0.04 ognized subspecies (D. p. morcomi for Washing 1 +2 6 ton, D. p. aestiva for Michigan and North Car C 74 5 +1 3 0 < 0.04 olina). Mitochondrial DNA markers identifying I +2 6 regional breeding locality may thus exist for tem D 73 4 +1 2.5 0 < 0.06 perate-zone populations ofthis species, although I +2 5 E 86 4 +1 2.5 0 < 0.01 more widespread sampling must be done to ver 4 +2 6.5 ify the possibility. I +3 9 Although all but one of the mtDNA lineages I +4 10 within the tropical distribution are endemic to the regions where sampled, they do not cluster in distinct regional groups. Several alternative Phylogenetic Relationships among Popula hypotheses can explain the lack of regional tions.- The haplotypes sampled form two groups monophyly in yellow warbler mtDNAs: the oc in the parsimony analysis: North American hap currence of past or present gene flow and in lotypes versus most of those from tropical resi trogression of mtDNA (Takahata and Slatkin dent populations. Differentiation of North 1984; Tegelstrom 1987), divergence between re American migratory populations from seden gional populations too recent for mtDNA lin tary, tropical populations is not surprising be eages to be sorted into monophyletic groups (Ta cause their breeding ranges do not overlap, and jima 1983; Avise et a1. 1984; Neigel and Avise gene flow between them is presumably negligible. 1986; Takahata 1989; Wu 1991), or parallel evo It is surprising, however, that the Baja California lution ofmtDNA (i.e., homoplasy ofrestriction haplotype (NA-G) is part ofthis assemblage, since site gains or losses) (Aquadro and Greenberg the population sampled is sedentary and tropical, 1983; Templeton 1983a; Moritz et a1. 1987). The with a plumage pattern nearly identical to that mtDNA data themselves do not allow testing of of Central American sedentary birds and differ the alternative hypotheses for mtDNA geneal ent from that ofNorth American migratory birds. ogic relationships that mayor may not be con Constraining the tree such that all haplotypes cordant with population phylogenies (Ball et a1. YELLOW WARBLER MTDNA PHYLOGENY 1925

1990). Additional data are necessary, for ex bats ofthe species Artibeusjamaicensis (Phillips ample, phylogeny estimates based on additional et al. 1989). The range of estimated sequence independent characters, or examination of geo divergences between Jamaican yellow warbler graphic patterns in phenotypic characters. Such haplotypes and those from other West Indian data exist for some instances of nonmonophyly localities (0.30-1.85%) either suggests that the in yellow warblers (see below), but not for others. population on that island has been isolated from Substantial ongoing gene flow among regions is those on some other islands for a long period or assumed to be negligible because ofthe marked that it was colonized by a different source pop morphological and behavioral differences among ulation than were some other islands. Although birds from the different regions, dispersal bar relationships among Jamaican haplotypes and riers between regions, and (with the exception of those from most Caribbean islands are unre CA-F) the lack ofany mtDNA haplotype being solved for yellow warblers, Jamaican popula sampled from more than one region. tions of many other organisms, including birds, The presence of haplotype CA-F on Hispan have been suggested as the colonization sources iola and on Caribbean islands off the coast of for some other islands (Bond 1963, 1978; Baker Panama (Bocas del Toro) is puzzling because these and Genoways 1978; Brown 1978). During Pleis localities are separated by 1500 km ofopen water tocene glaciations, Jamaica was much closer to and because phenotypic (especially plumage) dif the Central American mainland than at present, ferences between the two populations are marked. and it was the closest of any ofthe Greater An Possible explanations are introgression of tilles because of lower sea levels (Heatwole and mtDNA, retention ofancestral polymorphism in MacKenzie 1967; Buskirk 1985). mtDNA, parallel evolution of mtDNA, or mul The presence in the Lesser Antilles oftwo very tiple colonizations of Hispaniola coupled with divergent groups of haplotypes (WI-A, C, E, F rapid phenotypic evolution and/or selection vs. WI-B, G) that do not cluster as a monophy against males descended from mainland foun letic assemblage suggests multiple colonizations ders. Passage of hurricanes through the region of the West Indian archipelago and of at least could present the opportunity for unusual dis two individual islands. Haplotype WI-A is wide persal and thus introgression of mtDNAs. Re spread throughout the Caribbean region (on 9 of tention of ancestral polymorphism is also pos the 12 islands sampled). Most other haplotypes sible but unlikely for island populations, which sampled from localities where WI-A was found have probably experienced founder effects at col probably were derived from it because these hap onization, as well as post-colonization bottle lotypes (WI-C, WI-E, WI-F) differ from WI-A necks. It is estimated to take 4N (N=number of by only one ortwo restriction-site gains or losses. females in the founding population) generations Haplotype WI-G, from the Lesser Antilles, is for all individuals within a population to trace separated from WI-A by nine restriction-site the origin oftheir mtDNAs to a single ancestral changes (1.35% sequence divergence). WI-B dif haplotype (Tajima 1983; Avise et al. 1984; Nei fers from G by a single autapomorphous restric gel and Avise 1986). If a population bottleneck tion-site loss, and both are similar to the Vene (e.g., founder event) reduces the effective female zuelan haplotypes VE-A and VE-F (0.29-0.58% population size to a small number, time to sequence divergence). In the parsimony analysis monophyly of population mtDNAs would be WI-B and G form a monophyletic group with short and the presence of divergent mtDNAs VE-A and F, not with otherWest Indian lineages. within a population would be unlikely to rep Venezuelan populations (or their ancestors) were resent retention ofancestral polymorphism (Ta thus the most likely colonization sources ofsome jima 1983). It is interesting to note that in 71% ofthe present-day Lesser Antillean populations. ofthe shortest trees, CA-F and CA-H, which are The grouping ofWI-G and WI-B with VE-A and from the Caribbean coast ofPanama, cluster with VE-F and, thus, polyphyly ofWest Indies yellow haplotypes from coastal Venezuela instead of warblers, is supported by three site gains unique those from the Pacific coast ofCentral America. ly synapomorphic within yellow warblers. An Haplotypes from Jamaica (WI-H, WI-I, WI- hypothesis of multiple colonization is also sup J) form a monophyletic group within a larger ported by patterns of geographic variation in West Indian assemblage. Monophyly of Jamai plumage. Plumage color pattern of Martinique can taxa also has been demonstrated for Eleu birds (haplotype WI-G, subspecies D. p. ruficap therodactylus frogs (Hedges 1989), but not for ilia) is very similar to that of Venezuelan pop- 1926 N. K. KLEIN AND W. M. BROWN ulations at Morrocoy and Paraguana (haplotypes relatively high levels ofdivergence between east VE-A, VE-B, and VE-C, subspecies D. p. cie ern and western localities (mean estimated se negae and D. p. paraguanae) and different from quence divergence of1.80%) are (1) a past barrier that ofother West Indian birds (Hellmayr 1935; (vicariance event) divided an ancestral Vene Klein 1992). As discussed above, retention of zuelan population into two geographically iso ancestral polymorphism is an unlikely explana lated units corresponding to the two groups of tion for the presence ofverydivergent haplotypes lineages, which have subsequently diverged ge in the West Indies, because the populations in netically; or(2) an ancestral population was poly volved are on islands colonized by overwater morphic and the polymorphisms have since sort dispersal and, thus, are subject to founder effects ed along geographic lines. The present data do and population bottlenecks. not allow discrimination among these possibil The presence within islands ofhaplotypes be ities. The more basal position of one "western" longing to different clades suggests multiple col haplotype (VE-C) relative to the two "eastern" onizations ofsome individual islands. Both WI-G haplotypes suggests that haplotypes in the east and WI-A are found on Dominica and Guade ern localities may be derived from those in west loupe, the two islands north ofMartinique. Only ern localities. Currently all of these populations WI-A was found on St. Lucia, the adjacent island are isolated from each other by barriers of un south ofMartinique, and on Montserrat, the ad suitable habitat and, because plumage and mor jacent island north ofGuadeloupe. Even though phometric traits differ among them (Wiedenfeld sample sizes were small for St. Lucia (N = 4) and 1991; Klein 1992), present-day gene flow is prob Martinique (N = 4) and both haplotypes may ably negligible. exist on them, this hypothesis of historical re Evolution ofMigration. - Models of the evo lationship among island populations is support lution of migration systems in the New World ed by patterns of plumage variation among the avifauna invoke multiple causative factors (re individuals sampled (Klein 1992). Populations viewed in Cox 1985). All assume migratory be on Dominica and Guadeloupe have been de haviorevolved in sedentary populations, and that scribed as a recognizable taxon (subspecies D. p. each migratory species represents a separate data melanoptera) and many individuals have a point, that is, that migratory behavior evolved plumage pattern intermediate to that of most more than once, even within clades. Optimiza other West Indian birds and that of Martinique tion of migratory behavior on one of the ran birds (Klein 1992). St. Lucia birds (haplotype domly selected shortest trees depicting yellow WI-A) are more similar in appearance to pop warbler relationships (summarized in fig. 5, tree ulations on islands north of Guadeloupe (Klein I) suggests migratory behavior is the plesiorn 1992) and which also possess only WI-A and orphic state in this species and that sedentary closely related haplotypes. Multiple coloniza behavior evolved twice from migratory ances tions ofindividual islands and a similar pattern tors. Optimization on one of the randomly se in which one group ofrelated haplotypes is wide lected trees that is constrained to make migratory spread in the West Indies and another is confined populations form a monophyletic group (one step to the Lesser Antilles also have been demon longer than the shortest trees) suggests either ple strated for bats (Pumo et al. 1988; Jones 1989; siomorphy of migratory behavior with multiple Phillips et al. 1989). origins ofsedentary behavior (fig. 5, tree 2b), or In the phylogenetic analysis, Venezuelan hap a single origin ofmigratory behavior from a sed lotypes (VE-) also clustered into two groups. Each entary ancestor within yellow warblers (fig. 5, group formed a clade with a different set ofnon tree 2a). Venezuelan haplotypes, some ofwhich were de Caribbean Biogeography.- Although much has rived relative to those from Venezuela. Vene been published on the historical biogeography of zuelan haplotypes were thus polyphyletic and nonavian taxa in the Caribbean (Rosen 1975; paraphyletic. Thegroups are (I) haplotypes (VE-A Guyer and Savage 1986; Kluge 1988; Rauch and VE-F) from the two easternmost localities enberger 1988; Burgers and Franz 1989; Phillips (Mochima and Morrocoy) plus one (VE-C) from et al. 1989; Woods 1989; Hass 1991), compar the westernmost localities (Paraguana and An atively little has been written about the origins con de Iturre), and (2) the remaining haplotypes of West Indian bird species (Bond 1963, 1978; (VE-B, VE-D, and VE-E) from Paraguana and Cruz 1974; Olson 1976, 1978). Many exami Ancon de Iturre, Possible explanations for the nations of ecological factors and their effects on YELLOW WARBLER MTDNA PHYLOGENY 1927

bird distributions exist (Ricklefs and Cox 1972, 1978; Lack 1973, 1976; Terborgh et al. 1978; 1------troptcat Amertca Faaborg 1985), but no explicit hypotheses ofhis torical (phylogenetic) relationships among West Indian avian taxa have yet been published (but see McDonald and Smith 1990). A historical per ,- North America spective is essential for understanding present day distributions (Pregill and Olson 1981; Endler Baja Callfornta 1982). Although vicariance models have been pro North Amerl ca posed for the origins ofWest Indian taxa (Rosen 1- D. pensylvanlca 1975; Kluge 1988), these probably have little bearing on the distribution of yellow warblers, L- other outgroups as well as other passerines in the region. Fossil evidence for the time oforigin ofpasserine birds is scanty, but the group is thought to have ap peared at the beginning of the Miocene (Brod 2a korb 1971). The West Indies are thought to have assumed their present positions and configura ,- tropical America tion by the beginning ofthe Miocene (Donnelly 1988), leaving little possibility for a vicariant origin for taxonomically differentiated popula tions of warblers. .----__------North AmerIca Previous hypotheses for the origins of West Indian birds have been based on subjective anal L- Baja Callfornta yses of similarities among extant mainland and island taxa (Bond 1963, 1978), as well as offos silized remains (Olson 1978). Bond (1963, 1978) f------D. pensylvanlca postulatedtwo main sources ofcolonization: from L- other outgroups South America to the Lesser Antilles, and from North America (including Mexico and Central America) to the Greater Antilles via Jamaica or Cuba. These colonization routes have been pro 2b posed for nonavian taxa as well (Baker and Gen oways 1978; Brown 1978; Koopman 1989). Col 1------tropIcal AmerIca onization ofthe Greater Antilles by some North American avian taxa is also supported by pale ontological evidence (Olson 1976, 1978). The phylogenetic hypothesis based on yellow warbler .---- North America mtDNAs is consistentwith the model ofmultiple colonizations, and with colonization sources f------Baja CalifornIa proposed by Bond (1963, 1978) for widespread Caribbean birds, that is, colonization of some 1- D. pensylvanlca Lesser Antilles by Venezuelan birds, and colo nization ofthe GreaterAntilles by Central Amer L- otheroutgroups ican birds. Documentation of multiple colonizations of FIG. 5. Optimization of migratory behavior on tree topology (simplified for illustration) generated from archipelagoes in general, and the West Indies in analysis ofrestriction site and migratory behaviorchar particular, has several important implications. acters. Black bars, gain of migratory behavior. White bars, loss of behavior. Tree I: ACCTRAN and DEL TRAN optimization on shortest tree; Tree 2a: ACCT RAN optimization on tree one step longer than shortest t- tree; Tree 2b: DELTRAN optimization on tree one Basileuterus are migratory; all North American pop step longer than shortest tree. All outgroups except ulations are migratory except that in Baja California. 1928 N. K. KLEIN AND W. M. BROWN

The most general is the effect of multiple colo does notallow direct testing ofthestepping-stone nizations (and thus the potential for introgres colonization hypothesis for those islands. Nev sion ofcharacters) on phylogenetic inference. The ertheless, because yellow warbler populations on problem is particularly acute when the characters some adjacent islands are not each others' closest sampled are portions of DNA that are nonre relatives (e.g., St. Lucia vs. Martinique), our study combining and uniparentally transmitted across suggests a stepping-stone model is not always generations (Doyle 1993). The "foreign" mt correct, and this has implications for studies in DNA can be introgressed into a host population ecology and evolution. In tracing the evolution (Ferris et aI. 1983; Powell 1983; Tegelstrom ofparticular traits, the usual implicit assumption 1987), and this can occur without accompanying is that populations on adjacent islands are each introgression of phenotypic traits (Smith 1992). others' closest relatives and that interisland phe Through the stochastic lineage sorting process notypic variation therefore is due to different se (Tajima 1983; Avise et aI. 1984; Neigel and Av lection pressures on different islands, or to sto ise 1986), or through the active process ofselec chastic processes like genetic drift; that is, phe tion (Clark and Lyckegaard 1988), the intro notypic variation is assumed to have developed gressed DNA may become fixed within the host in situ (Ricklefs and Cox 1972). Previous authors population. Use ofthe introgressed characters to (Hellmayr 1935; Bond 1939), for example, as infer phylogeny may thus result in an erroneous sumed that Martinique yellow warblers were most reconstruction ofhistorical relationships among closely related to those on neighboring islands taxa. Although we were able to infer multiple and that the striking plumage differences among colonizations of individual islands by yellow island populations evolved in place. As indicated warblers, and have evidence ofintrogression (in by the phylogenetic relationships ofyellow war termediate phenotype), one's future ability to re bler mtDNAs sampled from these islands, this cover this history would be compromised ifone assumption is incorrect. This is not to say, how or the other of the "ancestral" haplotypes on ever, that stepping-stone colonization has not Dominica or Guadeloupe were to become lost occurred in the West Indies, only that the as from these populations. Whether other species sumption shouldbe tested in any individual case. have invaded the Caribbean archipelago multi ple times remains to be investigated. ACKNOWLEDGMENTS Documentation ofmultiple colonizations also Funding for this research was provided by a has implications for inferences about avian bio National Science Foundation Dissertation Re geography in the Caribbean. Ricklefs and Cox search grant BSR-881544S, the Frank M. Chap (1972, 1978) interpret patterns ofgeographic dis man Memorial Fund ofthe American Museum tribution and taxonomic differentiation ofWest ofNatural History, a Collection Study Grant from Indian birds as reflecting the passage of species the American Museum of Natural History, the through stages ofa taxon cycle. Widespread, un Sigma XI Scientific Research Society, the Hins differentiated species are suggested to be in the dale and Walker Fellowships of the Museum of beginning stage ofthe cycle, and endemic, well Zoology, University of Michigan, Block Grants differentiated species are considered to be at the from the Department of Biology, University of end ofthe cycle. They suggest the ultimate cause Michigan, a Dissertation Research Grant from ofsuch a cycle is competition-driven, increasing Rackham Graduate School, University ofMich specialization correlated with time since colo igan, and the Bird Division, Museum ofZoology, nization ofthe archipelago. However, any test of University of Michigan. this model must assume monophyly and thus a We thank the following for granting research/ single colonization of the West Indies for the collecting permits or for assistance and cooper species being investigated. ation in acquiring specimens: M. Sc. Guillermo Traditional avian biogeographic hypotheses for Canessa Mora, Servicio de Vida Sylvestre, Min island faunas have assumed a stepping-stone isterio de Recursos Naturales, Energia, y Minas, method of dispersal as most likely (Darlington Costa Rica; Dr. Giuseppe Colonello, Fundaci6n 1938; MacArthur and Wilson 1967; Ricklefs and La Salle de Ciencias Naturales, Venezuela; Ga Cox 1978), with colonization of an archipelago briel Charles, Department ofForestry, St. Lucia; occurring only once and to the island nearest the M. Benito-Espinal, Ministere de I'Equipement, mainland. The lack ofresolution ofrelationships du Logement, de l'Amenagement du Territoire among most yellow warbler island populations et des Transport, Ministere Charge de l'Envi- YELLOW WARBLER MTDNA PHYLOGENY 1929 ronnement, Guadeloupe, French West Indies; chondrial DNA polymorphisms among channel is Loic Beroud, Prefecture de la Martinique, Di land deer mice. Evolution 41:854-863. rection Departementale de l'Agriculture, Mar Avise, J. c., and W. S. Nelson. 1989. Molecular ge netic relationships ofthe extinct dusky seaside spar tinique, French West Indies; Felix Gregoire and row. Science 243:646-648. Colmore Christian, Forestry Division, Ministry Avise, J. c., J. E. Neigel, and J. Arnold. 1984. De of Agriculture, Commonwealth of Dominica; mographic influences on mitochondrial DNA lin Ministry of Agriculture, Trade, Lands, and eage survivorship in animal populations. Journal of Molecular Evolution 20:99-105. Housing, Montserrat; Alan Smith and David Avise, J. c., J. C. Patton, and C. F. Aquadro. 1980. Nellis, Department ofPlanning and Natural Re Evolutionary genetics of birds. III. Comparative sources, Division ofFish and Wildlife, U.S. Vir molecular evolution in New World warblers (Pa gin Islands; Emilio A. Bautista, Cecilia Hernan rulidae) and rodents (Cricetidae). Journal of He dez, and Tom s Vargas, Departamento de Vida redity 71:303-310. Ayala, F. J. 1986. On the virtues and pitfalls of the Silvestre, Secretaria de Estado de Agricultura, molecular evolutionary clock. Journal of Heredity Dominican Republic; Yvette Strong and Dr. So 77:226-235. mesh Sinha, Natural Resources and Conserva Baker, R. J., and H. H. Genoways. 1978. Zoogeog tion Division, Ministry ofAgriculture, Jamaica; raphy ofAntillean bats. Pp. 53-97 in F. B. Gill, ed. Zoogeography in the Caribbean. Proceedings ofthe Dr. Eduardo Cardona, Departamento de Recur Academy ofNatural Sciences, Philadelphia Special sos Naturales, Puerto Rico; the U.S. Fish and Publication 13. Wildlife Service; the Michigan Department of Ball, R. M., S. Freeman, F. C. James, E. Bermingham, Natural Resources; and the Division ofWildlife, and J. C. Avise. 1988. Phylogeographic structure Florida Game and Fresh-water Fish Division. ofred-winged blackbirds assessed by mitochondrial DNA. Proceedings ofthe National Academy ofSci We thank all those who have helped with field ences, USA 85:1558-1562. work, especially M. Bon Saint Come, E. Benito Ball, R. M., J. E. Neigel, and J. C. Avise. 1990. Gene Espinal, C. Gauveca, J. Gerwin, P. Hautcastel, genealogies within the organismal pedigrees ofran Y. Klein, S. Lanyon, P. Lau, G. Rosenberg, K. dom-mating populations. Evolution 44:360-370. Barrowclough, G. F., and K. W. Corbin. 1978. Ge Rosenberg, T. Schulenberg, T. Vargas, and J. netic variation and differentiation in the Parulidae. Wunderle. Specimens from Panama were pro Auk 95:691-702. vided by M. Braun, G. Morton, T. Parsons, and Berry, R. J. 1983. Diversity and differentiation: the D. Wiedenfeld; T. Peterson provided the speci importance ofisland biology for general theory. Oi kos 41:523-529. men from Baja California; S. Rohwer and C. Bibb, M. J., R. A. Van Etten, C. T. Wright, M. W. Wood provided specimens from Washington Walberg, and D. A. Clayton. 1981. Sequence and state. The computer program for calculating es gene organization of mouse mitochondrial DNA. timated sequence divergence was provided by R. Cell 26:167-180. Hagen. We thank P. Chu, J. Cracraft, M. Gach, Bledsoe, A. H. 1988. Nuclear DNA and evolution and phylogeny of the New World nine-primaried N. Johnson, F. Kraus, A. Larson, M. C. Mc oscines. Auk 105:504-515. Kitrick, P. Myers, R. B. Payne, G. Smith, and Bond, J. 1939. Notes on birds from the West Indies W. H. Wagner for their critical reviews of pre and other Caribbean islands. Notulae Naturae 13: vious drafts of the manuscript. This paper was 1-6. --. 1948. Origin of the bird fauna of the West submitted by N. Klein in partial fulfillment of Indies. Wilson Bulletin 60:207-229. the requirements for the Ph.D. in the Depart ---. 1961. Sixth supplement to the check-list of ment ofBiology, University ofMichigan. Klein's birds ofthe West Indies (1956). Academy of Nat dissertation chair, R. B. Payne, was especially ural Sciences, Philadelphia. helpful throughout all phases of this research. --. 1963. Derivation of the Antillean avifauna. Proceedings of the Academy of Natural Sciences, Philadelphia 115:79-98. LITERATURE CITED ---. 1978. Derivations and continental affinities Anderson, S., M. H. L. deBruijn, A. R. Coulson, J. C. of Antillean birds. Pp. 119-128 in F. B. Gill, ed. Eperon, F. Sanger, and I. G. Young. 1982. Com Zoogeography in the Caribbean. Proceedings ofthe plete sequence ofbovine mitochondrial DNA: con Academy ofNatural Sciences, Philadelphia Special served features of the mammalian mitochondrial Publications 13. genome. Journal of Molecular Biology 156:684 ---. 1979. Derivations ofLesser Antillean birds. 717. Proceedings of the Academy of Natural Sciences, Aquadro, C. F., and B. D. Greenberg. 1983. Human Philadelphia 131:89-103. mitochondrial DNA variation and evolution: anal ---. 1993. A field guide to the birds of the West ysis of nucleotide sequences from seven individu Indies. Houghton Mifflin, New York. als. Genetics 103:287-312. Brodkorb, P. 1971. Origin and evolution of birds. Ashley, M., and C. Wills. 1987. Analysis of mito- Avian Biology 1:19-55. 1930 N. K. KLEIN AND W. M. BROWN

Brown, F. M. 1978. The origins ofthe West Indian olution among four minnow species (Notropis: Cy butterfly fauna. Pp, 5-30 in F. B. Gill, ed. Zooge prinidae). Systematic Zoology 38:126-143. ography in the Caribbean. Proceedings ofthe Acad Dowling, T. E., C. Moritz, and J. D. Palmer. 1990. emy ofNatural Sciences, Philadelphia Special Pub Nucleic acids II: Restriction site analysis. Pp. 250 lications 13. 317 in D. M. Hillis and C. Moritz, eds. Molecular Brown, W. M. 1980. Polymorphism in mitochon systematics. Sinauer, Sunderland, Mass. drial DNA of humans as revealed by restriction Doyle, J. J. 1993. DNA, phylogeny, and the flowering endonuclease analysis. Proceedings ofthe National of plant systematics. BioScience 43:380-389. Academy ofSciences, USA 77:3605-3609. Eldredge, N., and J. Cracraft. 1980. Phylogenetic pat ---. 1983. Evolution of animal mitochondrial terns and the evolutionary process. Columbia Uni DNA. Pp. 62-88 in M. Nei and R. K. Koehn, eds. versity Press, New York. Evolution ofgenes and proteins. Sinauer, Sunder Endler, J. A. 1982. Alternative hypotheses in bio land, Mass. geography: Introduction and synopsis of the sym ---. 1985. The mitochondrial genome ofanimals. posium. American Zoologist 22:349-354. Pp. 95-130 in R. J. MacIntyre, ed. Molecular evo Faaborg, J. 1985. Ecological constraints on West In lutionary genetics. Plenum, New York. dian bird distributions. Pp. 621-653 in P. A. Buck Brown, W. M., and J. Vinograd. 1974. Restriction ley, M. S. Foster, E. S. Morton, R. S. Ridgely, and endonuclease cleavage maps of animal mitochon F. G. Buckley, eds. Neotropical ornithology. Or drial DNAs. Proceedings ofthe National Academy nithological Monographs 35. ofSciences, USA 71:4617-4621. Farris, J. S. 1981. Distance data in phylogenetic anal Burgers, G. H. T., and R. Franz. 1989. Zoogeography ysis. Advances in Cladistics 1:3-24. ofthe Antillean freshwater fish fauna. Pp. 263-304 Feduccia, A. 1977. A model for the evolution of in C. A. Woods, ed. Biogeography of the West In perching birds. Systematic Zoology 26:19-31. dies: past, present, and future. Sandhill Crane Press, Feduccia, A., and L. D. Martin. 1976. The Eocene Gainesville, Fla. zygodactyl birds of North America (Aves: Pici Buskirk, R. E. 1985. Zoogeographic patterns and tec formes). Smithsonian Contributions to Paleobiol tonic history of Jamaica and the northern Carib ogy 27:101-1 10. bean. Journal ofBiogeography 12:445-461. Ferris, S. D., R. D. Sage, C. Huang, J. T. Nielsen, U. Carr, S. M., S. W. Ballinger, J. N. Derr, L. H. Blan Ritte, and A. C. Wilson. 1983. Flow of mito kenship, andJ. W. Bickham. 1986. Mitochondrial chondrial DNA across a species boundary. Pro DNA analysis of hybridization between sympatric ceedings ofthe National Academy ofSciences, USA white-tailed deer and mule deer in west Texas. Pro 80:2290-2294. ceedings ofthe National Academy ofSciences, USA Fleischer, R. c., S. I. Rothstein, and L. S. Miller. 199I. 83:9576-9580. Mitochondrial DNA variation indicates gene flow Carr, S. M., A. J. Brothers, and A. C. Wilson. 1987. across a zone ofknown secondary contact between Evolutionary inferences from restriction maps of two subspecies ofthe brown-headed cowbird. Con mitochondrial DNA from nine taxa of Xenopus dor 93:185-189. frogs. Evolution 41:176-190. Gill, F. B., and B. Slikas. 1992. Patterns of mito Clark, A. G., and E. M. S. Lyckegaard. 1988. Natural chondrial DNA divergence in North American selection with nuclear and cytoplasmic transmis crested titmice. Condor 94:20-28. sion. III. Joint analysis ofsegregation and mtDNA Guyer, c., and J. M. Savage. 1986. Cladistic rela in Drosophila melanogaster. Genetics 118:471-481. tionships among anoles (Sauria: Iguanidae). Sys Cox, G. W. 1985. The evolution ofavian migration tematic Zoology 35:509-531. systems between temperate and tropical regions of Hare, M. P., and G. F. Shields. 1992. Mitochondrial the New World. American Naturalist 126:451-474. DNA variation in the polytypic Alaskan song spar Cruz, A. 1974. Distribution, probable evolution, and row. Auk 109:126-132. fossil record of West Indian woodpeckers (Family Hass, C. A. 1991. Evolution and biogeography of Picidae). Caribbean Journal of Science 14:183-188. West Indian Sphaerodactylus (Sauria: Gekkoni Darlington, P. J. 1938. The origin ofthe fauna ofthe dae): a molecular approach. Journal of Zoology, Greater Antilles, with discussion ofdispersal ofan London 225:525-561. imals over water and through air. Quarterly Review Heatwole, H., and F. MacKenzie. 1967. Herpeto ofBiology 13:274-300. geography ofPuerto Rico. IV. Paleogeography, fau Desjardins, P., and R. Morais. 1990. Sequence and nal similarity and endemism. Evolution 21:429 gene organization ofthe chicken mitochondrial ge 438. nome. Journal ofMolecular Biology 212:599-634. Hedges, S. B. 1989. Evolution and biogeography of DeSalle, R., and A. R. Templeton. 1987. Founder West Indian frogs of the genus Eleutherodactylus: effects accelerate the rate of mitochondrial DNA Slow-evolving loci and the major groups. Pp. 305 evolution in Hawaiian Drosophila. Evolution 42: 370 in C. A. Woods, ed. Biogeography ofthe West 1078-1084. Indies: past, present, and future. Sandhill Crane Donnelly, T. W. 1988. Geologic constraints on Ca Press, Gainesville, Fla. ribbean biogeography. Pp. 15-37 in J. K. Liebherr, Hellmayr, C. E. 1935. Birds of the Americas. Field ed. Zoogeography of Caribbean insects. Cornell Museum of Natural History Publications in Zool University Press, Ithaca, N.Y. ogy Vol. 13, part 8, pp. 363-385. Field Museum of Dowling, T. E., and W. M. Brown. 1989. Allozymes, Natural History, Chicago. mitochondrial DNA, and levels ofphylogenetic res- Hixson, J. E., and W. M. Brown. 1986. A comparison YELLOW WARBLER MTDNA PHYLOGENY 1931

ofsmall ribosomal RNA genes from the mitochon vance for population biology and systematics. An drial DNA ofthe great apes and humans: Sequence, nual Review of Ecology and Systematics 18:269 structure, evolution, and phylogenetic implica 292. tions. Molecular Biology and Evolution 3:1-18. Nei, M., and F. Tajima. 1983. Maximum likelihood Jones, J. K., Jr. 1989. Distribution and systematics estimation of the number of nucleotide substitu ofbats in the Lesser Antilles. Pp. 645-660 in C. A. tions from restriction site data. Genetics 105:207 Woods, ed. Biogeography ofthe West Indies: past, 217. present, and future. Sandhill Crane Press, Gaines Neigel, J. E., and J. C. Avise. 1986. Phylogenetic ville, Fla. relationships ofmitochondrial DNA under various Kessler, L. G., and J. C. Avise. 1985. A comparative demographic models ofspeciation. Pp, 515-534 in description of mitochondrial DNA differentiation E. Nevo and S. Karlin, eds. Evolutionary processes in selected avian and other vertebrate genera. Mo and theory. Academic Press, New York. lecular Biology and Evolution 2:109-125. Olson, S. L. 1976. Oligocene fossils bearing on the Klein, N. K. 1992. Historical processes and the evo origins ofthe Todidae and Momotidae (Aves: Cor lution ofgeographic variation patterns in the yellow aciiformes). Pp. 111-119 in S. L. Olson, ed. Col warbler (Dendroica petechia). Ph.D. diss. Univer lected papers in avian paleontology honoring the sity ofMichigan, Ann Arbor. 90th birthday ofAlexander Wetmore. Smithsonian Kluge, A. G. 1988. Parsimony in vicariance bioge Contributions to Paleobiology 27. ography: A quantitative method and a Greater An ---. 1978. A paleontological perspective ofWest tillean example. Systematic Zoology 37:315-328. Indian birds and mammals. Pp. 99-117 in F. B. Koopman, K. F. 1989. A review and analysis of the Gill, ed. Zoogeography in the Caribbean. Proceed bats of the West Indies. Pp, 635-644 in C. A. Woods, ings of the Academy of Natural Sciences, Phila ed. Biogeography ofthe West Indies: past, present, delphia Special Publications 13. and future. Sandhill Crane Press, Gainesville, Fla. ---. 1985. The fossil record of birds. Avian Bi Lack, D. 1973. The numbers and species of hum ology 8:80-238. mingbirds in the West Indies. Evolution 27:326 Ovendon, J. R., A. Wust-Saucy, R. Bywater, N. Broth 337. ers, and R. W. G. White. 1991. Genetic evidence --. 1976. Island biology. Blackwell Scientific, for philopatry in a colonially nesting seabird, the Oxford. Fairy Prion (Pachyptila turtur). Auk 108:688-694. Lansman, R. A., R. O. Shade, J. F. Shapira, and J. C. Phillips, C. J., D. E. Pumo, H. H. Genoways, and P. Avise. 1981. The use ofrestriction endonucleases E. Ray. 1989. Caribbean island zoography: A new to measure mitochondrial DNA sequence related approach using mitochondrial DNA to study Neo ness in natural populations. III. Techniques and tropical bats. Pp. 661-684 in C. A. Woods, ed. potential applications. Journal of Molecular Evo Biogeography of the West Indies: past, present, and lution 17:214-226. future. Sandhill Crane Press, Gainesville, Fla. MacArthur, R. H., and E. O. Wilson. 1963. An equi Phillips, C. J., D. E. Pumo, H. H. Genoways, P. E. librium theory of insular biogeography. Evolution Ray, andC. A. Briskie. 1991. Mitochondrial DNA 17:373-387. evolution and phylogeography in two Neotropical ---. 1967. The theory of island biogeography. bats, Artibeus jamaicensis and Artibeus Iituratus. Princeton University Press, Princeton, N.J. Pp. 97-123 in M. A. Mares and D. J. Schmidly, Mack, A. L., F. B. Gill, R. Colburn, and C. Spolsky. eds. Latin American mammalogy: history, biodi 1986. Mitochondrial DNA: a source of genetic versity, and conservation. University ofOklahoma markers for studies of similar passerine bird spe Press, Norman. cies. Auk 103:676-681. Powell, J. R. 1983. Interspecific cytoplasmic gene Maddison, W. P., M. J. Donoghue, and D. R. Mad flow in the absence of nuclear gene flow: Evidence dison. 1984. Outgroup analysis and parsimony. from Drosophila. Proceedings of the National Systematic Zoology 33:83-103. Academy ofSciences, USA 80:492-495. McDonald, M. A., and M. H. Smith. 1990. Specia Pregill, G. K. 1981. An appraisal of the vicariance tion, heterochrony, and genetic variation in His hypothesis of Caribbean biogeography and its ap paniolan Palm-tanagers. Auk 107:707-717. plication to West Indian terrestrial vertebrates. Sys Mengel, R. M. 1964. The probable history ofspecies tematic Zoology 30:147-155. formation in some northern wood warblers (Pa Pregill, G. K., and S. L. Olson. 1981. Zoogeography rulidae). Living Bird 3:9-44. ofWest Indian vertebrates in relation to Pleistocene Miller, L. D., and J. Y. Miller. 1989. The biogeog climatic cycles. Annual Review ofEcology and Sys raphy of West Indian butterflies (Lepidoptera: Pa tematics 12:75-98. pilionoidea, Hesperioidea): a vicariance model. Pp. Pumo, D. E., E. Z. Goldin, B. Elliot, C. J. Phillips, and 229-262 in C. A. Woods, ed. Biogeography of the H. H. Genoways. 1988. Mitochondrial DNA West Indies: past, present, and future. Sandhill Crane polymorphism in three Antillean island popula Press, Gainesville, Fla. tions of fruit bat, Artibeus jamaicensis. Molecular Moore, W. S., J. H. Graham, and J. T. Price. 1991. Biology and Evolution 5:79-89. Mitochondrial DNA variation in the Northern Raikow, R. J. 1978. Appendicular myology and the FlickeriColaptes auratus, Aves). Molecular Biology relationships ofthe New World nine-primaried os and Evolution 8:327-344. cines (Aves: Passeriformes). Bulletin of the Car Moritz, c., T. E. Dowling, and W. M. Brown. 1987. negie Museum of Natural History 7:1-43. Evolution of animal mitochondrial DNA: Rele- Rauchenberger, M. 1988. Historical biogeography of 1932 N. K. KLEIN AND W. M. BROWN

poeciliid fishes in the Caribbean. Systematic Zo tion endonuclease cleavage site maps with partic ology 37:356-365. ular reference to the evolution of humans and the Ricklefs, R. E., and G. W. Cox. 1972. Taxon cycle apes. Evolution 37:221-244. in the West Indian avifauna. American Naturalist ---. 1987. Genetic systems and evolutionary rates. 106:195-219. Pp. 218-234 in K. F. S Campbell and M. F. Day, ---. 1978. Stage of taxon cycle, habitat distri eds. Rates ofevolution. Australian Academy ofSci bution, and population density in the avifauna of ences, Canberra. the West Indies. American Naturalist 112:875-895. Terborgh, J. W., J. Faaborg, and H. J. Brockmann. Ridgely, R. S. 1976. A Guide to the birds ofPanama. 1978. Island colonization by Lesser Antillean birds. Princeton University Press, Princeton, N.J. Auk 95:59-72. Rosen, D. E. 1975. A vicariance model ofCaribbean Van Wagner, C. E., and A. J. Baker. 1990. Associ biogeography. Systematic Zoology 24:431-464. ation between mitochondrial DNA and morpho Seutin, G., J. Brawn, R. E. Ricklefs, and E. Bermingh logical evolution in Canada Geese. Journal ofMo am. 1993. Genetic divergence among populations lecular Evolution 31:373-382. of a tropical passerine, the streaked saltator (Sal Vawter, L., and W. M. Brown. 1986. Nuclear and tator albicollis). Auk 110:117-126. mitochondrial DNA comparisons reveal extreme Shields, G. F., and K. M. Helm-Bychowski. 1988. rate variation in the molecular clock. Science 234: Mitochondrial DNA ofbirds. Current Ornithology 194-196. 5:273-295. Wiedenfeld, D. A. 1991. Geographical morphology Shields, G. F., and A. C. Wilson. 1987. Subspecies of male yellow warblers. Condor 93:712-723. ofthe Canada Goose (Branta canadensis) have dis Wiley, E. O. 1981. Phylogenetics: the theory and tinct mitochondrial DNA's. Evolution41:662-666. practice of phylogenetic systematics. Wiley, New Sibley, C. G., and J. E. Ahlquist. 1990. Phylogeny York. and classification of birds: a study in molecular Williams, E. E. 1989. Old problems and new oppor evolution. Yale University Press, New Haven, Conn. tunities in West Indian biogeography. Pp, 1-46 in Smith, G. R. 1992. Introgression in fishes: signifi C. A. Woods, ed. Biogeography ofthe West Indies: cance for paleontology, cladistics, and evolutionary past, present, and future. Sandhill Crane Press, rates. Systematic Biology 41:41-59. Gainesville, Fla. Swofford, D. L. 1991. PAUP: Phylogenetic analysis Wilson, A. c., R. L. Cann, S. M. Carr, M. George, U. using parsimony, Version 3.0. Computer program Gyllensten, K. Helm-Bychowski, R. G. Higuchi, S. distributed by the Illinois Natural History Survey, R. Palumbi, E. M. Prager, R. Sage, and M. Sto Champaign, Ill. neking. 1985. Mitochondrial DNA and two per Swofford, D. L., and G. J. Olsen. 1990. Phylogeny spectives on evolutionary genetics. Biological Jour reconstruction. Pp. 41-50 in D. M. Hillis and C. nal of the Linnean Society 26:375-400. Moritz, eds. Molecular systematics. Sinauer, Sun Woods, C. A. 1989. The biogeographyofWest Indian derland, Mass. rodents. Pp. 741-798 in C. A. Woods, ed. Bioge Tajima, F. 1983. Evolutionary relationship ofDNA ography of the West Indies: past, present, and fu sequences in finite populations. Genetics 105:437 ture. Sandhill Crane Press, Gainesville, Fla. 460. Wright, J. M., C. Spolsky, and W. M. Brown. 1983. Takahata, N. 1989. Gene genealogy in three related The origin of the parthenogenetic lizard Cnemi populations: consistency probability between gene dophorus laredoensis inferred from mitochondrial and population trees. Genetics 122:957-966. DNA analysis. Herpetologica 39:410-416. Takahata, N., and M. Slatkin. 1984. Mitochondrial Wu, C.-I. 1991. Inferences of species phylogeny in gene flow. Proceedings ofthe National Academy of relation to segregation of ancient polymorphisms. Sciences, USA 81:1764-1767. Genetics 127:429-435. Tegelstrom, H. 1987. Transfer ofmitochondrial DNA Zink, R. M. 1991. The geography of mitochondrial from the northern Red-backed Vole (Clethriono DNA variation in two sympatric sparrows. Evo mys rutilus) to the Bank Vole (c. glareolus). Journal lution 45:329-339. of Molecular Evolution 24:218-227. Zink, R. M., W. L. Rootes, and D. L. Dittman. 1991. Templeton, A. R. 1983a. Convergent evolution and Mitochondrial DNA variation, population struc nonparametric inferences from restriction data and ture, and evolution of the common grackle (Quis DNA sequences. Pp. 151-179 in B. S. Weir, ed. calus quiscula). Condor 93:318-329. Statistical analysis ofDNA sequence data. Marcel Dekker, New York. ---. 1983b. Phylogenetic inference from restric- Corresponding Editor: A. Larson