386 ShortCommunications [Auk,Vol. 105

GeneticDifferentiation BetweenNorth AmericanKinglets and Comparisons with Three Allied Passerines

JAMESL. INGOLD,• LEE A. WEIGT, AND SHELDONI. GUTTMAN Departmentof Zoology,Miami University,Oxford, Ohio 45056 USA

The genusRegulus is composedof five species,two Rogers'genetic distance (Wright 1978)values (Fig. 1). of which are native to the Western Hemisphere We alsoanalyzed the allozymesas charactersto avoid (Clements1978). Mayr and Short (1970) discussedthe the problems and lossof information associatedwith possible relationshipsbetween the Ruby-crowned reducingelectrophoretic data setsto distancecoeffi- Kinglet (R. calendula)and the Golden-crownedKing- cients(Farris 1981,Felsenstein 1984). Branch lengths let (R. satrapa).They suggestedthat the Golden- of cladogramsderived in this manner have biological crowned Kinglet is most closelyrelated to the Gold- meaning. There are several ways to code and order crest (R. regulus)of the Palearcticfaunal region and allozyme characterstates, however, and no general that the Ruby-crownedKinglet is not closelyrelated concensusexists on the most appropriate approach to any of the other speciesof kinglet, even though it (reviewed by Buth 1984).We usedthe alleles as char- has hybridized with the Golden-crownedKinglet acters with the character statesbeing "presence" or (Gray 1958).We presentgenetic evidence that the two "absence";character coding in this manner acknowl- North American kinglets are not closelyrelated. edgesthe presence(or absence)of alleles rather than The birds usedin this study were mist-nettednear particular suites of alleles. The character-statedata Oxford,Butler Co., Ohio, and were collectedfor part were analyzedwith the PhylogeneticAnalysis Using of a larger studyon the historyof the North American Parsimony (PAUP) provided by Swofford (1984). avifauna. Yellow-breasted Chats (Icteriavirens; n = 7) Character stateswere weighted such that each locus and Common Yellowthroats (Geothlypistrichas; n = provided equal information; the tree (Fig. 2) was root- 22) were collected during late summer and autumn ed to the House Wren as the designatedancestor. 1980. Ruby-crowned Kinglets (n = 10), Golden- Only one locus, nonspecific protein (NSP), was crowned Kinglets (n = 14), and House Wrens (Trog- monomorphicacross all species(Table 1). The Yellow- lodytesaedon; n = 7) were collected during autumn breasted Chat was fixed for a different allele for AAT-1 1981.The birdswere broughtto the laboratoryalive, and AAT-2 than the other four species.The House were killed, and then were kept frozen whole at -70øC Wren was fixed for a different allele at MDH-2, while until used. the chat and yellowthroat shared an allele at this Heart, liver, muscle,and kidney sampleswere re- locus.We could not scoreMDH-2 in the kinglets. The moved from all individuals, homogenizedseparately, chat and yellowthroatwere fixed for a sharedallele centrifuged, and the supernatant stored at -70øC at MDH-1 that differed from the other three species. overnight. Standard electrophoretictechniques (In- Heterozygosity values (Table 2) ranged from 0.045 gold et al. 1984) were used to resolve 19 presumptive (House Wren) to 0.106 (Ruby-crownedKinglet). genetic loci for the House Wren, Common Yellow- Nei's (1972) genetic distanceand modified Rogers' throat, and Yellow-breasted Chat and 16 loci for the geneticdistance (Wright 1978)data are presentedin two kinglets. Proteins studied (Table 1) were: aspar- Table 3. The largestgenetic distance was betweenthe rate aminotransferase (AAT), aconitase (ACON), es- House Wren and the Yellow-breasted Chat, with a terase (EST), glucokinase(GK), glucose-3-phosphate modified Rogers'genetic distanceof 0.878. The two dehydrogenase(G-3-PDH), isocitratedehydrogenase kinglets showedthe lowest modified Rogers'genetic (ICD), lactate dehydrogenase(LDH), malate dehy- distance of 0.597. drogenase(MDH), malic enzyme (ME), nonspecific The conservatismof the supernatant (S-MDH, or protein (NSP), peptidase(PEP), phosphoglucomutase MDH-1 of this study) and mitochondrial (M-MDH, (PGM), 6-phosphogluconatedehydrogenase (6-PGD). or MDH-2 of this study) forms of malate dehydro- A matrix of Nei's (1972)genetic distance and modified genasein birds hasbeen well studied(Kitto and Wil- Rogers'genetic distance (Wright 1978)values was ob- son 1966, Kakizawa et al. 1982, Kuroda et al. 1982). tained for the 16 shared loci using the BIOSYS-1 pro- Most passefine birds possessalleles in common at gram (Swofford and Selander 1981). Both distance both MDH loci. In our hands, the two alleles of MDH-1 values were calculated to facilitate comparisonswith migrated to just slightly different placeson the gel. other electrophoreticstudies. A phenogram using the Because the difference was consistent for different unweighted pair-group algorithm with arithmetic individuals of the samespecies, we are confidentthat means (UPGMA) was prepared from the modified different allelesare present.The two allelesfor MDH-2 had distinct relative mobilities. Kitto and Wilson (1966) also found a slower allele for MDH-1 in the Cedar •Present address:Department of Biology, Ship- Waxwing (Bombycillacedrorum). The studies on pensburg University, Shippensburg, Pennsylvania MDH-2 (Kakizawa et al. 1982, Kuroda et al. 1982) did 17257 USA. not include North American passetines,and Kitto and April 1988] ShortCommunications 387

T^BI,E 1. Tissues,buffer systems,and enzymeloci for the 5 speciesexamined.

Tis- Buffer Regulus Regulus Geothlypis Icteria Troglody- sue systema Protein locus E.C. no. calendula satrapa trichas virens tesaedon Heart TC 8.0 AAT-1 2.6.1.1 A 1.00 1.00 1.00 1.00 B 1.00 AAT-2 2.6.1.1 A 1.00 B 1.00 1.00 1.00 1.00 G-3-PDH 1.2.1.12 A 0.98 B 1.00 C 0.02 D 0.29 E 0.71 ICD 1.1.1.42 A 0.04 B 0.96 C 0.05 D 0.05 E 0.95 F 0.95 G 1.00 H 1.00 LDH 1.1.1.27 A 1.00 1.00 B 1.00 MDH-1 1.1.1.37 A 1.00 1.00 1.00 B 1.00 1.00 Liver LiOH EST-2 3.1.1.1 A 1.00 1.00 0.77 B 0.14 C 0.02 1.00 D 0.21 0.14 E 0.72 EST-3 3.1.1.1 A 0.20 B 0.50 0.07 C 0.20 0.79 0.80 0.29 D 0.07 E 0.10 0.14 F 0.71 G 0.13 0.43 H 0.21 I 0.36 ME-1 1.1.1.40 A 1.00 B 1.00 0.03 C 0.97 D 0.10 E 0.90 F 1.00 ME-2 1.1.1.40 A 1.00 B 1.00 C 0.25 D 0.75 E 1.00 F 1.00 TC 8.0 GK 2.7.1.1 A 0.04 388 ShortCommunications [Auk, Vol. 105

TABLE 1. Continued.

Tis- Buffer Regulus Regulus Geothlypis Icteria Troglody- sue systema Protein locus E.C. no. calendula satrapa trichas virens tes aedon

B 0.96 C 0.14 D 0.57 E 0.05 F 1.00 G 0.95 0.29 H 1.00 PEP-1 3.4.13.1 A 0.98 B 0.07 C 0.07 D 0.29 E 0.93 F 0.42 G 1.00 0.86 H 0.02 I 0.29 J 0.07 PEP-2 3.4.13.1 A 1.00 B 0.90 0.96 C 0.05 D 0.90 E 0.10 0.04 F 1.00 G 0.05 MDH-2 1.1.1.37 A 1.00 1.00 B 1.00 6-PGD-1 1.1.1.44 A 0.70 0.05 B 0.30 0.07 0.95 1.00 C 0.29 D 0.93 0.71 6-PGD-2 1.1.1.44 A 1.00 0.50 0.20 B 0.50 1.00 1.00 0.80 RW PGM 2.7.5.1 A 0.04 0.07 B 0.05 C 1.00 0.96 0.93 0.93 D 0.02 1.00 NSP A 1.00 1.00 1.00 1.00 1.00

Liver LiOH ACON 4.2.1.3 A 0.29 B 0.05 C 0.71 D 0.95 E 0.30 0.04 F 1.00 G 0.70 H 0.96

' TC 8.0 and LiOH buffersfrom Selanderet al. 1971.RW buffer from Ridgway et al. 1970. April 1988] ShortCommunications 389

TABLE2. Intraspecific variation in 5 North American passerinespecies (standard error in parentheses). Geofhlypistrichos Per- Icterla vimns ce•n,tag.eMean heterozyg•osit¾ Mean poly- Hardy- alleles/ mor- Direct Weinberg locus phica count expectedb Fig. 1. Phenogrambased on modifiedRogers' ge- Regulus 1.4 31.3 0.106 0.117 netic distance(Wright 1978)values. The F-value (Far- calendula (0.2) (0.048) (0.054) ris 1972) is 0.153, the percentagestandard deviation Regulus 1.6 50.0 0.076 0.069 (Fitch and Margoliash 1967) is 2.88, and the cophe- satrapa (0.2) (0.030) (0.026) netic correlation is 0.953. Troglodytes 1.4 26.3 0.045 0.119 aedon (0.2) (0.038) (0.053) illus; ATs0H = -2.2) than either was to the Golden- Icteria 1.5 42.1 0.074 0.171 crownedKinglet (AT50H= -2.5). Although theseDNA virens (0.2) (0.039) (0.053) values are fairly high, they are within the range de- Geothlypis 1.8 63.2 0.073 0.090 termined for other passerinecongeners (Sibley and trichas (0.2) (0.023) (0.026) Ahlquist 1985).We proposethat the two kinglets are • A locus was consideredpolymorphic if more than one allele was the result of independent invasions from the Old detected. World. The DNA data suggestthat the Ruby-crowned b Unbiased estimate (Nei 1978). Kinglet is the most recentarrival. A comprehensive understandingof relationshipsamong the kinglets, Wilson (1966) suggestedthat the conservatismin however, would require a completestudy of all five MDH-2 is not as severe as it is in MDH-1. Regulus. We obtainedheterozygosity values consistent with The UPGMA tree (Fig. 1) shows that the yellow- other studies,although the Ruby-crownedKinglet throat and wren are equally similar to the kinglets. value(0.106) was higher than that obtainedby Avise The cladogram(Fig. 2), by providing systematicin- et al. (1980a;H = 0.048, n = 10). Sample sizes were ferencesbased on branch length, permits resolution the same in both studies, but few of the loci examined of the equalitiesapparent in Fig. 1. The sum of the were shared. branch lengths from the kinglet basal node to the The electrophoreticdata stronglysupport the con- wren (93) is shorter than to the yellowthroat (136) tention of Mayr and Short (1970) that the Ruby- and implies that the wrens are more similar to the crownedand Golden-crownedkinglets are not closely kingletsthan are the wood warblers,as represented related. The Nei (1972) geneticdistance of 0.50 is by the CommonYellowthroat or Yellow-breastedChat. larger than geneticdistance values for speciesin other Sibley and Ahlquist (1982a)suggested, based on un- passerine genera such as Dendroica(0.00-0.17; Avise published DNA-DNA hybridization data, that the et al. 1980b), Catharus(0.01-0.03; Avise et al. 1980a), kinglets and wren are mostclosely related becauseof Toxostoma(0.09; Avise et al. 1982), Vireo (0.03-0.51; a closerelationship of both groupsto the Old World Avise et al. 1982), and Parus (0.00-0.07; Braun and warblers and babblers. Robbins1986). The Nei distancefalls within the range Both figures demonstratethat the Yellow-breasted of 0.40-0.80,similar to thosereported for comparisons Chat, consideredto be a wood warbler (Sibley and between very closely related avian families or be- Ahlquist 1982b),is the mostdivergent memberof this tween highly divergent,confamilial genera (Avise et specificgroup of birds. The correctsystematic posi- al. 1982).The electrophoreticdata are supportedby tion of the Yellow-breastedChat has been widely DNA-DNA hybridization studies (Sibley and Ahl- debated (see Sibley and Ahlquist 1982b for review). quist 1985)that showedthe Ruby-crownedKinglet Avise et al. (1980b) showed that the chat was more was more closelyrelated to the Firecrest(R. ignicap- similar to the wood warblers than to either the Red-

TABLE3. Nei's(1972) genetic distance (above diagonal) and modifiedRogers' genetic distance (Wright 1978; below diagonal)estimates among 5 species.

! 2 3 4 5 1. Reguluscalendula -- 0.495 1.136 1.939 1.008 2. Regulussatrapa 0.597 -- 1.216 1.625 1.011 3. TroglodytesaedOn 0.773 0.796 -- 2.212 1.210 4. Icteria virens 0.860 0.843 0.878 -- 1.005 5. Geothlypistrichas 0.756 0.765 0.797 0.748 -- 390 ShortCommunications [Auk,Vol. 105

FITCHß W. M., & E. MARGOLIASH. 1967. Construction of phylogenetictrees. Science155: 279-284. GRAY,A.P. 1958. Bird hybrids. Farum RoyalßBucks, CommonwealthAgricultural Bureaux. Regulu•satrapa INGOLD, J. L., L. A. WEIGT, •a: $. I. GUTTMAN. 1984. Genetic differentiation within the avian genus Columba.Comp. Biochem.Physiol. 77B: 427-430. KAKIZAWA, R., N. KURODA, •a: $. UTIDA. 1982. Evo- lution of mitochondrialmalate dehydrogenase in birds (II). J. Yamashina Inst. Ornithol. 14: 131- 136. KITTO, G. B., & A. C. WILSON. 1966. Evolution of malatedehydrogenase in birds.Science 153: 1408- 1410. Fig. 2. The most parsimonious tree for 5 taxa of KURODA, N., R. KAKIZAWA, I-I. I-IORI, Y. (3SARA, N. passeriBesfrom PAUP packageusing allelesas char- USUDA, •a: $. UTIDA. 1982. Evolution of mito- acters and presence/absenceas character states.Con- chondrial malate dehydrogenasein birds. J. Ya- sistencyindex = 0.877;tree length = 407. mashina Inst. Ornithol. 14: 1-15. MAYR,E., &:L. L. SHORT. 1970. Speciestaxa of North eyed Vireo (Vireoolivaceus) or Swainson'sThrush (Ca- Americanbirds. Cambridgeß Massachusetts, Publ. tharusustulatus), two groupsfor which affinitieshave Nuttall. Ornithol. Club No. 9. been suggested.Sibley and Ahlquist (1982b) also NEI, M. 1972. Genetic distancebetween populations. showed that the chat is a wood warblerßalthough Am. Nat. 106: 283-292. probablya very early branchin the phylogenyof the ß 1978. Estimationof averageheterozygosity family. The branch length from the yellowthroatto and geneticdistance from a small number of in- either kinglet is shorterthan that to the chat;if the dividuals. Genetics 89: 583-590. chat is a wood warbler, it is quite distinct from the RIDGWAY,G. J., $. W. 5HERBURNE,&: R. D. LEWIS. 1970. yellowthroat. Polymorphism in the esterasesof Atlantic her- LITERATURE CITED ring. Trans. Am. Fish. Soc.99: 147-151. 5ELANDER,R. K., M. H. SMITHß5. YANG, W. E. JOHNSONß ARISE,J. C., J. C. PATTON,&: C. F. AQUADRO. 1980a. & R. B. GENTRY. 1971. Biochemicalpolymor- Evolutionary genetics of birds I. Relationships phism and systematicsin the genus Peromyscus. among North American thrushesand allies. Auk I. Variation in the old-field mouse (Peromyscus 97: 135-147. polionotus).Stud. Gen. VI, Univ. TexasPubl. 7103: --, & --. 1980b. Evolutionary ge- 49-90. netics of birds III. Comparative molecular evo- 5IBLEY,C. G., & J. E. AHLQUIST. 1982a. The relation- lution in New World warblers (Parulidae) and shipsof the swallows (Hirundinidae). J.Yamashi- rodents (Cricetinae). J. Hered. 71: 303-310. na Inst. Ornithol. 14: 122-130. --, & -- 1982. Evolutionary ge- ß & --. 1982b. The relationshipsof the netics of birds V. Genetic distances within Mim- Yellow-breasted Chat (Icteria virens) and the al- idae (mimic thrushes) and Vireonidae (vireos). leged slowdown in the rate of macromolecular Biochem. Gen. 20: 95-104. evolution in birds. Postilia 187: 1-18. BRAUN,M. J., & M. B. ROBBINS.1986. Extensive pro- ß& --. 1985. The phylogeny and classi- tein similarity of the hybridizing chickadeesPar- fication of the passefinebirdsß based on compar- us atricapillusand P. carolinensis.Auk 103: 667- isons of the genetic material, DNA. Proc. 18th 675. Int. Ornithol. Congr.: 83-121. BUTH,D.G. 1984. The applicationof electrophoretic SWOFFORD,D. L. 1984. Phylogeneticanalysis using data in systematicstudies. Annu. Rev. Ecol.Syst. parsimony(PAUP). Champaign,Illinois Nat. Hist. 15: 501-522. Surv. CLEMENTS,J.F. 1978. Birds of the world: a check list. ß•a: R. B. 5ELANDER. 1981. BIOSYS-I: a Fortran New YorkßTwo Continents Publ. Group. programfor the comprehensiveanalysis of elec- FARRIS,J.S. 1972. Estimatingphylogenetic trees from trophoreticdata in population geneticsand sys- distance matrices. Am. Nat. 106: 645-668. tematics. J. Hered. 72: 281-283. --. 1981. Distancedatainphylogeneticanalysis. WRIGHTß5. 1978. Evolutionand the geneticsof pop- Pp. 3-23 in Advances in cladistics,vol. 1 (V. A. ulations. Vol. 4, Variability within and among Funk and D. R. BrooksßEds.). New Yorkß New natural populations. Chicagoß Univ. Chicago York Botanical Garden. Press. FELSENSTEIN,J. 1984. Distancemethods for inferring ReceivedI July 1987,accepted 29 December1987. phylogenies:a justification.Evolution 38: 16-24.