EVOLUTIONARY RELATIONSHIPS WITHIN THE ENSATINA...

Syst. Biol. 41(3):273-291, 1992

EVOLUTIONARY RELATIONSHIPS WITHIN THE ENSATINAESCHSCHOLTZII COMPLEX CONFIRM THE RING

SPECIES INTERPRETATION

CRAIG MORITZ,1 CHRISTOPHER J. SCHNEIDER,2 AND DAVID B. WAKE2

^Department of Zoology, University of Queensland,St. Lucia, Queensland 4072, Australia

2Museum of Vertebrate Zoology, University of California,Berkeley, California 94720, USA

Abstract.—Sequences (644-681 bp) from the mitochondrial cytochrome b gene were obtainedfor 24 individuals representing the geographic range and morphological diversity of the poly-typic salamander ring species Ensatina eschscholtzii. These data were used to estimate the phy-logeny of components of the ring to test the biogeographic scenario underlying current inter-pretations of speciation in this complex. The analysis revealed high levels of nucleotide variationamong subspecies. Strong subdivision was evident within the subspecies platensis and oregonensis.The phylogenetic hypothesis of minimum length that is best supported by the data containsone monophyletic group that includes populations from the southern Sierra Nevada and moun-tains of southern California (croceater, klauberi, and southern platensis) and another that includespopulations of the southern and central coastal regions (xanthoptica and eschscholtzii). Samplesof oregonensis were typically basal, but their precise branching order was unstable. Both oregonensisand platensis were paraphyletic, with several disparate lineages in oregonensis and a strong north-south dichotomy in platensis. The data were incompatible with a biogeographic model thatrequired all subspecies to be monophyletic but were compatible with slightly modified predic-tions of a model assuming stepwise colonizations from north to south down the Sierra Nevadaand independently down the coastal ranges. These features provide strong support for thebiogeographic scenario central to the interpretation of Ensatina eschscholtzii as a ring species.Division of this complex into separate species on the basis of the observed patterns of monophylyfor mitochondrial DNA (mtDNA) is unwarranted because further sampling could reveal addi-tional instances of paraphyly across subspecies and, more generally, because mtDNA aloneshould not be used to infer species boundaries. [DNA sequences; mitochondrial DNA; cytochromeb; ring species; phylogenetics; geographic variation; speciation; Caudata.]

Salamanders of the Ensatina eschscholtzii Stebbins (1949) recognized seven sub-

complex provide a particularly clear ex- species distributed in the mountains en-ample of a ring species (Stebbins, 1949). circling the central valley of California (Fig.Ring species are circularly arranged poly- 1). Unblotched subspecies occur on thetypic taxa with gradual transitions be- coastal margins and replace each other fromtween adjacent components but abrupt southern {eschscholtzii), through centralchanges where the terminal races come into (xanthoptica), to northern California (ore-contact. They are supposed to exemplify gonensis and picta) into Oregon and Wash-speciation through gradual divergence ington. Blotched subspecies are distributedamong large population units (Rensch, down the inland side of the valley on the1929; Mayr, 1963) and, thus, the linkage mesic western slopes ©f the Sierra Nevada,between microevolutionary processes and The subspecies platensis ranges widelyspeciation. This view contrasts with recent through the Sierra but is replaced at themodels of speciation that emphasize ge- southern end by populations of croceater.netic changes in small peripheral popula- The third blotched form, klauberi, is dis-tions (Mayr, 1982; White, 1982; Carson and junct and occurs as isolated populations inTempleton, 1984). It is also contrary to the the mountains of southern California. Theopinion that the process of speciation is blotched and unblotched forms are in con-qualitatively distinct from intraspecific mi- tact in three places: across the top of thecroevolution (e.g., Goldschmidt, 1933; ring, in the central Sierran foothills whereGould and Eldredge, 1977). a disjunct series of populations of xanthop-

273

at North C

arolina State University on M

arch 25, 2014http://sysbio.oxfordjournals.org/

Dow

nloaded from

http://sysbio.oxfordjournals.org/


274 SYSTEMATIC BIOLOGY VOL. 4 1

115°

35°-

E 3 PictaMM °r09E23 esch

klaucrocplatxan

200 km

FIGURE 1. Map of western USA showing the dis-tribution of localities from which Ensatina mtDNAswere sequenced. The ranges of the subspecies aredefined on the basis of information in Stebbins (1949)and extensive recent collections (Wake, unpubl. data).Abbreviations, localities, and MVZ catalog numbersare in the Appendix.

tica is in contact with platensis, and in themountains of southern California whereklauberi and eschscholtzii meet at four dis-junct sites.

Stebbins (1949) proposed that the speciesoriginally occurred in the Douglas fir andredwood forests of southern Oregon andnorthern California and that during peri-ods of greater humidity, salamanders dis-persed down the coastal and inland rangesto meet again in southern California. Withincreasing aridity, the populations became

less continuous, particularly in the south-ern part of the range. The Sierran popu-lations of xanthoptica were thought to berecently derived from the coastal popula-tions via a transvalley leak during the moremesic periods of the Pleistocene.

Based on this model of historical bio-geography, Stebbins proposed that thesouthward colonizations were each accom-panied by differentiation such that inter-breeding was precluded when the two lin-eages came into contact again in themountains of southern California. The keyobservation (Stebbins, 1949:506) that sug-gested the gradual accumulation of repro-ductive isolation was that

progressive genetical divergence with progres-sively increasing reproductive incompatibility isobserved from north to south between the coastaland interior series of races—from free interbreed-ing in northern California, through occasional hy-bridization in the Sierra in central California, tosympatry with reproductive isolation in southernCalifornia.

Additional collecting and analyses (Steb-bins, 1957; Brown and Stebbins, 1964;Brown, 1974; Wake et al., 1986, 1989) havesupported the basic interpretation.

An analysis of allozyme variation (Wakeand Yanev, 1986) revealed strong differ-entiation throughout the complex. Somesubspecies were remarkably diverse for al-lozymes (e.g., platensis, oregonensis), where-as others (e.g., xanthoptica, eschscholztii)showed little differentiation. The presenceof fixed differences between local popu-lations of the same subspecies was contraryto the view (Dobzhansky, 1958) that thecontinuity of the ring was the result ofongoing gene flow. Given such strong geo-graphic structuring, the initial allozymestudy did not sample sufficient popula-tions to determine the discreteness of sub-species or their relationships. Wake andYanev (1986:713), however, suggested thatthe blotched and unblotched forms "mightrepresent separate adaptive responses toprovincial selection pressures rather thandiscrete historical entities or incipient spe-cies."

In this paper, we used sequences frommitochondrial DNA (mtDNA) to estimate

at North C



Dow

nloaded from



1992 RELATIONSHIPS WITHIN A RING SPECIES 275

phylogenetic relationships among com-ponents of the ring and to evaluate thehistorical biogeographic scenario basic toStebbins's (1949) evolutionary model. Spe-cifically, we evaluated the phylogeneticpredictions from two alternative scenarios:(1) Stebbins's model, which suggests in-dependent stepwise colonizations downthe coastal and inland ranges from north-ern Californian populations, with a recenttransvalley leak of xanthoptica, and (2) amodel based on the premise that the cur-rent distribution of subspecies is the resultof secondary contacts between subspeciesthat had diverged considerably in geo-graphic isolation (cf. Stebbins, 1949:503).

We compared mtDNAs to infer popu-lation history. The mtDNA "gene tree"should reasonably reflect population his-tory where local populations are recipro-cally monophyletic for their mtDNAs anddiversity within populations is much lessthan that among populations (Avise, 1989).These conditions should be met where lo-cal populations have been small and havehad restricted gene flow over long periods(ca. 4A^ generations; Neigel and Avise,1986; Birky et al., 1989). These conditionsare likely to be met by Ensatina eschscholtziisubspecies, given their extreme subdivi-sion for allozyme alleles (Wake and Yanev,1986) and presumed low vagility. Prompt-ed by controversy over the taxonomy ofEnsatina (Frost and Hillis, 1990), we pointout the limitations of mtDNA for definingspecies boundaries.

MATERIALS AND METHODS

Sampling Design

The analysis covers multiple, geograph-ically separated representatives of eachnamed subspecies, concentrating on mor-phologically "pure" populations as de-fined by Stebbins (1949) or our own ob-servations. Sampling and analysis was donein three phases (see Baverstock and Moritz,1990). The first was a pilot study, based on378 bp of mitochondrial cytochrome b se-quence from 15 samples of Ensatina andtwo outgroup samples, to determine thesuitability of the gene. Aneides lugubris and

Plethodon elongatus were selected as out-groups on the basis of other studies (Wake,1963; Larson et al., 1981) that suggestedthese species as members of the sister cladeof Ensatina, itself a monotypic genus. Thepilot study revealed appropriate levels ofvariation (up to 12% observed sequence di-vergence within Ensatina), although phy-logenetic analysis failed to resolve thebranching order at the base of the tree.

In the second phase, we added nine moresamples of Ensatina to replicate represen-tation of some subspecies and to attemptto split long branches, a potential cause ofinstability or error in parsimony analysis(Felsenstein, 1978; Swofford and Olsen,1990). We also extended the sequences by303 bp in an attempt to increase resolutionat the base of the phylogeny. The final dataset included 24 individuals from 22 local-ities (Fig. 1) and one each from the out-groups, Aneides and Plethodon. In the thirdphase of sampling, we screened multiplesamples from three populations for diag-nostic restriction fragment length poly-morphisms to assess the distribution ofidentified mtDNA lineages within andamong localities.

DNA Preparation, Amplification, andSequencing

DNA was prepared by phenol-chloro-form extraction (Maniatis et al., 1982) orby boiling minute amounts (<5 mg) of fro-zen tissue in a 5% (w/v) solution of Chelex(BioRad). Both approaches yielded DNAsuitable for amplification by the polymer-ase chain reaction (PCR).

PCR amplifications were done usingprimers (Fig. 2) that spanned approxi-mately two-thirds of the cytochrome b geneand yielded a sequence homologous to co-dons 7-234 in the published sequence ofXenopus (Roe et al., 1985). Primers cyt-b2(Kocher et al., 1989) and MVZ 16 (C. Or-rego and D. Irwin, pers. comm.) amplifymtDNA from a wide variety of vertebrates.Primers MVZ 15,18, and 25 were designedto match sequences from Ensatina and, for18 and 25, also from Xenopus.

For double-strand • reactions, template

at North C



Dow

nloaded from



276 SYSTEMATIC BIOLOGY VOL. 41

m15

100 200 300

^ ^ Cytochro

•

25

400 500 600 700 600

- - - • 18b2

Sequence

(b)

Primer

31 EndWithin Xenopusgene sequence

19352

MVZ15 G A A C T A A T G G C C C A C A C J T T A C G N A A

MVZ 25 TAAAGAAACTTGAAA J AT J GG T GT

cyt-b2 AAACTGCAGCCCCTCAGAATGATATTTGTCCTCA 405

MVZ18 GTCTTTGTATGAGAAGTATG 667

MVZ16 AAATAGGAAgTATCAjTCTGGTTT^AT 804

1626816601166541691617053

FIGURE 2. Location (a) and definition (b) of oligonucleotide primers used for amplification and directsequencing of cytochrome b. The scale represents the distance from the start of the gene in base pairs. Thelocation of the 3' end of the primers is defined relative to the 3' end of the gene and to the published sequenceof Xenopus mtDNA (Roe et al., 1985). Arrows indicate the region sequenced from each primer. Primers 15and 25 generated light strand (antisense) sequence.

and primers were annealed at 45-50°C, us-ing 0.5 pmol of each primer, 0.75 mMdNTPs, and 1.5 mM MgCl2 in a pH 8.4buffer with 50 mM KC1 and 10 mM Tris-HC1 (final concentrations). Reactions weretypically run for 35 cycles in a total volumeof 12.5 or 25 jul, using 0.3 or 0.6 units ofTaq polymerase (Cetus), respectively. Ali-quots of 5 n\ were run on 2-4% low-melt-ing-point agarose gels from which a smallplug was taken and diluted 1:100 in 10 mMTris-0.1 mM EDTA to provide template forsingle-strand reactions.

Single-strand template was prepared byasymmetric PCR (Gyllensten and Erlich,1988) with 1:50 primer ratios and the samereaction profiles as above in 50-^1 reac-tions with 1.2 units of Taq polymerase. Theyield and purity of single-strand productswere assessed by electrophoresis of 5-̂ 1 ali-quots through 4% agarose (1 x TAE) gels.For successful reactions, the remaining 45/A was purified and concentrated to ca. 20jiil on disposable centrifugal filters (Milli-

pore MC30). Dideoxy chain terminationsequencing (Sanger et al., 1977) was donewith the U.S. Biochemicals Sequenase ver-sion 2.0 kit and 35S-labeled dATP. All dou-ble- and single-strand PCR reactions in-cluded negative controls to guard againstcontamination of reagents with DNA. Nosuch contamination was detected.

Data Analysis

Sequences were read from both strandswith >90% overlap and aligned by eye toeach other and, using amino acid transla-tions, to the same region of Xenopus mt-DNA. The distribution and amount of vari-ation among the sequences were analyzedwith the MOSY program written in C forDOS PCs by Chris Meacham (Univ. Cali-fornia, Berkeley). Measures of actual di-vergence (cf. observed difference) were es-timated using the correction of Brown etal. (1982).

For phylogenetic analysis, the generalstrategy was to minimize the noise gen-

at North C



Dow

nloaded from




K I H P L L K I I N N S F I D L P T P S N L S Y L W Nxan SR AAAATTCACCCACTACTAAAAATCATTAATAATTCCTTCATTGACTTACCTACTCCATCAAACTTATCCTACCTATGAAA

xan Ororeg WAoreg Inoreg ORoreg Broreg Meoreg Heplat Blplat Arplat Haplat Macroc Tecroc Veklau Crklau CW1klau CW2klau SJSesch SBaesch SBeesch CWpi etaPlethodonAneides

7777777777

???77?9979779777QQ

T..T..

C ,

C

CC

7 7 7 7 7 7

9997997779

rTTrTTTTTT

TTTTTTTTTTTTTTT7 ??

TT

7 7 7 7 7 7 7 7 7 7 7 y7797777777

??????????'7799777777'

j

.T...

.T...

.T...

.G...

.A...

.T...

.T...

.T...

.T...

.T...• T....T....T....T...?T?

.T...

...T.77777

T.

T.T.

???...

T.T.T.

???

T.

T.??7??G

C

ccccc.c.c.c.

.c

.c

.c

.C..A

.c

.c

.c

.c

.c

.c

T....C.C.c.c.C..T..T...

T C

..T

..T..G.

..T

..T

..T

..T

..T

..T

..TC.

..T

..T

..T

..T

..T

..TCC

...C.T.

.c

.c

.cT.T.

.C.C

.CG

.c

.. G

..GC ..C

.c

.c

.c

.c

.A

.c

.c

.c

.A..C.GT..

.CG.C

.T

.T

.T

.T

.T

.T

.T

.T

.T

.T

.T•TC.G..TC..TC..TC.G.

.T

.T

.TC.

.T

T

T

TTT

.A.A.

.TT

.TT

.TT

..T,.T.TT.TT.TT..T,.T.TT.TT..T..T..T..T

.T

.TT

..TG

.TT

F G S L L G I C L I S Q I M T G L F L A M H Y T A Dxan SR CTTTGGGTCACTACTAGGAATCTGTTTAATTTCACAAATTATAACTGGCCTCTTTCTAGCAATACATTATACAGCAGACAxan Avxan Or C.T.TC Coreg WA A G C.T T T.oreg In A..GT G C.T..C T...Toreg OR A G.T G C.T Toreg Br A T C C.T..C T...Toreg Me A T G C CT..C T...Toreg He A G G C.T..C T Cplat Bl A...T C.T T Cplat Ar A..CT C.T..C T Cplat Ha A T C A C.G..C T...T T.plat Ma A T A C.G T...T T.croc Te A T G C.G..C A...Tcroc Ve A GT G C.G..C A...Tklau Cr T A C.T T...Tklau CU1 T A..T...T C.T T..T...T Cklau CW2 T A T C.T T..T...T Cklau SJS T A C.T T...T Cesch SBa CC.Cesch SBe A CC.Cesch CW A CC.Cpi eta A C . G C.T T...TPlethodon ...CA...T..T T CC AAT C..T..A G C T.Aneides T G AT..C T T..A..T..A G C.C T

FIGURE 3. Sequences from codon 7 to codon 234 of cytochrome b from 23 individuals of Ensatina andrepresentatives of the outgroups (Aneides lugubris and Plethodon elongatus). A complete nucleotide sequenceand its amino acid translation is given for a xanthoptica from Santa Rosa. Other sequences with nucleotidesidentical to the xan SR sequence are represented by periods (.). Some sequences could not be reliably readclose to the primers; ambiguities are represented by question marks (?). The sequence found in both individualsof picta is represented once. Abbreviations are defined in the Appendix.

at North C



Dow

nloaded from




T T S A F S S V V H I C R D V N Y G W I M A N I H A Nxan SR CCACTTCCGCATTCTCATCAGTTGTACATATCTGCCGCGATGTAAATTACGGATGAATTATACGAAATATTCACGCCAACxan Av Cxan Ororeg WA C..T T G G . . . G . . T Toreg In T T A T C G Toreg OR C..T T G . . . G . . T Toreg Br T A C T T . .C T Toreg Me T T A T . . C . . . G . . T Toreg He C..T A T T . . G . . . G Tp la t Bl C A T C T G Tp l a t Ar . . T . C . . T A T C..G T G Tp la t Ha C A T T Tp la t Ma C A T T Tcroc Te C A T G Tcroc Ve C A T G T Tklau Cr C A.C.C T C C . . T . . C Tklau CW1 C A . . . C T C C . . T . . C Tklau CW2 C A . . . C T C C . . T . . C Tklau SJS C A.C.C T C C . . T . . C Tesch SBa T . . Gesch SBe . C Aesch CW C Api eta C T A T . .T C..T G T TPlethodon T C ?. .C C..T T.A CAneides .T.T T CA A. .C T . . T . . . T . A G T . . T . . T

G A S F F F I C I Y M H I G R G I Y Y G S Y M F K K Txan SR GGAGCCTCATTTTTCTTTATCTGCATCTACATACATATTGGACGTGGTATTTATTACGGATCCTATATATTTAAAAAAACxan Av G T C Gxan Or G Goreg WA T T . . T . . T T A A T Goreg In C . . T . . T . . T . . T C A G Goreg OR T T . . T . . T T A A T Goreg Br A T . .T C G. .A. .C T . .C Goreg Me A C A C T T . .C Goreg He T T T..TT C A A T A G Gp l a t Bl T A A T T . .C Gp l a t Ar T . .T A A T T . .C Gp l a t Ha T T C A . . A . . A T TC..G Gpla t Ma C T T C A . . A . . A T T Gcroc Te T A A T G.G. .croc Ve T A A T G.G. .klau Cr C..T T . . T . . T . . T A A T. .G G.G. .k lau CW1 C..T T . . T . . T . . T C A A T Gklau CW2 C . . T . . . . . T . . T . . T . . T C A A T Gklau SJS C..T T . . T . . T . . T A A T G.G. .esch SBa T . .T A T C Gesch SBe T. .T C A T C C . . .Gesch CW T. .T A T C C . . .Gpi eta TT A . . C . . A T T C . . .GPlethodon T C..T T . . T . . T C . T A. .AGCA. .C . .T A..A GAneides . . T C..T T . .T C A..CT.A T A A.. .GG

FIGURE 3. Continued.

erated by homoplasy through a prioriidentification of classes of variation ex-pected to be saturated (F. Villablanca, W.K. Thomas, and A. C. Wilson, submitted)and the use of appropriate weighting mea-sures and constraints. Phylogenies were

estimated using PAUP (version 3.01; Swof-ford, 1991). Unless otherwise noted, allanalyses used a 10:1 step matrix to weighttransversions over transitions. Bootstrapanalyses with the step matrix applied weredone using PAUP 3.0s.

at North C



Dow

nloaded from




W N I G V L L L F L V M A T A F V G Y V L P W G Q Mxan SR ATGAAACATCGGAGTACTATTACTGTTTCTTGTTATAGCAACAGCATTTGTAGGTTACGTTCTTCCATGAGGACAAATATxan Avxan Ororeg WA T T . .T A T . A . . . T A Toreg In T T . .T A . G . . . T . A A C C..T Coreg OR T T . .T A T . A . . . T A T . .Coreg Br C T . .T TT A C..Toreg Me C T . .T TT T.A C..Toreg He T T A T . A . . . T G T. .CPla t Bl T T . .T TA C C T . .Cp l a t Ar T T . .T TA C A T . .Cp l a t Ha T T A . T . . . T . A A C T . .Cp la t Ma T T A . T . . . T . A . . C A C T . . C . . Ccroc Te T T . .T A . T . . . T . A A T T . . C . . Ccroc Ve T T . .T A . T . . . T . A A C . . T . . C . . C . . Gklau Cr T T . .T A..C.CT.A A..G G C C..Cklau CW1 T T . .T A. .C.TT.A A..G C T . . C . . Cklau CW2 T T . .T A..C.CT.A A..G C T . . C . . Cklau SJS T T . . T . ? A..C.CT.A A..G G C T . . C . . Cesch SBa C G T.A C T . . C . . Cesch SBe T ? G ?A C T . . C . . Cesch CW T ?? G A C T . . C . . CPi e ta T T . .T A T.A A C . . T . . CPlethodon C T TA.TC.CT.A.. .T.A T . . A . . T CAneides C A . T . . . T . A . .C. .A A . . T . . C . . C . . T

S F U G A T V I T N L L S A I P Y M G D M L V Q U I Uxan SR CATTCTGAGGAGCTACAGTCATCACAAACCTCTTATCAGCAATCCCTTATATAGGAGATATATTAGTTCAATGAATTTGAxan Av Cxan Ororeg WA C..C T G A C.C Goreg In T A. .C Coreg OR C . C T G A C.C Goreg Br .C T . .T T A C C.Coreg Me .C T T A. .C Coreg He T C . C T G T. .A TCC Gpla t Bl T C A A C Gp l a t Ar C A T G C.C Gp l a t Ha T G A.. .GC C.Cp l a t Ma T G A..CGC Ccroc Te T A..CGC C.Ccroc Ve T A..CGC C.Cklau Cr T T C C.Cklau CW1 T T C C.Ck l a u CW2 T T C C.Cklau SJS T T C C.Cesch SBa T C.C Cesch SBe T C.C Cesch CW T T C.C Cpi eta T . .T T C C . . .Plethodon T . . G . . T . . C . . T . . A . . T . . C TCC T. .A C A C . C .Aneides . T . . T T . . C . . C T . . C . . T . . T C . T T . . A . . C G C A


RESULTS

Sequence Variation

Sequences 644-681 bases in length wereobtained from 24 individuals of Ensatinaand one individual each of Plethodon elon-gatus and Aneides lugubris (Fig. 3). The mi-

nor differences in length are due to vari-able sequence readability close to theprimers. No substitutions causing frame-shift or nonsense mutations were present.The mean base content of the light strandacross all taxa (0.13 G, 0.31 A, 0.35 T, 0.20C) is biased toward A and T, with a marked

at North C



Dow

nloaded from




G G F S V D K A T L T R F F T F H F I L P F I V V G Ixan SR GGAGGCTTTTCAGTCGACAAAGCAACCTTAACCCGATTTTTTACCTTCCACTTCATTCTACCATTTATTGTAGTAGGAATxan Av . . G . . Txan Ororeg UA A. .T C.T G . T . . T . . T . . T A . G . . . G .oreg In T . . A . . T C G . T . . T . . T . . T Goreg OR A. .T C.T G . T . . T . . T . . T . . C A.G. .TG.oreg Br G A. .T C.T T . .T T T.Coreg Me A A. .T C.T G.T. .T T Coreg He A. .T C.T. .T G . T . . T . . T . . T . . C G.p la t Bl . .G A. .T C.T G . T . . T . . T . . T . . . T A.T.Cp l a t Ar . .G A . .T C.T G.T . .T . . T . . T . . . T A.CAC G.p l a t Ha A A. .T C G . T . . T . . T . . Tp l a t Ma A A. .T C G.T. . T . . T . . T . . .Tcroc Te A . .T C G . T . . T . . T Gcroc Ve A . .T C G . T . . T . . T ?klau Cr A. .T C G . T . . T . . T . . T . . Cklau CW1 A. .T C G . T . . T . . T . . T . . C Cklau CW2 A. .T C G . T . . T . . T . . T . . C Cklau SJS A . .T C G . T . . T . . T . . T . . C Cesch SBa T . .T G T T . . . T G G.esch SBe T . .T G T T . . . Tesch CW T . .T G . . . . T T . . . Tpi eta . .G A. .T C.C CG.T. . T . . T . . T . . .T CPlethodon . .TT.A A C.T. .T G T T. .G A.CTCG...GCAneides . . T . . T TA.T. . T . . G . . C . . .C.C G T . .T . .CT C A.TA TG.

S M I H L L F L H E T G S N N P T G L Y S D M D K Ixan SR TAGTATAATTCACCTATTATTCCTACATGAGACTGGGTCA7ATAACCCAACAGGACTTTATTCTGACATAGATAAAATTTxan Av C . . . Axan Or C . . . A Toreg WA . . . C . T T T . .G A . . C . . C . . C A T A.T.Coreg I n . . . C . T T G A . . C . C . . . A T C A.T.Coreg OR . . . C . T T T. .G A . . C . . C C A . . . . T A.T.Coreg Br . . .C .C .C .T T..C A C.A....T A Coreg Me . . . C . C C . T T A C..A T T A Coreg He . . . C . T TT G A . . C . C . A . . . . T Aplat Bl . . . C . T T A..C.C...A T.A A.T Cplat Ar . . . C . T T G A . . C C C A T.A A.Tplat Ha . . . C . T T..G A . . C . C . A . C . T Aplat Ma . . . C . T A . . C . C . A . C . T Acroc Te . . . C . T G A..C.C.. .A.C.T C Acroc Ve . . . C . T G A. . C . C .A.?..T C Aklau Cr . . . C . T C G A . . C C C A T A A.T Cklau CU1 . . . C . T G A . . C C C A T A...T A.T Cklau CU2 . . . C . T G A . . C C C A T A...T A.T Cklau SJS . . . C . T C G A . . C C C A T A A.T Cesch SBa . . . C . C G A C.. .A Cesch SBe . . . C . T T G A C...A Aesch CU . . . C . T G A C.A Api eta T T..C A..C.C.CA C.A.T..GPlethodon CC..C. .C TC T .A..C.C.TA CT.AACC . .A. .TC CAneides . . .C . .TG.A C T..CA..A..C.A T....T.AA CAG.CCT C


deficit of G's. This bias in nucleotide com-position is typical of vertebrate mtDNAgenes (Brown, 1985).

The magnitude of variation is high. In-cluding both Ensatina and the outgroups,250 (37%) of 681 nucleotides and 51 of 227(22%) amino acids vary (Fig. 3). Observed

sequence differences are between 0.3 and19.2%. Corrected for multiple hits (Brownet al., 1982), the values range from 0.8 to39.7% (Table 1). Within Ensatina, observedsequence difference ranges from 0.3 to12.1%, and corrected values are between0.8 and 16.2%.

at North C



Dow

nloaded from




S F H P Y F S Y K D L F G Fxan SR CATTCCACCCATACTTCTCATACAAAGACTTATTTGGATTTxan Avxan Or ? Aoreg UA Toreg In T Toreg OR T G . . .oreg Br T Coreg Me T Coreg He "...? Tp la t Bl Tp l a t Ar Tp l a t Ha T TTp l a t Ma T TT Gcroc Te .G T TTcroc Ve T TTklau Cr Tklau CW1 Tklau CW2 Tklau SJS Tesch SBaesch SBeesch CWpi eta T T APlethodon T T. .T C . . C . G . . C . . .Aneides T? TCT ?????????????


Estimates of corrected DN A sequence dif-ference are relatively low within five sub-species: picta (0.0%, n = 2), xanthoptica (1.6-2.5%, n = 3), eschscholtzii (1.0-2.7%, n = 3),klauberi (0.6-2.4%, n = 4), and croceater (1.2%,n = 2). The other two subspecies are strong-ly differentiated; sequence differenceswithin oregonensis range from 1.3 to 14.5%{n = 5) and those within platensis are 2.4-14.0% (n = 4). The variation in sequencedifferences within subspecies does not ap-pear to be due to different scales of sam-pling. Some widely separated sampleswithin oregonensis and within eschscholtziihave similar sequences, whereas some geo-graphically close samples within oregonen-sis and within platensis have very differentsequences (cf. Table 1 and Fig. 1).

Phylogenetic Analysis: Character Evaluation

For phylogenetic analysis, each nucleo-tide was treated as a character with up tofour unordered states. Phylogenetically in-formative variation is present at 184 of the250 variable positions (Fig. 3). For the 250variable positions, 52 changes (21%) are inthe first codon position, 25 (10%) in thesecond, and 173 (69%) in the third. Tran-

sitions, especially between C and T, out-weigh transversions; biases > 10:1 are com-mon among the more similar genomes.

To test for saturation of base substitu-tions (i.e., multiple hits), transitions andtransversions at each codon position wereplotted against maximum-likelihood esti-mates of relative divergence (Fig. 4; F. Vil-lablanca, W. K. Thomas, and A. C. Wilson,submitted). These divergence estimates areanalogous but not equivalent to sequencedivergence and take into account devia-tions from even base composition, differ-ences in substitution rate among codon po-sitions, and any bias toward transitions(Felsenstein, 1990). Values for these pa-rameters derived from the data are 2:1:7changes at the lst:2nd:3rd codon positionsand a 10:1 transition/tranversion ratioamong similar sequences. Nonlinearity inthe plots indicates increasing saturation ofsubstitutions for that class of characters.

The plots (Fig. 4) indicate that increasesin first and third codon position transitionsare nonlinear in comparison to relative di-vergence estimates. This nonlinearity, in-dicative of saturation effects, is particularlyobvious for comparisons between Ensatina

at North C



Dow

nloaded from




TABLE 1. Corrected (Brown et al., 1982) estimates of sequence divergence among samples of Ensatina andthe outgroups. The two samples of picta are identical over the 681 bp analyzed and are represented once here.Localities are shown in Figure 1 and defined in the Appendix.

1. xan SR2. xan Av3. xan Or4. oreg OR5. oreg In6. oreg WA7. oreg Br8. oreg Me9. oreg He

10. plat Bl11. plat AT12. plat Ha13. plat Ma14. croc Te15. croc Ve16. klau Cr17. klau CW118. Jt/flw CW219. klau SJS20. escb SBa21. esch SBe22. esch CW23. picfo24. Plethodon25. Aneides

1

1.82.6

13.912.914.716.214.314.312.915.614.514.013.212.813.715.114.813.67.36.96.5

12.839.736.9

xanthoptica

2

1.613.012.213.815.212.813.611.614.213.512.812.412.013.214.414.113.26.56.25.8

11.537.834.9

3

—

12.011.112.914.412.612.410.713.312.812.412.011.612.313.212.812.35.75.65.3

11.438.035.3

4

—

8.51.3

13.511.05.58.98.4

11.511.510.610.19.7

10.710.310.010.810.59.97.7

33.434.0

5

—

9.414.311.910.410.410.911.010.39.28.5

10.311.210.710.410.29.28.99.0

38.534.7

oregonensis

6

—

14.511.85.99.79.2

12.412.511.110.69.9

11.010.510.311.411.110.58.3

34.432.9

7

—

3.614.412.114.816.015.616.214.715.716.515.915.513.714.213.215.039.235.4

8

—11.510.513.213.713.214.013.113.914.613.913.812.012.511.712.537.434.9

9

—

8.811.112.612.511.811.111.812.611.911.810.910.39.6

10.134.231.5

10

3.812.612.211.610.910.711.611.110.810.48.88.49.2

32.733.7

platensis

11

—

14.014.012.812.111.210.811.511.112.010.810.79.4

34.437.1

12

—

2.47.17.7

11.112.111.611.511.410.910.312.434.340.1

13

6.87.3

10.711.610.911.010.910.510.211.635.238.9

and the outgroups. The high level of first-base transitions appears to be due to thelarge proportion (11.2%) of leucines in thesequence. For this base, C <-> T transitionsat the first position are silent. The otherclasses of characters (first-base transver-sions, second-base transitions and trans-versions, and third-base tranversions) ap-pear to increase linearly with themaximum-likelihood distance estimate.

Phylogenetic Analysis: Estimation ofRelationships

Because of the evidence for saturation offirst and third codon position transitionsamong the more divergent sequences, weestimated the most-parsimonious tree intwo steps. First, the obviously saturatedclasses of information were excluded to es-timate the root and the branching patternat the base of the tree. To do this, we useda 0:1 step matrix to discount transitions atthe first and third codon positions (Swof-

ford and Olsen, 1990). Second, the shortesttopology obtained was used to constrainthe basal branching order in an analysisusing all classes of characters, but with a10:1 step matrix to weight transversions(the more "conservative" substitutions)over transitions. This two-step approachminimizes noise at the base of the tree dueto character homoplasy while using all ofthe information to resolve relationshipsamong more similar sequences (Swoffordand Olsen, 1990).

The analysis excluding first- and third-base transitions produced a single shortesttree of 163 steps (Fig. 5b). Sequences fromthe northern populations of platensis arebasal, followed by southwestern oregonen-sis (Branscomb/Navarro), then northernoregonensis and picta, and then a group con-sisting of one oregonensis (Ingot) and theother subspecies. The number of unambig-uous changes defining the basal branchesis small, particularly for those involving

at North C



Dow

nloaded from




croceater

14 15 16

klauberi

17 18

TABLE 1.

19

Extended.

eschsclwltzii

20 21 22

picta

23

Plethodon

24

Aneides

25

1.29.7

10.710.210.110.710.09.3

11.336.334.8

—10.211.410.710.510.09.58.9

10.337.534.3

—2.41.90.6

10.910.69.5

10.035.436.5

—0.82.2

11.811.410.310.935.538.1

—1.5

11.311.09.9

10.534.937.4

—11.110.69.5

10.234.836.7

—2.72.4

10.735.834.4

—1.09.6

38.034.7

—9.0

36.933.6

_36.135.2 32.0

oregonensis (Fig. 5b). Exchanging positionsamong basal taxa (oregonensis and northernplatensis) results in trees one to three stepslonger.

In the second step of the analysis, whichuses all classes of characters, the first threebranches of the ingroup were constrainedas described above. The single shortest tree(Fig. 5a) had 1,839 steps, counting eachtransversion as 10 steps. The only substan-tial difference between this and the "basal"tree (Fig. 5b) is in the location of oregonensisfrom Ingot. Relationships within clades aremore clearly resolved, with evidence formonophyly of eschscholtzii, clearer resolu-tion within klauberi, and the exclusion ofpicta from the clade consisting of north-coastal oregonensis. The same topology isobtained using the same constraint with-out differential weighting of transversionsand transitions.

Using no constraints results in twoshortest trees of 1,839 steps. One is the same

as the single shortest tree above (Fig. 5a).The alternative topology (Fig. 5c) definesthe same major clades, except that com-ponents of oregonensis (and picta) are dif-ferently located. We regard the topologyin Figure 5a as a better estimate of phy-logeny because the branching order of bas-al mtDNAs is not compromised by classesof information that can have high homo-plasy (saturation) among such highly di-vergent sequences.

Aside from instability in the branchingorder of the oregonensis groups, most cladesare strongly supported by a large numberof unambiguous character states, have highrepeatability in bootstrap replications, andare consistent among analyses (Fig. 5).

The major features of the "best-estimate"tree (Fig. 5a) are (1) monophyly of mtDNAsfrom each of the subspecies xanthoptica,eschscholtzii, klauberi, croceater, and picta, (2)identification of a monophyletic southerncoastal clade (xanthoptica and eschscholtzii)

at North C



Dow

nloaded from




60 1(a)

60(b) o 1st position

A 2nd position• 3rd position

0.00 0.05 0.10

Maximum-likelihood DNA distance0.00 0.05 0.10

Maximum-likelihood DNA distance

FIGURE 4. Plots of the numbers of transitions (a) and transversions (b) at the first, second, and third codonpositions against maximum-likelihood indices of corrected DNA divergence (ML option in DNADIST, PHY-LIP). Linear regressions are shown for all but first- and third-base transitions. For these, second-order poly-nomial regressions are shown. Partial F statistics for second-order polynomials (with degrees of freedomcorrected for nonindependence) are F20 = 4.12, 0.1 > P > 0.05, for first-position transitions and F20 = 26.914,P < 0.001, for third-position transitions. All other P values are >0.1, indicating that a linear model is moreappropriate for transitions at the second codon position and transversions at all positions. The separate cloudsof points for large divergences are the comparisons between Ensatina and the outgroups.

and a southern Sierran clade (klauberi, cro-ceater, and southern platensis), and (3) para-phyly of mtDNAs from both platensis andoregonensis (see Fig. 6).

Phylogenetic Analysis:Tests of A Priori Models

An alternative way to use the data is totest whether they are consistent with apriori predictions based on competing evo-lutionary hypotheses. If the ring is actuallya montage of secondary contacts amongstrongly differentiated subspecies, theneach of the subspecies should, barring in-trogression and assuming isolation longerthan ANf generations, be monophyletic formtDNA. To test this prediction, wesearched for the shortest trees satisfyingthe constraint that the mtDNAs from eachsubspecies are monophyletic. The relation-ships among subspecies are free to vary.The single shortest tree produced underthis constraint (Fig. 7a) has 1,918 steps,which is 79 steps longer than the best-es-

timate phylogeny (Fig. 5a). The best-esti-mate phylogeny requires fewer steps for33 characters, whereas the monophylymodel is more parsimonious for 10 char-acters. Using a winning sites test (Pragerand Wilson, 1988), we can reject the mono-phyly model (x2 = 12.3, P < 0.001).

To test the scenario proposed by Steb-bins (1949), a simple model based on in-dependent and stepwise colonization of theSierra Nevada and the southern Coastalranges from appropriately located ances-tral populations of oregonensis was con-structed (Fig. 7b). This model assumes thatpopulations were established in a gradualstepping-stone manner, with each propa-gule bearing derived states from its sourcepopulation. This model requires 1,913 steps,74 steps more than the minimum-lengthtrees. Whereas the shorter tree is more par-simonious for 24 characters, the coloniza-tion model provides a better explanationfor 14 characters. On this basis the modelcannot be rejected (x2 = 2.63, 0.10 > P >

at North C



Dow

nloaded from




(a)

22

456

19100

464

20100

561

21100

8100

23

p-dz3 1

91 1—-—

5 n -99 L_L_9 1 4

100 L_210 1 9

100 1—£—

240_

0_

4_

14

9

5

55

53

xan Or

xan Av

xan SR

esch CW

esc/7 SBe

esc/> SBa

Wau CW2

Wau CW1

Wau SJS

Wau Cr

croc Ve

croc Te

p/af Ma

plat Ha

crep In

oreg WA

oreg OR

oreg He

picta WC

p/cfa LD

oreg Me

oreg Br

p/af Ar

plat Bl

Aneides

Plethodon

(b)

oreg WA

oreg OR

p/cfa WC

picta LD

oreg He

oreg Me

oreg Br

p/af Ar

plat Bl

Aneides

Plethodon

FIGURE 5. Minimum-length phylogenetic trees for mtDNAs from Ensatina. Abbreviations follow the Ap-pendix, (a) "Best-estimate" tree based on all classes of characters, with the three most basal branches of theingroup constrained and a 10:1 step matrix weighting transversions over transitions. The number above eachbranch is the number of unambiguous characters defining the branch. The number below is the proportionof 100 bootstrap replicates (with 10:1 step matrix applied) in which the clade to the right was present. Dashedlines indicate poorly resolved branches, (b) Shortest tree obtained using step matrices to exclude first- andthird-base transitions to resolve the branching pattern at the base of the tree. The number above each branchis the number of unambiguous characters defining the branch, (c) Alternative equally parsimonious topologyobtained from an analysis of all classes of characters with a 10:1 step matrix but no constraint on the orderof basal branches.

at North C



Dow

nloaded from




0.05), although the difference approachesstatistical significance.

A closer fit between the data and thecolonization model proposed by Stebbinsis achieved by allowing mtDNAs fromsouthern platensis and croceater to be mono-phyletic, leaving the other details the same(Fig. 7c). This model requires 1,879 steps,40 steps more than the shortest tree, andis reasonably consistent with the data (16vs. 13 sites; x2 = 0.31, P > 0.10).

Phylogenetic Analysis: Evaluation ofSampling Effects

Except for two individuals of klauberifrom Camp Wolahi, the above analyses arebased on one individual per locality. Totest whether a single individual is repre-sentative or whether sampling of multipleindividuals is likely to reveal divergentmtDNAs from the same locality, restrictionenzymes were used to screen multiple sam-ples of amplified DNA (using primers 15-18; Fig. 2) from three populations of ore-gonensis from northern California, the geo-graphic region with the highest mtDNAdiversity. Scanning the available sequenc-es (Fig. 3) revealed that digestion with Hmfland Mspl can diagnose each of the majormtDNA lineages identified in this region(Table 2). Each of the six restriction sitesscreened included one or more phyloge-netically informative nucleotide positions.These sites are uniform for each of thewithin-population samples (In, Me, Salyer[Trinity Co., CA]; Table 2), suggesting thatthe bulk of the variation is distributedamong rather than within populations. Thishypothesis is consistent with similarity ofsequences from some geographically prox-imal populations (cf. Table 1 and Fig. 1).Thus, a sequence from a single individualis apparently representative for a popula-tion of these salamanders, at least for thepurposes of broad-scale biogeographic andphylogenetic comparisons.

DISCUSSION

Biogeographic Implications

Stebbins (1949) proposed that Sierran andsouth-coastal populations of Ensatina were

FIGURE 6. Simplified interpretation of the mtDNAphylogeny illustrating the subspecies with mono-phyletic mtDNAs (circled) and those with potentiallyor demonstrably paraphyletic mtDNA (*). mtDNAsfrom croceater and picta are enclosed by dashed linesto indicate the samples were taken from small geo-graphic areas.

separately derived from northern popu-lations, perhaps from those in the redwoodand Douglas fir forests of northern Cali-fornia and southwestern Oregon. This pre-diction was largely based on the observa-tions that these populations have the mostgeneralized color pattern and the highestpopulation densities. Stebbins regardedpicta as a close approximation of the an-cestral state.

The evidence from mtDNA sequencessubstantiates three critical components ofStebbins's (1949) biogeographic model and,therefore, of his evolutionary scenario forEnsatina. The first component is that pop-ulations from northern California were an-cestral to the Sierran and south-coastal ra-diations. Populations from the top of thering show the highest regional diversityin mtDNAs; these sequences are consis-tently basal in the phylogenetic analyses.In the best-estimate phylogeny (Fig. 5a),mtDNAs from oregonensis are paraphyleticwith respect to other Ensatina. The veryinstability of the mtDNA relationshipsamong the three groups of oregonensis andother Ensatina is to be expected if oregonen-sis was ancestral, with extant lineages hav-ing mainly primitive or uniquely derivedcharacter states. These mtDNAs are eachseparated from others by long branches thatare expected to complicate phylogeneticanalyses (Felsenstein, 1978; Swofford andOlsen, 1990). Increasing the number of

at North C



Dow

nloaded from




oreg In

oreg Br

oreg Me

oreg WA

oreg OR

oreg He

plat Ha

plat Ma

p/af Bl

p/af Ar

klau Cr

Wau SJS

klau CW1

Wat/ CW2

p/cfa WC

picta LD

xan SR

xan Av

xan Or

esc/) SBa

esc/7 CW

esc/7 SBe

croc Te

croc Ve

Plethodon

Aneides

(b) Wau Cr

Wau SJS

klau CW1

Wau CW2

croc Ve

croc Te

p/af Ma

plat Ha

p/af Ar

plat Bl

oreg! In

esc/) SBa

esc/) CW

esc/) SBe

xan SR

xan Av

xan Or

oreg Br

oreg Me

oreg WA

oreg OR

oreg He

r— picta WC

p/'efa LD

Plethodon

Aneides

(c)Wau Cr

klau SJS

Wau CW1

klau CW2

croc Ve

croc Te

plat Ma

p/af Ha

plat Ar

p/af Bl

oreg In

esch SBa

esc/7 CW

esch SBe

xan SR

xan Av

xan Or

oreg Br

oreg Me

oreg WA

oreg OR

oreg He

p/cfa WC

p/cfa LD

Plethodon

Aneides

FIGURE 7. Shortest trees consistent with (a) all subspecies being monophyletic, (b) a simple model ofstepwise colonization down each side of the central valley of California, and (c) a modification of the colo-nization model in which southern platensis populations (plat Ma, plat Ha) were the sister group to samples ofcroceater. The phylogeny shown in (c) was the most consistent with the data. Abbreviations follow theAppendix.

at North C



Dow

nloaded from




TABLE 2. Distribution of restriction sites for Hind and Mspl within the 681-bp segment of cytochrome bfrom northern Californian Ensatina. Localities are shown in Figure 1 and defined in the Appendix.

Sample Typea

Restriction siteb

Hmfl

98 457 473 614

Mspl

125 593

Among populationspicta Aoreg He Boreg Br/Me Coreg In Dplat Bl Eplat Ar F

Within populations

oreg In (« = 4) Doreg Me (n = 12) Coreg Salyer (n = 12) F

010000

000

1—1

00000

000

000100

100

011101

111

010111

01

a Scanning the sequences (Fig. 3) for Hinil and Mspl sites revealed that each of the major lineages could be diagnosed bydigestion with these two enzymes (types A-F). Digestion profiles of multiple individuals from each of three localities revealedone type per locality.

b Restriction sites are numbered from the first base of the sequences (Fig. 3).

samples and the length of each sequencefailed to improve resolution at the base ofthe tree. However, the mtDNA from picta,thought to be ancestral by Stebbins (1949),is unremarkable. It simply appears as oneamong many basal and strongly differen-tiated lineages in the area.

The second component supported by themtDNA evidence is that the blotched sal-amanders of the inland ranges and the un-blotched coastal forms represent indepen-dent radiations. As expected, mtDNAs fromthe coastal subspecies, xanthoptica and esch-scholtzii, form a strongly supported mono-phyletic group; croceater, klauberi, andsouthern platensis also form a monophy-letic group, but it is less well supported.This observation precludes two alternativebiogeographic scenarios that are not com-patible with the ring-species concept. Oneis that xanthoptica could be derived fromthe inland races, having spread from eastto west to make recent secondary contactwith coastal forms of Ensatina. The other isthat klauberi could be derived from south-ern eschscholtzii.

However, the simplest biogeographicscenario, which assumes stepwise coloni-zation down each side of the valley (Fig.7b), is not well supported (although a mi-nor modification of it [Fig. 7c] is). This dis-

crepancy is largely due to the diversity ofmtDNA within platensis. Two very distinctmtDNA lineages were found: one in north-ern platensis and the other in the southernsamples. The former was consistently basalin the phylogenetic analyses, whereas thelatter was highly derived and more closelyrelated to croceater than to the northernclade.

One hypothesis to explain the dichoto-my of mtDNA within platensis is that thenorthern and southern groups stem fromtwo divergent lineages of mtDNA inde-pendently derived from southern oregon-ensis. One lineage has persisted and diver-sified during the colonization of thesouthern Sierras. The other is (as yet) foundonly as far south as the central Sierra Ne-vada. This interpretation requires lineagesorting (Avise et al., 1984; Neigel and Av-ise, 1986) but with unusually high levelsof mtDNA differentiation within the an-cestor. This is precisely the situation in ex-tant populations of oregonensis from north-ern California. This hypothesis predictsthat further sampling of oregonensis willreveal mtDNAs that are closely related tothe divergent lineages found within pla-tensis.

The third essential component of Steb-bin's (1949) model was that diversification

at North C



Dow

nloaded from




within Ensatina was gradual, as suggestedby the increasing levels of isolation be-tween blotched and unblotched salaman-ders from north to south, but there was noindication of the time involved. The largedivergences among Ensatina mtDNAs, aswell as the large intraspecific genetic dis-tances for allozymes (Wake and Yanev,1986), are consistent with a long evolu-tionary history. The great diversity ofmtDNA within oregonensis (estimated di-vergences up to 14.5%; Table 1) suggeststhat this is an ancient group. The levels ofmtDNA differentiation of croceater +southern platensis from klauberi (10.2-11.6%;Table 1) and of eschscholtzii from xanthoptica(5.8-7.3%; Table 1) suggest that each of thesouthern radiations has been divergingover a long period. However, the low di-vergence of both mtDNA and allozymes(Wake and Yanev, 1986) among geograph-ically distant samples of eschscholtzii andbetween coastal and inland populations ofxanthoptica suggests that the spread of thesesubspecies has been relatively recent. Thus,both of the well-documented secondarycontacts within Ensatina, i.e., between xan-thoptica and platensis and between esch-scholtzii and klauberi (reviewed in Wake etal., 1989), may be relatively recent.

The general support from the mtDNAevidence for the biogeographic scenarioproposed by Stebbins (1949) substantiatesEnsatina as a ring species and thus as anexemplar of gradual microevolutionary di-vergence leading toward speciation. ThemtDNA evidence confirms that klauberi andeschscholtzii are the end points of two in-dependently evolving lineages that arenow genetically and in some places (Brown,1974; Wake et al., 1986) reproductively iso-lated.

Implications for Systematics: mtDNA andSpecies Boundaries

The number and distribution of speciesin Ensatina has long been contentious(Stebbins, 1949); recent discussion has fo-cused on the implications of different spe-cies concepts (e.g., Frost and Hillis, 1990).The observed monophyly of mtDNAs fromsome subspecies, notably klauberi, could be

taken as support for the recognition as in-dependent evolutionary lineages, i.e., spe-cies, under some definitions. However, be-cause of differences in the behavior oforganellar and nuclear genes, the use oforganellar genes alone to identify speciesboundaries can lead to errors.

Because of its clonal and maternal in-heritance, mtDNA has a lower effectivepopulation size and can be strongly dif-ferentiated geographically, whereas nucle-ar genes cannot (Birky et al., 1983, 1989).This discrepancy will be exaggerated wherefemales disperse less often than males, asappears to be the case in Ensatina (Stebbins,1954; Wake, unpubl. data). Thus, in old,geographically subdivided taxa, phyloge-netic analysis could reveal geographicallydiscrete monophyletic groups of mtDNAs(i.e., phylogeographic category 1 of Aviseet al., 1987). However, such an analysiscannot identify species borders becauseputative taxa may be united by nuclear geneflow. We have uncovered such cases in En-satina, e.g., among populations of oregonen-sis along a transect in northern Californiaand within platensis in the central SierraNevada (Jackman and Wake, in prep.). Theopposite situation, where otherwise dis-crete taxa are paraphyletic for mtDNA, iswell known (e.g., Ferris et al., 1983; Whitte-more and Schall, 1991) and appears to bethe case for platensis and oregonensis, whichwere regarded by Stebbins (1954) as mor-phologically separable but are shown hereto be paraphyletic for mtDNA. The use ofmtDNA patterns without corroborationfrom other evidence would lead to incor-rect conclusions about both genetic isola-tion and cohesion within Ensatina. Gener-ally, we caution against using mtDNApatterns as the sole criterion for determin-ing species boundaries.

ACKNOWLEDGMENTS

We thank C. Orrego for excellent technical advice;D. A. Good for technical assistance; J. L. Patton, M.F. Smith, F. X. Villablanca, and C. A. Meacham forassistance with data analysis; A. Larson, D. M. Hillis,S. V. Edwards, C. Orrego, F. X. Villablanca, D. A.Good, and T. R. Jackman for comments on the manu-script; and R. C. Stebbins for his continuing interest,encouragement, and discussion. This study was sup-

at North C



Dow

nloaded from




ported by grants to D.B.W. from the Berkeley CampusBiomedical Support Program, the Berkeley Commit-tee on Research, and the National Science Foundation(BSR 9019810) and by a travel grant (to CM.) fromthe University of Queensland.

REFERENCES

AVISE, J. C. 1989. Gene trees and organismal phy-logenies: A phylogenetic approach to populationbiology. Evolution 43:1192-1208.

AVISE, J. C, J. ARNOLD, R. M. BALL, E. BERMINGHAM,T. LAMB, J. E. NEIGEL, C.A. REEB, AND N. C. SAUNDERS.1987. Intraspecific phylogeography: The mito-chondrial DNA bridge between population genet-ics and systematics. Annu. Rev. Ecol. Syst. 18:489-522.

AVISE, J. C, J. E. NEIGEL, AND J. ARNOLD. 1984. De-mographic influences on mitochondrial DNA lin-eage survival in animal populations. J. Mol. Evol.20:99-105.

BAVERSTOCK, P. R., AND C. MORITZ. 1990. Samplingdesign. Pages 13-24 in Molecular systematics (D. M.Hillis and C. Moritz, eds.). Sinauer, Sunderland,Massachusetts.

BlRKY, C. W., P. FUERST, AND T. MARUYAMA. 1989.

Organelle gene diversity under migration, muta-tion, and drift: Equilibrium expectations, approachto equilibrium, effects of heteroplasmic cells, andcomparisons to nuclear genes. Genetics 121:613-627.

BIRKY, C. W., T. MARUYAMA, AND P. FUERST. 1983. An

approach to population and evolutionary genetictheory for genes in mitochondria and chloroplasts,and some results. Genetics 103:513-527.

BROWN, C. W. 1974. Hybridization among the sub-species of the plethodontid salamander Ensatinaeschscholtzii. Univ. Calif. Publ. Zool. 98:1-57.

BROWN, C. W., AND R. C. STEBBINS. 1964. Evidence

for hybridization between blotched and un-blotched subspecies of the salamander Ensatina esch-scholtzii. Evolution 18:706-707.

BROWN, W. M. 1985. The mitochondrial genome ofanimals. Pages 95-130 in Molecular evolutionarygenetics (R. J. Maclntyre, ed.). Plenum, New York.

BROWN, W. M., E. M. PRAGER, A. WANG, AND A. C.

WILSON. 1982. Mitochondrial DNA sequence ofprimates: Tempo and mode of evolution. J. Mol.Evol. 18:225-239.

CARSON, H. L., AND A. R. TEMPLETON. 1984. Genetic

revolutions in relation to speciation phenomena:The founding of new populations. Annu. Rev. Ecol.Syst. 15:97-132.

DOBZHANSKY, T. 1958. Species after Darwin. Pages19-55 in A century after Darwin (S. A. Barnett, ed.).Heinemann, London.

FELSENSTEIN, J. 1978. Cases in which parsimony andcompatibility methods will be positively mislead-ing. Syst. Zool. 27:27-33.

FELSENSTEIN, J. 1990. PHYLIP manual, version 3.3.University Herbarium, Univ. California, Berkeley.

FERRIS, S. D., R. D. SAGE, C.-H. HUANG, J. T. NIELSEN,

U. RITTE, AND A. C. WILSON. 1983. Flow of mito-

chondrial DNA across a species boundary. Proc.Natl. Acad. Sci. USA 80:2290-2294.

FROST, D. R., AND D. M. HILLIS. 1990. Species inconcept and practice: Herpetological applications.Herpetologica 46:87-104.

GOLDSCHMIDT, R. 1933. Some aspects of evolution.Science 78:539-547.

GOULD, S. J., AND N. ELDREDGE. 1977. Punctuatedequilibria: The tempo and mode of evolution re-considered. Paleobiology 3:115-151.

GYLLENSTEN, V., AND H. ERLICH. 1988. Generation ofsingle-stranded DNA by the polymerase chain re-action and its applications to direct sequencing ofthe HLA DQa locus. Proc. Natl. Acad. Sci. USA 85:7652-7656.

KOCHER, T. D., W. K. THOMAS, A. MEYER, S. V. ED-WARDS, S. PAABO, F. X. VlLLABLANCA, AND A. C.WILSON. 1989. Dynamics of mitochondrial DNAevolution in animals: Amplification and sequenc-ing with conserved primers. Proc. Natl. Acad. Sci.USA 86:6196-6200.

LARSON, A., D. B. WAKE, L. R. MAXSON, AND R. HIGHTON.

1981. A molecular phylogenetic perspective on theorigins of morphological novelties in the salaman-ders of the tribe Plethodontini (Amphibia, Pleth-odontidae). Evolution 35:405-422.

MANIATIS, T., E. F. FRISCH, AND J. SAMBROOK. 1982.

Molecular cloning: A laboratory manual. ColdSpring Harbor Laboratory Publications, Cold SpringHarbor, New York.

MAYR, E. 1963. Animal species and evolution. Bel-knap Press, Cambridge, Massachusetts.

MAYR, E. 1982. Processes of speciation in animals.Pages 1-19 in Mechanisms of speciation (C. Bari-gozzi, ed.). Alan R. Liss, New York.

NEIGEL, J. E., AND J. C. AVISE. 1986. Phylogenetic

relationships of mitochondrial DNA under variousdemographic models of speciation. Pages 515-534in Evolutionary processes and theory (E. Nevo andS. Karlin, eds.). Academic Press, New York.

PRAGER, E. M., AND A. C. WILSON. 1988. Ancient

origin of lactalbumin from lysozyme: Analysis ofDNA and amino acid sequences. J. Mol. Evol. 27:326-335.

RENSCH, B. 1929. Das Prinzip geographisher Ras-senkreise und des Problem der Artbildung. Born-traeger, Berlin.

ROE, B. A., D.-P. MA, R. K. WILSON, AND J. F.-H. WONG.

1985. The complete nucleotide sequence of theXenopus laevis mitochondrial DNA genome. J. Biol.Chem. 260:9759-9774.

SANGER, F., S. NICKLEN, AND A. R. COULSON. 1977.

DNA sequencing with chain-terminating inhibi-tors. Proc. Natl. Acad. Sci. USA 74:5463-5467.

STEBBINS, R. C. 1949. Speciation in salamanders ofthe plethodontid genus Ensatina. Univ. Calif. Publ.Zool. 48:377-526.

STEBBINS, R. C. 1954. Natural history of the sala-manders of the plethodontid genus Ensatina. Univ.Calif. Publ. Zool. 54:47-124.

STEBBINS, R. C. 1957. Intraspecific sympatry in thelungless salamander Ensatina eschscholtzii. Evolution11:265-270.

at North C



Dow

nloaded from




SWOFFORD, D. L. 1991. PAUP manual, version 3.0.Illinois Natural History Survey, Champaign.

SWOFFORD, D. L., AND G. J. OLSEN. 1990. Phylogenyreconstruction. Pages 411-501 in Molecular system-atics (D. M. Hillis and C. Moritz, eds.). Sinauer,Sunderland, Massachusetts.

WAKE, D. B. 1963. The comparative osteology of theplethodontid salamander genus Aneides. J. Mor-phol. 113:77-118.

WAKE, D. B., AND K. P. YANEV. 1986. Geographicvariation in allozymes in a "ring species," the pleth-odontid salamander Ensatina eschscholtzii of westernNorth America. Evolution 40:702-715.

WAKE, D. B., K. P. YANEV, AND C. W. BROWN. 1986.Intraspecific sympatry in a "ring species," theplethodontid salamander Ensatina eschscholtzii ofsouthern California. Evolution 40:866-868.

WAKE, D. B., K. P. YANEV, AND M. M. FRELOW. 1989.Sympatry and hybridization in a "ring species":The plethodontid salamander Ensatina eschscholtzii.Pages 134-157 in Speciation and its consequences(D. Otte and J. A. Endler, eds.). Sinauer, Sunder-land, Massachusetts.

WHITE, M. J. D. 1982. Rectangularity, speciation andchromosome architecture. Pages 74-104 in Mecha-nisms of speciation (C. Barigozzi, ed.). Alan R. Liss,New York.

WHITTEMORE, A. T., AND B. A. SCHAAL. 1991. Inter-specific gene flow in symaptric oaks. Proc. Natl.Acad. Sci. USA 88:2540-2544.

Received 7 September 1991; accepted 11 January 1992

APPENDIX

Abbreviations, collecting localities, and MVZ cat-alog numbers for subspecies of Ensatina eschscholtziifrom which mtDNA was sequenced (see Fig. 1): 1:oreg WA: Granite Falls, Snohomish Co., Washington,MVZ 168659; 2: oreg OR, Wolf Creek Rd., Lane Co.,Oregon, MVZ 168668; 3: picta LD: Low Divide Rd.,Del Norte Co., California, MVZ 172513; 4: picta WC:Wilson Creek Rd., Del Norte Co., California, MVZ168709; 5: oreg He: Helena, Trinity Co., California,MVZ (S)10938; 6: oreg Me: Navarro, Mendocino Co.,California, MVZ 194896; 7: oreg Br: Branscomb Rd.,Mendocino Co., California, MVZ 194708; 8: oreg In:Ingot, Shasta Co., California, MVZ 182000; 9: xan SR:Santa Rosa, Sonoma Co., California, MVZ 205681; 10:xan Or: Orinda, Contra Costa Co., California, MVZ163850; 11: esch SBa: Zaca Creek, Santa Barbara Co.,California, MVZ 167654; 12: esch SBe: Millard Canyon,San Bernardino Mtns., Riverside Co., California, MVZ181460; 13: esch CW: Alpine (MVZ 178729) and klauCWl and klau CW2: Camp Wolahi (MVZ 191684, MVZ194908), San Diego Co., California; 14: plat Bl: Blod-gett, El Dorado Co., California, MVZ 172459; 15: platAT: Arnold, Calaveras Co., California, MVZ 158006;16: xan Av: Avery, Calaveras Co., California, MVZ202316; 17: plat Ma: Gooseberry Flat Rd., Madera Co.,California, MVZ 169033; 18: plat Ha: Hartland, TulareCo., California, MVZ 169165; 19: croc Te: TehachapiMtn. Park, Kern Co., California, MVZ 202330; 20: crocVe: Alamo-Little Mutau Creek, Ventura Co., Califor-nia, MVZ 195607; 21: klau Cr: Crystal Creek, San Ber-nardino Co., California, MVZ 185823; 22: klau SJS:Santa Rosa Mountain Rd., Riverside Co., California,MVZ 185844.

at North C



Dow

nloaded from



EVOLUTIONARY RELATIONSHIPS WITHIN THE ENSATINA...

Documents

Transcript of EVOLUTIONARY RELATIONSHIPS WITHIN THE ENSATINA...