MOLECULARPHYLOGENETICSAND

Molecular Phylogenetics and Evolution 32 (2004) 375–382

EVOLUTION

www.elsevier.com/locate/ympev

Molecular phylogeny of the antitropical genusPseudolabrus (Perciformes: Labridae): evidence

for a Southern Hemisphere origin

Kohji Mabuchi,a,* Tetsuji Nakabo,b and Mutsumi Nishidaa

a Ocean Research Institute, University of Tokyo, 1-15-1 Minamidai, Nakano, Tokyo 164-8639, Japanb The Kyoto University Museum, Kyoto University, Yoshida, Sakyo, Kyoto 606-8501, Japan

Received 1 October 2003; revised 6 January 2004

Available online 1 March 2004

Abstract

The genus Pseudolabrus comprises 11 species of marine nearshore fishes which are antitropically distributed: two species occur in

East Asia, the remaining nine species being distributed in the Southern Hemisphere, mainly in the temperate Pacific. The distri-

butions of their closely allied genera, collectively called ‘‘pseudolabrines’’ are, however, restricted to the Australia-New Zealand

region. The molecular phylogeny of six of the 11 Pseudolabrus species from both Hemispheres and four of the five other pseudo-

labrine genera was reconstructed from nucleotide sequence data from mitochondrial DNA 12S rRNA, tRNAVal, and 16S rRNA

genes. Both parsimony and Bayesian analyses were performed. Results are not consistent with a previous phylogenetic hypothesis

based on osteological data, particularly in the relationship between Pseudolabrus and Notolabrus, indicating a probable need for

reviewing the status of Notolabrus (or the delimitations of both Pseudolabrus and Notolabrus). The two Northern Hemisphere

species of Pseudolabrus were monophyletic and nested deep into the clade of the Southern Hemisphere pseudolabrines, which in-

dicates that both pseudolabrines and Pseudolabrus originated in the Southern Hemisphere. A dispersal rather than vicariance ex-

planation for the antitropical distribution of Pseudolabrus is more parsimonious given the number of dispersal events, extinctions,

and evolutionary adaptations required under the phylogeny. Based on molecular clock calibrations, the transequatorial divergence

was suggested to be early to mid Pliocene at the earliest.

� 2004 Elsevier Inc. All rights reserved.

Keywords: Antitropical distribution; Labridae; Pseudolabrus; Notolabrus; Mitochondrial DNA

1. Introduction

The occurrence of the same taxa to north and south

of the tropics but not the intervening region is one of themost interesting phenomena of biogeography, and such

disjunct distributions are observed for a variety of taxa,

both terrestrial and aquatic (Briggs, 1995; Randall,

1981). There are many variants to this kind of distri-

bution, especially related to latitude: polar, boreal, cool

temperature, warm temperature, and subtropical. For

all these patterns collectively, Hubbs (1952) proposed

the encompassing term ‘‘antitropical,’’ and the term hascome into general use (Randall, 1981). As seen in the

* Corresponding author. Fax: +81-3-5351-6822.

E-mail address: mabuchi@ori.u-tokyo.ac.jp (K. Mabuchi).

1055-7903/$ - see front matter � 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.ympev.2004.01.008

fact that Darwin (1859) already mentioned it in his ‘‘On

the origin of species,’’ these disjunct distribution patterns

have long-attracted biologists as interesting problems to

resolve.Several hypotheses have been proposed for antitro-

pical distributions of marine taxa (Burridge, 2002).

Species intolerant of tropical temperatures may have

dispersed across equatorial waters at shallow depths

during periods of cooler climate, such as glaciations

(Ekman, 1953; Lindberg, 1991), or by descent into

cooler waters at tropical latitudes (Ekman, 1953; Hubbs,

1952). Dispersal may have also been facilitated overlonger time scales via island integration (Rotondo et al.,

1981; Springer, 1982). Alternatively, vicariance mecha-

nisms may have isolated Northern and Southern

Hemisphere taxa by extirpation of intervening tropical

Fig. 1. The antitropical distribution of Pseudolabrus, based on Russell (1988) and Mabuchi and Nakabo (1997). Regions with single asterisks indicate

overlapping distributions and species with double asterisks were not examined in the present study.

376 K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382

relatives. Possible causes of such extirpation include

competition with younger, superior tropical species(Briggs, 1987), the submergence of equatorial islands

(Rehder, 1980), and a mid-Miocene rise in equatorial

temperatures (Valentine, 1984; White, 1986). More

ancient vicariant events, such as fragmentation of

Pangea or Pacifica (Crame, 1993; Nelson, 1985) are also

proposed to explain the origin of some antitropicality.

Reconstruction of inter- and intraspecific phyloge-

netic relationships within antitropical taxa can providemuch information about the origin of such distributions.

For example, phylogenetic relationships enable the in-

ference of transequatorial dispersal with its probable

direction, frequency, and location (Burridge, 2002;

Burridge and White, 2000). Application of molecular

genetic techniques to the study of antitropical biogeo-

graphy is attractive for the following two reasons. First,

these techniques may better resolve phylogenetic rela-tionship, particularly when taxa are morphologically

conserved. Second, levels of molecular variation can

provide relative and absolute estimates of the divergence

time, which facilitates the evaluation of specific dispersal

and vicariance hypotheses (Burridge and White, 2000).

The present study addressed questions relating to the

origin of the antitropical labrid fish genus Pseudolabrus,

using molecular genetic techniques. The 11 species ofPseudolabrus (Perciformes: Labridae) are marine near-

shore fishes with an overall antitropical distribution

(Fig. 1; Mabuchi and Nakabo, 1997; Russell, 1988); two

species are restricted to the waters of the East Asia,

while the remaining nine species occur in the temperate

waters of the Southern Hemisphere from Western

Australia to Juan Fernandez in the East Pacific. Based

on the osteological data, this labrid genus is recognized

to form a monophyletic assemblage together with thefollowing five genera: Austrolabrus, Dotalabrus, Eupe-

trichthys, Notolabrus, and Pictilabrus (Russell, 1988).

Collectively, these fishes are termed the ‘‘pseudolab-

rines,’’ with all but Pseudolabrus restricted to Australia

and New Zealand. DNA sequence data from the mito-

chondrial 12S and 16S rRNA genes, plus the intervening

tRNAVal, were used to elucidate the phylogenetic in-

terrelationship within Pseudolabrus and among thegenera of pseudolabrines. Based on the phylogeny,

probable mechanisms for the development of the Pseu-

dolabrus antitropicality were inferred, together with its

timing.

2. Materials and methods

2.1. Materials examined

Fishes used in this study were collected via a variety

of methods, including the use of both hand- and gill-nets

while SCUBA diving, angling, and from museum spec-

imens. Tissue samples were obtained from six of the 11

recognized species of the antitropical genus Pseudola-

brus (for their geographic distributions, see Fig. 1),including both of the Northern Hemisphere represen-

tatives P. sieboldi and P. eoethinus. Three of the seven

recognized species of Notolabrus (Russell, 1988), and

single representatives of Austrolabrus, Eupetrichthys,

and Pictilabrus were included in the phylogenetic anal-

yses as possible sister-groups of Pseudolabrus. Both

Austrolabrus and Eupetrichthys are monotypic (Russell,

Table 1

List of species used in the study, with DDBJ/GenBank/EMBL accession numbers

Classification Species Accession No.

Order Perciformes

Suborder Labroidei

Family Pomacentridae Abudefduf vaigiensis AB121234

Family Labridae

Tribea Hypsigenyins Choerodon azurio AB121235

Tribea Julidines Halichoeres tenuispinnis AB121236

Tribea Labrichthyines Ophthalmolepis lineolata AB121237

Suezichthys gracilis AB121238

Tribea Pseudolabrines Notolabrus gymnogenis AB121239

Notolabrus parilus AB121240

Notolabrus tetricus AB121241

Pseudolabrus biserialis AB121242

Pseudolabrus eoethinus AB121243

Pseudolabrus fuentesi AB121244

Pseudolabrus guentheri AB121245

Pseudolabrus miles AB121246

Pseudolabrus sieboldi AB121247

Pictilabrus laticlavius AB121248

Austrolabrus maculatus AB121249

Eupetrichthys angustipes AB121250

aLabrid tribes after Gomon (1997).

K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382 377

1988), while Pictilabrus consists of three species

(Hutchins and Morrison, 1996). The two species of

Dotalabrus were not examined. To root correctly the

phylogenetic network within the pseudolabrines, the

following four labrid species, Suezichthys gracilis, Oph-

thalmolepis lineolata, Halichoeres tenuispinnis, andChoerodon azurio were also included in the analysis,

because the former three species belong to the genera

which were recognized to be closely related to the

pseudolabrines based on osteological data (Russell,

1988), and the last species is classified into the Hypsig-

enyini which is believed to be most primitive among

labrid tribes (Gomon, 1997). In addition, the poma-

centrid, Abudefduf vaigiensis was used as an outgroup ofthe labrid species, on the basis of traditionally accepted

classification: Pomacentridae and Labridae both belong

to the suborder Labroidei (Nelson, 1994). Species used

in this study were listed in Table 1.

2.2. DNA amplification and sequencing

Genomic DNA was extracted from muscle tissuepreserved in 99.5% ethanol, using the Qiagen DNeasy

tissue kit. An approximately 1500 bp fragment of the

mitochondrial genome including the latter half of 16S

rRNA, the entire tRNAVal, and former half of the 12Sr

RNA gene was amplified using the following two sets of

primers: L1083-12S (ACA AAC TGG GAT TAG ATA

C)+H1903-16S (GTA GCT CGT YTA GTT TCG GG)

and L1803-16S (AGT ACC GCA AGG GAA AGCTGA AA)+H2590-16S (ACA AGT GAT TGC GCT

ACC TT) (Miya and Nishida, 2000). The two primer-

sets amplified two different fragments partially over-

lapping each other. PCR proceeded for 30 cycles on a

Model 9700 thermal cycler (Applied Biosystems), with

denaturation at 94 �C for 10 s, annealing at 45 �C for

10 s, and extension at 72 �C for 30 s, the final cycle beingfollowed by an extension at 72 �C for 5min. The PCR

products were electrophoresed on a 1% agarose gel,

purified using a Pre-Sequencing Kit (USB), and then

sequenced with dye-labeled terminators (Applied Bio-

systems). Primers used were the same as those for PCR

and each of the two partially overlapping fragments was

sequenced for both strands. All sequencing reactions

were performed according to the manufacturer�s in-structions. Labeled fragments were analyzed on a Model

3100 DNA sequencer (Applied Biosystems). All

sequences are available from DDBJ, EMBL, and Gen-

Bank under accession numbers shown in Table 1.

2.3. DNA alignment

DNA sequences were edited using the computer pro-grams EditView ver. 1.01 (Applied Biosystems), Auto-

Assembler ver. 2.1 (Applied Biosystems), and DNASIS

ver. 3.2 (Hitachi Software Engineering). The accuracy of

base determination was checked by comparison of light

and heavy complementary DNA strands. The two rRNA

and single tRNA genes, together with their secondary

structures, were identified by homology to other verte-

brates (for rRNA, European ribosomal RNA database:http://oberon.fvms.ugent.be:8080/rRNA/index.html; for

A

B

Fig. 2. Scatter-plots of the number of mutations versus corrected

distance using Kimura�s (1980) two-parameter model in pairwise

comparisons. Open triangles, transitions; solid circles, transversions.

378 K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382

tRNA, Kumazawa and Nishida, 1993). Initial alignment

of the sequences for all taxa was made using the Clustal

X (Thompson et al., 1997) with default settings. Loop

and stem regions of the secondary structure were iden-tified as separate categories of data for analyses. The

stem regions were checked for base pair complementarity

and the alignments in the loop region were manually

adjusted (Kjer et al., 1994). Bulges (bases within stem

regions not involved in base pair complementarity) were

included with the stems category for analyses. The

number of variable and parsimony-informative sites, and

frequencies of transitions and transversions were calcu-lated separately for stem and loop regions, using PAUP*

4.0b8a (Swofford et al., 2001). Kimura�s (1980) two-pa-rameter corrected genetic distances were calculated for

all pairwise comparisons among taxa for total aligned

sequences. Site saturation in loop and stem region was

assessed by plotting the corrected distances against the

total number of transitions and transversions for each

pairwise comparison (Fig. 2).

2.4. Phylogenetic analyses

Parsimony analyses were performed using the heu-

ristic search option, 1000 random sequence addition

replicates, and the tree bisection and reconnection

algorithm in PAUP* 4.0b8a (Swofford et al., 2001).

A. vaigiensis was used as the most distant outgroup of

Pseudolabrus, with C. azurio as the second outgroup.Forty-nine nucleotides from loop regions of the 16S

rRNA gene fragment were removed from all analyses

because homology of these characters could not be de-

termined. Two parsimony analyses were performed, the

first with all nucleotides weighted equally and the second

with loop regions analyzed for transversions only. This

weighting was done to account for evidence of satura-

tion of transitions in the more variable loop regions.Parsimony trees were evaluated using summary statistics

reported by PAUP* (e.g., tree length, ensemble consis-

tency index). Support for nodes was evaluated by cal-

culating decay indices (Bremer, 1988, 1994) and

nonparametric bootstrap values (Felsenstein, 1985) us-

ing PAUP*, the latter employing a heuristic search and

1000 bootstrap pseudoreplications.

Modeltest ver. 3.06 (Posada and Crandall, 1998) wasused to determine a model of sequence evolution closely

approximating that responsible for the data. Maximum

likelihood analysis was not performed because of com-

putational constrains caused by the size of the dataset.

To estimate phylogeny under the sequence evolution

model determined by Modeltest, a Bayesian analysis was

performed. Analysis of the dataset was carried out using

Mr. Bayes v3.0b4 (Huelsenbeck and Ronquist, 2001).Monte Carlo Markov chains were run for 2,000,000

generations, and trees were sampled every 100 genera-

tions where the Markov chain reached stationarity,

leaving 19,960 trees for analysis. The posterior proba-

bilities of trees and tree parameters were estimated from

this distribution.

3. Results

Mitochondrial DNA sequence data were obtained for

a total 1442 aligned bases including 447 bp of the 12S

rRNA gene, 74 bp of the tRNAVal gene, and 784 bp of

the 16S rRNA for 17 taxa (DDBJ/GenBank/EMBL

accession numbers in Table 1). Stem regions contained

658 nucleotides in total, of which 202 were variable and108 were parsimony informative. Loop regions con-

tained 784 nucleotides of which 330 were variable; 209

were parsimoniously informative in the unweighted

analysis and 191 in the weighted analysis (transversions

only).

Scatter-plots of corrected distances versus number of

mutations are presented in Fig. 2. A linear relationship

between the number of substitutions and genetic dis-tance suggests that there was no site saturation for

transitions or transversions in the stem regions (Fig. 2A).

Similarly there was no evidence of site saturation of

transversions in loop regions (Fig. 2B). A plateau for the

plot of loop transitions versus genetic distance suggests

that loop transitions are saturated and thus possibly

inappropriate for phylogenetic analysis (Fig. 2B).

Fig. 3. A single most-parsimonious tree with loop regions analyzed for transversions only. Abudefduf vaigiensis was used as the most distant out-

group. Numbers above nodes indicate bootstrap values from 1000 pseudoreplications (only values 50% are shown), numbers below nodes indicate

decay indices. TL¼ 915, CI¼ 0.53, RI¼ 0.52. Posterior probabilities from 19,960 trees produced by a Bayesian analysis under TrN+ I+G are also

indicated in the parentheses along the branches (only values 80% are shown).

K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382 379

Parsimony analysis of the nucleotide data excluding

loop transitions resulted in a single most-parsimonious

tree (Fig. 3: TL¼ 915, CI excluding uninformative

characters¼ 0.53, RI¼ 0.52). The species of pseudolab-

rines were monophyletic with a bootstrap value of 84and a decay index of 3. The pseudolabrines clade was

divided into two clades, Pseudolabrus+Notolabrus clade

and Austrolabrus+Eupetrichthys+Pictilabrus clade, the

former clade supported by a bootstrap value of 97 and a

decay index of 7, and the latter clade by a bootstrap

value of 98 and a decay index of 9. Pseudolabrus is pa-

raphyletic with respect to Notolabrus: Pseudolabrus

biserialis and P. miles form a monophyletic assemblagetogether with three Notolabrus species, with a bootstrap

value of 80 and a decay index of 2. The two Northern

Hemisphere species, Pseudolabrus eoethinus and P. sie-

boldi, were monophyletic with a bootstrap value of 94

and a decay index of 5, and the Northern Hemisphere

clade was sister to an Eastern Pacific species, P. fuentesi

with a bootstrap value of 99 and a decay index of 8.

Pseudolabrus guentheri was sister to the latter threePseudolabrus species with a bootstrap value of 67 and a

decay index of 1. The pseudolabrines clade was sister to

S. gracilis with a bootstrap value of 99 and a decay in-

dex of 13, and the pseudolabrines +S. gracilis clade was

sister to O. lineolata+H. tenuispinnis clade with a

bootstrap value of 98 and a decay index of 11.

Parsimony analysis of the unweighted nucleotide data

resulted in three equally most-parsimonious trees (TL ¼1192, CI excluding uninformative characters¼ 0.53,

RI¼ 0.54). One of the three most-parsimonious trees

had the same topology as the single most-parsimonious

tree resulting from the weighted parsimony analysis.

Among the three equally most-parsimonious trees, both

of the Pseudolabrus+Notolabrus clade and Austrola-

brus+Eupetrichthys+Pictilabrus clade were repro-

duced, and, within the former clade, the monophyly of

the two Northern Hemisphere species of Pseudolabrus,

its sister relationship to the East Pacific congenericspecies, and the monophyly of P. miles, P. biserialis,

and monophyletic three Notolabrus species were also

recovered.

Results from Modeltest (Posada and Crandall, 1998)

indicated that the model TrN+ I+G provided the best fit

to the data as determined by the likelihood ratio test. The

Bayesian analysis produced 20,000 trees of which 40 were

discarded as burn-in. This left 19,960 trees to estimatetopology and tree parameters. The percentage of times a

node occurs within these 19,960 trees was interpreted as

the posterior probability of the node (see Fig. 3). Esti-

mated model parameters were� ln l ¼ 7292:298� 4:737,A ¼ 0:360� 0:011, C ¼ 0:249� 0:010, G ¼ 0:188�0:009, T ¼ 0:202� 0:009, a ¼ 0:555� 0:126. Topologyof 50% consensus tree was identical to that of the most-

parsimonious tree from weighted analysis (see Fig. 3),except Pseudolabrus biserialis and P. miles were mono-

phyletic only in the former. The other major clades de-

scribed above were all supported by 100% posterior

probability (see Fig. 3).

4. Discussion

4.1. Taxonomic implications

The results of this study support pseudolabrines� re-lationships that do not agree with the previous hy-

pothesis based on osteological data (Russell, 1988)

Dotalabrus*

Notolabrus

Pseudolabrus

Pictilabrus

Austrolabrus

Eupetrichthys

Fig. 4. Alternative phylogenetic hypothesis for pseudolabrines based

on comparative osteology (Russell, 1988). The genus with asterisk was

not examined in the present study.

380 K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382

(Fig. 4). The most notable difference between the rela-

tionships based on molecular data and that on osteo-

logical data involves the phylogenetic relationships

between Pseudolabrus and Notolabrus. In Russell�s(1988) hypothesis, Pseudolabrus is monophyletic, andphylogenetically separated from Notolabrus: Pseudola-

brus is sister to a Austrolabrus+Pictilabrus clade, and

the Pseudolabrus+Austrolabrus+Pictilabrus clade is

sister to Eupetrichthys, Notolabrus being sister to the

clade consisting of these four genera. It was based on

these relationships that Russell (1988) erected Notola-

brus, and its constituent species were previously allo-

cated to Pseudolabrus. Our molecular analyses, however,demonstrated that Pseudolabrus was paraphyletic to

Notolabrus, and the Pseudolabrus+Notolabrus clade

was sister to the Austrolabrus+Pictilabrus+Eupetrich-

thys clade (Fig. 3). Although a final decision cannot be

reached now owing to incomplete taxon sampling, the

present molecular phylogeny seems to indicate that the

generic status of Notolabrus (or generic delimitations of

both Notolabrus and Pseudolabrus) requires review.

4.2. Probable mechanisms for the development of the

Pseudolabrus antitropicality

Phylogenetic analyses based on the molecular data

demonstrated that the two Northern Hemisphere species

of Pseudolabrus were monophyletic and nested deep

within the clade of Southern Hemisphere pseudolabrines(Fig. 3). Mapping species distributions onto the MP tree

indicated that both pseudolabrines and Pseudolabrus

originated in the Southern Hemisphere. Although some

Pseudolabrus species were not examined here, this

inference will not be upset by any addition of the yet to

be examined species of Pseudolabrus, because such spe-

cies all occur in the Southern Hemisphere. However, the

possibility of independent southern origins of theNorthern Hemisphere species cannot be discounted

(e.g., Burridge and White, 2000; Grant and Bowen,

1998).

Several mechanisms have been proposed for the de-velopment of antitropical distributions, divisible into

two classes, dispersal and vicariance (Burridge, 2002).

The dispersal mechanism assumes that a taxon was

originally represented on one side of the tropics, and

subsequently moved across the equator to colonize the

opposite hemisphere. On the other hand, the vicariance

mechanism assumes that taxa once occupied the tropics

but were later extirpated, resulting in isolated Northernand Southern Hemisphere populations. In the present

case, the vicariance mechanism with tropical origin of

Pseudolabrus seems to be unlikely, because such a

mechanism requires both multiple independent adapta-

tions to temperate waters and highly biased allotment of

species to each Hemisphere for explaining the distribu-

tion of Pseudolabrus. The vicariance mechanism with the

Southern Hemisphere origin of Pseudolabrus also seemsto be unlikely, because it requires the common ancestor

of the Northern Hemisphere species, once adapted to

tropical waters, to colonize temperate Asia without

leaving any descendants in the tropical region.

The dispersal mechanism with the Southern Hemi-

sphere origin of Pseudolabrus, seems more likely than

the above two vicariance mechanisms. In our result, the

two Northern Hemisphere species were monophyleticand derived relatively recently during the radiation of

Pseudolabrus (Fig. 3). Therefore, the dispersal mecha-

nism can explain the Pseudolabrus antitropicality by

only a single (two even at maximum) transequatorial

dispersal(s).

Within the antitropical marine fish groups studied to

date using molecular data, nearly half of the estimates of

transequatorial divergences fall within Pleistocene, per-haps implicating cooler temperatures during glaciations

for transequatorial dispersal (Burridge, 2002). However,

transequatorial divergence during recent glaciations

does not appear to be the case for Pseudolabrus. The

transequatorial divergence in partial 16S rRNA gene of

Pseudolabrus (divergences between Pseudolabrus fuentesi

and each of the two Asian Pseudolabrus species) ranges

from 3.8 to 4.1% [corrected by Kimura�s (1980) two-parameter model]. Applying a 16S rRNA molecular

clock calibration of 1.0% sequence divergence (for all

base substitutions) per million years derived from gem-

inate amphi-American Centropomus (Tringali et al.,

1999), the transequatorial divergence of Pseudolabrus is

estimated as 3.8–4.1 million years ago, corresponding to

early to mid Pliocene, which is older than Pleistocene or

late Pliocene. However, more recent transequatorialdivergences may become apparent with the analysis of

remaining five Pseudolabrus species.

Although some Southern Hemisphere species of

Pseudolabrus were not examined in this study, it is very

interesting that the monophyletic two Northern Hemi-

sphere species of Pseudolabrus were sister not to Aus-

tralian or New Zealand species, but to an East Pacific

K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382 381

species, Pseudolabrus fuentesi, from Easter Island (Figs.1 and 3). A similar phenomenon was found in the de-

mersal nearshore fish subgenus Goniistius (family

Cheilodactylidae, order Perciformes) (Burridge and

White, 2000), which is similar to Pseudolabrus in dis-

tribution and habitat. Within the antitropical subgenus,

one of the three Asian species, Cheilodactylus (Goniis-

tius) zebra showed the highest sequence similarity not

with other Japanese or Australian-New Zealand species,but with a species from Easter Island, Cheilodactylus

(G.) plessisi (Burridge and White, 2000). This fact to-

gether with our result for Pseudolabrusmay indicate that

some transequatorial dispersals undertaken by Asian

species might have occurred from the East Pacific rather

than the West Pacific, although Burridge (2002) still

considered Goniistius a West Pacific divergence. Inter-

estingly, Parrish et al. (1989) suggested that transequa-torial divergences were more predominant in the coastal

East Pacific, owing to cooler water temperature, nar-

rower tropical zone, etc. The timing of the transequa-

torial divergence between C. (G.) zebra and C. (G.)

plessisi was estimated as late Miocene-early Pliocene

(Burridge and White, 2000), which is, interestedly,

comparable to that of transequatorial divergence for

Pseudolabrus.

Acknowledgments

We thank Australian Museum, Osaka Aquarium

KAIYUKAN, H. Ikeda, and R. Robertson for pro-

viding tissue or fresh whole specimens. We also thankthe following individuals and institutions for their help

in collecting the specimens: Y. Yanagisawa, J. Paxton,

M. McGrouther, S. Reader, M. Uno, J.B. Hutchins,

S.M. Morrison, Weekly Sunday Fishing, T. Kudo, K.

Matsumoto, T. Mukai, and The University of Queens-

land. Thanks are also due to C.P. Burridge, B.C. Rus-

sell, and J.G. Inoue for providing useful information

and for discussions of the results. We are grateful toC.P. Burridge for correcting English. This study was

supported partly by Research Fellowships of the Japan

Society for the Promotion of Science for Young Scien-

tists (No. 72503), and Grants-in-Aid from the Ministry

of Education, Culture, Sports, Science and Technology,

Japan (12NP0201).

References

Bremer, K., 1988. The limits of amino acid sequence data in

angiosperm phylogenetic reconstruction. Evolution 42, 795–803.

Bremer, K., 1994. Branch support and tree stability. Cladistics 10, 295–

304.

Briggs, J.C., 1987. Antitropical distribution and evolution in the Indo-

West Pacific Ocean. Syst. Zool. 36, 237–247.

Briggs, J.C., 1995. Global Biogeography. Elsevier, New York.

Burridge, C.P., 2002. Antitropicality of Pacific fishes: molecular

insights. Environ. Biol. Fish. 65, 151–164.

Burridge, C.P., White, R.W.G., 2000. Molecular phylogeny of the

antitropical subgenus Goniistius (Perciformes: Cheilodactylidae:

Cheilodactylus): evidence for multiple transequatorial divergences

and non-monophyly. Biol. J. Linn. Soc. 70, 435–458.

Crame, J.A., 1993. Bipolar molluscs and their evolutionary implica-

tions. J. Biogeogr. 20, 145–161.

Darwin, C., 1859. On the origin of species by means of natural

selection, or the preservation of forward races in the struggle for

life. London.

Ekman, S., 1953. Zoogeography of the Sea. Sidwick and Jackson,

London.

Felsenstein, J., 1985. Confidence limits on phylogenetics: an approach

using the bootstrap. Evolution 39, 783–791.

Gomon, M.F., 1997. Relationships of fishes of the labrid tribe

Hypsigenyini. Bull. Mar. Sci. 60, 789–871.

Grant, W.S., Bowen, B.W., 1998. Shallow population histories in

deep evolutionary lineages of marine fishes: insights from

sardines and anchovies and lessons for conservation. J. Hered.

89, 415–426.

Hubbs, C.L., 1952. Antitropical distribution of fishes and other

organisms. Symposium on the Problems of Bipolarity and of

Pantemperate Faunas. In: Proceeding of the Seventh Pacific

Science Congress (Pacific Science Association), vol. 3, pp. 324–

329.

Huelsenbeck, J.P., Ronquist, F.R., 2001. MRBAYES: Bayesian

inference of phylogenetic trees. Bioinformatics 17, 754–755.

Hutchins, J.B., Morrison, S.M., 1996. Pictilabrus brauni, a new species

of labrid fish from South-western Australia. Revue Fr. Aquariol.

23, 39–42.

Kimura, M., 1980. A simple method for estimating evolutionary rate

of base substitutions through comparative studies of nucleotide

sequences. J. Mol. Evol. 16, 111–120.

Kjer, K.M., Baldridge, G.D., Fallon, A.M., 1994. Mosquito large

subunit ribosomal RNA: simultaneous alignment of primary and

secondary structure. Biochem. Biophys. Acta 1217, 147–155.

Kumazawa, Y., Nishida, M., 1993. Sequence evolution of mitochon-

drial tRNA genes and deep-branch animal phylogenetics. J. Mol.

Evol. 37, 380–398.

Lindberg, D.R., 1991. Marine biotic interchanges between Northern

and Southern Hemispheres. Paleobiology 17, 308–324.

Mabuchi, K., Nakabo, T., 1997. Revision of the genus Pseudolabrus

(Labridae) from the East Asian waters. Ichthyol. Res. 44, 321–334.

Miya, M., Nishida, M., 2000. Use of mitogenomic information in

teleostean molecular phylogenetics: a tree-based exploration under

the maximum-parsimony optimality criterion. Mol. Phylogenet.

Evol. 17, 437–455.

Nelson, G.J., 1985. A decade of change: the future of biogeography.

Earth Sci. Hist. 4, 187–196.

Nelson, J.S., 1994. Fishes of the World, third ed. Wiley, New York.

Parrish, R.H., Serra, R., Grant, W.S., 1989. The monotypic sardines,

Sardina and Sardinops: their taxonomy, distribution, stock struc-

ture, and zoogeography. Can. J. Fish. Aquat. Sci. 46, 2019–

2036.

Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of

DNA substitution. Bioinformatics 14, 817–818.

Randall, J.E., 1981. Examples of antitropical and antiequatorial

distribution of Indo-West-Pacific fishes. Pac. Sci. 35, 197–209.

Rehder, H.A., 1980. The marine mollusks of Easter Island (Isla de

Pascua) and Sala y G�omez. Smithsonian Contrib. Zool. 289, 1–167.

Rotondo, G.M., Springer, V.G., Scott, G.A.J., Schlander, S.O., 1981.

Plate movement and island integration—a possible mechanism in

the formation of endemic biotas, with special reference to the

Hawaiian Islands. Syst. Zool. 30, 12–21.

Russell, B.C., 1988. Revision of the labrid fish genus Pseudolabrus and

allied genera. Rec. Aust. Mus., Suppl. 9, 1–72, pls. 1–4.

382 K. Mabuchi et al. / Molecular Phylogenetics and Evolution 32 (2004) 375–382

Springer, V.G., 1982. Pacific Plate biogeography, with special reference

to shorefishes. Smithsonian Contrib. Zool. 367, 1–182.

Swofford, D.L., Waddell, P.J., Huelsenbeck, J.P., Foster, P.G., Lewis,

P.O., Rogers, J.S., 2001. Bias in phylogenetic estimation and its

relevance to the choice between parsimony and likelihood methods.

Syst. Biol. 50, 525–539.

Thompson, J.D., Gibson, T.J., Plewniak, E., Jeanmougin, F., Higgins,

D.G., 1997. The ClustalX-Windows interface: flexible strategies for

multiple sequence alignment aided by quality analysis tools.

Nucleic Acids Res. 25, 4876–4882.

Tringali, M.D., Bert, T.M., Seyoum, S., Bermingham, E., Bartolacci,

D., 1999. Molecular phylogenetics and ecological diversification of

the transisthmian fish genus Centropomus (Perciformes: Centrop-

omidae). Mol. Phylogenet. Evol. 13, 193–207.

Valentine, J.W., 1984. Neogene marine climate trends: implications for

biogeography and evolution of shallow-sea biota. Geology 12, 647–

650.

White, B.N., 1986. The isthmian link, antitropicality and American

biogeography: distributional history of the Atherinopsinae (Pisces:

Atherinidae). Syst. Zool. 35, 176–194.