doi:10.1111/j.1095-8649.2009.02332.x, available online at www.interscience.wiley.com
Phylogeography of the flathead mullet Mugil cephalus
in the north-west Pacific as inferred from the mtDNA
control region
B. W. Jamandre*, J.-D. Durand† and W. N. Tzeng*‡
*Institute of Fisheries Science, College of Life Science, National Taiwan University, No. 1, Sec. 4, Roosevelt Rd, Taipei 10617, Taiwan and †Institut de Recherche pour le D´eveloppement (IRD),
UR5119 ECOLAG, Campus IRD/ISRA de Bel Air, BP 1386, CP 18524 Dakar, S´en´egal (Received 24 June 2008, Accepted 20 April 2009)
The population genetic structure and historical demography of the flathead mullet Mugil cephalus were investigated using the mtDNA control region (CR) sequences (909–1015 bp) of 126 individuals collected from seven locations in the north-west Pacific between 2005 and 2007. Haplotype diversity (h= 0·9333–1·000) and nucleotide diversity (π = 0·0046–0·1467) varied greatly among the sampling locations. Phylogenetic analysis of the CR sequences indicated that M. cephalus in the north-west Pacific belongs to two highly divergent lineages (lineages 1 and 2), with the inferred population structure being closely associated with the distribution of both lineages. Two populations were identified, one from the East China Sea and the other from the South China Sea. The former samples were obtained from Taiwan and Qingdao of north China and associated with lineage 1 haplotypes. The latter samples were collected from the Philippines, Pearl River of South China and two samples from Japan, all of which were associated with lineage 2. Japanese samples from Okinawa and Yokosuka had different degrees of mixing between lineages 1 and 2. Historical demographic variables in both populations indicated that Pleistocene glaciations had a strong impact on M. cephalus in the north-west Pacific, resulting in a recent demographic decline of the East China Sea population but in demographic equilibrium for the South China Sea population. Japan appears to be a contact zone between lineages 1 and 2, but it may also be indicative of coexistence between resident and migratory populations. Further global studies are required to clarify the taxonomic status of this cosmopolitan species. © 2009 The Authors
Journal compilation© 2009 The Fisheries Society of the British Isles
Key words: geological events; mtDNA control region; Mugil cephalus; north-west Pacific; phylogeography; population genetics.
INTRODUCTION
The flathead mullet Mugil cephalus L. is a euryhaline marine fish with a global distribution in coastal and estuarine waters of subtropical and tropical regions (42◦ N – 42◦ S). This species occurs at water temperatures of 12–25◦ C and tolerates a wide range of salinities from freshwater to hypersaline conditions of 100 (Thomson,
‡Author to whom correspondence should be addressed. Tel.:+886 2 3366 2887; fax: +886 2 2363 9570; email: wnt@ntu.edu.tw
1966; Kuo et al., 1973; Collins, 1985; Vega-Cendejas et al., 2004). The M. cephalus fishery is one of the most important coastal fisheries in Taiwan, China and Japan. The gonads of ripe M. cephalus are considered a delicacy by local residents, with high market prices being attained, particularly in Taiwan and Japan (Liao, 1981).
Mugil cephalus makes an annual spawning migration from the coastal waters
of northern China to the offshore waters of south-western Taiwan. This migration occurs around the winter solstice, especially between 10 days before and after 21 or 22 December (Tung, 1981; Huang & Su, 1989). The females become mature at about 3 years of age and spawn once a year for at least 2 or more years thereafter (Tung, 1981). Newly hatched larvae drift with coastal currents towards estuarine nursing areas along the western coast of Taiwan, where they grow for 1–2 months. The juveniles then migrate to the coastal waters of northern China to be recruited into the adult stock (Chang et al., 2004).
The life history and population dynamics of M. cephalus in Taiwan, such as age, growth, reproduction, recruitment, mortality and fishing conditions have been intensively studied (Liao, 1981; Tung, 1981; Su & Kawasaki, 1995; Chang et al., 2000; Chang et al., 2004; Hsu et al., 2007). However, genetic structuring of the north-west Pacific populations remains unclear (Huang et al., 2001). The worldwide genetic structure of M. cephalus has been studied using allozyme (Rossi et al., 1998) and mtDNA techniques (Crosetti et al., 1994). These studies revealed marked genetic divergences among different geographical populations, raisings questions about its true taxonomic status. Analysis of mtDNA control region (CR) sequences also indicated that M. cephalus had high genetic divergence between Atlantic and Pacific populations, but there were few samples collected at regional scales, with only a single sample from the Gulf of Mexico in the north-east Atlantic (Rocha-Olivares et al., 2000). At present, because the CR sequences show exceptionally high divergences among populations on a geographical scale, Rocha-Olivares et al. (2005) have suggested that M. cephalus should be considered a species complex. Resolving the taxonomic status of M. cephalus requires, however, detailed investigations of genetic structures on regional scales (Crosetti et al., 1994; Rossi et al., 1998; Rocha-Olivares et al., 2000).
Control region sequences have been successfully used by Rocha-Olivares et al. (2000, 2005) to reveal extensive polymorphic sequences that were useful in evaluating the population structures and historical demography of M. cephalus. The current study used this gene to investigate the genetic structure of M. cephalus in the north-west Pacific region, to infer its present-day migration pattern in relation to the past climatic and geological evolution within the region, and to contribute information towards future global assessment of the taxonomic status of this cosmopolitan species.
MATERIALS AND METHODS F I S H C O L L E C T I O N
A total of 126 adult M. cephalus were collected from seven locations within the East China Sea and South China Sea during the winters (December to February) of 2005 to 2007. The distribution of samples was: Qingdao in northern China (22 individuals), Pearl River in southern China (17), Keelung in northern Taiwan (22), Kaohsiung in southern Taiwan (26),
Qingdao
East China sea YSWC CCC Okinawa Is. Keelung Pacific Ocean TWC Luzon Is. KC PHILIPPINES South China Sea
CHINA Pearl river SCSWC Kaohsiung TAIWAN JAPAN Yokosuka 40° 30° 20° 110° 120° 130° 140° 0 250 500 km
FIG. 1. Sampling locations ( ) and ocean currents affecting the distribution and dispersal of M. cephalus in the north-west Pacific (KC, Kuroshio Current; SCSWC, South China Sea warm current; TWC, Taiwan warm current; CCC, China coastal current; YSWC, Yellow Sea warm current).
Okinawa in southern Japan (10), Yokosuka in central Japan (13) and Luzon in the Philippines (16) (Fig. 1). The specimens from Taiwan and Okinawa were collected from offshore waters, while those from Pearl River, Yokosuka and Luzon were collected from estuaries.
D N A E X T R A C T I O N A N D S E Q U E N C I N G
For each specimen, muscle tissues and fin clips were collected and preserved in the 95% ethanol–water solution and returned to the laboratory. Genomic DNA was then extracted from a small piece of either muscle tissue or fin clips, using a commercially available total genomic DNA extraction kit (Bioman Scientific Co. Ltd.; www.bioman.com.tw).
The whole CR of the mitochondrial DNA (CR) was amplified using modified primers MulPro (McepCR-F - CCAAGGCCAAGATTTTTACATT) and MulPhe (TCTTGA-CATCTTCAGCGTTCGC) of Rocha-Olivares et al. (2000) primers. Polymerase chain reac-tion (PCR) amplificareac-tion was carried out in a 25μl total volume containing 0·1 ng DNA, 1·25 pmol reverse primer, 1·25 pmol forward primer, 5 mM deoxyribonucleotide triphos-phate (dNTP), 1·5 mM MgCl2 and 0·5 U Taq polymerase (Bioman) at the temperature of
94◦ C for 2 min. This was followed by 35 temperature cycles, with each cycle comprising denaturation at 94◦ C for 30 s, annealing at 58◦C for 45 s, extension at 70◦C for 45 s and then a final elongation at 70◦ C for 10 min. PCR products were checked on a 1% agarose gel stained with ethidium bromide and then sequenced with forward and reverse primers using dye-terminator cycle sequencing using an ABI 373A instrument (Applied Biosystems; www.appliedbiosystems.com).
P H Y L O G E N E T I C A N A LY S I S
The CR sequences determined in this study and those previously obtained from Hawaii and the Atlantic–Gulf of Mexico (Rocha-Olivares et al., 2000) from GenBank (www.ncbi.nlm.nih.gov) were aligned, using the software MAFFT ver. 6 (Katoh et al., 2002) with default parameters and certain manual improvements. This software is a multiple sequence alignment programme based on fast Fourier transformation (FFT), that implements
two different computing methods, the progressive method and the iterative refinement method (Katoh et al., 2002).
The phylogenetic relationships among haplotypes were explored using the neighbour-joining (NJ) method (Saitou & Nei, 1987) in conjunction with the Tamura–Nei model. Trees were constructed with MEGA 4.02 (Kumar et al., 2004) and their statistical significance was assessed using the bootstrap resampling technique (Felsenstein, 1985).
According to the recommendations of McMillan & Palumbi (1997) and Rocha-Olivares et al. (2005) regarding estimations of net average distances between lineages, the gamma corrected version of Tamura & Nei’s (1993) model (TrN-G) was used and the gamma distribution (a) was set at 0·604. Population genetic diversity was measured within each sample as the number of distinct haplotypes (n) and mean nucleotidic divergence (π ), using the software DnaSP 4.10.9 (Rozas et al., 2003) and Arlequin ver. 3.01 (Excoffier et al., 2005).
P O P U L AT I O N S T R U C T U R E A N D H I S T O R I C A L D E M O G R A P H Y
To investigate the genetic structure of M. cephalus in the north-west Pacific a hierarchical analysis of molecular variance (AMOVA) was performed (Excoffier et al., 2005), imple-mented in Arlequin ver. 3.01. Geographic partitionings, consistent with different scenarios of dispersal (historical v. contemporary), were tested. Historical delineation of coastlines were based on results from Liu et al. (2007), who showed that in the north-west Pacific region there were three biogeographic areas during the Pleistocene era. Therefore, the geographic partitioning tests were conducted with respect to (1) Pleistocene period separation, (2) north-west Pacific marginal seas separation, (3) Present oceanic currents system separation and (4) population structure separation as inferred from the STmatrix.
The historical demography of M. cephalus was inferred from mismatch distribution analysis as performed by Arlequin software. A unimodal pattern of mismatch distribution indicates that the population has experienced a rapid expansion in the recent past, while multimodal distributions indicate populations that are in demographic equilibrium (Rogers & Harpending, 1992). The parameter τ was estimated from the data and calculations were carried out using the method outlined in Rogers (1995). The goodness of fit of the observed data to a simulated model of expansion was tested using the raggedness index (Harpending, 1994) and a corresponding P -value was also obtained.
RESULTS G E N E T I C D I V E R S I T Y
The sizes of the 126 CR sequences from the north-west Pacific varied between 909 and 1015 base pairs (bp). The sequences have been deposited at GenBank under accession numbers EU663629–EU663754. Multiple alignments of the sequences made numerous indels (insertion and deletion) at the 3 end of the CR due to the presence of tandem repeats. To avoid large variance and limit homoplasy, the hyper variable 3 region was excluded, so that the CR sequence sizes used in the phylogenetic and further analyses were 901 bp.
For the 126 sequences, 105 haplotypes were observed, of which seven haplotypes were shared by the Philippines, Pearl River and Okinawa, and 14 haplotypes by Pearl River, Japan, Taiwan and Qingdao. The haplotype diversity (h) was high and ranged between 0·933 and 1·000 (Table I). Samples from Taiwan (KL, KS) and northern China (QD) presented the lowest genetic diversity (π= 0·005–0·008) (Table I), close to that of samples from the Atlantic (0·013). On the other hand, the Philippines (PH) and southern China (PR) samples presented a higher genetic diversity (π = 0·032and 0·066, respectively) comparable to that observed from
TABLEI. Genetic diversity of Mugil cephalus and historical demographic parameters at seven
of the sampling locations
Diversity index Demographic parameter
Location code N n h π t RI P Luzon PH 16 16 1·000 0·032 32·12 0·005 0·870 Pearl River PR 17 15 0·978 0·066 Okinawa OK 10 7 0·933 0·112 — — — Yokosuka YK 13 12 0·987 0·147 Qingdao QD 22 22 1·000 0·008 3·46 0·010 0·999 Keelung KL 22 20 0·987 0·005 Kaohsiung KS 26 22 0·979 0·007 Hawaii* HA 19 19 1·000 0·031 28·70 0·011 0·810 Atlantic* AT 95 94 1·000 0·013 5·40 0·003 0·660
N, sample size; n, number of haplotypes; h, haplotype diversity; π , nucleotide diversity; τ , divergence
time; RI, Raggedness index.
∗Data from Rocha-Olivares et al. (2000).
Hawaii (π = 0·031). The Japanese samples (OK, YK) exhibited the highest genetic diversity observed (π = 0·112–0·147) as both lineages occurred sympatrically.
The NJ tree reconstructed from the CR sequences illustrated the divergences among populations of M. cephalus on a global and regional scale (Fig. 2). There were four lineages observed in the tree: lineages 1 and 2 for the specimens from the north-west Pacific, and lineages 3 and 4 from Hawaii and Atlantic–Gulf of Mexico. All individuals from Qingdao, Kaohsiung and Keelung belonged to lineage 1, whereas all individuals from Luzon and Pearl River (with exception of one haplotype) belonged to lineage 2. Yokosuka and Okinawa samples consisted of a mixture of haplotypes belonging to both lineages 1 and 2 (Figs 2 and 3). Within each individual lineage there was no significant phylogenetic relationship observed. The NJ tree suggested a rapid ancestral radiation except for the Atlantic lineage. The genetic distances among the Hawaii lineage and lineages 1 and 2 were similar, ranging between 45·6 and 48·0% for corrected sequence divergences (TrN-G). Net average genetic distances (Tamura & Nei, 1993), with gamma correction between lineages 1 and 2, was 48·0% (Table II).
G E N E T I C S T R U C T U R E A N D H I S T O R I C A L D E M O G R A P H Y
Significant differences in genetic structure were detected in the CR haplotypes (ST= 0·7280; P < 0·05) among the samples from different geographical areas.
Within the north-west Pacific area, two geographic populations were identified from the pair-wise ST matrix (Table III): the East China Sea population (ECS,
Qingdao, Keelung, Kaohsiung) and the South China Sea population including Japan (SCS, Philippines, Okinawa, Yokosuka). Between these two populations, there was little or no genetic heterogeneity (ST= 0·153, P = 0·009, ST= 0·079 NS,
respectively) but with extreme interpopulation differentiations (CT= 0·77, P =
0·000). Differentiation of these two populations was mainly due to the distribution of lineages 1 and 2, with more genetic heterogeneity present in the SCS population when compared to the ECS population where only lineage 1 occurred (Fig. 3).
QD7 KL31 QD23 QD17 QD18 KS11 QD20 QD21 QD24 KL8 QD19 KL14 KS16 AP002930 KL3 KL30 QD12 YK14 QD16 KL16 QD9 YK13 QD6 KL7 KL29 YK18 KS12 KL33 YK15 YK17 KS26 QD25 KL32 QD8 KS5 KL4 QD13 PR25 QD4 KS36 KS37 QD14 KL5 YK16 KS35 KL13 KS27 KL36 KS39 QD10 QD22 QD5 KS40 QD11 KL37 KS6 KS4 KS8 KS25 KL10 KS33 KS32 KS24 KS30 KS29 KS3 KL11 QD15 KL35 Lineage 1 PR6 OK1 PH16 OK3 PR9 PR13 YK2 YK7 PH4 PR5 PR3 PH14 PH15 PH2 PH12 PR21 YK1 PR4 YK5 PR2 PR22 PR27 PH11 OK5 PH1 PR8 PR11 PR12 PH5 OK9 PH3 PH8 PR10 PH10 OK2 YK3 PH7 Lineage 2 Hawaii Atlantic–Gulf of Mexico 61 100 56 100 59 60 62 99 87 87 82 96 65 100 55 95 99 64 66 70 63 57 64 0⋅05
40° 30° 20° 110° 120° 130° 140° Qingdao CHINA JAPAN Yokosuka Okinawa Is. East China Sea
Keelung TAIWAN Kaohsiung
Luzon Is.
South China Sea PHILIPPINES
Pearl river
1 2
0 250 500 km
FIG. 3. Geographical distribution of lineages among seven sampling locations of Mugil cephalus in the north-west Pacific ( , land areas at the lonorth-west sea level during the last glacial maximum of the Pleistocene period).
TABLEII. Mean TrN-G uncorrected (below diagonal) and TrN-G corrected (above diagonal)
sequence divergences (%) among Mugil cephalus CR haplotypes
Lineage 1 Lineage 2 Hawaii Atlantic
Lineage 1 0·5 /0 ·5 48 45·9 74·7
Lineage 2 29·5 2·6 /2 ·8 45·6 77·8
Hawaii 28·3 27·4 2·5 /2 ·7 71·1
Atlantic 38·7 38·9 35·8 1·1 /1 ·1
Diagonal: within-region mean uncorrected/corrected per cent sequence divergences.
The genetic structure of the M. cephalus in the north-west Pacific was further investigated with hierarchical AMOVA that took into account different environmental scenarios (Historical–Pleistocene v. Contemporary–Sea currents) (Table II). The best subdivision that would limit the intrapopulation variation and maximize the interpopulation differentiation of the samples was found to accord with the present
FIG. 2. Neighbour-joining tree inferred from the mtDNA CR sequences of Mugil cephalus in the north-west Pacific, Hawaii and Atlantic–Gulf of Mexico (QD, Qingdao of China; PR, Pearl River of China; YK, Yokosuka of Japan; OK, Okinawa of Japan; KL, Keelung of Taiwan; KS, Kaohsiung of Taiwan; PH, Luzon of the Philippines).
T ABLE III. Pair-wis e ST of Mugil cephalus collected from seven locations South C hina Sea E ast C hina Sea ST PH OK YK PR QD KL KS HA
North- west Pacific
OK 0· 0851 YK 0· 374*** 0· 057 PR 0· 0153 − 0· 029 0· 220* QD 0· 941*** 0· 809*** 0· 556*** 0· 869*** KL 0· 948*** 0· 819*** 0· 567*** 0· 876*** 0· 002 KS 0· 945*** 0· 826*** 0· 584*** 0· 879*** 0· 021* 0· 016 HA 0· 891*** 0· 730*** 0· 643*** 0· 808*** 0· 951*** 0· 960*** 0· 955*** AT 0· 945*** 0· 861*** 0· 802*** 0· 896*** 0· 978*** 0· 983*** 0· 979*** 0· 951*** P H , the P h ilippines; OK, Okinawa; YK, Y okosuka; P R, P earl R iver; QD, Qingdao; KL, K eelung; KS , K aohsiung; HA, Hawaii; A T , A tlantic. ∗P< 0· 05; ∗∗P< 0· 01; ∗∗∗ P< 0· 001; Bold values denote signifi cant v al ue after B onferroni correction.
TABLEIV. Hierarchical AMOVAs
Source of variation % of variation Phi()-stat P
Pleistocene period separation [PH,PR, KS][KL, QD][OK, YK]
Between groups 11·81 CT= 0·118 0·463
Among pops. or groups 64·06 SC= 0·726 0
Within populations 24·13 ST= 0·759 0
North-west Pacific marginal–seas separation [PH,PR,KS][KL,OK,QD,YK]
Among groups −8·17 CT= −0·082 0·39
Among pops. or groups 81·02 SC= 0·749 0
Within populations 27·14 ST= 0·729 0
Oceanic currents system separation: Kuroshio Current
[PH,OK,YK][PR,KS,KL,QD]
Among groups 38·86 CT= 0·389 0·106
Among pops. or groups 39·92 SC= 0·653 0
Within populations 21·22 ST= 0·788 0
Population structure inferred from the st matrix
[PH,PR,OK,YK][KS,KL,QD]
Among groups 79·4 CT= 0·794 0·026
Among pops. or groups 3·92 SC= 0·190 0·002
Within populations 16·68 ST= 0·833 0
PH, Luzon, Philippines; PR, Pearl River; KS, Kaohsiung; KL, Keelung; QD, Qingdao; OK, Okinawa; YK, Yokosuka.
oceanic currents in the area where Philippines, Okinawa and Yokosuka would be preferentially connected through the Kuroshio Current (Fig. 1 and Table IV). The grouping that took into account a potential Pleistocene separation of South China Sea locations (Luzon, Kaohsiung and Pearl River) from more northern locations (Fig. 1), as suggested by the Chelon haematocheilus (Temminck & Schlegel) phylogeographic structure (Liu et al., 2007), created a poor fit in relation to the observed geographic diversity in M. cephalus.
The mismatched distribution pattern provides an insight into the past popula-tion demography. Because Japanese samples belonged to a contact zone where the genetic diversity was strongly influenced by recent dispersion of both SCS and ECS populations, Japanese samples were excluded from the mismatch anal-yses. The mismatch distribution pattern of the SCS population was polymodal, with the main mode at a pair-wise difference of 34 nucleotides [Fig. 4(a)]. The result of the raggedness test rejected a model of sudden expansion (Harpending’s raggedness index= 0·0045, P = 0·87). In contrast, the ECS population was uni-modal with the mode at a pair-wise difference of four nucleotides [Fig. 4(b)]. The raggedness test indicated a significant fit of the observed mismatch distri-bution to a model of sudden expansion (Harpending’s raggedness index= 0·01,
45 40 35 25 15 5 30 20 10 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 Pairwise differences Frequency 600 500 400 300 200 100 0 60 40 20 0 0 150 250 Observed Simulated (a) (b)
FIG. 4. Mismatch distributions from the mtDNA CR sequences of Mugil cephalus from the seven sampling locations: (a) lineage 2; (b) lineage 1; , observed distributions; , expected distributions from the sudden expansion model.
DISCUSSION
U S E F U L N E S S O F T H E C R S E Q U E N C E S I N P H Y L O G E N E T I C A N A LY S I S
High genetic divergences were found in the CR sequences among geographical populations of M. cephalus in this study. Rocha-Olivares et al. (2000, 2005) reported an unexpectedly high level of CR sequence divergence between populations separated by a wide geographic range, as is the case between the Atlantic and the Pacific populations where there is up to 70% corrected sequence divergence (TrN-G). This study has revealed two highly divergent lineages at a regional scale in the north-west Pacific (48·0% TrN-G). Lineage 1 has already been described by Rocha-Olivares
et al. (2005) using a Japanese M. cephalus sequenced by Miya et al. (2001), whereas
lineage 2 was identified during this study. Atlantic lineages appear to be the most divergent (71·1–77·8% TrN-G), with Pacific lineages less so (45·6–48·0% TrN-G). The above studies have demonstrated that intra and interoceanic divergences in M.
cephalus greatly exceed the expected level of divergence within a species, as already
emphasized by Rocha-Olivares et al. (2000). More recently, an investigation into the genetic structure of another mugilid in the north-west Pacific, C. haematocheilus,
showed that its CR sequence divergent levels ranged from 0·53 to 0·63% TrN-G (Liu
et al., 2007) and were considerably lower than those observed for M. cephalus.
The high level of genetic divergence in M. cephalus led Rocha-Olivares et al. (2005) to question whether the CR sequences were based on NUMT (nuclear mitochondrial DNA) pseudogene material, since it had been reported that the NUMT pseudogene evolves faster than their mitochondrial copies. However, upon analysing the CR sequences and using nested PCR, the results obtained in this study demonstrated that the M. cephalus material was mitochondrial in origin and not pseudogenetic. Therefore, rapid evolution of the CR gene could be responsible for the high levels of divergence in M. cephalus. This is in agreement with Caldara et al. (1996) who suggested that, among the species of Mugilidae, the Mugil lineage might exhibit a higher rate of evolution. However, it remains unclear as to why levels of divergence in M. cephalus exceed the usual intraspecific range, even when compared to species with rapid evolution such as within the Chaetodontidae (McMillan & Palumbi, 1997). Nevertheless, this cannot be used to determine the taxonomic rank, because considerable heterogeneity for evolutionary rate of the mitochondrial CR has been described among species (Bowen et al., 2006). Using less variable markers than the CR gene, Fraga et al. (2007) and then Heras et al. (2008) found much more limited inter M. cephalus lineage divergences [1–2% for 16S ribosomal RNA (rRNA) and 4–5% for cytochrome b]. These results lead to conflicting interpretations, with Heras et al. (2008) concluding there was a species complex due to the allopatry of M. cephalus lineages but Fraga et al. (2007) proposed that M. cephalus is a single species which includes Mugil liza Valenciennes and Mugil platanus G¨unther as separate M. cephalus lineages.
Because of its high evolutionary rate, the M. cephalus CR gene is not reliable as a molecular clock to date lineage divergence. Indeed, parallel mutations quickly blur the relationship between the genetic distance and divergence time. As mentioned by McMillan & Palumbi (1997), the utility of CR sequences to derive phylogenetic information on divergences is restricted to very short periods of time. Therefore, the divergence estimates among the lineages within the geographical populations in the
M. cephalus phylogenetic tree reconstructed from the CR sequences (Fig. 2) should
be considered with caution. Furthermore, this study could not provide a divergence date for M. cephalus due to ambiguities in the evolution rate of the species.
H I S T O R I C A L D E M O G R A P H Y O F MUGIL CEPHALUS I N T H E N O RT H - W E S T PA C I F I C
The two north-west Pacific lineages exhibited contrasting historical demographics, suggesting different evolutionary histories. Lineage 2 has a mismatch distribution that was significantly different from that expected from a model of sudden expansion. The broad, polymodal distribution is typical for a population under long-term equilibrium. A similar distribution was observed for the Hawaiian population of M. cephalus and, to a lesser extent, the Gulf of Mexico and Atlantic populations (Rocha-Olivares et al., 2000). By contrast, lineage 1 has a unimodal mismatch distribution typical of an expanding population. The raggedness index reflected the degree of departure from a smooth distribution typical of an expanding population. The index was significant for lineage 2 but not for lineage 1 suggesting that, in contrast to lineage 2, lineage 1 experienced a recent bottleneck (demographic crash) and is now in a phase of
population expansion. Since the distribution range of lineage 1 was in the temperate zone, further north than lineage 2 in the tropical zone, it can be assumed that Pleistocene glaciations had a greater effect on lineage 1 than on lineage 2. This result concurs with findings of Liu et al. (2007) for C. haematocheilus, which found a similar trend in historical demography, i.e. population expansion in the northern region (East China Sea) and equilibrium in the southern region (South China Sea). Pleistocene glaciations are known to have had a major effect on the genetic diversity and structure of temperate species (Hewitt, 1996, 2004).
Coastal habitats in the tropical region were also affected by the glaciations. Lowering of the sea level had a profound effect on these ecosystems during Pleistocene glaciations, either promoting differentiation or creating bottlenecks for populations (Chenoweth et al., 1998; Fauvelot et al., 2003; Durand et al., 2005). These geological events limited estuarine connectivity, sometimes drying out coastal lagoons, which increased genetic structuring of fish populations (Chenoweth et al., 1998) and decreased their genetic diversity (Fauvelot et al., 2003). Our results indicate that such effects were more drastic in high than low latitudes, and suggest that lineages 1 and 2 of M. cephalus in the north-west Pacific were isolated during the Pleistocene era.
E V O L U T I O N A RY S I G N I F I C A N C E O F T H E MUGIL CEPHALUS
L I N E A G E S
In this study, a high genetic heterogeneity was found among samples of M.
cephalus in the north-west Pacific, as expressed by the presence of two lineages
(Fig. 2). Lineage 1 dominates samples located within the continental shelf of the East China Sea (Qingdao, Keelung and Kaohsiung), whereas lineage 2 (Philippines) was found outside continental shelf waters (with the exception of Pearl River). Both lineages (1 and 2) were found in samples from Japan.
Questions regarding the taxonomic status of M. cephalus have been raised in many genetic studies (Crosetti et al., 1994; Rossi et al., 1998; Rocha-Olivares et al., 2000, 2005; Fraga et al., 2007; Heras et al., 2008). These studies have revealed the presence of highly isolated populations but were unable to determine whether the absence of gene flow was due to genetic incompatibility or geographic discontinuities, primarily because the sampling locations covered widely separated areas. In this study, the sampling locations were concentrated within the distributional range of M. cephalus in the north-west Pacific. A mixing of the two lineages (1 and 2) in Japan was considered to be the result of a secondary contact of previously geographically isolated lineages. The northward dispersion of both lineages with the Kuroshio Current from the Philippines, Taiwan and Okinawa to Japan may have facilitated the mixing (Fig. 1). Finally, the absence of allopatric distribution of north-west Pacific
M. cephalus lineages promotes the concept of a single M. cephalus species in the
region.
Despite the above genetic evidence, the question of the taxonomic status of M.
cephalus remains largely unresolved. Two forms of M. cephalus, with different
reproductive behaviours, have been described in the north-west Pacific (Liu, 1986; Su & Kawasaki, 1995). In the coastal waters of Taiwan, both resident and migratory forms of M. cephalus have been reported (Huang et al., 2001). The resident form lives in estuarine environments all year and reproduces at sea during late winter
(late December to late January). The migratory form reproduces in mid-winter (October to December) and then migrates from Taiwanese to Chinese coastal waters (Liu, 1986). If reproductive periods of the two forms overlapped, then gene flow between the resident and migratory forms would be possible (Huang et al., 2001). However, Huang et al. (2001) described extensive genetic differentiation at the glycosylphosphatidylinositol (GPI) enzyme, suggesting an absence of gene flow or strong selective pressure towards GPI polymorphism. Samples used in the present study could not be used to compare resident and migratory forms based on the ECS or SCS population. Samples from Taiwan and China, except those from the Pearl River, were collected from offshore waters during their spawning migration, while the other samples were collected from estuarine waters. Therefore, the resident and migratory forms of M. cephalus might be sympatric in the Pearl River, Okinawa and Yokosuka, as indicated by the mixing of the two lineages. Past geological events and contemporary environmental conditions might have favoured the divergence of resident and migratory forms of M. cephalus.
Further studies of life history, reproductive behaviour, recruitment and genetic architecture are required to determine whether the inferred secondary contact of previously isolated lineages of M. cephalus has resulted in either the formation of introgressed populations (e.g. two lineages within one population) or the coexistence in sympatry of genetically isolated races (e.g. each population/race/species is composed of only one lineage). A combined analysis of mtDNA and biparentally inherited markers could determine the level of genetic isolation of mtDNA lineages.
We would like to express our thanks to C. W. Chang, S. Chow, X. P. Nie, H. Imai, J. X. Liu, C. C. Hsu and M. P. Garcia for their assistance and effort in obtaining samples from Japan, Taiwan and the Philippines. We would like to thank A. K. Whitfield, F. Lecomte and D. McKenzie for their helpful comments and corrections of the English. This research project was funded by the National Science Council of the Executive Yuan, Taiwan ROC (Contract Nos. NSC95-2915-I-002-108 and 96-2923-I-002-001 awarded to W. N. Tzeng).
References
Bowen, B. W., Muss, A., Rocha, L. A. & Grant, W. S. (2006). Shallow mtDNA coalescence in Atlantic pygmy angelfishes (genus Centropyge) indicates a recent invasion from the Indian Ocean. Journal of Heredity 97, 1–12.
Caldara, F., Bargelloni, L., Ostellari, L., Penzo, E., Colombo, L. & Patarnello, T. (1996). Molecular phylogeny of grey mullets based on mitochondrial DNA sequence analysis: evidence of a differential rate of evolution at the intrafamily level. Molecular Phylogenetics and Evolution 6, 416–424.
Chang, C. W., Tzeng, W. N. & Lee, Y. C. (2000). Recruitment and hatching dates of grey mullet (Mugil cephalus L.) juveniles in the Tanshui estuary of North-west of Taiwan. Zoological Studies 39, 99–106.
Chang, C. W., Lizuka, Y. & Tzeng, W. N. (2004). Migratory environmental history of the grey mullet Mugil cephalus as revealed by otolith Sr:Ca ratios. Marine Ecology-Progress Series 269, 277–288.
Chenoweth, S. F., Hughes, J. M., Keenan, C. P. & Lavery, S. (1998). When oceans meet: a teleost shows secondary intergradation at an Indian-Pacific interface. Proceedings of the Royal Society B 265, 415–420.
Collins, M. L. (1985). Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (South Florida), striped mullet. US Fish and Wildlife Service Biology Report 82.
Crosetti, D., Nelson, W. S. & Avise J. C. (1994). Pronounced genetic structure of mitochon-drial DNA among populations of the circumglobally distributed grey mullet (Mugil cephalus). Journal of Fish Biology 44, 47–58.
Durand, J.-D., Tine, M., Panfili, J., Thiaw, O. T. & Lae, R. (2005). Impact of glaciations and geographic distance on the genetic structure of a tropical estuarine fish, Ethmalosa fimbriata (Clupeidae, S. Boudich, 1825). Molecular Phylogenetics and Evolution 36, 277–287.
Excoffier, L., Laval, G. & Schneider, S. (2005). Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1, 47–50.
Fauvelot, C., Bernardi, G. & Planes, S. (2003). Reductions in the mitochondrial DNA diversity of coral reef fish provide evidence of population bottlenecks resulting from Holocene sea-level change. Evolution 57, 1571–1583.
Felsenstein, J. (1985). Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39, 783–791.
Fraga, E., Schneider, H., Nirchio, M., Santa-Brigida, E., Rodrigues-Filho, L. F. & Sam-paio, I. (2007). Molecular phylogenetic analyses of mullets (Mugilidae, Mugiliformes) based on two mitochondrial genes. Journal of Applied Ichthyology 23, 598–604. Harpending, H. C. (1994). Signature of ancient population growth in a low-resolution
mitochondrial DNA mismatch distribution. Human Biology 66, 591–600.
Heras, S., Roldan, M. I. & Gonzalez Castro, M. (2008). Molecular phylogeny of Mugilidae fishes revised. Reviews in Fish Biology and Fisheries 19, 217–231.
Hewitt, G. M. (1996). Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society 58, 247–276.
Hewitt, G. M. (2000). The genetic legacy of the Quaternary ice ages. Nature 405, 907–913. Hewitt, G. M. (2004). Genetic consequences of the climatic oscillations in the Quaternary.
Philosophical Transactions of the Royal Society B 359, 183–195.
Hsu, C. C., Han, Y. S. & Tzeng, W. N. (2007). Evidence of flathead mullet Mugil cephalus L. spawning in waters northeast of Taiwan. Zoological Studies 46, 717–725.
Huang, C. S. & Su, W. C. (1989). Analysis on the fishing condition of grey mullet in Taiwan, 1984–1985. In Study on the Resource of Grey Mullet in Taiwan, 1983–1985 (Su, W. C. ed.), pp. 35–48. Kaohsiung: Taiwan Fisheries Research Institute.
Huang, C. S., Weng, C. F. & Lee, S. C. (2001). Distinguishing two types of gray mullet, Mugil cephalus L. (Mugiliformes: Mugilidae), by using glucose-6-phosphate isomerase (GPI) allozymes with special reference to enzyme activities. Journal of Comparative Physiology B 171, 387–394.
Katoh, K., Misazawa, K., Kuma, K. I. & Takashi, M. (2002). MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Research 30, 3059–3066.
Kumar, S., Tamura, K. & Nei, M. (2004). MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics
5, 150–163.
Kuo, C. M., Nash, C. E. & Shenandeh, Z. H. (1973). A procedural guide to induce spawning in grey mullet (Mugil cephalus L.). Aquaculture 3, 1–14.
Liao, I. C. (1981). Cultivation methods. In Aquaculture of Grey Mullets, IBP 26 (Oren, O. H., ed.), pp. 361–390. Cambridge: Cambridge University Press.
Liu, C. H. (1986). Survey of the spawning grounds of gray mullet. In Study on the Resource of Gray Mullet in Taiwan 1983–1985 (Su, W. C. ed.), pp. 63–72. Kaohsiung: Taiwan Fisheries Research Institute.
Liu, J. X., Gao, T. X., Wu, S. F. & Zhang, Y. P. (2007). Pleistocene isolation in the North-Western Pacific marginal seas and limited dispersal in a marine fish, Chelon haematocheilus (Temminck & Schlegel, 1845). Molecular Ecology 16, 275–288. McMillan, W. O. & Palumbi S. R. (1997). Rapid rate of control-region evolution in Pacific
Butterflyfishes (Chaetodontidae). Journal of Molecular Evolution 45, 473–484. Miya, M. A., Kawaguchi, A. & Nishida, M. (2001). Mitogenomic exploration of higher
38 newly determined complete mitochondrial DNA. Molecular Biology and Evolution
18, 1993–2009.
Rocha-Olivares, A., Garber, N. M. & Stuck, K. C. (2000). High genetic diversity, large inter-oceanic divergence and historical demography of the striped mullet. Journal of Fish Biology 57, 1134–1149.
Rocha-Olivares, A., Garber, N. M., Garber, A. F. & Stuck K. C. (2005). Structure of the mitochondrial control region and flanking region tRNA genes of Mugil cephalus. Hydrobiologica 15, 139–149.
Rogers, A. R. (1995). Genetic evidence for a Pleistocene explosion. Evolution 49, 608–615. Rogers, A. R. & Harpending, H. C. (1992). Population growth makes waves in the distribu-tion of pairwise genetic differences. Molecular Biology and Evoludistribu-tion 49, 552–569. Rossi, A. R., Capula, M., Crosetti, D., Campton, D. E. & Sola, L. (1998). Genetic divergence
and phylogenetic inferences in five species of Mugilidae (Pisces: Perciformes). Marine Biology 131, 213–218.
Rozas, J., Sanchez-Delbarrio, J. C., Messeguer, X. & Rozas, P. (2003). DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19, 2496–2497.
Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4, 406–425.
Su, W. C. & Kawasaki, T. (1995). Characteristics of the life history of gray mullet from Taiwan waters. Fisheries Science 61, 377–381.
Tamura, K. & Nei, M. (1993). Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 12, 512–526 .
Thomson, J. M. (1966). The grey mullets. Oceanography and Marine Biology Annual Review
4, 301–335.
Tung, I. H. (1981). On the Fishery of gray mullet, Mugil cephalus L., in Taiwan. Reports of the Institute of Fishery Biology of the Ministry of Economic Affairs, National Taiwan University 3, 38–102.
Vega-Cendejas M. E. & Hernandez de Santillana M. (2004). Fish community structure and dynamics in a coastal hypersaline lagoon: Rio Lagartos, Yucatan, Mexico. Estuarine Coastal and Shelf Science 60, 285–299.
