Quick viewing(Text Mode)

Comprehensive Phylogeny of the Family Sparidae (Perciformes: Teleostei) Inferred from Mitochondrial Gene Analyses

Comprehensive Phylogeny of the Family Sparidae (Perciformes: Teleostei) Inferred from Mitochondrial Gene Analyses

Genes Genet. Syst. (2009) 84, p. 153–170 Comprehensive phylogeny of the family (: Teleostei) inferred from mitochondrial gene analyses

Satoru N. Chiba1, Yukio Iwatsuki2, Tetsuo Yoshino3 and Naoto Hanzawa1* 1Graduate school of Science and Engineering, Yamagata University, 1-4-12 Kojirakawa-machi, Yamagata 990-8560, Japan 2Division of Fisheries Science, Faculty of Agriculture, University of Miyazaki, 1-1 Gakuen-Kibanadainishi, Miyazaki 889-2192, Japan 3Department of Marine Sciences, Faculty of Science, University of the Ryukyus, 1 Senbaru, Nishihara-cho, Okinawa 903-0213, Japan

(Received 13 October 2008, accepted 3 February 2009)

Sparid consist of approximately 115 in 33 genera that are broadly distributed in tropical and temperate coastal waters. Although several phyloge- netic analyses were conducted based on specific molecular markers, their classifi- cation remains unresolved. Here, we present the most comprehensive molecular phylogeny of the family Sparidae to date, based on cytochrome b (cyt-b) genes. We determined 18 sequences of sparids and conducted phylogenetic analyses among 72 individuals representing 66 sparids with 23 outgroup species. Phyloge- netic trees were constructed according to partitioned Maximum Likelihood (ML) and Bayesian methods. The phylogenetic analyses were conducted on two dif- ferent data sets (including all positions; RY-coding). The phylogenetic trees showed monophyly of the family Sparidae with a different taxon, centracanthid . The subfamilies in the Sparidae in all trees are non-monophyletic and do not agree with current classification of the subfamilies. The genera , , , , , , and Spicara are also non-monophyletic and their classifications should be revised based on the phylogenetic relationships and reinvestigation of morphological characters. The sparids are divided into three major clades, A, B and C, respectively in the ML tree based on all codon positions, whereas clade C was paraphyletic in the other trees. The species in clade C are known to be present in the eastern Pacific to western Atlantic, whereas those in clades A and B are distributed in various oceanic regions. Some sub-clades in clades A and B consist of species that are dis- tributed in defined local regions. We further investigated evolutionary patterns of 87 morphological characters by ancestral character-state reconstruction accord- ing to the parsimony criteria. The results suggested high evolutionary plasticity of the characters in sparids, indicating that it causes species-diversity and taxo- nomic confusion at various taxonomic levels, and that such convergent evolution may occur more frequently also in other coastal fishes.

Key words: sequence variation, ML, Bayes, character state reconstruction, convergent evolution

because similar morphological characters among species INTRODUCTION are derived from convergent evolution, and speciation In most taxa of coastal fishes that are distributed patterns are quite complicated (Duftner et al., 2007; worldwide, it is difficult to infer accurate phylogenetic Westneat et al., 2005). One such coastal family, the relationships and to classify the fish appropriately Sparidae (Perciformes), which consists of approximately 115 species belonging to 33 genera, is the most diverse of Edited by Toshihiko Shiroishi the sparoid families (Nelson, 2006). The species in the * Corresponding author. E-mail: [email protected] Sparidae inhabit tropical and temperate coastal waters and 154 S. N. CHIBA et al. occasionally occur in estuaries as nurseries (Carpenter, feeding strategies and dentition. However, since they 2001). used only S. maena as the outgroup, their analysis The Sparidae, Nemipteridae, and were rec- showed incorrect phylogenetic relationships among ognized their osteological close relationships and termed sparids and of the sparids to related families. Orrell et them as “spariform” fish (Akazaki, 1962). Additionally it al. (2002) and Orrell and Carpenter (2004) inferred the was found that the three families were also closely related phylogeny of representatives of all 33 sparid genera and osteologically to the family Centracanthidae and classi- a number of percoid outgroups using the mitochondrial fied all of them into the new superfamily Sparoidea cytochrome b (cyt-b) gene and 16S rRNA gene sequences. (Johnson, 1980). The phylogenetic position of the cent- They also found that the currently defined subfamilies racanthids, however, is still uncertain with respect to are not monophyletic and showed Spicara to be a member sparid genera (Carpenter and Johnson, 2002; Day, 2002; of the sparid ingroup. Day (2002) analyzed a numerical Orrell et al., 2002; Orrell and Carpenter, 2004). The four phylogeny of the interrelationships among most extant families form a monophyletic clade and can be placed in genera in the sparid using morphological data. The the superfamily Sparoidea by a cladistic analysis of 54 morphological data support their hypothesis that the morphological characters (Carpenter and Johnson, 2002). currently defined subfamilies are non-monophyletic and However, the Lethrinidae was a sister taxa to the that Spicara is nested deeply within the ingroup, but a Sparidae (Orrell et al., 2002). Therefore the superfamily few analyzed sparid genera are known to be non- Sparoidea was not recognized as the formal superfamily monophyletic based on molecular studies to date. in classification pending more study of other percoid fam- Because mitochondrial cyt-b genes are relatively easy ilies in hopes of presenting a comprehensive and mono- to amplify and sequence (Nishida et al., 1998), they con- phyletic classification of the entire group (Nelson, 2006). tain both slowly and rapidly evolving codon positions (e.g. Sparid species are classified based on morphological Saitoh et al., 2006; Yamanoue et al., 2007), and a signif- characters, such as dentition, spinous and soft fin ray icant amount of sequence data is available in GenBank counts, scalation and body color (e.g. Akazaki, 1962; and other DNA databases, these genes are very useful Bauchot and Hureau, 1986; Smith and Smith, 1986). In and convenient markers to clarify the evolutionary his- particular, dentition has been used by many researchers tory of fishes at various phylogenetic levels (e.g. Sasaki et to distinguish genera (e.g. Cuvier, 1817, 1829; Smith, al., 2006; Kuriiwa et al., 2007; Pérez et al., 2007; Timm et 1938; Munro, 1948, 1949; Akazaki, 1962) and to provide al., 2008). Most recent studies have focused on resolving subfamily classification (Smith, 1938; Smith and Smith, phylogeny conflicts at a higher taxonomic level mostly by 1986; Akazaki, 1962). Smith and Smith (1986) erected adding characters rather than taxa, and this has not four subfamilies, Sparinae, Denticinae, Pagellinae, and improved our understanding of lower taxonomic level Boopsinae, using dentitional characters, which are con- relationships. Mitogenomic data, for example, is known sidered to reflect trophic specialization. Akazaki (1962) to be a powerful genetic marker to resolve uncertainties further subdivided the Sparinae into two new subfami- at various phylogenetic levels. However, mitogenomic lies, Diplodinae and Pagrinae. Fiedler (1991) distin- data is available only for five sparid species despite the guished only three subfamilies, Sparinae (includes Smith fact that sparid classification needs a comprehensive phy- and Smith’s Pagellinae and Sparinae), Denticinae, and logenetic study that focuses on various phylogenetic Boopsinae. levels. These data well provide better tools for detailed In previous molecular studies of the relationships evolutionary studies, and are necessary to test existing among sparids and between the family Sparidae and morphological phylogenetic hypotheses, and current other percoids deduced from isozymes and their expres- sparid classification. Furthermore, adding more taxa, sion patterns and sequence analysis, taxon sampling was together with adding characters to molecular phyloge- limited mainly to sparids from the eastern Atlantic and/or netic data, can be an efficient way to overcome phyloge- Mediterranean or Japanese waters (Taniguchi et al., netic uncertainty (Graybeal, 1998; Hillis, 1996, 1998; 1986; Basagila, 1991; Garrido-Ramos et al., 1995, 1998, Hillis et al., 2003; Pollock et al., 2002; Zwickl and Hillis, 1999; Jean et al., 1995; Allegrucci et al., 1999; Hanel and 2002; but see Miller and Hormiga, 2004; Rosenberg and Sturmbauer, 2000; Summerer et al., 2001; de la Herrán Kumar, 2001, 2003; Rokas and Caroll, 2005). However, et al., 2001). Hanel and Sturmbauer (2000) analyzed it is obvious that combining multiple genes is beneficial mitochondrial 16S rDNA (16S) sequences of all 24 sparid to any phylogenetic problem. species described from the northeastern Atlantic and Here, we determined 18 cyt-b gene sequences for 15 Mediterranean, and used the centracanthid Spicara species of the 68 sparid taxa and estimated their phylog- maena as the outgroup. Their analysis indicated that eny with 23 outgroups. This data set covers approxi- some sparid genera and the currently defined subfamilies mately 60% of sparid species. The protein coding cyt-b are not monophyletic and that there are three major mito- gene that was employed is a useful marker for phyloge- chondrial lineages, each comprising species with different netic analysis among many species. The codon parti- Molecular phylogeny of sparid fishes 155 tioned Maximum likelihood (ML) and Bayesian methods Analysis of sequence variation The sequences were were used to analyze the phylogeny of sparids and other multiply aligned by using Clustal X (Thompson et al., related percoids. In addition, evolutionary patterns of 87 1997). We dealt with the cyt-b gene sequences, as fol- morphological characters were traced onto the Bayesian lows, to construct two different data sets: (1) all positions tree using an ancestral states reconstruction method. included (designated as 123n; subscript “n” denotes nucle- otides); and (2) third codon positions converted into purine (R) and pyrimidine (Y) (RY-coding; designated as MATERIALS AND METHODS 12n3r; subscript “r” denotes RY-coding; Phillips and Taxonomic sampling We determined cyt-b gene sequ- Penny, 2003; Harrison et al., 2004). For the RY-coding, ences for 18 individuals of 15 sparid species including one we used arbitrarily “A” and “C” instead of “R” and “Y” for putative distinct Acanthopagrus species. A list of the 72 third codon positions in 12n3r to account only transver- sparids and 23 outgroups examined in this study is pro- sions avoiding estimation of transitional changes. vided in Table 1, along with the DDBJ/EMBL/GenBank Because RAxML is not allowed that base frequencies are accession numbers and references. The new additional zero at each partition, we used “G” and “T” instead of “A” species identification and nomenclature followed Hayashi and “C” for a few invariable sites of third codon positions. (2002), Iwatsuki and Carpenter (2006), Iwatsuki et al. Pairwise comparison and statistical information from (2007), Munro (1949), Kume and Yoshino (2008). The the cyt-b gene sequences were obtained using PAUP two South African sarba were identified 4.0b10 (Swofford, 2002). To examine patterns of sequ- by regional taxonomic experts. The present study ence variations for the 1st, 2nd and 3rd codon positions includes the putative distinct Acanthopagrus species separately, we plotted pairwise nucleotide differences (Acanthopagrus sp.), which is morphologically similar to (sorted into transitional [TS] and transversional [TV] A. butcheri. To examine monophyly and the interrela- difference) to evolutionary distances as a substitute for tionships of sparids, species belonging to all 33 recognized absolute geological time. The maximum-likelihood (ML) genera in the Sparidae were used for phylogenetic distances with the GTR + I + Γ model derived from each analyses. Additionally, to obtained a rooted phyloge- of the codon positions were used as evolutionary dis- netic tree, we used species belonging to other families of tances. the Sparoidea, Centracanthidae, Lethrinidae and Nemip- teridae and the related outgroups, Moronidae, Cae- Phylogenetic analyses The data sets were analyzed sionidae, Lutjanidae, Chaetodontidae, Pomacanthidae, using codon positions partitioned in both the ML and and Lateolabracidae in the Parciformes, and Bayesian methods. The six substitution types-ML model the non-Perciformes Satyrichthys amiscus. Previous (“nst = 6”) was selected as the best-fit model for base- molecular studies used relatively few outgroups. For a substitutions by Modeltest 3.7 (Posada and Crandall, more robust phylogeny of sparids, we sampled 23 out- 1998), for each codon partition of 123n data set on the group taxa. basis of Akaike Information Criteria (AIC; Akaike, 1973). ML analysis was performed using RAxML v7.0.4 DNA extraction A small piece of skeletal muscle was (Stamatakis et al., 2008) with the following settings. The excised from fresh or frozen specimens and immediately maximum likelihood model employed GTR + I + Γ, with preserved in 99.5% ethanol. Total DNA was purified base frequencies estimated from the data. As substitu- from the tissues according to phenol/chloroform method. tion frequencies differed among the first, second and third Using the total DNA and the primer pair AJG15 (Akihito, codon positions in the cyt-b gene (see Fig. 1), each codon 2000) and H15913 (Minegishi et al., 2005), approximately position was treated separately for 123n data set. The 1.2 kbp of mtDNA fragments including the cyt-b gene program is not allowed a single rate category. Then, we were amplified under the following PCR conditions. The also applied GTR + I + Γ model for each codon position of 30 amplification cycles were performed at 94°C for 45 sec, 12n3r data set. The significance of each branch of the 54°C for 30 sec, and 72°C for 1 min. Complete cyt-b gene trees was evaluated by bootstrapping (Felsenstein, 1985) sequences (1,141 bp) were determined by using an auto- with 1,000 replications. mated DNA sequencer (3100, Applied Biosystems) using Bayesian analysis was performed using MrBayes v3.12 the following primers. AJG15, H15913, L15411 (Miya (Ronquist and Huelsenbeck, 2003) with the following and Nishida, 1999), H15149 (Inoue et al., 2000), H15560 settings. We set two different partition (123n and 12n3r) (Miya and Nishida, 1999), SPL377 (5’-cag ctt tcg tag ggt depending on the data sets. Each codon position was atg tc-3’), SPL570 (5’-cgc cat aac tat act aca cc-3’), SPH451 treated separately (charset 1st_pos = 1-.¥3; charset (5’-aca gaa ggt tgg taa tga ca-3’) and SPH506 (5’-gaa aat 2nd_pos = 2-.¥3; charset 3rd_pos = 3-.¥3; partition ccg cct caa att ca-3’). Ambiguous sequences were recon- by_codon = 3:1st_pos,2nd_pos,3rd_pos; set partition = firmed from clones of their cyt-b gene fragments isolated by_codon;). Rate variation across sites was modeled by TA-cloning methods. using a γ-distribution (rates = “invgamma”). All of the 156 S. N. CHIBA et al.

Table 1. Collection data for samples used in this study

Taxon Accession No. Voucher Collection Locality References & Remarks Sparidae, Sparoidea, Perciformes Boopsinae boops X81567 Unknown Unknown Cantatore et al. (1994) crenidens AF240699 No voucher Qatif Market, eastern Saudi Arabia Orrell et al. (2002) Gymnocrotaphus curvidens AF240700 RUSI 49447 Kenton-on-Sea, South Africa Orrell et al. (2002) Oblada melanura AF240701 No voucher Azohia, Bay of Cartagena, Spain Orrell et al. (2002) Pachymetopon aeneum AF240702 RUSI 49672 Kenton-on-Sea, South Africa Orrell et al. (2002) Polyamblyodon germanum AF240703 RUSI 49690 Kenton-on-Sea, South Africa Orrell et al. (2002) Sarpa salpa AF240704 RUSI 49456 Kenton-on-Sea, South Africa Orrell et al. (2002) cantharus AF240705 ODU 2782 Fiumicino Fish Market, Italy Orrell et al. (2002) Denticinae Argyrozona argyrozona AF240706 RUSI 58449 Durban Fish Market, South Africa Orrell et al. (2002) Cheimerius matsubarai AB458408 KPM-NI20688 Iyumachi Fish Market, Okinawa pref., Japan This study Cheimerius nufar AF240707 RUSI 49443 Kenton-on-Sea, South Africa Orrell et al. (2002) Dentex abei AB458400 KPM-NI20690 Awase, Okinawa pref., Japan This study Dentex angolensis DQ197939 Unknown Unknown Unpublished Dentex canariensis DQ197940 Unknown Unknown Unpublished Dentex dentex AF143197 Unknown Unknown Allegrucci et al. (1999) Dentex gibbosus DQ197941 Unknown Unknown Unpublished Dentex hypselosomus AB458401 KPM-NI20728 Kute, Shimane pref., Japan This study Dentex macrophthalmus EF455993 Unknown Unknown Unpublished Dentex maroccanus DQ197942 Unknown Unknown Unpublished Orrell et al. (2002); They treated as Dentex spariformis AF240708 AMS I.36450-002 Nelson Bay, Dentex tumifrons AF240709 RUSI 49684 Kenton-on-Sea, South Africa Orrell et al. (2002) praeorbitalis AF240710 RUSI 49686 Kenton-on-Sea, South Africa Orrell et al. (2002) hasta AF240734 ODU 2783 Shuwaik Market, Kuwait City, Kuwait Orrell et al. (2002) Diplodinae probatocephalus AF240716 VIMS 010192 Chesapeake Bay, Virginia Orrell et al. (2002) EU036432 Unknown Unknown Unpublished Diplodus argenteus AF240721 NSMT-P48013 Sea Life Park Tokyo, origin: Argentina Orrell et al. (2002) Diplodus bermudensis AF240722 No voucher Bermuda Orrell et al. (2002) AF240723 RUSI 49680 Kenton-on-Sea, South Africa Orrell et al. (2002) Diplodus holbrookii AF240724 ODU 2789 Atlantic, South Carolina Orrell et al. (2002) EU036434 Unknown Unknown Unpublished Diplodus sargus cadenati DQ197946 Unknown Unknown Unpublished Diplodus sargus sargus EF439523 Unknown Unknown Unpublished EU036438 Unknown Unknown Unpublished Bahia Honda- side, Florida Keys, rhomboides AF240726 No voucher Orrell et al. (2002) Florida Pagellinae Boopsoidea inornata AF240711 ODU 2791 St. Sebastian Bay, South Africa Orrell et al. (2002) mormyrus AF240712 ODU 2784 Fiumicino Fish Market, Italy Orrell et al. (2002) DQ197970 Unknown Unknown Unpublished AF240714 ODU 2792 R/V African, Station 17491, South Africa Orrell et al. (2002) Pagellus bogaraveo AB305023 Unknown Tarifa, Cádiz, Spain Ponce et al. (2008) Pagellus erythrinus EU036466 Unknown Unknown Unpublished Pagrinae bleekeri AB458407 KPM-NI20686 Awase, Okinawa pref., Japan This study AF240717 AMS I.36447-001 N. Territory, Australia Orrell et al. (2002) Xia et al. (2007); They treated as Evynnis cardinalis EF107158 Unknown Guangdong Province, China Parargyrops edita Evynnis tumifrons AB458402 KPM-NI20792 Meitsu, Miyazaki pref., Japan This study Orrell et al. (2002); They treated as Evynnis tumifrons AF240725 NSMT-P 47497 Miyazaki pref., Japan Evynnis japonica Pagrus auratus AF240727 No voucher Sydney Fish Market, New Zealand Orrell et al. (2002) Pagrus auriga AB124801 Unknown Unknown Unpublished Pagrus caeruleostictus DQ197975 Unknown Unknown Unpublished AP002949 Unknown Unknown Miya et al. (2001) Pagrus pagrus AF240729 ODU 2790 Atlantic, North Carolina Orrell et al. (2002) Continued Molecular phylogeny of sparid fishes 157

Continued

Taxon Accession No. Voucher Collection Locality References & Remarks Sparinae AB458391 AMS I.42803-001 Sydney Fish Markets, Australia This study AB458392 MUFS-23434 Yaeyama Isls., Okinawa pref., Japan This study AB458393 AMS I.43204-019 Port Hacking, Point Danger, N.S.W, Aus- This study tralia AB458394 MUFS-23437 Kamikawaguchi, Kouchi pref., Japan This study AB458395 KPM-NI20489 Kasado Is., Yamaguchi pref., Japan This study Acanthopagrus sivicolus AB458396 URM-P40972 Awase, Okinawa pref., Japan This study Acanthopagrus taiwanensis AB458399 MUFS-22859 Tonkou, Taiwan This study Acanthopagrus chinsira AB458397 URM-P40886 Awase, Okinawa pref., Japan This study Acanthopagrus sp. AB458398 KPM-NR0043453 Queen Victoria Market, Victoria, Australia This study; Picture voucher nodosus AF240718 No voucher Atlantic, Charleston, SC Orrell et al. (2002) Chrysoblephus cristiceps AF240719 RUSI 49441 Kenton-on-Sea, South Africa Orrell et al. (2002) Cymatoceps nasutus AF240720 RUSI 49445 Kenton-on-Sea, South Africa Orrell et al. (2002) Porcostoma dentata AF240730 RUSI 58450 Durban, South Africa Orrell et al. (2002) Pterogymnus laniarius AF240731 No voucher Plettenberg Bay, South Africa Orrell et al. (2002) Rhabdosargus thorpei AF240732 RUSI 49683 Ponta do Ouro, Mozambique Orrell et al. (2002) AB458403 AMS I.41877-022 Spooky point, N. S. W., Australia This study Rhabdosargus sarba AB458404 AMS I.40872-004 Durban area, South Africa This study Rhabdosargus sarba AB458405 AMS I.41050-001 Mngazi Estuary, South Africa This study Rhabdosargus sarba AB458406 NSMT-P77931 Tanegashima Is., Kagoshima pref., Japan This study durbanensis AF240733 RUSI 49673 Kenton-on-Sea, South Africa Orrell et al. (2002) Sparus aurata AF240735 ODU 2787 Fiumicino Fish Market, Italy Orrell et al. (2002) Stenotomus chrysops AF240736 No voucher Chesapeake Bay Orrell et al. (2002) Nemipteridae, Sparoidea, Perciformes Manila Fish Market, Luson, Manila, Orrell et al. (2002) Nemipterus marginatus AF240754 USNM 345202 Philippines Scolopsis ciliatus AF240753 USNM 346853 Gunimaras Island, Philippines Orrell et al. (2002) Lethrinidae, Sparoidea, Perciformes obsoletus NC_009855 Unknown Unknown Yamanoue et al. (2007) Wattsia mossambica AF381261 AMS I.36447-005 Australia Lo Galbo et al. (2002) Centracanthidae, Sparoidea, Perciformes Spicara alta AF240738 ODU 2793 Angola Orrell et al. (2002) Spicara flexuosa EU036503 Unknown Unknown Unpublished AP009164 Unknown Unknown Yamanoue et al. (2007) EU036507 Unknown Unknown Unpublished Lateolabracidae, Perciformes japonicus AF240741 VIMS 10381 Market sample, Japan Orrell et al. (2002); Picture voucher Lateolabrax latus AF240743 MTUF 27451 Sasebo, Nagasaki Pref., Japan Orrell et al. (2002) Moronidae, Perciformes labrax X81566 Unknown Unknown Cantatore et al. (1994) VIMS Trawl Survey, Chesapeake Bay, americana AF240744 No voucher Orrell et al. (2002) Virginia Haemulidae, Perciformes Diagramma pictum NC_009856 Unknown Unknown Yamanoue et al. (2007) Parapristipoma trilineatum NC_009857 Unknown Unknown Yamanoue et al. (2007) Lujanidae, Perciformes Lutjanus rivulatus NC_009869 Unknown Unknown Yamanoue et al. (2007) Lutjanus stellatus DQ900662 Unknown Unknown Unpublished Caesionidae, Perciformes Caesio cuning AF240749 USNM 345193 Fish market, Iloilo Panay, Philippines Orrell et al. (2002) Pterocaesio tile NC_004408 Unknown Unknown Miya et al. (2003) Chaetodontidae, Perciformes auripes AP006004 Unknown Unknown Yamanoue et al. (2007) Heniochus diphreutes AP006005 Unknown Unknown Yamanoue et al. (2007) Pomacanthidae, Perciformes Centropyge loriculus AP006006 Unknown Unknown Yamanoue et al. (2007) Chaetodontoplus septentrionalis AP006007 Unknown Unknown Yamanoue et al. (2007) Triglidae, Satyrichthys amiscus AP004441 Unknown Unknown Miya et al. (2003) 158 S. N. CHIBA et al.

Fig. 1. Base substitution patterns in cytochrome b (cyt-b) genes of 72 sparid and 23 outgroup species. Transitional (TS) and trans- versional (TV) substitutions per site are plotted against pairwise evolutionary distances. The patterns for 1st, 2nd, and 3rd codons are shown respectively. model parameters were taken as unlinked and the rate tions in 12n3r data set, we set a single rate category (lset multipliers variable across partitions, unlinking substitu- nst = 1) instead of 6 (lset nst = 6) to allow the program to tion rates of the GTR model, the proportion of invariable estimate only transversional changes between purine (R) sites, the gamma shape parameters and base frequencies and pyrimidine (Y) nucleotides. The Markov chain across all partitions (“unlink revmat = (all) pinvar = (all) Monte Carlo search was run with four chains for shape = (all) statefreq = (all)”) and rate multipliers vari- 6,000,000 generations, sampling the Markov chain every able across partitions (“prset ratepr = variable”). We 1000 generations, and the sample points of the first 2,000 used the default setting for the priors on the proportion generations were discarded as “burn-in”, after which the of invariable sites (0–1) and the gamma shape parameter chain reached a stationarity state for each of the two data (0.1–50.0). A Dirichlet distribution was assumed for the sets. rate matrix and base frequency. For third codon posi- Molecular phylogeny of sparid fishes 159

Fig. 2. Phylogenetic tree of maximum likelihood analyses conducted on 123n data set under the GTR + I + Γ model. Numbers at branches indicate support values greater than 50 on 1,000 bootstrap replicates. Scientific names with asterisks indicate our new data. Major distribution legend is as follows: WA, Western Atlantic; EA, Eastern Atlantic; SeA, Southeastern Atlantic; M, Mediterranean; WI, Western Indian Ocean; I-WP, Indo-West Pacific; WP: Western Pacific. The distribution referred to FishBase (Froese and Pauly, 2008). 160 S. N. CHIBA et al.

Fig. 3. Phylogenetic tree of Bayesian analysis conducted on 123n data set under the GTR + I + Γ model. The topology of this tree fundamentally agrees with that of Maximum likelihood (on 12n3r data set) and Bayesian trees (on12n3r data set). Numbers above branches are more than 70 of posterior probabilities (an asterisk represent PP = 100) based on 123n (left) and 12n3r data sets (right). Numbers below branches indicate bootstrap supports greater than 50 on 1,000 bootstrap replicates in ML analysis of the 12n3r data set. Scientific names with asterisks indicate our new data. Molecular phylogeny of sparid fishes 161

Divergence time estimation The mcmctree program RESULTS in the phylogenetic package PAML v4 (Yang, 2007) was used by assuming a topological relationship thus Patterns of sequence variations DNA fragments con- obtained, but without assuming the molecular clock. For taining 1,141 bp of the complete cyt-b gene were success- the divergence time analysis based on multiple con- fully determined for 18 individuals from 15 sparid species straints, we newly constructed a topology, because we (Table 1). The cyt-b, protein-cording gene sequences have not yet been able to identify the sparid fossils cor- with 77 gene sequences searched from DNA databases rectly due to taxonomic confusion (see discussion) and the were aligned without any gaps. However, we only program does not allow multiple branching. The topology aligned 1,140 bp of sequences for the following analyses, was arranged based on ML tree (Fig. 2) for phylogenetic because some of the sequences from DNA databases relationships in sparid clades A and B, Bayesian tree lacked the last nucleotide of the complete cyt-b gene. (Fig. 3) for phylogenetic relationships in sparid genera, Both pairwise transitional (TS) and transversional (TV) Archosargus, Lagodon, Calamus and Stenotomus, and differences for each codon position are shown in Fig. Sparoidea families, and recent phylogenetic studies 1. The differences increased linearly with the evolu- (Yamanoue et al., 2006; Mabuchi et al., 2007; Azuma et tionary distance, except the TS at the 3rd codon position al., 2008) for non-Sparoidea and overall. Upper and/or in which marked saturation was observed in earlier lower time constraints at selected nodes were set for the stages of evolution (< 1 evolutionary distances) with no Bayesian MCMC processes to estimate divergence time. increases being observed thereafter. The mcmctree was run using the HKY85 substitution model that is the most parameter-rich model available in Phylogenetic inference by Maximum Likelihood this program. A total of 75,000 generations were sam- analysis ML analysis of the 123n data set was carried pled at every 2 steps after discarding 50,000 initial steps out according to the following statistics: nucleotide substi- as burn-in. We used seven time constraints based on tution rate matrix (a–c: 0.678, a–g: 5.064, a–t: 1.379, c–g: both fossil evidence and biogeographical assumptions for 0.467, and c–t: 10.096), (a–c: 0.469, a–g: 3.246, rate a–t: the divergences of continental cichlid groups (Azuma et 0.743, c–g: 3.652, and c–t: 4.440) and (a–c: 0.164, a–g: al., 2008) (Table 2). 9.966, a–t: 0.278, c–g: 0.676, and c–t: 3.170); assumed pro- portion of invariable sites 0.475, 0.540, and 0.005; and the Ancestral character-state reconstruction Day (2002), shape of the estimated gamma parameter (α) = 0.568, in his study of the phylogenetic relationships of the 0.580, and 2.573 for 1st, 2nd and 3rd codon positions, Sparidae, evaluated 87 morphological characters of the respectively. The analysis was conducted under the majority of sparid genera and representatives of the GTR + I + Γ model and produced a maximum likelihood sparoid families. Patterns of evolution of his 87 morpho- tree with log likelihood = –29,599.825 (Fig. 2). logical characters were investigated by the ancestral The Sparidae, Centracanthidae, Nemipteridae and character-state reconstruction using the parsimony Lethrinidae belonging to the Sparoidea were placed in a criterion in Mesquite ver 2.01 (Maddison and Maddison, monophyletic clade in the ML tree derived from 123n data 2007). Day (2002) omitted six genera, (Cymatoceps, set. However, its bootstrap support (BS) was considered Gymnocrotaphus, Peterus, Pterogymnus, Taius, and low (< 50). The family Moronidae was placed in a sister- Sparidentex) due to the unavailability of specimens, and group of the Sparoidea. The analysis showed monophyly considered that Taius and Sparidentex are subgenera of of Sparidae plus Centracanthidae with strong statistical Dentex. Therefore, we applied characters of Sparidentex to support (BS = 100). The sparid species were divided into those of Dentex and omitted Cymatoceps, Gymnocrotaphus, three major clades, A, B, and C, and some sub-clades. Peterus and Pterogymnus from the analysis. The charac- The three major clades were supported by 95, 86 and 75 ters were traced onto the Bayesian tree. BS, respectively.

Table 2. Maximum (U) and minimum (L) time constrains (MYA) used for dating at nodes in Fig. 4

Node Constraint Reference information 1 U221 Estimated divergence time appear (Yamanoue et al., 2006) 2 L50 Pleuronectiform fossil from the Ypresian (Patterson, 1993) 3 L112, U145 Continental breakup between Africa + South Amrica and Madagascar + Indo/Sri Lanka (Azuma et al., 2008) 4 L100, U120 Continental breakup between Africa and South Amrica (Azuma et al., 2008) 5 L85, U95 Continental breakup between Madagascar and Indo/Sri Lanka (Azuma et al., 2008) 6 L50 Sparid and Centracanthid fossils from the Ypresian (Patterson, 1993) 7 L32 Estimated divergence time between Takifugu and Tetraodon (Benton and Donoghue, 2007) 162 S. N. CHIBA et al.

However, none of the six subfamilies in the Sparidae Sparidentex hasta was a sister species to Acanthopagrus was monophyletic in the ML tree. Four species of the latus and belonged to the Acanthopagrus clade (B-6) with centracanthid Spicara were additionally analyzed strong support (BS = 100). Acanthopagrus schlegelii, A. in this study. The genus Spicara was paraphyletic and sivicolus and A. taiwanensis were closely related and was placed in both major sparid major clades A and their cyt-b gene sequences were almost identical. B. The genera Acanthopagrus, Cheimerius, Dentex, Acanthopagrus sp. was a sister species of A. butcheri and Diplodus, Pagellus and Pagrus were not monophyletic. had genetically diverged from other Acanthopagrus

Fig. 4. Divergence time for each lineage in Sparoidea estimated from the partitioned Bayesian analysis. Posterior divergence time with 95% credibility intervals (shaded rectangles) was obtained based on cyt-b sequences (1,132 bp). Numbers at nodes indicate max- imum and/or minimum time constraints (see table 2 for details of the individual constraints). Molecular phylogeny of sparid fishes 163 species. This study also included four individuals of Table 3. Number of parsimony steps investigated by ancestral Rhabdosargus sarba collected from South Africa, Australia character-state reconstruction based on Day’s (2002) 87 and Japan. The Australian R. sarba was closely related morphological characters and the Bayesian tree (Fig. 3) to the Japanese and South African strains, which were Character Steps Character Steps placed at a sister position of the two. Oblada melanura Braincase Oral Jaws (continued) was placed in the Diplodus clade (B-4). 1 9 infraorbitals Another ML analysis of the 12n3r data set was carried 2 12 48 4 out according to the following statistics: nucleotide sub- 3 15 49 3 stitution rate matrix (a–c: 0.682, a–g: 5.084, a–t: 1.452, c– 5 9 palatine arch g: 0.436, and c–t: 10.290), (a–c: 0.5430, a–g: 3.583, rate a– 6 7 50 9 t: 0.800, c–g: 3.932, and c–t: 4.532) and (a–c: 0.540, a–g: 8 7 51 9 0.000, a–t: 0.000, c–g: 0.000, and c–t: 0.000); assumed pro- 9 12 52 2 portion of invariable sites 0.473, 0.535, and 0.307; and the 10 4 53 9 shape of the estimated gamma parameter (α) = 0.561, 11 3 opecular bones and lower 0.607, and 1.898 for 1st, 2nd and 3rd codon positions, 12 4 part of the hyoid arch 13 11 57 9 respectively. The analysis was conducted under the 14 8 58 13 GTR + I + Γ model and produced a maximum likelihood 15 8 59 10 tree with log likelihood = –13,124.290 (only BSs are 16 7 60 2 shown in Fig. 3). 17 1 62 1 Topology of the ML tree derived from the 12n3r data set 18 10 Pharyngeal jaws fundamentally agreed with that derived from the 123n 19 15 lower pharyngeal jaws data set. However, the major topological differences are 20 1 63 5 the position of the genera, Archosargus, Lagodon, Calamus 21 3 65 17 and Stenotomus, were diverged from the other sparid spe- 22 7 66 15 cies as basal lineages (Calamus (Stenotomus (Archosargus, 23 7 67 6 Lagodon)(Clade A, Clade B))). 24 12 68 12 25 10 upper pharyngeal jaws Phylogenetic inference by Bayesian analysis As Oral Jaws 69 9 shown in Fig. 3, a 50% majority rule consensus tree was upper jaw 70 10 obtained from the resultant 4,001 trees by Bayesian anal- 26 19 71 6 ysis of the cyt-b 123n data set. The Bayesian analysis 27 16 72 9 gave the following statistics posteriorly: the nucleotide 28 1 73 5 substitution rate matrix (a–c: 0.324, a–g: 4.169, a–t: 29 3 74 10 1.592, c–g: 0.211, and c–t: 6.803), (a–c: 0.658, a–g: 3.192, 30 14 Pectral and Pelvic girdles a–t: 0.808, c–g: 3.589, and c–t: 4.452) and (a–c: 0.169, a–g: 31 7 76 1 10.277, a–t: 0.308, c–g: 0.646, and c–t: 2.985); assumed 32 5 77 1 proportion of invariable sites 0.405, 0.526, and 0.008; and 33 4 Axial skeleton mandible 78 3 the shape of the estimated gamma parameter (α) = 0.467, 34 9 Dorsal and anal fins 0.671, and 2.690 for 1st, 2nd and 3rd codon positions, 35 16 79 2 respectively. 36 7 80 5 Topology of the Bayesian tree fundamentally agreed 37 12 81 2 with that of the ML tree derived from RY-coding. The 38 14 82 2 major topological differences between the ML tree based 39 13 Caudal fin on 123n data set and the Bayesian tree are the positions 40 11 84 15 of the family Moronidae and the genera, Archosargus, 41 11 85 14 Lagodon, Calamus and Stenotomus. In the Bayesian dentition 86 1 tree, Moronidae was placed into the Sparoidea with 42 14 87 1 strong posterior probability (PP = 1.00). The genera, 43 7 Archosargus, Lagodon, Calamus and Stenotomus, were 44 11 diverged from the other sparid species as basal lineages 45 12 (Calamus (Stenotomus (Archosargus, Lagodon)(Clade A, 46 9 Clade B))). 47 12 Another Bayesian tree based on the 12n3r data set was See Day (2002) for the character numbers. similar to that of 123n data set (only PPs are shown in Characters 4, 7, 54, 55, 56, 61, 64, 75 and 83 are not shown, Fig. 3). The Bayesian analysis gave the following statis- because they don’t have any parsimony steps. 164 S. N. CHIBA et al. tics posteriorly: the nucleotide substitution rate matrix Timing of sparid divergences Divergence time esti- (a–c: 0.323, a–g: 4.154, a–t: 1.600, c–g: 0.200, and c–t: mation was conducted among 94 bony fishes, including 72 8.092), (a–c: 0.662, a–g: 3.577, a–t: 0.944, c–g: 3.169, and sparids (Fig. 4). The estimated divergence time between c–t: 4.732), and (single rate parameter); assumed propor- Nemipteridae (Nemipterus marginatus) and Sparidae was tion of invariable sites 0.432, 0.546, and 0.309; and the 115.1–186.2 million years ago (MYA). It was estimated shape of the estimated gamma parameter (α) = 0.510, that the genera, Calamus, Stenotomus, and Archosargus 0.175, and 2.470 for 1st, 2nd and 3rd codon positions, + Lagodon, and other sparids diverged in order during respectively. 71.8–146.9 MYA. It was similarly estimated that clades

Fig. 5. Continued Molecular phylogeny of sparid fishes 165

A and B diverged 64.9–104.9 MYA. needed to explain the evolution of characters in the Bayesian tree is shown in Table 3. Many parsimony Patterns of evolution of morphological characters steps were needed to explain the evolution of the brain- Patterns of evolution of morphological characters among case, oral jaws, pharyngeal jaws, and caudal fin charac- genera were presumed based on Day’s (2002) 87 mor- ters. Only four patterns of evolution of dentition phological characters. The number of parsimony steps character, caniniform, conical, molariform, and incisiform

Fig. 5. Evolution of dentition patterns reconstructed on the Bayesian tree. The patterns of caniniform, conical, molariform, and incisiform are shown in A, B, C and D, respectively. The red, blue, white, and green arrowheads indicate the estimated point that the character was lost, gained, equivocally changed, and crenulated, respectively. 166 S. N. CHIBA et al. teeth, which have been used to distinguish genera and to Archosargus, and Lagodon in the ML tree derived from provide a subfamily classification, were shown in Fig. 5- the 123n data set (Fig. 2). These genera diverge from A, B, C and D, respectively. The patterns of evolution of the other sparid species as basal lineages in the ML tree caniniform, conical, molariform, and incisiform teeth derived from the 12n3r data set and the Bayesian trees were explained parsimoniously by 7, 11, 12 and 9 steps, (Fig. 3). Orrell et al. (2002) and Orrell and Carpenter respectively. (2004) also pointed out that the phylogenetic positions of those genera are uncertain. The genera positions agreed completely among the Bayesian tree derived from the DISCUSSION 123n data set and two methods tree (ML and Bayesian) Inter-family relationships The Bayesian trees derived derived from the data set excluded the TS at 3rd codons from two different data sets (123n and 12n3r) showed by RY-coding. The result suggests that the topological monophyly of the superfamily Sparoidea with the family difference between the ML tree derived from the 123n Moronidae, which consists of the genera Morone and data set and the other trees is mainly due to saturation Dicentrachus (Fig. 3). The ML tree derived from the of 3rd codon position in TS. We accordingly hypothe- 123n data set also showed that Sparoidea is monophyletic sized that the Bayesian trees derived from the two data and is positioned as a sister clade of the Moronidae (Fig. sets (123n and 12n3r) and the ML tree derived from 2). These results partially agree with the superfamily 12n3r represented better estimation of phylogenies on the proposed by Johnson (1980), whereas their statistical genera. From these reliable phylogenies, it is thought support was somewhat low (BS < 50). Both ML and that common ancestors of the genera diverged from the Bayesian trees derived from the 123n data set showed other sparid species and migrated to the eastern Pacific that families Haemulidae, Lutjanidae, Caesionidae, Cha- and western Atlantic in early stages of their evolutionary etodontidae and Pomacanthidae were positioned in a history. monophyletic clade (Figs. 2, 3). Nelson (2006) provision- The estimated divergence time between the genera, ally put the Asian genus Lateolabrax in the family Moron- Calamus, Stenotomus, Archosargus + Lagodon, and other idae, although the independent family Lateolabracidae, sparids corresponded to (Fig. 4). Akazaki which includes the genus, is supported by Eschmeyer (1970), in his study on the biogeography of sparid fishes, (2008) and Springer and Johnson (2004). However, our speculated that the ancestors of sparids probably origi- analyses strongly indicate that the species of Lateolabrax nated in a Mesozoic ocean of Tethys Sea from the Upper are far from those of the Moronidae and the family Lateo- Cretaceous to Paleocene and migrated to the coastal labracidae is independent from the Moronidae. region of the New world during the early Paleocene. This All ML and Bayesian trees derived from the two data speculation generally coincides with our hypothesis. sets (123n and 12n3r) also showed that the family Clades A and B consist of species that are distributed Sparidae is monophyletic with strong support and that in various oceanic regions at present (Fig. 2). Some sub- the family Centracanthidae is contained within the clades, however, consist of species that are distributed in Sparidae clade (Figs. 2, 3). This evidence agrees with defined local regions (Fig. 2). Moreover, the estimated recent molecular analyses using fewer species based on divergence time between clades A and B corresponded to the combined data of cyt-b and 16S rRNA (Orrell and Upper Cretaceous (Fig. 4). Therefore, it is considered Carpenter, 2004). The family Centracanthidae includes that divergence between the common ancestors of clade A two genera, Centracanthus and Spicara, with eight species and B occurred in the Tethys Sea. Subsequently, these (Nelson, 1994). Because only four species of Spicara were ancestors have been moving around tropical and tem- analyzed in this study, the problem of the Centracanthidae perate coastal waters. has not been completely resolved. Additional phyloge- netic analyses using all species in the Centracanthus and Disagreement between molecular phylogeny and Spicara will clarify the validity of the family Centra- current classification in the Sparidae There are no canthidae. monophyletic subfamilies in either phylogenetic tree (Figs. 2, 3). The results fundamentally agree with those Intra-family relationships and phylogeography in of recent molecular analyses (Hanel and Sturmbauer, the Sparidae Sparid species were divided into three 2000; Orrell et al., 2002; Orrell and Carpenter, 2004). major clades in the ML tree derived from the 123n data Dentition is the most important character that has been set (Fig. 2). The another ML tree derived from the 12n3r used to classify subfamilies (Smith, 1938; Smith and data set and the Bayesian trees derived from the two data Smith, 1986; Akazaki, 1962). However, as mentioned sets (123n and 12n3r) also indicated similar intra-family below, our results showed that the evolutionary changes relationships to the ML tree derived from the 123n data in dentition occurred independently in some lineages of set. However, there is a major topological difference for the Bayesian tree (Fig. 5). clade C, which consists of the genera Calamus, Stenotomus, Although nine sparid and one centracanthid genera Molecular phylogeny of sparid fishes 167 including multiple species were analyzed in this study, pendently from the ancestors. The results suggested only three of the genera were monophyletic (Figs. 2, high evolutionary plasticity of many morphological char- 3). The currently recognized genera Dentex, Spicara, acteristics in sparid evolution. Therefore, not only den- Pagrus, Cheimerius, and Pagellus were divided into phy- tition characters but also many morphological characters logenetically diverged clades (Figs. 2, 3). have multiple independent origins. This information The genus Acanthopagrus was placed into the mono- helps to explain the causes of species-diversity and taxo- phyletic sub-clade, B-6, with S. hasta (Figs. 2, 3). In fact, nomic confusion at various taxonomic levels from species the body and fin color of S. hasta resemble those of to family in the Sparidae rather than other Johnson‘s Acanthopagrus species (Munro, 1948). Moreover, recent (1980) sparoid families. molecular studies suggested that S. hasta is closely As mentioned above, common ancestors of the major related to Acanthopagrus spp. (Orrell et al., 2002; Orrell sparid clades, recognized in this study, might have differ- and Carpenter, 2004). Because 12 species are currently entiated in the Tethys Sea. According to the hypothesis, regarded as valid for Acanthopagrus (Kume and Yoshino, sparids have undergone adaptive radiation in order to 2008) and it is still unclear how many species belong to take advantage of various trophic and environmental con- Sparidentex (Iwatsuki and Carpenter, 2006; Kume and ditions by their high evolutionary plasticity. For exam- Yoshino, 2008), additional phylogenetic analysis using all ple, the molariform teeth probably developed to eat hard species in Acanthopagrus and Sparidentex will be neces- preys, such as Crustacea and armored Mollusca (Akazaki, sary to clarify the validity of the classification. A genet- 1962). The molariform teeth were possibly absent in the ically distinct species, Acanthopagrus sp. was found in sparid common ancestor, gained at least five times inde- Australia. Although the species was morphologically pendently, and were subsequently lost at least four times similar to A. butcheri, it was genetically divergent from independently in the sparid lineages (Fig. 5C). As a all other Acanthopagrus species (Figs. 2, 3). result, many sparid species evolved and were distributed The sub-clade B-3 in Figs. 2 and 3 consists of Sparus, world wide, and the adaptive characters resemble each Rhabdosargus and Sparodon. The generic name that other among the distinct clades due to convergent evolu- should be applied to Sparus sarba Forsskål, 1775 is very tion. Therefore, phylogenetic studies based on morpho- confusing among fish taxonomists. Some researchers logical characters were very confused and generated proposed that the genus Rhabdosargus Fowler, 1933 incorrect phylogenetic hypotheses. should be applied to the species (e.g. Smith and Smith, To obtain further supporting evidence for the hypothe- 1986; Dor, 1984; Bauchot and Skelton, 1986). Although sis, it is necessary to estimate the more accurate diver- our phylogenetic trees agree with this proposal, some gence time with multiple constraints based on fossil Japanese taxonomists have still insisted that the genus records within sparids. Although many fossil records of Sparus should be applied to the species (Akazaki, 1962, sparids have been reported (e.g. Patterson, 1993; Day, 1984; Hayashi, 2002). Our results clearly show the 2003), we have not yet been able to assign the fossils to genetic divergence between S. aurata and R. sarba. the phylogenetic positions correctly due to taxonomic Four specimens of R. sarba collected from South Africa, confusion. Further molecular phylogenetic and taxo- Japan and Australia were analyzed in this study. The nomic studies using many more species will enable an results clearly show that the species are highly diverged accurate estimation of the divergence time among species genetically between the Indian and Pacific (Figs. and could clarify the complicated evolutionary features of 2, 3). Moreover, our other study shows that genetic sparids. divergence occurs among populations in some western Pacific regions (Chiba et al., in preparation). It is We thank Y. Hara and H. Tamate, Yamagata University, K. urgently necessary to revise the various taxonomic levels Kuriiwa, National Museum of Nature and Science, and T. from species to family because there are many problems Sasaki, Tokyo Institute of Technology, for their helpful advice and discussions. We also thank H. Motomura, The Kagoshima in the classification of the Sparidae as shown here. University Museum, M. McGrouther, The Australian Museum, and R. Hoshina, Ritsumeikan University, for providing Austra- Convergent evolution of morphology Previous stu- lian and South African specimens. M. Katoh, Seikai National dies showed that trophic types of sparids had multiple Fisheries Research Institute, Y. Sakurai, Okinawa Environmen- independent origins (Hanel and Sturmbauer, 2000; Day, tal Research Co., Ltd., H. Ishimori, and R. Fujii, University of the Ryukyus, for providing and assisting with other specimen 2002), which is further supported by our results. Namely collections. We also thank H. Senou and R. Takahashi, the evolutionary patterns of four dentition characters Kanagawa Prefectural Museum of Natural History, for register- have multiple independent origins (Fig. 5). Many parsi- ing and storing specimens. We also thank two anonymous mony steps were also needed to explain the evolution of reviewers for their constructive comments on an earlier draft. the braincase, oral jaws, pharyngeal jaws, and caudal fin This work was partially supported by a grant from Ito Grant for Ichthyology, Fujiwara Natural History Foundation awarded to characters (Table 3). The number of parsimony steps the first author. indicates the number of times of a character evolved inde- 168 S. N. CHIBA et al.

isation, pour servir de base à l’histoire naturelle des ani- REFERENCES maux et d’introduction à l’anatomie comparée. Edition 2. Règne (ed. 2) 2, Deterville, Paris. Akazaki, M. (1962) Studies on the perciform fishes: anatomy, Day, J. D. (2002) Phylogenetic relationships of the Sparidae phylogeny, ecology, and (in Japanese with (Teleostei: Percoidei) and implications for convergent English summary). Misaki Mar. Biol. Inst. Kyoto Univ. trophic evolution. Biological Journal of the Linnean Soci- Spec. Rep. 1, Kosugi Ltd., Osaka. ety 76, 269–301. Akazaki, M. (1970) Specific divergence of spariform fishes when Day, J. D. (2003) Evolutionary relationships of the Sparidae considered from their geographical distribution. Honyurui- (Teleostei: Percoidei): integrating fossil and Recent data. kagaku 20–21, 131–146. Transactions of the Royal Society of Edinburgh. Earth Sci- Akazaki, M. (1984) Sparidae. In: The fishes of the Japanese ences 93, 333–353. archipelago (eds.: H. Masuda, K. Amaoka, C. Araga, T. de la Herrán, R., Rejón, C. R., Rejón, M. R., and Garrido-Ramos, Uyeno and T. Yoshino), pp. 170–173. pls 166–167. Tokai M. A. (2001) The molecular phylogeny of the Sparidae University Press, Tokyo. (Pisces, Perciformes) based on two satellite DNA families. Akaike, H. (1973) Information theory as an extension of the Heredity 87, 691–697. maximum likelihood principle. In: Second international Dor, M. (1984) Checklist of the fishes of the (CLOF- symposium on information theory (eds.: B. N. Petrov and F. RES). pp. 437. The Israel Academy of Sciences and Human- Caski), pp. 267–281. Akademia Kiado, Budapest. ities, Jerusalem. Akihito, Iwata, A., Kobayashi, T., Ikeo, K., Imanishi, T., Ono, H., Duftner, N., Sefc, K. M., Koblmuller, S., Salzburger, W., and Umehara, Y., Hamamatsu, C., Sugiyama, K., Ikeda, Y., et Taborsky, M. (2007) Parallel evolution of facial stripe pat- al. (2000) Evolutionary aspects of gobioid fishes based upon terns in the Neolamprologus brichardi/pulcher species com- a phylogenetic analysis of mitochondrial cytochrome B plex endemic to Lake Tanganyika. Mol. Phylogenet. Evol. genes. Gene 259, 5–15. 45, 706–715. Allegrucci, G., Caccone, A., and Sbordoni, V. (1999) Cytochrome Eschmeyer, W. N. (2008) electronic version b sequence divergence in the European sea (updated 23 April 2008): http://www.calacademy.org/research/ (Dicentrarchus labrax) and phylogenetic relationships ichthyology/catalog/fishcatsearch.html among some Perciformes species. J. Zoolog. Syst. Evol. Felsenstein, J. (1985) Confidence limits on phylogenies: an Res. 37, 149–156. approach using the bootstrap. Evolution 39, 783–791. Azuma, Y., Kumazawa, Y., Miya, M., Mabuchi, K., and Nishida, Fiedler, K. (1991) Family Sparidae. In: Lehrbuch der speziellen N. (2008) Mitogenomic evaluation of the historical biogeog- zoologie, Teil2: Fische, (ed.: D. Starck), pp. 354–355, Gustav raphy of cichlids toward reliable dating of teleostean Fischer Verlag, Jena. divergences. BMC Evol. Biol. 8, 215. Froese, R., and Pauly, D. (2008) FishBase. World Wide Web elec- Basagila, F. (1991) Interspecific gene differences and phylogeny tronic publication. www..org, version (06/2008). of the Sparidae family (Perciformes, Teleostei), estimated Garrido-Ramos, M. A., Jamilena, M., Lozano, R., Cardenas, S., from electrophoretic data on enzymatic tissue-expression. Rejon, C. R., and Rejon, M. R. (1995) Phylogenetic relation- Comp. Biochem. Physiol., B 99, 495–508. ships of the Sparidae family (Pisces, Perciformes) inferred Bauchot, M. L., and Hureau, J. C. (1986) Sparidae. In: Fishes of from satellite-DNA. Hereditas 122, 1–6. the Northern-Eastern Atlantic and the Mediterranean, Vol. Garrido-Ramos, M. A., de la Herrán, R., Rejón, C. R., and Rejón, 2. (eds.: P. J. P. Whitehead, M. L. Bauchot, J. C. Hureau, J. M. R. (1998) A satellite DNA of the Sparidae family (Pisces, Nielsen and E. Tortonese), pp. 883–907, United Nations Perciformes) associated with telomeric sequences. Cytoge- Educational Scientific and Cultural Organization, Paris. net. Cell Genet. 83, 3–9. Bauchot, M.-L., and Skelton, P. H. (1986) Sparidae. In: Check- Garrido-Ramos, M. A., de la Herrán, R., Jamilena, M., Lozano, list of the freshwater fishes of Africa (CLOFFA) vol. 2 (eds.: R., Rejón, C. R., and Rejón, M. R. (1999) Evolution of centro- J. Daget, J.-P. Gosse and D.F.E. Thys van den Audenaerde), meric satellite DNA and its use in phylogenetic studies of pp. 331–332. Tervuren; and ORSTOM, Paris. the Sparidae family (Pisces, Perciformes). Mol. Phylo- Benton, M. J., and Donoghue, P. C. (2007) Paleontological evi- genet. Evol. 12, 200–204. dence to date the tree of life. Mol. Biol. Evol. 24, 26–53. Graybeal, A. (1998) Is it better to add taxa or characters to a dif- Cantatore, P., Roberti, M., Pesole, G., Ludovico, A., Milella, F., fcult phylogenetic problem? Syst. Biol. 47, 9–17. Gadaleta, M. N., and Saccone, C. (1994) Evolutionary ana- Hanel, R., and Sturmbauer, C. (2000) Multiple recurrent evolu- lysis of cytochrome b sequences in some Perciformes: evi- tion of trophic types in northeastern Atlantic and Mediter- dence for a slower rate of evolution than in mammals. J. ranean seabreams (Sparidae, Percoidei). J. Mol. Evol. 50, Mol. Evol. 39, 589–597. 276–283. Carpenter, K. E. (2001) Family Sparidae. In: The living marine Harrison, G. L., McLenachan, P. A., Phillips, M. J., Slack, K. E., resources of the western central Pacific, FAO species identi- Cooper, A., and Penny, D. (2004) Four new avian mitochon- fication guide for fishery purposes. vol. 5, part 3 (eds.: K. E. drial genomes help get to basic evolutionary questions in the Carpenter and V. H. Niem), pp. 2990–3003. FAO, Rome. late Cretaceous. Mol. Biol. Evol. 21, 974–983. Carpenter, K. E., and Johnson, G. D. (2002) A phylogeny of spa- Hayashi, M. (2002) Sparidae. In: Fishes of Japan with pictorial roid fishes (Perciformes, Percoidei) based on morphology. keys to species, English edition (ed: T. Nakabo), pp. 856– Ichthyol. Res. 49, 114–127. 859. Tokai University Press, Tokyo. Cuvier, G. (1817) Le Règne Animal distribué d’après son organ- Hillis, D. M. (1996) Inferring complex phylogenies. Nature 383, isation pour servir de base à l’histoire naturelle des ani- 130–131. maux et d’introduction à l’anatomie comparée. Les reptiles, Hillis, D. M. (1998) Taxonomic sampling, phylogenetic accuracy, les poissons, les mollusques et les annélides. Edition 1. and investigator bias. Syst. Biol. 47, 3–8. Règne Animal (ed. 1) 2, Deterville, Paris. Hillis, D. M., Pollock, D. D., McGuire, J. A., and Zwick, D. J. Cuvier, G. (1829) Le Règne Animal, distribué d’après son organ- (2003) Is sparse taxon sampling a problem for phylogenetic Molecular phylogeny of sparid fishes 169

inference? Syst. Biol. 52, 124–126. name for Chrysophrys cuvieri Day. Copeia, 275–280. Inoue, J. G., Miya, M., Tsukamoto, K., and Nishida, M. (2000) Munro, I. S. R. (1949) Revision of australian silver breams Mylio Complete mitochondrial DNA sequence of the Japanese sar- and Rhabdosargus. Mem. Qld. Mus. 12, pls. xvi–xxiii, pp. dine Sardinops melanostictus. Fisheries science 66, 924– 182–223. 932. Nelson, J. S. (1994) Fishes of the world, 3rd ed., John Wiley & Iwatsuki, Y., and Carpenter, K. E. (2006) Acanthopagrus Sons, Inc., New York. taiwanensis, a new sparid fish (Perciformes), with compari- Nelson, J. S. (2006) Fishes of the world, 4th ed., John Wiley & sons to Acanthopagrus berda (Forsskal, 1775) and other Sons, Inc., Hoboken, New Jersey. nominal species of Acanthopagrus. Zootaxa 1202, 1–19. Nishida, M., Ohkawa, T., and Iwata, H. (1998) Methods of anal- Iwatsuki, Y., Akazaki, M., and Taniguchi, N. (2007) Review of ysis of genetic population structure with mitochondrial DNA the species of the genus Dentex (Perciformes: Sparidae) in markers. Suisanikusyu 26, 36–55. the western Pacific defined as the D. hypselosomus complex Orrell, T. M., Carpenter, K. E., Musick, J. A., and Graves, J. E. with the description of a new species, Dentex abei and a (2002) Phylogenetic and biogeographic analysis of the redescription of Evynnis tumifrons. Bull. Natl. Mus. Nat. Sparidae (Perciformes: Percoidei) from cytochrome b sequ- Sci., Ser. A, Suppl. 1, 29–49. ences. Copeia, 618–631. Jean, C. T., Hui, C. F., Lee, S. L., and Chen, C. T. (1995) Varia- Orrell, T. M., and Carpenter, K. E. (2004) A phylogeny of the tion in mitochondrial DNA and phylogenetic relationships of fish family Sparidae (porgies) inferred from mitochondrial fishes of subfamily Sparinae (Perciformes: Sparidae) in the sequence data. Mol. Phylogenet. Evol. 32, 425–434. Coastal waters of Taiwan. Zoological Studies 34, 270–280. Patterson, C. (1993) Osteichthyes: Teleostei. In: The fossil record Johnson, G. D. (1980) The limits and relationships of the Lutjan- 2 (ed.: M. J. Benton), pp. 621–656. Chapman & Hall, London. idae and associated families. Bull. Scripps Inst. Oceanogra- Pérez, G. A. C., Řĺčan, O., Ortĺ, G., Bermingham, E., Doadrio, I., phy, University of California press, Ltd., London. and Zardoya, R. (2007) Phylogeny and biogeography of 91 Kume, M., and Yoshino, T. (2008) Acanthopagrus chinshira, a species of heroine cichlids (Teleostei: Cichlidae) based on new sparid fish (Perciformes: Sparidae) from the East Asia. sequences of the cytochrome b gene. Mol. Phylogenet. Evol. Bull. Natl. Mus. Nat. Sci., Ser. A, Suppl. 2, 47–57. 43, 91–110. Kuriiwa, K., Hanzawa, N., Yoshino, T., Kimura, S., and Nishida, Phillips, M. J., and Penny, D. (2003) The root of the mammalian M. (2007) Phylogenetic relationships and natural hybridi- tree inferred from whole mitochondrial genomes. Mol. zation in rabbitfishes (Teleostei: Siganidae) inferred from Phylogenet. Evol. 28, 171–185. mitochondrial and nuclear DNA analyses. Mol. Phylo- Pollock, D. D., Zwichkl, D. J., McGuire, J. A., and Hillis, D. M. genet. Evol. 45, 69–80. (2002) Increased taxon sampling is advantageous for phylo- Lo Galbo, A. M., Carpenter, K. E., and Reed, D. L. (2002) Evolu- genetic inference. Syst. Biol. 51, 664–671. tion of trophic types in emperor fishes (Lethrinus, Lethrin- Ponce, M., Infante, C., Jimenez-Cantizano, R. M., Perez, L., and idae, Percoidei) based on cytochrome B gene sequence Manchado, M. (2008) Complete mitochondrial genome of the variation. J. Mol. Evol. 54, 754–762. , Pagellus bogaraveo (Perciformes: Mabuchi, K., Miya, M., Azuma, Y., and Nishida, M. (2007) Inde- Sparidae), with high levels of length heteroplasmy in the pendent evolution of the specialized pharyngeal jaw appara- WANCY region. Gene 409, 44–52. tus in cichlid and labrid fishes. BMC Evol. Biol. 7, 10. Posada, D., and Crandall, K. A. (1998) Modeltest: testing the Maddison, W. P., and Maddison, D. R. (2007) Mesquite: A mod- model of DNA substitution. Bioinformatics 14, 817–818. ular system for evolutionary analysis. version 2.0. http:// Rokas, A., and Caroll, S. B. (2005) More Genes or more taxa? mesquiteproject.org. The relative contribution of gene number and taxon number Miller, J. A., and Hormiga, G. (2004) Clade stability and the to phylogenetic accuracy. Mol. Biol. Evol. 22, 1337–1344. addition of data: a case study from erigonine spiders (Ara- Ronquist, F., and Huelsenbeck, J. P. (2003) MrBayes 3: neae: Linyphiidae, Erigoninae). Cladistics 20, 385–442. Bayesian phylogenetic inference under mixed models. Bio- Minegishi, Y., Aoyama, J., Inoue, J. G., Miya, M., Nishida, M., informatics 19, 1572–1574. and Tsukamoto, K. (2005) Molecular phylogeny and evolu- Rosenberg, M. S., and Kumar, S. (2001) Incomplete taxon sam- tion of the freshwater eels genus Anguilla based on the pling is not a problem for phylogenetic inference. Proc. whole mitochondrial genome sequences. Mol. Phylogenet. Natl. Acad. Sci. USA 98, 10751–10756. Evol. 34, 134–146. Rosenberg, M. S., and Kumar, S. (2003) Taxon sampling, bioin- Miya, M., Kawaguchi, A., and Nishida, M. (2001) Mitogenomic formatics, and phylogenomics. Syst. Biol. 52, 119–124. exploration of higher teleostean phylogenies: a case study Saitoh, K., Sado, T., Mayden, R. L., Hanzawa, N., Nakamura, for moderate-scale evolutionary genomics with 38 newly K., Nishida, M., and Miya, M. (2006) Mitogenomic evolution determined complete mitochondrial DNA sequences. Mol. and interrelationships of the Cypriniformes (: Biol. Evol. 18, 1993–2009. Ostariophysi): the first evidence toward resolution of higher- Miya, M., and Nishida, M. (1999) Organization of the mitochon- level relationships of the world’s largest freshwater fish drial genome of a deep-sea fish, Gonostoma gracile (Teleo- clade based on 59 whole mitogenome sequences. J. Mol. stei: Stomiiformes): first example of transfer RNA gene Evol. 63, 826–841. rearrangements in bony fishes. Mar. Biotechnol. 1, 416– Sasaki, T., Kartavtsev, Y. P., Chiba, S. N., Uematsu, T., Sviri- 426. dov, V. V., and Hanzawa, N. (2006) Genetic divergence and Miya, M., Takeshima, H., Endo, H., Ishiguro, N. B., Inoue, J. G., phylogenetic independence of Far Eastern species in sub- Mukai, T., Satoh, T. P., Yamaguchi, M., Kawaguchi, A., family Leuciscinae (Pisces: Cyprinidae) inferred from mito- Mabuchi, K., et al. (2003) Major patterns of higher teleo- chondrial DNA analyses. Genes Genet. Syst. 82, 329–340. stean phylogenies: a new perspective based on 100 complete Smith, J. L. B. (1938) The south African fishes of the families mitochondrial DNA sequences. Mol. Phylogenet. Evol. 26, Sparidae and Denticidae. Transactions of the Royal society 121–138. of south Africa 26, pls. xviii–xxix, pp. 225–305. Munro, I. S. R. (1948) Sparidentex hasta (Valenciennes), a new Smith, J. L. B., and Smith, M. M. (1986) Sparidae. In: Smiths’ 170 S. N. CHIBA et al.

sea fishes, (eds.: M. M. Smith and P. C. Heemstra), pls. (Amphiprioninae, Pomacentridae) in the centre of marine lviii–lxii, pp. 580–594, Springer-Verlag, Berlin. biodiversity. Mol. Phylogenet. Evol. 49, 268–276. Springer, V. G., and Johnson, G. D. (2004) Study of the dorsal Westneat, M. W., Alfaro, M. E., Wainwright, P. C., Bellwood, D. gill-arch musculature of teleostome fishes, with special ref- R., Grubich, J. R., Fessler, J. L., Clements, K. D., and erence to the Actinopterygii. Bull. Bio. Soc. Wash. 11, 1– Smith, L. L. (2005) Local phylogenetic divergence and global 260. evolutionary convergence of skull function in reef fishes of Stamatakis, A., Hoover, P., and Rougemont, J. (2008) A Rapid the family Labridae. Proc. R. Soc. B 272, 993–1000. Bootstrap Algorithm for the RaxML Web-Servers. Syst. Xia, J., Xia, K., Gong, J., and Jiang, S. (2007) Complete mito- Biol. 57, 758–771. chondrial DNA sequence, gene organization and genetic Summerer, M., Hanel, R., and Sturmbauer, C. (2001) Mitochon- variation of control regions in Parargyrops edita. Fisheries drial phylogeny and biogeographic affinities of sea breams of science 73, 1042–1049. the genus Diplodus (Sparidae). J. Fish Biol. 59, 1638– Yamanoue, Y., Miya, M., Matsuura, K., Yagishita, N., Mabuchi, 1652. K., Sakai, H., Katoh, M., and Nishida, M. (2007) Phyloge- Swofford, D. L. (2002) PAUP*: Phylogenetic analysis using par- netic position of tetraodontiform fishes within the higher simony. (* and other methods), Ver. 4.0b10. Sinauer Associ- : Bayesian inferences based on 44 whole mitochon- ates, Sunderland, MA. drial genome sequences. Mol. Phylogenet. Evol. 45, 89– Taniguchi, N., Fujita, M., and Akazaki, M. (1986) Genetic diver- 101. gence and systematics in Sparid fish from Japan. In: Proc. Yamanoue, Y., Miya, M., Inoue, J. G., Matsuura, K., and 2nd. Int’l Conf. Indo-Pacific Fishes, (eds.: T. Uyeno, R. Arai, Nishida, M. (2006) The mitochondrial genome of spotted T. Taniuchi and K. Matsuura), pp. 849–858, Ichthyological green pufferfish Tetraodon nigroviridis (Teleostei: Tetra- Society of Japan, Tokyo. odontiformes) and divergence time estimation among model Thompson, J. D., Gibson, T. G., Plewniak, F., Jeanmougin, F., organisms in fishes. Genes Genet. Syst. 81, 29–39. and Higgins, D. G. (1997) The clustalX windows interface: Yang, Z. H. (2007) PAML: aprogram package for phylogenetic flexible strategies for multiple sequence alignment aided by analysis by maximum likelihood. Comput. Appl. Biosci. 13, quality analysis tools. Nucletic Acids Res. 24, 4876–4882. 555–556. Timm, J., Figiel, M., and Kochzius, M. (2008) Contrasting pat- Zwickl, D. J., and Hillis, D. M. (2002) Increased taxon sampling terns in species boundaries and evolution of anemonefishes greatly reduces phylogenetic error. Syst. Biol. 51, 588–598.

Top View