JOURNAL OF EXPERIMENTAL ZOOLOGY (MOL DEV EVOL) 308B:642–654 (2007) Phylogenetic Relationships of Danio Within the Order Cypriniformes: A Framework for Comparative and Evolutionary Studies of a Model Species RICHARD L. MAYDEN1, KEVIN L. TANG1, KEVIN W. CONWAY1, JÖRG FREYHOF2, SARAH CHAMBERLAIN1, MIRANDA HASKINS1, LEAH SCHNEIDER1, MITCHELL SUDKAMP1, ROBERT M. WOOD1, MARY AGNEW1, ANGELO BUFALINO1, ZOHRAH SULAIMAN3, MASAKI MIYA4, KENJI SAITOH5, AND SHUNPING HE6 1 Department of Biology, Saint Louis University, St. Louis, Missouri 2 Leibniz Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany 3 Department of Biology, Faculty of Science, Universiti Brunei Darussalam, Tungku BE, Brunei Darussalam 4 Department of Zoology, Natural History Museum & Institute, Chiba, Chuo-ku, Chiba, Japan 5 Tohoku National Fisheries Research Institute, Fisheries Research Agency, Shiogama-shi, Miyagi, Japan 6 Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, China ABSTRACT The evolutionary relationships of species of Danio and the monophyly and phylogenetic placement of the genus within the family Cyprinidae and subfamily Rasborinae provide fundamentally important phyloinformatics necessary for direct evaluations of an array of pertinent questions in modern comparative biology. Although the genus Danio is not one of the most diverse within the family, Danio rerio is one of the most important model species in biology. Many investigations have used this species or presumed close relatives to address specific questions that have lasting impact on the hypothesis and theory of development in vertebrates. Largely lacking from this approach has been a holistic picture of the exact phylogenetic or evolutionary relationships of this species and its close relatives. One thing that has been learned over the previous century is that many organismal attributes (e.g., developmental pathways, ecologies, behaviors, speciation) are historically constrained and their origins and functions are best explained via a phylogenetic approach. Herein, we provide a molecular evaluation of the phylogenetic placement of the model species Danio rerio within the genus Danio and among hypothesized closely related species and genera. Our analysis is derived from data using two nuclear genes (RAG1, rhodopsin) and five mitochondrial genes (ND4, ND4L, ND5, COI, cyt b) evaluated using parsimony, maximum likelihood, and Bayesian analyses. The family Cyprinidae is resolved as monophyletic but the subfamily Rasborinae (priority over Danioinae) is an unnatural assemblage. Danio is identified as a monophyletic group sister to a clade inclusive of the genera Chela, Microrasbora, Devario, and Inlecypris, not Devario nor Esomus as hypothesized in previous studies. Danio rerio is sister Grant sponsor: USA National Science Foundation Assembling the Tree of Life Program; Grant number: EF-0431326; Grant sponsor: USA NSF Research Experience for Undergraduate (REU) supplement in 2005. Correspondence to: Dr. Richard L. Mayden, Department of Biology, Saint Louis University, 3507 Laclede Avenue, St. Louis, MO 63103. E-mail: cypriniformes@gmail.com r 2007 WILEY-LISS, INC. Received 12 March 2007; Revised 8 April 2007; Accepted 14 April 2007 Published online 6 June 2007 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/jez.b.21175 PHYLOGENETICS OF DANIO 643 to D. kyathit among the species of Danio evaluated in this analysis. Microrasbora and Rasbora are non-monophyletic assemblages; however, Boraras is monophyletic. J. Exp. Zool. (Mol. Dev. Evol.) 308B:642– 654, 2007. r 2007 Wiley-Liss, Inc. How to cite this article: Mayden RL, Tang KL, Conway KW, Freyhof J, Chamberlain S, Haskins M, Schneider L, Sudkamp M, Wood RM, Agnew M, Bufalino A, Sulaiman Z, Miya M, Saitoh K, He S. 2007. Phylogenetic relationships of Danio within the order Cypriniformes: a framework for comparative and evolutionary studies of a model species. J. Exp. Zool. (Mol. Dev. Evol.) 308B:642–654. Danio rerio, commonly referred to as the zebrafish or zebra danio, is a small cyprinid fish native to the streams of South-eastern Himalayan region (Talwar and Jhingran, ’91; Menon, ’99). Owing to its small body size, short life span, and its ability to reproduce in captivity, the zebrafish has become one of the most important model organisms for vertebrate developmental biology and genetics (Detrich et al., ’99). Despite its importance as a model organism, the phylogenetic relationships of the zebrafish, and the composition and relationships of the entire genus Danio, remain unclear. This is a major gap in our knowledge because the phylogenetic placement of any model organism is of great importance if it is to be used in comparative biological studies. Danio is a member of the order Cypriniformes, a large group of freshwater fishes distributed throughout North America, Africa, and Eurasia. Cypriniforms are in turn placed within the series Otophysi (a subgroup of the larger superorder Ostariophysi), a clade of freshwater fishes that also includes the tetras (order Characiformes), South American knifefishes (order Gymnotiformes), and catfishes (order Siluriformes). Members of the Otophysi are characterized by the presence of the Weberian apparatus (Rosen and Greenwood, ’70; Fink and Fink, ’81). This structure comprises modified perilymphatic and endolymphatic spaces of the inner ear, which are connected to a modified swimbladder through a series of modified anterior vertebral elements, termed the Weberian ossicles (Chranilov, ’27). The Otophysi accounts for about 30% of the known fish species and 64% of all freshwater species (Nelson, 2006).The order Cypriniformes comprises six families: Cyprinidae, Catostomidae, Gyrinocheilidae, Psilorhynchidae, Cobitidae, and Balitoridae (Nelson, 2006). Danio is placed in the family Cyprinidae. Cyprinidae is the largest clade of freshwater fishes and the second largest vertebrate family with approximately 2,420 currently recognized species (Nelson, 2006). Although cyprinid monophyly has been substantiated (Siebert, ’87; Cavender and Coburn, ’92) the intrarelationships of the family remain uncertain (Howes, ’91; Nelson, 2006). The family is believed to be composed of between seven and ten nominal subfamily groupings (Howes, ’91; Nelson, 2006), one of which, subfamily Rasborinae, includes Danio. The name Rasborinae appears to have been proposed first by Weber and de Beaufort (’16) to include several cyprinid genera (including Danio) composed of small-sized to medium-sized species, and species lacking features thought to be characteristic of other groups. Other authors have chosen to use the name ‘Danioinae’ in the same meaning (e.g., Banarescu, ’68; Rainboth, ’91). As the name Rasborinae has priority over Danioinae, we chose to use the former, following Gosline (’75). The intrarelationships and interrelationships of the Rasborinae have yet to be thoroughly analyzed and several researchers (Howes, ’91; Cavender and Coburn, ’92; Nelson, 2006) have suggested that the Rasborinae may represent a non-monophyletic grouping. As with the poorly understood relationships within the Cyprinidae, relationships of Rasborinae to other subfamilies and the relationships of purported members of this subfamily, including Danio and D. rerio, remain unresolved. The genus Danio was originally described as a subgenus of Cyprinus by Hamilton (1822). At present, the genus consists of 66 nominal species, of which about 45 are considered valid (Fang, 2003). Eight junior synonyms of Danio are currently recognized (Eschmeyer et al., ’98). The name Brachydanio was later generally adopted as the valid genus name for the smaller, slenderbodied danios (including D. rerio), particularly in the aquarium literature (Axelrod, ’85; Riehl and Baensch, ’91), with the name Danio reserved only for the larger, deeper-bodied Danio species. Chu (’81) considered Brachydanio as a synonym of Danio, based on overlapping meristic (fin ray and scale) counts between species of the two genera, and placed the small, slender-bodied species of J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 644 R.L. MAYDEN ET AL. Brachydanio back into Danio. Despite this change, the generic name Brachydanio was still used often in combination with the specific name rerio (e.g., Westerfield, ’89). However, the name Danio rerio is currently accepted as the correct name for the zebrafish (Eschmeyer et al., ’98). Early studies of the phylogenetic relationships of Danio appeared rapidly when D. rerio was identified as a model organism species; however, these studies did not include a very thorough sampling of species that needed to be evaluated for a more comprehensive test of the monophyly of the genus, likely relationships of sister taxa, and the monophyly of the subfamily Rasborinae to which these fishes belong. Meyer et al. (’93) presented the first phylogenetic analysis inclusive of Danio rerio using 12S and 16S rDNA. That study, which only included a few species of Danio and related genera, was expanded by Meyer et al. (’95), which examined more genes for more species. Though the number of taxa and the number of genes had doubled between studies, the outcome was the same: D. rerio (1D. ‘frankei’, a synonym of D. rerio) was identified as the sister group to all other small-bodied species, with this ‘slenderbodied’ clade collectively forming the sister group of the ‘deep-bodied’ danios. The same result has been recovered frequently using different molecular and morphological data sets (Fig. 1A). From Meyer et al. (’93) to Sanger and McCune (2002), all previous phylogenetic studies of the zebrafish demonstrated monophyly for Danio. It was not until Fang (2003) included a wider assortment of small Southeast Asian ‘rasborin’ cyprinids, including species that had never been analyzed with Danio before, that different phylogenetic relationships were inferred. In Fang’s (2003) analysis, Esomus Swainson, 1839 was found to be the sister group of the ‘slender-bodied’ species of Danio, rendering Danio as an unnatural, polyphyletic assemblage (Fig. 1B). Though Fang (2003) did not resolve the relationships of the smaller species of Danio (including D. rerio), she suggested for the first time that Danio was not monophyletic. Fang (2003) reserved the name Danio, for the slenderbodied clade that included D. dangila (the type species of Danio), and employed the name Devario for the clade composed of the deeper-bodied species of Danio. Taxon and character sampling are critical components for the development of robust hypotheses of evolutionary relationships (Hillis, ’98). Limited taxon sampling and character evaluation can both result in an incomplete interpretation of J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b Fig. 1. Previous hypotheses of zebrafish relationships. (A) The phylogenetic position of Danio rerio and its close relatives based on 16S and 12S mitochondrial ribosomal genes (modified from Meyer et al., ’95; Fig. 2). Similar topologies have been obtained by Meyer et al. (’93), Zardoya et al. (’96), Sanger and McCune (2002), Quigley et al. (2004, 2005); (B) Cladogram of rasborin taxa representing a strict consensus tree of two equally parsimonious trees of 97 steps (modified from Fang, 2003; Fig. 1). ‘Slender-bodied’ (gray) and ‘deepbodied’ (black) clades are highlighted in each. sister group relationships, and can consequently mislead the scientific community as to the evolutionary relationships of species. This can have a dramatic impact on the construction of predictive hypotheses for these species in diverse areas of biology when researchers rely on a phylogenetic hypothesis to propose developmental, evolutionary, ecological, physiological, and other hypotheses of genetically mediated attributes. PHYLOGENETICS OF DANIO Using molecular sequence data from a variety of nuclear and mitochondrial genes and phylogenetic analyses, we provide an evaluation of the phylogenetic placement of the zebrafish within the genus Danio, evaluate the general relationships of species within Danio and many closely related genera and species. We also provide a much broader survey of taxa that have been traditionally allied to the subfamily Rasborinae to test the monophyly of this group. Using this more comprehensive phylogenetic perspective for D. rerio, it is hoped that the evolutionary hypothesis will provide a variety of scientific communities interested in this species and its relatives with a more powerful predictive framework with which to develop and test hypotheses regarding the features discovered in this model species. MATERIALS AND METHODS Taxon sampling Representative taxa from four of the 11 subfamilies of Cyprinidae (Cyprininae, Gobioninae, Leuciscinae, and Rasborinae) belonging to 28 genera were included in this analysis. In an effort to represent the diversity of the Cyprinidae, taxon sampling focused on including members of genera that previously have been hypothesized as the sister group of Danio (Sanger and McCune, 2002; Fang, 2003), in addition to a representative sampling of rasborin taxa. Outgroup taxa were drawn from the families Balitoridae, Catostomidae, and Cobitidae, as these families, along with the Cyprinidae, form the bulk of cypriniform diversity. A total of 56 cyprinid species and six non-cyprinid cypriniform species were examined for this study; a full list of species examined is given in Table 1. DNA amplification and sequencing Genomic extractions were taken from muscle tissue or fin clips either frozen at 801C or preserved in ethanol (495% concentration), using DNeasy tissue extraction kits (Qiagen, Valencia, CA, USA). Target regions of mitochondrial and nuclear DNA were amplified using polymerase chain reaction (PCR) and the primers listed in Table 2. Target loci included cytochrome b (1,141 bp), a 712-bp fragment of cytochrome c oxidase subunit I, nicotinamide adenine dinucleotide (reduced form) (NADH) dehydrogenase subunits 4L and 4 (1,748 bp), a 894-bp fragment of 645 NADH dehydrogenase subunit 5, a 1,518-bp fragment of exon III of recombinase activating gene 1, and a 905-bp fragment of rod opsin (protein component of the rhodopsin photoreceptor). Amplification used the following thermal cycling profiles: 941C denaturing (30–60 sec), 45–551C annealing (30–60 sec), and 721C extension (2 min 30 sec), for 30–40 cycles; an initial heating step at 941C for 30–60 sec preceded cycling and some profiles included a final extension step at 721C for 2–5 min after cycling was complete. Amplified products were either directly purified using QIAquick PCR purification kits (Qiagen) or loaded onto agarose gels and electrophoresed, followed by excision of the target DNA from the gel and purification using QIAquick gel extraction kits (Qiagen). Purified PCR samples were then sent to a commercial sequencing facility at Macrogen Inc. for cycle sequencing on an Applied Biosystems 3730xl automated sequencer, using the primers indicated in Table 2. Both strands were sequenced for all target gene regions. A consensus light strand sequence for each gene region for each taxon was assembled from the two complementary sequences; these consensus sequences have been deposited with GenBank (see Table 2 for accession numbers). An initial alignment was made by eye, comparing our sequence data with the previously published sequences of Danio rerio; due to the protein-coding nature of these loci, alignment was relatively straightforward. Some loci for some taxa could not be successfully sequenced, these base positions were coded as missing for those taxa. Additional sequence data for the outgroup taxa as well as the non-rasborin cyprinids were obtained from GenBank (Table 2). Phylogenetic analyses Three separate types of searching methods were performed on the compiled data matrix: parsimony, maximum likelihood, and Bayesian. Parsimony searches were carried out using PAUP 4.0b10 (Swofford, 2002) in conjunction with PAUPRat (Sikes and Lewis, 2001), which implements a parsimony ratchet search (Nixon, ’99) in PAUP. PAUPRat was used to perform 1,000 parsimony ratchet replicate searches in PAUP using 5, 10, 15, 20, and 25% perturbation, with 200 replicates for each percentage value. Gaps were treated as a fifth character state. The best trees recovered via the ratchet searches were then used as the starting trees for a heuristic parsimony search in PAUP J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 646 R.L. MAYDEN ET AL. TABLE 1. Species of Cypriniformes sequenced for nuclear and mitochondrial genes, with GenBank accession numbers Taxon Order Cypriniformes Family Balitoridae Crossostoma lacustre Lefua echigonia Family Catostomidae Carpiodes carpio Myxocyprinus asiaticus Family Cobitidae Cobitis sinensis Cobitis striata Family Cyprinidae Subfamily Cyprininae Carassius auratus Carassius carassius Cyprinus carpio Sawbwa resplendens Subfamily Gobioninae Sarcocheilichthys variegatus Subfamily Leuciscinae Campostoma anomalum Hemitremia flammea Luxilus chrysocephalus Nocomis biguttatus Notemigonus crysoleucas Notropis atherinoides Opsopoeodus emiliae Phoxinus erythrogaster Richardsonius balteatus Subfamily Rasborinae Boraras brigittae Boraras maculatus Boraras merah Boraras sp. cf. micros Boraras urophthalmoides Chela cachius Chela dadiburjori Danio choprai Danio erythromicron Danio feegradei Danio kyathit Danio nigrofasciatus Danio sp. ‘‘hikari’’ Danio sp. ‘‘panther’’ Danio rerio Danio roseus Danionella sp. Devario devario Esomus sp. cf. ahli Inlecypris auropurpurea Microrasbora kubotai Microrasbora rubescens Opsaridium sp. Opsariichthys uncirostris Rasbora argyrotaenia Rasbora brittani Rasbora caudimaculata Rasbora cephalotaenia Rasbora daniconius COI Cyt b ND4 ND5 RAG1 Rhod NC001727 NC004696 NC001727 NC004696 NC001727 NC004696 NC001727 NC004696 N/A EF458305 N/A N/A NC005257 NC006401 NC005257 NC006401 NC005257 NC006401 NC005257 NC006401 N/A N/A N/A N/A NC007229 NC004695 NC007229 NC004695 NC007229 NC004695 NC007229 NC004695 N/A EF458303 N/A N/A NC006580 NC006291 NC001606 EF452895 NC006580 NC006291 NC001606 N/A NC006580 NC006291 NC001606 EF452824 NC006580 NC006291 NC001606 N/A N/A N/A EF458304 N/A N/A N/A N/A N/A NC004694 NC004694 NC004694 NC004694 N/A N/A EF452850 EF452851 EF452852 EF452853 EF452854 EF452855 EF452856 EF452857 EF452858 AF452079 AY281054 AF117167 AY486057 U01318 AY281062 U17270 AY281055 AY096011 EF452784 EF452785 EF452786 EF452787 EF452788 EF452789 EF452790 EF452791 N/A EF452751 EF452752 EF452753 EF452754 EF452755 EF452756 EF452757 EF452758 EF452759 EF452827 EF452828 EF452829 EF452830 EF452831 EF452832 EF452833 EF452834 EF452835 EF452898 EF452899 EF452900 EF452901 EF452902 EF452903 EF452904 EF452905 EF452906 N/A EF452859 EF452884 EF452885 EF452886 EF452891 EF452892 EF452879 EF452867 EF452861 EF452862 EF452863 EF452860 EF452864 NC002333 EF452865 EF452887 EF452866 EF452888 EF452889 EF452868 EF452890 EF452893 EF452894 EF452880 EF452869 EF452870 EF452881 EF452872 N/A N/A N/A N/A N/A EF452745 EF452746 EF452740 EF452737 EF452732 EF452733 N/A EF452731 EF452734 NC002333 EF452735 EF452741 EF452736 EF452742 EF452743 EF452738 EF452744 EF452747 EF452748 N/A N/A N/A N/A N/A EF452792 EF452793 EF452815 EF452816 EF452817 EF452821 N/A EF452810 EF452800 EF452795 EF452796 N/A EF452794 EF452797 NC002333 EF452798 EF452818 EF452799 N/A EF452819 EF452801 EF452820 EF452822 EF452823 EF452811 EF452802 EF452803 EF452812 EF452805 EF452760 EF452761 EF452778 EF452779 EF452780 N/A N/A N/A EF452768 EF452763 EF452764 N/A EF452762 EF452765 NC002333 EF452766 EF452781 EF452767 EF452782 N/A EF452769 EF452783 N/A N/A EF452776 EF452770 EF452771 N/A EF452773 N/A N/A EF452838 EF452839 EF452840 EF452845 N/A N/A N/A N/A N/A N/A N/A N/A NM131389 N/A EF452841 N/A EF452842 EF452843 N/A EF452844 EF452846 EF452847 EF452836 N/A N/A N/A N/A N/A N/A EF452909 EF452910 EF452911 EF452914 EF452915 N/A N/A N/A N/A N/A N/A N/A NM131084 N/A N/A N/A N/A EF452912 N/A EF452913 N/A EF452916 EF452907 N/A N/A N/A N/A J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 647 PHYLOGENETICS OF DANIO TABLE 1. Continued Taxon COI Rasbora dorsiocellata Rasbora kottelati Rasbora rubrodorsalis Rasbora sp. cf. bankanensis Rasbora sumatrana Rasbora trilineata Rasbora vaterifloris Rasbora vulcanus Sundadanio axelrodi Trigonostigma espei Trigonostigma hengeli Zacco platypus Zacco temminckii EF452873 N/A EF452874 EF452871 EF452882 EF452883 EF452876 EF452875 N/A EF452877 EF452878 EF452896 EF452897 Cyt b ND4 N/A N/A N/A N/A N/A N/A N/A N/A EF452739 N/A N/A EF452749 EF452750 EF452806 N/A EF452807 EF452804 EF452813 EF452814 N/A N/A EF452808 N/A EF452809 EF452825 EF452826 ND5 EF452774 EF452775 N/A EF452772 EF452777 N/A N/A N/A N/A N/A N/A N/A N/A RAG1 Rhod N/A N/A N/A N/A EF452837 N/A N/A N/A N/A N/A N/A EF452848 EF452849 N/A N/A N/A N/A EF452908 N/A N/A N/A N/A N/A N/A EF452917 EF452918 Abbreviations: COI, cytochrome c oxidase subunit I; Cyt b, cytochrome b; ND4, NADH dehydrogenase subunit 4; ND5, NADH dehydrogenase subunit 5; RAG1, recombination activating gene 1; Rhod 5 opsin. TABLE 2. List of primers and primer sequences used in reconstructing relationships of Cypriniformes species Primer Cytochrome b LA-danio HA-danio Cytochrome c oxidase subunit I LCO1490 HCO2198 NADH dehydrogenase subunit 4 & 5 L10444 L11427-ND4-C L12328-Leu-C H11618a H12296-Leu-C H13393-ND5-C Opsin Rh193 Rh1073r RAG1 2533F 4090R Sequence (50 –30 ) GACTYGAARAACCACYGTTG CTCCGATCTTCGGATTACAAG Source This study This study GGTCAACAAATCATAAAGATATTGG TAAACTTCAGGGTGACCAAAAAATCA Folmer et al. (’94) Folmer et al. (’94) AAGACCTCTGATTTCGGCTCA CCWAAGGCSCATGTWGARGC AACTCTTGGTGCAAMTCCAAG TGRCTKACSGAKGAGTAGGC CAAGAGTTTTTGGTTCCTAAG CCTATTTTKCGGATGTCTTGYTC This study Miya et al. (2006) Miya et al. (2006) This study Miya et al. (2006) Miya et al. (2006) CNTATGAATAYCCTCAGTACTACC CCRCAGCACARCGTGGTGATCATG Chen et al. (2003) Chen et al. (2003) CTGAGCTGCAGTCAGTACCATAAGATGT CTGAGTCCTTGTGAGCTTCCATRAAYTT Lopez et al. (2004) Lopez et al. (2004) with TBR branch swapping. The most-parsimonious trees recovered were evaluated using summary values reported by PAUP (e.g., tree length, consistency index). Branch support was evaluated by calculating decay index values (Bremer, ’88) for each clade using TreeRot v. 2 (Sorenson, ’99). Bootstrap values were calculated in PAUP using a simple heuristic search and 10,000 bootstrap replicates (1,000 heuristic searches, each with ten random addition replicates). Maximum likelihood searches were conducted with GARLIv0.941 (Zwickl, unpublished data; http://www.bio.utexas.edu/grad/zwickl/web/garli. html). Twenty individual runs were performed using the default search settings (5,000,000 generations) and termination criteria (genthreshfortopoterm 5 20,000; scorethreshforterm 5 0.05) with random starting topologies. Hierarchical likelihood ratio tests (hLRTs) performed with MrModelTest v.2.2 (Nylander, 2004) and PAUP found that the best-fit model for a majority of loci was GTR1I1G see Results). As this is the default model implemented by GARLI the GTR1I1G model was used for the likelihood searches and gaps were treated as missing. Resulting topologies were optimized in PAUP to calculate log likelihood scores, using the rates matrix, base composition frequencies, proportion of invariable sites, and gamma distribuJ. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 648 R.L. MAYDEN ET AL. tion estimated by the individual GARLI searches. The lnL scores estimated by GARLI versus PAUP differed by less than 0.00001 when they differed at all. The topology with the best likelihood score was retained. Bootstrap values were calculated in GARLI using 100 bootstrap replicates and the termination criteria were relaxed slightly (genthreshfortopoterm 5 10,000). Bayesian analyses were conducted with the MPI version (Altekar et al., 2004) of MrBayes v.3.1.2 (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003) on a parallel computing cluster with 12 Apple Xserve dual processor nodes running UNIX. Cobitis striata was designated as the single outgroup taxon. Before the analysis, the sequence data were partitioned by codon position and loci, resulting in 15 separate data partitions (one for each codon position for each of the five loci). MrModelTest v.2.2 (Nylander, 2004) and PAUP were used to perform hLRTs on each partition to determine the best-fit model of nucleotide substitution. These models were then applied to the appropriate partitions during the subsequent MrBayes analyses. Two independent Bayesian searches were conducted using the same search parameters: 10,000,000 generations and 24 chains (with one cold chain and 23 heated chains), with sampling every 10,000 generations. The distribution of log likelihood scores was examined to determine burn-in time for each of the analyses. Trees recovered after stationarity had been achieved were retained. Branch support for each clade was based on clade credibility values, indicated by the frequency of occurrence of each clade among the trees retained after the initial burn-in topologies were discarded. The trees retained from the two independent analyses were combined into a single pool for the purposes of calculating clade credibility values, which was accomplished by constructing a 50% majority-rule consensus of these trees in PAUP. RESULTS A total of 6,921 aligned base positions were obtained for 62 taxa, 56 cyprinid species and six species representing other cypriniform families. Our searches utilizing different optimality criteria converged on very similar hypotheses of relationships. The parsimony analysis recovered 18 most-parsimonious trees with a total length of 21,861 steps (CI 5 0.270, RI 5 0.431, RC 5 0.116); the strict consensus of these trees is shown in J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b Figure 2. The family Cyprinidae is recovered as monophyletic, as are the subfamilies Cyprininae and Leuciscinae. The subfamily Rasborinae, however, is rendered polyphyletic by the placement of Opsariichthys and Zacco in a clade with species of the subfamilies Gobioninae and Leuciscinae. The genus Danio is found to be monophyletic within a clade containing the remaining Rasborinae representatives, with Danio rerio occupying an apical position within the group and sister to D. kyathit. Danio erythromicron is clearly placed within Danio and not Microrasbora, as has been previously hypothesized for this species (Kottelat and Witte, ’99). The sister taxon of Danio is a clade composed of Chela, Inlecypris, Microrasbora, and most notably Devario, a previously hypothesized sister group of Danio (sensu Sanger and McCune, 2002). The genus Esomus is recovered as the basal sister group to the larger clade containing Danio, and is not the sister group of the genus as previously hypothesized by Fang (2003). A monophyletic group containing Sundadanio sister to Danionella plus Opsaridium forms the sister group to the remaining members of this clade, excluding Esomus. This large clade containing Danio and relatives (Chela, Danionella, Devario, Esomus, Inlecypris, Microrasbora, Opsaridium, and Sundadanio) is recovered as the sister group to a clade that includes Rasbora, Rasboroides, Boraras, and Trigonostigma. The genus Rasbora is polyphyletic relative to Rasboroides vaterifloris, a monophyletic Boraras, and species of Trigonostigma, with some Rasbora more closely related to members of the other three genera than to other Rasbora. The largely unresolved relationships among some species of Rasbora in the RasboraRasboroides-Trigonostigma clade is a result of limited sequence variability in these species for the genes examined, not conflicting relationships. The optimal topology recovered by the maximum likelihood searches has a lnL score of 92,351.0953 (Fig. 3). The results of the maximum likelihood analysis are largely congruent with the results of the parsimony analysis. The genus Danio is found to be monophyletic and a member of a clade which contains most of the putative members of the Rasborinae, which is again not monophyletic because Opsariichthys and Zacco are recovered in a clade with the gobionins and leuciscins. Danio rerio is again recovered in a derived position in the genus, with D. kyathit as its sister species. A clade comprised Chela, Devario, Inlecypris, and Microrasbora appears again as the sister taxon of Danio. In these analyses the genus PHYLOGENETICS OF DANIO 649 Fig. 2. Strict consensus of 18 equally most-parsimonious trees resulting from a parsimony analysis with parsimony ratcheting; TL 5 21,861 steps, CI 5 0.270, RI 5 0.431, RC 5 0.116. Decay index support (above) and bootstrap values for 10,000 replicates (below) are reported at each node. Rasbora is resolved as polyphyletic relative to monophyletic Boraras and Trigonostigma. In this analysis, however, Esomus forms the basal sister group to the Rasbora-Rasboroides-Boraras-Trigo- nostigma clade, not the clade containing Danio and relatives as described above (Fig. 2). Some resolution of species relationships within Rasbora is elucidated in this analysis, but supJ. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 650 R.L. MAYDEN ET AL. Fig. 3. Tree topology recovered with the best log likelihood score of 20 independent maximum likelihood searches; lnL 5 92,351.0953. Bootstrap values for 100 replicates are reported at each node (bootstrap values were less than 50% where no value is given). the best-fit model for a majority of loci was GTR1I 1G. The following data partitions had GTR1G: cyt b third positions, RAG1 third positions, and rhodopsin third positions. HKY1I1G was found as the best-fit model for second positions of RAG1; F811I1G was the best-fit model for first positions in rhodopsin; and F811G was the best fit for rhodopsin second positions. In both Bayesian searches, it appeared that stationarity had been reached after approximately 50,000 generations. The first 100,000 generations (11 trees) were discarded to ensure all burn-in trees were excluded from the results, leaving 990 of the original 1,001 saved trees (from 10,000,000 generations) from each of the independent searches. This yielded a total of 1,980 trees from a combined 20,000,000 generations that were included in the 50% majority-rule consensus tree (Fig. 4). The relationships found in the Bayesian tree mostly correspond with those seen in the parsimony and likelihood trees. Danio rerio and D. kyathit are recovered as sister species in a monophyletic Danio, with the clade of Chela, Devario, Inlecypris, and Microrasbora as the sister group of Danio. One of the major differences between the Bayesian tree and the other two trees is that Sundadanio axelrodi is recovered as the sister group of the subfamily Cyprininae, not as part of the ‘‘rasborin’’ clade (putative rasborins minus Opsariichthys, and Zacco), as seen in the other two searches. This highlights one of the only major areas of disagreement between the three different search methods: Sundadanio, along with Danionella, Esomus, and Opsaridium, are found in different positions in all three trees. This instability is a result of insufficient sequence data for appropriate genes for these four taxa, as is also observed in species of Rasbora. One other notable difference in the Bayesian tree is that the sister group of the ‘‘rasborin’’ clade is the Cyprininae (and Sundadanio axelrodi) not the leuciscin1 gobionin1Opsariichthys1Zacco clade seen in the parsimony and likelihood trees. DISCUSSION port for these relationships is poor and branch lengths are very short, again illustrating the general lack of variation among these taxa for the genes examined. As in the parsimony analyses, the genus Microrasbora is not a monophyletic group and does not include Danio erythromicron. For the Bayesian analyses, the hLRTs performed with MrModelTest and PAUP found that J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b Although the order Cypriniformes contains the world’s most diverse group of freshwater fishes, including important species for biological and evolutionary studies as well as many species of great cultural and economic importance, the diversity and relationships of these species are far from being well understood. To some degree, the paucity of information regarding these species PHYLOGENETICS OF DANIO 651 Fig. 4. The 50% majority-rule consensus of 1,980 trees recovered by Bayesian inference, with clade credibility values reported at each node. is a result of the overwhelming diversity within the group, as scientists traditionally have had difficulty identifying these species. The Cypriniformes Tree of Life Initiative (www.cypriniformes.org) is a worldwide consortium of researchers that have recently focused, in a collaborative partnerJ. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 652 R.L. MAYDEN ET AL. ship, to jointly resolve the phylogenetic relationships of these important fishes and better understand their diversity, biogeography, anatomy, ecology, and evolution. This collaborative effort has also involved an important opportunity for systematic ichthyologists to interact with scientists involved in primary research on other aspects of the biology of Danio rerio, facilitating a better understanding of the many different biological properties of this species and its relatives. With the current and impending threats to most of our planet’s biodiversity and their ecosystems, more focused and collaborative international efforts involving scientists from many different countries are needed to study a group as diverse as the Cypriniformes. Although several researchers have worked toward improving our understanding of the evolutionary relationships of Danio rerio, only through this collaborative international effort have we been able to obtain and evaluate a larger number of species of Cypriniformes, providing better insight into the phylogenetic position and biology of this important model species. Most developmental and genetic studies involving Danio rerio make comparisons with sticklebacks or fugu, or even human or mouse. These comparisons are important for understanding evolution, but such comparisons can also aid our understanding of evolutionary developmental biology if they are cast in a phylogenetic framework. A synthetic approach to the evolutionary origins of diversity and/or control mechanisms of morphology can be the focus of research in purely theoretical fields, as well as in more applied fields like medicine, agriculture, and aquaculture. Goldschmidt (’40, pp 205–206) noted that: ‘‘The change from species to species is not a change involving more and more additional atomistic changes, but a complete change of the primary pattern or reaction system into a new one, which afterward may again produce intraspecific variation by micromutation.’’ Goldschmidt’s hypothesis was universally rejected and widely ridiculed within the biological community of his day, which favored the Neo-Darwinian explanations of R.A. Fisher, J.B.S. Haldane, and Sewall Wright. The Neo-Darwinian paradigm, however, has yet to provide a convincing mechanism for the origin of species that has withstood testing within a phylogenetic framework, largely because adherents of this paradigm have lacked the basic data or testable hypotheses which would allow one to study causal relationships between descent, genetics, and morphology. Model organisms provide clear opportunities J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b for the scientific community to investigate many evolutionary and developmental questions. However, a well-corroborated phylogeny of the model organism and its closest relatives is required before the evaluation of pertinent questions can begin. In all of our analyses, the genus Danio is recovered as a monophyletic group and D. rerio is sister to D. kyathit. With more sampling of species of Danio in the future, this sister group relationship may change and species not yet sampled may be identified as the closest relative. However, predictive comparative analyses of different aspects of the biology, anatomy, physiology, genetics, and development of D. rerio can begin to focus on those species identified as close relatives by this study, which include D. kyathit, D. nigrofasciatus, D. sp. ‘‘panther,’’ D. roseus, D. sp. ‘‘hikari,’’ D. choprai, and D. erythromicron (Quigley et al., 2004, 2005). Our phylogenetic analyses indicate that comparisons of Danio with species of Devario and/or Esomus may be misleading because they are not as closely related as once thought. Danio erythromicron, a species originally described in Microrasbora, is found to be more closely related to D. rerio than the putative species of the non-monophyletic Microrasbora resolved herein. Although the resulting placement of this species in Danio may be somewhat surprising to some, this species was recently referred to Danio by Kottelat and Witte (’99). These authors reassigned D. erythromicron on the basis of a suite of morphological characters, but also indicated that this change should be further verified with additional data; herein their conclusion is strongly supported by our molecular sequence data and analyses. The non-monophyly of the subfamily Rasborinae is not too surprising as this grouping has traditionally been used as a ‘‘catch-all’’ for taxa not easily assignable to other subfamilies (Howes, ’91). Further evaluation of additional cypriniform taxa will provide further tests of the monophyly of this subfamily and other subfamilies. Formal reclassification of species from Rasborinae or other subfamilies should await more comprehensive analyses following from the Cypriniformes Tree of Life initiative or other investigations (Saitoh et al., 2006). Other important findings include the close relationship between Devario and Inlecypris, two clades with similar general coloration and morphology, and the observation that the genus Esomus is not closely related to the ‘‘slender’’ PHYLOGENETICS OF DANIO group of Danio as previously hypothesized. We have demonstrated the non-monophyly of the genus Rasbora for the first time. Although this conclusion is also one that is not too surprising given the tremendous anatomical, morphological, and ecological diversity in this genus, it clearly indicates that Rasbora, a highly speciose group of southeast Asian cyprinids, is in need of significant systematic attention. The monophyly of the genus Boraras as proposed by Conway (2005), based on morphological characters in a phylogenetic analysis, is corroborated with molecular sequence data herein. The monophyly of the genus Trigonostigma is supported in both likelihood and Bayesian analyses, whereas the relationships of species in this genus are unresolved in the parsimony analysis. Regardless of the specifics of relationships within those two genera, the placement of Boraras and Trigonostigma (and Rasboroides, in the parsimony tree) within Rasbora reveals Rasbora as a unnatural assemblage in need of additional phylogenetic study. The collaborative efforts initiated recently with the Cypriniformes Tree of Life (www.cyprinifor mes.org), and between members of the Tree of Life project and molecular geneticists and developmental biologists using D. rerio as a model system, provide far-reaching opportunities to advance many areas of comparative and evolutionary biology. Knowledge of the phylogenetic relationships of D. rerio provide a fundamental framework in which scientists can begin to better understand the evolution of genetic mechanisms central to morphological and developmental changes in evolutionary time, morphological adaptations to different environments, and the nature of mutants, all major areas of inquiry that are of interest to a broad audience of comparative and evolutionary biologists. ACKNOWLEDGMENTS This research would not have been possible without the close collaboration among many scientists around the world taking part in the Cypriniformes Tree of Life; we acknowledge their commitment to this important project and thank them for their vision and creative energies to make this initiative a success and a model project for future large-scale systematic initiatives. We also wish to acknowledge the financial support for this research from the USA National Science Foundation Assembling the Tree of Life Program (EF-0431326) and the USA NSF Research Experi- 653 ence for Undergraduates (REU) supplement. S. Chamberlain, M. Haskins, L. Schneider, and M. Sudkamp were all REU students, whose efforts on the sequencing and alignment of the data that they generated made this study possible. We thank C. Dillman and N. Lang for their assistance in conducting various analyses. Finally, we thank J. Webb and T. Schilling for organizing the SICB symposium wherein these findings were presented and extending the invitation for participation. This research was supported by the USA National Science Foundation Assembling the Tree of Life Program (EF-0431326) and the USA NSF Research Experience for Undergraduates (REU) supplement in 2005. LITERATURE CITED Altekar G, Dwarkadas S, Huelsenbeck JP, Ronquist F. 2004. Parallel Metropolis-coupled Markov chain Monte Carlo for Bayesian phylogenetic inference. Bioinformatics 20: 407–415. Axelrod HR. 1985. Dr. Axelrod’s Atlas of Freshwater Aquarium Fishes. Neptune City, NJ: T.F.H. Publications. Banarescu P. 1968. Remarks on the genus Chela HamiltonBuchanan (Pisces, Cyprinidae) with description of a new subgenus. Ann Mus Civ Stor Nat ‘Giacomo Doria’ 77:53–64. Bremer K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795–803. Cavender TM, Coburn MM. 1992. Phylogenetic relationships of North American Cyprinidae. In: Mayden RL, editor. Systematics, Historical Ecology, and North American Freshwater Fishes. Stanford, CA: Stanford University Press. p 293–328. Chen W-J, Bonillo C, Lecointre G. 2003. Repeatability of clades as a criterion of reliability: a case study for molecular phylogeny of Acanthomorpha (Teleostei) with larger number of taxa. Mol Phylogenet Evol 26:262–288. Chranilov NS. 1927. Beiträge zur Kenntnis des Weber’schen Apparates der Ostariophysi. 1. vergleichend-anatomische Übersicht der Knochenelemente des Weber’schen Apparates bei Cypriniformes. Zool Jb Anat 49:501–597. Chu XL. 1981. A preliminary revision of fishes of the genus Danio from China. Zool Res 2:145–156. Conway KW. 2005. Monophyly of the genus Boraras (Teleostei: Cyprinidae). Ichthyol Explor Freshw 16:249–264. Detrich WH, Westerfield M, Zon L. 1999. The zebrafish, genetics and genomics, Vol. 60. San Diego, CA: Academic Press. Eschmeyer WN, Ferraris CJ, Hoang MD, Long DJ. 1998. Species of fishes. In: Eschmeyer WN, editor. Catalog of fishes. San Francisco, CA: California Academy of Sciences. p 25–1820. Fang F. 2003. Phylogenetic analysis of the Asian cyprinid genus Danio (Teleostei, Cyprinidae). Copeia 2003: 714–728. Fink SV, Fink WL. 1981. Interrelationships of the ostariophysan fishes (Teleostei). Zool J Linn Soc 72:297–353. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. 1994. DNA primers for amplification of mitochondrial cytochrome c J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b 654 R.L. MAYDEN ET AL. oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3:294–299. Goldschmidt RB. 1940. The material basis of evolution. New Haven, CT: Yale University Press. Gosline W. 1975. The cyprinid dermosphenotic and the subfamily Rasborinae. Occas Pap Mus Zool Univ Mich 673: 1–13. Hamilton F. 1822. An account of the fishes found in the River Ganges and its branches. Edinburgh: Archibald Constable and Co. Hillis DM. 1998. Taxonomic sampling, phylogenetic accuracy, and investigator bias. Syst Biol 47:3–8. Howes GJ. 1991. Systematics and biogeography: an overview. In: Winfield IJ, Nelson JS, editors. Cyprinid fishes: systematics, biology and exploitation. London: Chapman and Hall. p 1–33. Huelsenbeck JP, Ronquist F. 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17:754–755. Kottelat M, Witte KE. 1999. Two new species of Microrasbora from Thailand and Myanmar, with two new generic names for small Southeast Asian cyprinid fishes (Teleostei: Cyprinidae). J South Asian Nat Hist 4:49–56. López JA, Chen W-J, Ortı́ G. 2004. Esociform phylogeny. Copeia 2004:449–464. Menon AGK. 1999. Check list—fresh water fishes of India. Rec Zool Surv India Misc Publ Occas Pap 175:1–366. Meyer A, Biermann CH, Ortı́ G. 1993. The phylogenetic position of the zebrafish (Danio rerio), a model system in developmental biology: an invitation to the comparative method. Proc R Soc Lond B 252:231–236. Meyer A, Ritchie PA, Witte KE. 1995. Predicting developmental processes from evolutionary patterns: a molecular phylogeny of the zebrafish (Danio rerio) and its relatives. Philos Trans R Soc Lond B 349:103–111. Miya M, Saitoh K, Wood RM, Nishida M, Mayden RL. 2006. New primers for amplifying and sequencing the mitochondrial ND4/ND5 gene region of the Cypriniformes (Actinopterygii: Ostariophysi). Ichthyol Res 53:75–81. Nelson JS. 2006. Fishes of the World. New York, NY: John Wiley and Sons. Nixon KC. 1999. The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15:407–414. Nylander JAA. 2004. MrModeltest v2.2. Uppsala: Evolutionary Biology Centre, Uppsala University. Quigley IK, Turner JM, Nuckels RN, Manuel JL, Budi E, MacDonald EI, Parichy DM. 2004. Pigment pattern evolution by differential deployment of neural crest and postembryonic melanophore lineages in Danio fishes. Development 131:6053–6069. J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b Quigley IK, Turner JM, Nuckels RN, Manuel JL, Budi E, MacDonald EI, Parichy DM. 2005. Evolutionary diversification of pigment pattern in Danio fishes: differential fms dependence and stripe loss in D. albolineatus. Development 132:89–104. Rainboth WJ. 1991. Cyprinids of South East Asia. In: Winfield IJ, Nelson JS, editors. Cyprinid fishes: systematics, biology and exploitation. London: Chapman and Hall. p 156–202. Riehl R, Baensch HA. 1991. Aquarien Atlas. Band. 1. Melle: Mergus, Verlag für Natur- und Heimtierkunde. Ronquist F, Huelsenbeck JP. 2003. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19:1572–1574. Rosen DE, Greenwood PH. 1970. Origin of the Weberian apparatus and the relationships of the ostariophysan and gonorynchiform fishes. Am Mus Novit 2428:1–25. Saitoh K, Sado T, Mayden RL, Hanzawa N, Nakamura K, Nishida M, Miya M. 2006. Mitogenomic evolution and interrelationships of the Cypriniformes (Actinopterygii: Ostariophysi): The first evidence towards resolution of higher-level relationships of the world’s largest freshwater-fish clade based on 59 whole mitogenome sequences. J Mol Evol 63:826–841. Sanger TJ, McCune AR. 2002. Comparative osteology of the Danio (Cyprinidae: Ostariophysi) axial skeleton with comments on Danio relationships based on molecules and morphology. Zool J Linn Soc 135:529–546. Siebert DH. 1987. Interrelationships among families of the order Cypriniformes (Teleostei). Unpubl. Ph.D. dissertation. New York, NY: City University of New York. Sikes DS, Lewis PO. 2001. PAUPRat: PAUP implementation of the parsimony ratchet. Storrs, CT: Department of Ecology and Evolutionary Biology, University of Connecticut. Sorenson MD. 1999. TreeRot, version 2. Boston, MA: Boston University. Swofford DL. 2002. PAUP. Phylogenetic analysis using parsimony (and Other Methods). Version 4. Sunderland, MA: Sinauer Associates. Talwar PK, Jhingran AG. 1991. Inland fishes of India and adjacent countries, vol 1. Rotterdam: A. A. Balkema. Weber M, de Beaufort LF. 1916. The fishes of the IndoAustralian archipelago. III. Ostariophysi: II Cyprinoidea, Apodes, Synbranchi. Leiden: E. J. Brill. Westerfield M. 1989. The zebrafish book: a guide for the laboratory use of zebrafish (Brachydanio rerio). Eugene, OR: University of Oregon Press. Zardoya R, Abouheif E, Meyer A. 1996. Evolutionary analyses of hedgehog and Hoxd-10 genes in fish species closely related to the zebrafish. Proc Natl Acad Sci USA 93:13036–13041.