Phylogeny and biogeography of the extant species of triplespine fishes (Triacanthidae, Tetraodontiformes) Blackwell Science Ltd FRANCESCO SANTINI & JAMES C. TYLER Accepted: 7 November 2001 Santini, F. & Tyler, J. C. (2002). Phylogeny and biogeography of the extant species of triplespine fishes (Triacanthidae, Tetraodontiformes). — Zoologica Scripta, 31, 321– 330. A new phylogenetic hypothesis for the living species of triplespine fishes of the Indo-Western Pacific family Triacanthidae (Tetraodontiformes, Teleostei) is proposed. A data set of 55 morphological characters (34 osteological and 21 morphometric) was constructed. A cladistic analysis of the osteological data set yielded a single most-parsimonious tree. This cladogram does not support the monophyly of one of the four genera, Tripodichthys, but Bremer values for this analysis are low. The osteological data set was then combined with a data set of 21 morphometric characters that had previously been used to diagnose the four genera. The analysis of the combined data set produced the same phylogenetic hypothesis, but with greater nodal support. The biogeographical distribution of the living species is then interpreted with the use of this new phylogenetic information. Francesco Santini, Department of Zoology, University of Toronto, Ramsay Wright Zoological Laboratories, 25 Harbord Street, Toronto, M5S 3G5, Canada. E-mail: fsantini@zoo.utoronto.ca James C. Tyler, National Museum of Natural History, Smithsonian Institution, Washington DC 20560-0106, USA. E-mail: tyler.jim@nmnh.si.edu Introduction The family Triacanthidae ( Tetraodontiformes, Teleostei) is composed of seven extant species of benthic shallow water fishes inhabiting many open bottom and coral reef regions of the Indo-Western Pacific region, plus six fossil taxa, two from the Middle Eocene of Monte Bolca (Italy), three from the Lower Oligocene of Canton Glarus (Switzerland) and one from the Lower Oligocene of Romania. Triacanthids form a basal clade of Tetraodontiformes, and are of great interest because they do not yet exhibit most of the dramatic losses or simplifications of osteological features that characterize the morphology of the more derived tetraodontiform groups, even though triacanthids do show a few of the reductive trends (i.e. in the dorsal-fin spines posterior to the first spine, as seen in Fig. 1, and the associated pterygial elements) that are also found in the balistoid and tetraodontoid clades of Tetraodontiformes [see Tyler (1980)]. The family Triacanthidae is composed of four Recent genera: Triacanthus Cuvier, with two species, T. biaculeatus (Bloch) and T. nieuhofi Bleeker; the monotypic Trixiphichthys Fraser-Brunner, with T. weberi (Chaudhuri); the monotypic Pseudotriacanthus Fraser-Brunner, with P. strigilifer (Cantor); and Tripodichthys Tyler, with three species, T. blochi (Bleeker), T. angustifrons (Hollard) and T. oxycephalus (Bleeker). Detailed descriptions of the extant species and a key to their identification are available in Tyler (1968), while Fig. 1 Pseudotriacanthus strigilifer exhibits the typical morphological shape of recent Triacanthidae. a detailed description of their osteology can be found in Tyler (1980). Historically, the ‘Familia Triacanthoidei’ of Bleeker (1859) was created to include taxa of both triacanthids and triacanthodids, which Bleeker recognized as subfamilies, the Triacanthiformes and Paratriacanthiformes in, respectively, 1865 and 1866. Hollard (1860) erected the ‘Tribus Triacanthiens’, which is equivalent to the Triacanthoidei of Bleeker. As © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 4, September 2002, pp321–330 321 Phylogeny of Triacanthidae • F. Santini & J. C. Tyler presently recognized, the family Triacanthidae Bleeker (1859) is divided into two subfamilies (Tyler 1968, 1980; Tyler & Winterbottom 1999; Tyler & Santini 2001; Baciu & Tyler 2002; Tyler & Santini 2002): the Middle Eocene Protacanthodinae Tyler, 1968, which includes only the genus Protacanthodes Gill, from the Middle Eocene of Monte Bolca, with two species, P. ombonii (Zigno) and P. nimesensis Tyler & Santini; and the Triacanthinae Tyler, 1968, which includes the monotypic Lower Oligocene genus Cryptobalistes Rath, the three species of the Lower Oligocene genus Acanthopleurus Agassiz, and the four Recent genera with seven extant species. Tyler (1968, 1980) proposed a phylogenetic hypothesis for the basal Tetraodontiformes. He hypothesized that the triacanthodid Hollardinae gave rise to the Triacanthidae, and then to the rest of the Tetraodontiformes. His hypothesis was supported by the presence, in both the Triacanthidae and the Hollardinae, of a dome-shaped supraoccipital, which excludes the epiotics from meeting on the dorsal surface of the skull, and of the shaft-like form of the pelvis posterior to the pelvic spines. Within the Triacanthidae, Tyler (1968, 1980) removed several taxa from the genus Triacanthus, recognized as valid the genera Pseudotriacanthus and Trixiphichthys, and erected a new genus, Tripodichthys. He also hypothesized a close relationship between Tripodichthys and Pseudotriacanthus on the basis of the tapering pelvis, ending at a point, and between Trixiphichthys and Triacanthus on the basis of the nontapering pelvis. The extinct Oligocene Acanthopleurus serratus Agassiz and the Miocene Marosichthys huismani (de Beaufort), later shown to be an acanthuroid surgeonfish by Tyler (1997), were identified as belonging to this group, even if the relationships could not be further determined. The group of extant plus fossil triacanthids was hypothesized to be related to Protacanthodes. The relationship of Cryptobalistes to the other triacanthoids was not resolved. Winterbottom (1974), based mainly on myological data from the extant species, hypothesized that the Triacanthidae form a monophyletic group composed of two lineages: one formed by Protacanthodes and the other by the Recent taxa, Acanthopleurus serratus, and Marosichthys. The Triacanthidae was then supposed, on the basis of the presence of a frictional locking mechanism for the first dorsal spine against the first pterygiophore and of a complex pelvic structure, to represent the sister group of the Triacanthodidae (which, according to Winterbottom, includes the clades Hollardinae and Triacanthodinae + Cryptobalistinae). Myological data could not firmly indicate if the Triacanthidae are more closely related to the Triacanthodidae or to the other Tetraodontiformes. Winterbottom made no attempt to resolve the relationships within the Triacanthidae. Tyler & Winterbottom (1999) identified five synapomor322 phies for the subfamily Triacanthinae, composed in their analysis of Cryptobalistes, Acanthopleurus serratus and A. colletti (Tyler), and by the seven extant species. Monophyly of this subfamily was subsequently corroborated by Santini & Tyler (unpublished observation). Tyler & Winterbottom (1999) also identified five synapomorphies that support the monophyly of the clade of the seven extant species of Triacanthidae, but they could not resolve the relationships among Cryptobalistes, Acanthopleurus, and this modern clade. Ongoing work on the higher level genealogical relationships of the Tetraodontiformes (Santini & Tyler unpublished observation) suggests that the clade formed by Protacanthodes ombonii and the newly described P. nimesensis Tyler & Santini (2001) represents the sister group of the Triacanthinae. Additionally, the enigmatic Middle Eocene (Monte Bolca) Protobalistum imperiale (Massalongo) and Spinacanthus cuneiformis (Blainville), which Tyler (1968) hypothesized as being basal to the split between Triacanthodidae and Triacanthidae and which Winterbottom (1974) hypothesized as being related to the Balistidae + Monacanthidae clade, could be related to the Aracanidae and Ostraciidae. The purpose of this paper is to propose a new phylogenetic hypothesis for the relationships of the extant Triacanthidae using osteological and morphometric data. Materials and methods Species examined The materials examined are arranged below alphabetically by genus, species, and museum abbreviation. The museum number, number of specimens in each lot, standard length (SL), and locality of capture of specimens are given. The museum abbreviations follow Leviton et al. (1985). Parahollardia lineata ( Longley). ANSP 93375, 1, 62.2 mm, Florida; ANSP 97637, 3, 79.4 –84.4 mm, Louisiana; ANSP 100473, 1, 45.7 mm, Florida; ANSP 102144, 1, 82.0 mm, Florida; ANSP 102145, 1, 86.1 mm, South Carolina. Pseudotriacanthus strigilifer (Cantor). ANSP 89387, 4, 119 – 145 mm, Thailand; SU 41732, 1, 79.0 mm, India. Triacanthus biaculeatus ( Bloch). ANSP 76585, 1, 124 mm, China; ANSP 89387, 7, 89.6–110 mm, Thailand; ROM 39688, 1, 104 mm, Australia; SU 27746, 2, 113–118 mm, Borneo; SU 41730, 1, 25.6 mm, India; USNM 71055, 1, 71.8 mm, Japan; USNM 147904, 1, 145.2 mm, Saudi Arabia. Triacanthus nieuhofi Bleeker. ANSP 102982, 1, 108 mm, Australia; ZMA 101832, 1, 99.9 mm, Indonesia. Tripodichthys angustifrons (Hollard). ANSP 98719, 1, 137 mm, Australia; AMS 3696, 1, 63.9 mm, Australia; ROM 39651, 1, Zoologica Scripta, 31, 4, September 2002, pp321– 330 • © The Norwegian Academy of Science and Letters F. Santini & J. C. Tyler • Phylogeny of Triacanthidae 85.6 mm, Australia; USNM 357414, 1, 34.4 mm, Papua New Guinea; USNM 357415, 1, 33.9 mm, Myanmar. Tripodichthys blochi (Bleeker). ANSP 63421–25, 5, 24.3– 105 mm, Philippines; ANSP 63429–54, 11, 18.4– 50.9 mm, Philippines; ANSP 63476, 1, 81.0 mm, Philippines; ANSP 77348, 1, 80 mm, Philippines; ANSP 103298, 3, 55.1– 88.9 mm, locality unknown; ANSP 103301, 3, 69.4–97.3 mm, locality unknown; ROM 22983, 2, 77 – 78 mm, Singapore; SU 26930, 2, 92.0 –107 mm, Philippines; USNM 102932, 1, 43.1 mm, Philippines. Tripodichthys oxycephalus ( Bleeker). ANSP 102325, 1, 125 mm, Bay of Bengal. Tryxiphichthys weberi (Chaudhuri). ANSP 101389, 3, 101– 119 mm, Bay of Bengal; ANSP 102136, 1, 29.6 mm, Bay of Bengal; USNM 280327, 1, 91.3 mm, Philippines; USNM 280329, 1, 104.7 mm, Philippines; USNM 357416, 1, 98.2 mm, Philippines; USNM 357417, 1, 37.2 mm, Philippines. Osteological preparations Most osteological observations are based on cleared and stained material. Many cleared and stained materials were obtained on loan; other specimens were cleared with trypsin and stained at the Royal Ontario Museum using the methods described by Dingerkus & Uhler (1977) and Taylor & Van Dyke (1985). The specimens were subsequently dissected to various degrees using a Wild M5A binocular dissecting microscope. Drawings of the cleared and stained osteological materials were made using a Wild Typ 181300 1.25× camera lucida. Radiographs were used for particularly large specimens, which would probably not clear and stain well, and for specimens of infrequently collected species that are unavailable for such processes. Phylogenetic analysis In order to generate hypotheses of relationships, osteological and morphometric characters were analysed following the principles of phylogenetic systematics (cladistics), as first outlined by Hennig (1966) and subsequently modified by several authors, most notably Wiley (1981) and Farris (1983). A data set of 34 putatively informative osteological characters was obtained from the examination of the available specimens. Twenty-one of the 24 morphometric characters used by Tyler (1968) for his comparative descriptions of the genera of Triacanthidae were included in the analysis. Three of the original 24 morphometric characters are autapomorphic and, hence, were excluded. We are aware of the fact that the use of morphometric characters to infer phylogenetic relationships is contentious (Pimentel & Riggins 1987; Cranston & Humphries 1988; Bookstein 1994; Zelditch et al. 2000). However, we decided to include the morphometric characters used by Tyler in his taxonomic revision of the family for two reasons: the first is consistency, given that such features had been employed within an evolutionary taxonomic framework by Tyler (1968), as opposed to the cladistic one that we are adopting here; the second reason is that if we decide to adopt a ‘total evidence’ approach, as advocated by Kluge (1989), we should include all conceivable sources of phylogenetic information, leaving it to our analysis to determine the fit of these features with the evolutionary history of the organisms, without a priori exclusion of any kind of character. The level of polymorphism in our data set is very low, with the striking exception of character 24 (number of soft dorsal and anal fin pterygiophores) in the examined specimens of Tripodichthys blochi. Tripodichthys blochi is one of the most common species of triplespines in museum collections, and we were able to obtain many specimens for osteological examination (most of which had already been cleared and stained, and dissected), in addition to the ones prepared at the Royal Ontario Museum. Unfortunately, for many of these specimens the exact locality of capture is not known. We believe that this large range of variability in this particular taxon could be evidence of the fact that these organisms are representatives of organismal lineages that have been separated long enough to have attained a status of reproductive isolation, and we might be dealing with cryptic species, a common problem that plagues marine systematists (Knowlton 1993). However, in the absence of more accurate information regarding the distribution of these groups of organisms, for the time being we prefer to maintain the taxonomic status of T. blochi for all these specimens. The data matrix (shown in Table 1) was analysed using PAUP* 4 (Swofford 1999). Multistate characters were run unordered, given that their direction of evolution is not known. The outgroup method was used to polarize the characters (Watrous & Wheeler 1981). Following the protocol recommended by Nixon & Carpenter (1993), a single outgroup was selected in order to root the phylogenetic network produced by PAUP. The triacanthodid Parahollardia lineata was chosen as the outgroup because it is considered a relatively basal member of that family, the sister group to the triacanthids. The consistency index (CI), the retention index (RI) and the rescaled consistency index (RC) (Kluge & Farris 1969; Farris 1989) are provided for each analysis. Character evolution was studied using ACCTRAN and DELTRAN optimization options of MACCLADE 3.07 (Maddison & Maddison 1992). MACCLADE was also used to produce the cladograms. To assess the degree of support for the individual clades in the cladogram, the number of extra steps that would have to be added before a clade is lost from the strict consensus tree of the near © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 4, September 2002, pp321– 330 323 Phylogeny of Triacanthidae • F. Santini & J. C. Tyler minimum length cladograms was calculated [Bremer 1988; ‘Bremer support’ of Källersjö et al. (1992)]. Character list Skull ( Fig. 3) 1 Parasphenoid: straight, not dorsally arched (0); dorsally arched (1). 2 Ventral flange of parasphenoid: very shallow (0); deep, at least twice the depth of the shaft-like part of the parasphenoid (1). 3 Ectopterygoid, shape of dorsal edge: straight (0); convex (1). 4 Mesopterygoid: present (0); absent (1). 5 Metapterygoid, shape of dorsal edge: straight (0); convex (1). 6 Hyomandibular: no flange present (0); flange present, horizontal with respect to body axis (1); flange present, oblique to body axis (2). 7 Hyomandibular, when flange present: lower part two or more times longer than the upper part (0); lower part significantly less than twice the length of the upper part (1); not applicable (—). 8 Symplectic: never sutured or attached to metapterygoid (0); fully sutured to metapterygoid along all of its length (1); only partially sutured to metapterygoid, free from it posteriorly (2). 9 Front edge of opercle: slightly concave (0); moderately to deeply convex (1). Paired fin girdles 10 Ventral shape of shaft-like pelvis: V-shaped to roundish (0); flat and expanded (1). 11 Dorsal shape of shaft-like pelvis: flat and broadly expanded (0); V-shaped and narrowly rounded (1); moderately expanded, wider than the middle region, roundish to flattened (2). 12 Ventral surface of pelvis: wider anteriorly than posteriorly, tapering to a point posteriorly (0); about the same width posteriorly as anteriorly, not tapering to a point posteriorly (1). 13 Lateral surface of pelvis: no crest present (0); one oblique crest present (1). 14 Flange on rear base of pelvic fin spine: somewhat rounded posteriorly (0); offset and more angular posteriorly (1). Vertebral column 15 Epineurals: thin (0); thick (1). 16 Point of insertion of first epineural: second abdominal vertebra (0); third abdominal vertebra (1); fourth abdominal vertebra (2). 17 Point of insertion of last epineural: seventh abdominal vertebra (0); third caudal vertebra (1); second caudal vertebra (2). 18 Number of abdominal vertebrae in front of first pterygiophore of soft dorsal fin: eight (0); five (1); four (2). 19 Haemal spines: from seventh abdominal vertebra (0); from eighth abdominal vertebra (1). 324 20 Caudal peduncle length, as measured by number of vertebrae behind last haemal spine supporting anal fin pterygiophores: five (0); six (1); seven (2). Median fins ( Fig. 4) 21 Number of pterygiophores in spiny dorsal fin: five (0); four (1); three (2). 22 Number of pterygiophores in soft dorsal fin: 14 (0); 22 (1); 23 (2); 24 (3); 25 (4). 23 Number of pterygiophores in anal fin: 13 (0); 15 (1); 16 (2); 17 (3); 18 (4). 24 Number of anal fin pterygiophores between haemal spines of second and third caudal vertebrae: two (0); three (1); four (2). 25 Distal pterygiophores of soft dorsal and anal fins: ossified (0); unossified (1). 26 Lateral flange on last centrum and hypural plate: absent (0); present (1). 27 Caudal peduncle, fleshy, transverse indentation just in front of caudal fin base: absent (0); present (1). 28 Caudal fin shape: rounded (0); deeply forked (1). 29 Uroneurals, when present: a series of long, thin, and very small bones (0); a single bone, approximately one-half the size of the epural (1); a single bone, or sometimes two bones, of very reduced size, much less than one-half the size of the epural (2). 30 Hypurals: all five free from each other and from last centrum (0); hypurals consolidated, 1 + 2 and 3 + 4 fused to each other and to last centrum, 5 free (1). 31 Hypural 5: supported by indentation of last centrum and hypurals 1 –4 (0); no indentation of last centrum and hypurals 1 –4 (1); not applicable (—). 32 Hypural plate, indentation between hypurals 2 and 3 (whether fused or free): relatively deep, more than one-third the length of PU1 centrum and hypurals (0); shallow, less than one-third the length of PU1 centrum and hypurals (1). Miscellaneous ( Fig. 5) 33 Scales: bearing small upright individual spinules (0); bearing a cruciform ridge (1); bearing a series of parallel vertical ridges (2). 34 Sagitta (otolith), gross morphology: deep indentation on both sides (0); indentation on one side, other side relatively straight (1). Morphometrics All measurements are as defined in Tyler (1968: 2 –5) and are given as % SL. 35 Head length: ≥ 35 (0); 29–32 (1). 36 Snout length: < 17 (0); 17–19 (1); 20–21 (2); 24 (3). 37 Eye diameter: ≥ 14 (0); 8 –10 (1). 38 Postorbital length of head: 9 (0); 8 (1); 7 (2); 6 (3). 39 Mouth width: ≥ 7 (0); 5 –6 (1); < 5 (2). Zoologica Scripta, 31, 4, September 2002, pp321– 330 • © The Norwegian Academy of Science and Letters F. Santini & J. C. Tyler • Phylogeny of Triacanthidae 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Gill opening depth: ≥ 8 (0); 6 – 7.5 (1). Snout to spiny dorsal origin: ≥ 50 (0); 39 – 42 (1); 44 –46 (2). Body depth: ≥ 57 (0); 35– 40 (1); 45 (2); 31 (3). Second dorsal fin spine length: ≥ 28 (0); 7 –9 (1); 16 –21 (2). Soft dorsal fin base length: 19 (0); 27– 30 (1); 24–26 (2). Soft dorsal fin height: ≥ 16 (0); 7 – 10 (1); 6 (2). Anal fin base length: 15– 18 (0); 19– 21 (1); 12 (2). Anal fin height: 15 (0); 10 – 13 (1); 6– 8 (2). Caudal fin length: 27– 30 (0); 17– 20 (1); 22– 25 (2). Caudal peduncle depth: 10 (0); 2– 3 (1). Caudal peduncle length: ≤ 16 (0); 24– 27 (1); 21 (2). Pelvic girdle width: ≥ 3.7 (0); 2.8 – 3.2 (1); 2.5 (2); 2.0 (3). Pelvic girdle length (of posterior process): 27 (0); 19 –21 (1). Pelvic girdle width in length: 6 – 7 (0); 8 – 10 (1). Pectoral fin length: ≥ 15 (0); 9 – 11 (1). Uppermost pectoral fin ray length: 3 (0); < 2 (1). 3 15 Triacanthus biaculeatus 4 Triacanthus nieuhofi 8 Trixiphichthys weberi 1 Pseudotriacanthus strigilifer 4 1 Tripodichthys oxycephalus 12 1 Tripodichthys blochi 1 Tripodichthys angustifrons Fig. 2 A most-parsimonious cladogram obtained from the analysis of Results and discussion Phylogenetic analysis The cladistic analysis of the 34 osteological characters produced one most-parsimonious tree, with a tree length of 70 steps (CI = 0.757, RI = 0.575, RC = 0.435). Within the Triacanthinae, two monophyletic clades are supported by our analysis: a clade formed by the two species of Triacanthus, and one formed by the three species of Tripodichthys and the single species of Trixiphichthys and Pseudotriacanthus. In this latter clade, Tripodichthys angustifrons is the first lineage to split off; the next taxon to diverge is Tripodichthys blochi. Tripodichthys oxycephalus represents the sister taxon to the clade formed by Pseudotriacanthus strigilifer + Trixiphichthys weberi. The Bremer support for most of the nodes in this cladogram is relatively low, three for the node supporting the Triacanthus clade, and one for all of the other nodes. Analysis of the full data set of 55 morphological characters (osteological and morphometric) yielded one mostparsimonious tree, with a tree length of 114 steps (CI = 0.798, RI = 0.603, RC = 0.482). The topology of the cladogram is the same as that obtained from the analysis of the osteological data alone (Fig. 2). The monophyly of Tripodichthys is still not supported. The only difference resulting from the inclusion of the morphometric characters is an increase to four in the Bremer support for the Triacanthus biaculeatus + Triacanthus nieuhofi clade. Because the putative non-monophyly of Tripodichthys is indicated by nodes with only a weak Bremer value of one, we suggest that it continue to be recognized along with the other three genera until such a time that additional characters (e.g., myological, neural, genetic) of triacanthids can be included in an analysis. Character polarization was performed using MACCLADE (Maddison & Maddison 1992) in ACCTRAN. This procedure favours reversals (including losses of features) over parallel- the data set. The branch lengths are indicated above the branches, Bremer values are indicated below the branches. isms. In working with Tetraodontiformes, a clade which exhibits a very high number of simplifications and losses of anatomical features, this assumption seems to be justified. The monophyly of the Triacanthidae is supported by 34 synapomorphies: 1(1) parasphenoid arched dorsally; 2(1) ventral flange of parasphenoid at least twice as deep as the shaftlike part of the parasphenoid; 4(1) mesopterygoid absent; 9(1) front edge of opercle moderately to deeply convex; 10(1) ventral shape of shaft-like pelvis flat and expanded; 13(1) lateral surface of pelvis bearing one oblique crest; 14(1) flange on pelvic spine offset and more angular posteriorly; 15(1) epineurals thick; 16(1) first epineurals insert on third abdominal vertebra; 18(1) five abdominal vertebrae present in front of first pterygiophore of soft dorsal fin; 20(1) six vertebrae present in caudal peduncle behind last haemal spine supporting anal fin; 22(3) 24 pterygiophores present in soft dorsal fin; 23(4) 18 pterygiophores present in anal fin; 24(1) three anal fin pterygiophores present between second and third caudal vertebrae; 26(1) lateral flange present on last centrum and hypural plate; 28(1) caudal fin deeply forked; 30(1) hypurals consolidated, 1 + 2 and 3 + 4 fused to each other and to last centrum, 5 free; 33(1) scales bear a cruciform ridge; 35(1) head length 29 –32% SL; 37(1) eye diameter 8 – 10% SL; 39(1) mouth width 5 –7% SL; 40(1) gill opening depth 6 –7.5% SL; 41(1) distance between snout and origin of spiny dorsal fin 39–42% SL; 42(1) body depth 35 – 40% SL; 43(1) length of second dorsal fin spine 7 –9% SL; 44(1) length of soft dorsal fin base 27 –30% SL; 45(1) soft dorsal fin height 7 –10% SL; 47(1) anal fin height 10 –13% SL; 48(2) caudal fin length 22 –25% SL; 49(1) caudal peduncle depth 2 –3% SL; 50(1) caudal peduncle length 24 –27% SL; © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 4, September 2002, pp321– 330 325 Phylogeny of Triacanthidae • F. Santini & J. C. Tyler Fig. 3 A–C. Lateral view of the skull of representative species of Triacanthidae. — A. Triacanthus biaculeatus. — B. Pseudotriacanthus strigilifer. — C. Trixiphichthys weberi. The numbers refer to the character states described in the text. 52(1) pelvic girdle length 19 – 22% SL; 54(1) pectoral fin length 9 – 11% SL; 55(1) uppermost pectoral fin ray length < 2% SL. The monophyly of Triacanthus is supported by 15 characters: 6(1) hyomandibular with a horizontal flange; 7(0) lower part of hyomandibula two or more times longer than the upper part; 8(1) symplectic fully fused to metapterygoid; 11(1) dorsal profile of shaft-like pelvis V-shaped and narrowly rounded; 12(1) ventral surface of pelvis of same width posteriorly as anteriorly, and not tapering to a point; 17(1) third caudal vertebra point of insertion of last epineural; 19(1) haemal spines start from eighth abdominal vertebra; 21(2) three pterygiophores present in spiny dorsal fin; 27(1) caudal peduncle fleshy transverse indentation present; 29(1) a single uroneural bone present, approximately one-half the size of the epural; 34(1) sagitta (otolith) with an indentation on one side only, other side relatively straight; 36(1) snout length 17 – 19% SL; 38(1) postorbital length 8% SL; 46(1) anal fin base length 19– 21% SL; 51(1) pelvic girdle width 2.8–3.2% SL. 326 The monophyly of the Tripodichthys + Trixiphichthys + Pseudotriacanthus clade is supported by 12 characters: 3(1) dorsal shape of ectopterygoid convex; 5(1) shape of dorsal edge of metapterygoid convex; 6(2) hyomandibular flange oblique to body axis; 7(1) hyomandibula lower part significantly less than twice the length of the upper part; 11(2) dorsal shape of shaft-like pelvis moderately expanded, wider than the middle region, roundish to flattened; 17(2) second caudal vertebra point of insertion of last epineural; 21(1) four pterygiophores present in spiny dorsal fin; 29(2) a single uroneural bone, or sometimes two bones, of very reduced size; 36(2) snout length 20 –21% SL; 38(3) postorbital length 6% SL; 51(3) pelvic girdle width 2% SL; 53(1) pelvic girdle width in length eight to 10 times. The monophyly of Tripodichthys blochi + Tripodichthys oxycephalus + Pseudotriacanthus + Trixiphichthys is supported by four characters: 16(2) fourth abdominal vertebra point of insertion of first epineural; 23(2) 16 pterygiophores present Zoologica Scripta, 31, 4, September 2002, pp321– 330 • © The Norwegian Academy of Science and Letters F. Santini & J. C. Tyler • Phylogeny of Triacanthidae Fig. 4 A– C. The caudal fin structure of representative species of Triacanthidae. — A. Triacanthus nieuhofi. — B. Pseudotriacanthus strigilifer. — C. Tripodichthys oxycephalus. The numbers refer to the character states described in the text. in anal fin; 24(2) four anal fin pterygiophores present between second and third caudal vertebrae; 32(1) indentation between hypurals 2 and 3 (whether fused or free) in hypural plate shallow, less than one-third the length of PU1 centrum and hypurals. The monophyly of Tripodichthys oxychephalus + Pseudotriacathus + Trixiphichthys is supported by three characters: 18(2) four abdominal vertebrae present in front of first pterygiophore of soft dorsal fin; 27(1) fleshy transverse indentation of caudal peduncle present; 41(2) distance between snout and spiny dorsal fin origin 44– 46% SL. The monophyly of Pseudotriacanthus + Trixiphichthys is supported by eight characters: 25(1) distal pterygiophores of soft dorsal and anal fin unossified; 29(1) a single uroneural bone, approximately one-half the size of the epural; 43(2) second dorsal fin spine 16 –21% SL; 44(2) soft dorsal fin base length 24 – 26% SL; 45(2) soft dorsal fin height 6% SL; 47(2) anal fin height 6 –8% SL; 48(1) caudal fin length 17 –20% SL; 51(2) pelvic girdle width 2.5% SL. Autapomorphic characters for the taxa are as follows. For Triacanthus biaculeatus: 14(0) flange on pelvic spine rounded posteriorly; 22(2) 23 pterygiophores in soft dorsal fin. For Triacanthus nieuhofi: 20(2) seven vertebrae in caudal peduncle behind last haemal spine supporting anal fin; 25(1) distal pterygiophores of soft dorsal and anal fins unossified; 42(2) body depth 45% SL; 48(0) caudal fin 27 –30% SL. For Tripodichthys angustifrons: 31(1) hypural 5 not supported by indentation of last centrum plus hypurals 1 –4. For Tripodichthys blochi: 8(1) symplectic fully fused to metapterygoid; 22(1) 22 pterygiophores in soft dorsal fin; 23(1 / 2 / 3) 15 – 17 © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 4, September 2002, pp321– 330 327 Phylogeny of Triacanthidae • F. Santini & J. C. Tyler Table 1 Data set for the eight taxa analysed. Polymorphic characters are indicated as P in the data set, not applicable characters are indicated as —. The polymorphic characters in our analysis were: for Triacanthus biaculeatus, character 19, states 0 and 1; for Tripodichthys blochi, character 23, states 1, 2 and 3, character 24, states 1 and 2. Parahollardia lineata Triacanthus biaculeatus Triacanthus nieuhofi Trixiphichthys weberi Pseudotriacanthus strigilifer Tripodichthys blochi Tripodichthys angustifrons Tripodichthys oxycephalus Parahollardia lineata Triacanthus biaculeatus Triacanthus nieuhofi Trixiphichthys weberi Pseudotriacanthus strigilifer Tripodichthys blochi Tripodichthys angustifrons Tripodichthys oxycephalus 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 0 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 0 1 1 0 1 1 1 1 0 0 0 1 1 1 1 1 0 1 1 2 1 2 2 2 — 0 0 1 1 1 1 1 0 1 1 2 0 1 0 0 0 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 2 2 2 2 2 0 1 1 1 0 0 0 0 0 1 1 1 1 1 1 1 0 0 1 0 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 2 2 2 1 2 0 1 1 2 1 2 2 2 0 P 1 1 0 0 0 0 0 2 2 1 1 1 1 1 0 2 3 2 3 1 3 4 0 4 4 2 2 P 4 4 0 1 1 2 2 P 1 2 0 1 1 1 1 2 2 2 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 0 1 1 1 1 1 1 1 — 0 0 0 1 0 1 0 0 0 0 0 1 1 0 1 0 1 1 1 2 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 1 0 1 1 3 2 2 2 2 0 1 1 1 1 1 1 1 0 1 1 3 3 3 3 2 0 1 1 2 1 2 1 1 0 1 1 1 1 1 1 1 0 1 1 2 1 1 1 2 0 1 2 1 3 1 1 1 0 1 1 2 2 1 1 1 0 1 1 2 2 1 1 1 0 1 1 2 2 1 1 1 0 1 1 0 2 0 0 0 0 2 0 1 1 2 2 2 0 1 1 1 1 1 1 2 0 1 1 2 2 3 3 0 0 1 1 1 1 1 1 1 0 0 0 1 1 1 1 0 Fig. 5 A, B. View of sagitta (otolith) of representative species of Triacanthidae. —A. Triacanthus biaculeatus. — B. Pseudotriacanthus strigilifer. The numbers refer to the character states described in the text. pterygiophores in anal fin; 24(1/2) three or four anal fin pterygiophores between haemal spines of second and third caudal vertebrae; 39(2) mouth width < 5% SL. For Tripodichthys oxycephalus: 20(0) five vertebrae in caudal peduncle behind last haemal spine supporting anal fin; 22(4) 25 pterygiophores in soft dorsal fin; 23(4) 18 pterygiophores in anal fin; 38(2) postorbital length 7% SL; 50(2) caudal peduncle length 21% SL; 51(0) pelvic girdle width ≥ 3.7% SL; 53(0) pelvic girdle width in length six to seven times. For Pseudotriacanthus strigilifer: 6(1) hyomandibular flange present, and horizontal with respect to body axis; 17(1) third caudal vertebra point of insertion of last epineural; 31(1) hypural 5 not supported by indentation of last centrum plus hypurals 1 –4; 33(2) scales bear a series of parallel ridges; 41(1) distance from snout to spiny dorsal fin origin 39 – 42% SL; 42(3) body depth 31% SL; 46(2) length of anal fin base 12% SL. For Trixiphichthys weberi: 3(0) dorsal shape of ectopterygoid straight; 4(0) mesopterygoid present; 8(2) symplectic only 328 0 1 1 2 2 1 1 2 0 1 1 2 2 1 1 1 0 1 2 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 1 0 0 1 0 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 partially attached to metapterygoid; 12(1) ventral surface of pelvis with about the same width posteriorly as anteriorly, not tapering to a point; 14(0) flange on pelvic fin spine rounded posteriorly; 19(1) haemal spines begin from eighth abdominal vertebra; 22(2) 23 pterygiophores in soft dorsal fin; 32(0) indentation between hypurals 2 and 3 relatively deep, more than one-third the length of PU1 centrum and hypurals; 35(0) head length ≥ 35% SL; 36(3) snout 24% SL; 39(2) mouth width 4–5% SL. Using the DELTRAN method of optimization, which favours parallel evolution over reversals, some characters have different distributions than above, as follows. Under DELTRAN, it is ambiguous if character 14(1) supports the monophyly of the clade Tripodichthys + Trixiphichthys + Pseudotriacanthus, with convergent evolution in Triacanthus nieuhofi, or if it represents a synapomorphy of the Triacanthidae. Character 19(1) becomes an autapomorphy for Triacanthus nieuhofi and Trixiphichthys weberi. Character 23(2) supports the clade Trixiphichthys + Pseudotriacanthus. Character 24(2) supports the clade Tripodichthys oxycephalus + Trixiphichthys + Pseudotriacanthus. Character 41(2) evolves convergently in Trixiphichthys weberi and Tripodichthys oxycephalus. Character 51 is ambiguous in its optimization, and it is not clear if it evolves state (3) in the ancestor of the clade Tripodichthys + Trixiphichthys + Pseudotriacanthus, or independently twice, once in Tripodichthys angustifrons and once in Tripodichthys blochi. Biogeography All of the living species of Triacanthinae are distributed in the Indo-Western Pacific, an area characterized by the highest Zoologica Scripta, 31, 4, September 2002, pp321– 330 • © The Norwegian Academy of Science and Letters F. Santini & J. C. Tyler • Phylogeny of Triacanthidae diversity in marine fauna of all the world’s oceans (Briggs 1995), whereas the several fossil taxa of triacanthids known to date have been found only in Europe (Italian and Swiss Alps, and the Carpathians in Romania). These fossil and Recent distributions would be concordant with a tethyan origin of this group, and a subsequent colonization of the Indian and Western Pacific oceans. Triacanthus biaculeatus is rather uniformly distributed across the whole Indo-Western Pacific area, ranging in the east to Japan and New South Wales in Australia, and in the west to the Persian Gulf, and is known from China, Indonesia (including Vietnam), the Philippines, Malaya, the Bay of Bengal all the way to Ceylon, the coast of the Arabian Sea from India to the Persian Gulf and the Gulf of Oman. However, Triacanthus biaculeatus is not known from the coasts of Arabia or Madagascar. Triacanthus nieuhofi has a much more restricted distribution, with the majority of records being from the Indonesian region. Only one specimen, whose identification is reported as dubious by Tyler (1968), is known from the Bay of Bengal, and two other specimens have been reported from Australia [even though Tyler (1968) indicated uncertainty about this locality]. Triacanthus biaculeatus is known from only very shallow waters ( 0 – 22 m), while the depth range for T. nieuhofi is unknown. Tripodichthys blochi is only known from the Western Pacific, commonly along the coast of Southern Asia, in the Philippines and Indonesia, but only occasionally from Japan and China. Tripodichthys angustifrons is known only from Australia (New South Wales, Queensland, Northern Territories and Western Australia) and Indonesia. Tripodichthys oxycephalus occurs from Indonesia and the Gulf of Thailand through the Bay of Bengal to the eastern coast of India. Tripodichthys blochi and Tripodichthys angustifrons are known from very shallow waters, usually from 0 to 15 m, with only sporadic catches of Tripodichthys blochi below 20 m, while Tripodichthys oxycephalus is known from 20 to 35 m. Trixiphichthys weberi is known from the Philippines, the Gulf of Thailand, Indonesia, and on both sides of the Bay of Bengal. Pseudotriacanthus strigilifer is known from the Gulf of Oman to the Arabian Sea coast of India, the Bay of Bengal, the Gulf of Thailand, Indonesia and the Philippines. Both Pseudotriacanthus strigilifer and Trixiphichthys weberi are recorded as being collected at depths between 20 and 60 m, with a few specimens caught as deep as approximately 110 m (Tyler 1968). Based on the results of our phylogenetic analysis, it seems that the clade of extant Triacanthidae had its origin from an ancestral taxon that lived in very shallow depths of the late palaeotethys. The lineage leading to Tripodichthys oxycephalus, Pseudotriacanthus and Trixiphichthys later developed deeper water habitats. Acknowledgements We wish to acknowledge the collaboration of the following individuals who made the material of their institutions avail- able for examination: R. Winterbottom, Royal Ontario Museum; S. Jewett, National Museum of Natural History, Smithsonian Institution; W. Saul, Academy of Natural Sciences of Philadelphia; I. Insbrucker, Zoological Museum Amsterdam; M. McGrouther, Australian Museum, Sydney. This work has been made possible by a Short-term Visitor Fellowship awarded to F. Santini by the Smithsonian Institution for collaborative research with the second listed author of this publication. F. Santini also received financial support from The International Council for Canadian Studies and from R. Winterbottom, Royal Ontario Museum and University of Toronto. The comments of two anonymous reviewers helped improve the manuscript. References Baciu, D. S. & Tyler, J. C. (2002). Cephalacanthus trispinosus Ciobanu 1977 from the Oligocene of Romania, a valid species of the triplespine fish genus Acanthopleurus (Tetraodontiformes: Triacanthidae). Review Romain (in press). Bleeker, P. (1859). Enumeratio specierum piscium hucusque in archipelago Indico observatarum, etc. Acta Societatis Scientiarum Indo-Neêrlandicae, 6, 1 – 276. Bleeker, P. (1865). Atlas Ichthyologique des Indes Orientales Neerlandaises, Vol. 5. Baudroides, Ostracions, Gymnodontes, Balistes. Amsterdam: F. Muller. Bleeker, P. (1866). Systema Balistidorum, Ostracionidorum, Gymnodontidorumque revisum. Nederlandsch Tijdschrift Dierkunde, 3, 8– 19. Bookstein, F. L. (1994). Can biometrical shape be a homologous character? In B. K. Hall (Ed.) Homology: the Hierarchical Basis of Comparative Biology ( pp. 198 – 227). New York: Academic Press. Bremer, K. (1988). The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution, 42, 795–803. Briggs, J. C. (1995). Global Biogeography. Amsterdam: Elsevier Science. Cranston, P. S. & Humphries, C. J. (1988). Cladistics and computers: a chironomid conundrum? Cladistics, 4, 72 – 92. Dingerkus, G. & Uhler, L. D. (1977). Enzyme clearing of alcian blue stained whole small vertebrates for demonstration of cartilage. Staining Technology, 51, 229 – 232. Farris, J. S. (1983). The logical basis of phylogenetic analysis. In N. I. Platnick & V. A. Funk (Eds) Advances in Cladistics (pp. 7–36). New York: Columbia University Press. Farris, J. S. (1989). The retention index and rescaled consistency index. Cladistics, 5, 417 – 419. Hennig, W. (1966). Phylogenetic Systematics. Urbana: University of Illinois Press. Hollard, H. (1860). Mémoire sur le squelette des poisons Plectognathes étudié au point de vue des caractères qu’il peut fornir pour la classification. Annales Sciences Naturelle ( Paris), 4, 5 – 46. Källersjö, M., Farris, J. S., Kluge, A. G. & Bult, C. (1992). Skewness and permutation. Cladistics, 8, 275 – 287. Kluge, A. G. (1989). A concern for evidence and a phylogenetic hypothesis of relationships among Epicrates (Boidae: Serpentes). Systematic Zoology, 38, 246 – 259. Kluge, A. G. & Farris, J. S. (1969). Quantitative phyletics and the evolution of anurans. Systematic Zoology, 18, 1 – 32. © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 4, September 2002, pp321– 330 329 Phylogeny of Triacanthidae • F. Santini & J. C. Tyler Knowlton, N. (1993). Sibling species in the sea. Annual Review of Ecology and Systematics, 24, 189 – 216. Leviton, A. E., Gibbs, R. H., Heal, E. & Dawson, C. E. (1985). Standards in herpetology and ichthyology: part I. Standard symbolic codes for institutional resource collections in herpetology and ichthyology. Copeia, 3, 802 – 832. Maddison, W. P. & Maddison, D. R. (1992). MacClade: Analysis of Phylogeny and Character Evolution, Version 3.07. [Computer Software and Manual]. Sunderland, MA: Sinauer Associates. Nixon, K. C. & Carpenter, J. M. (1993). On outgroups. Cladistics, 9, 413–426. Pimentel, R. A. & Riggins, R. (1987). The nature of cladistic data. Cladistics, 3, 201–209. Swofford, D. L. (1999). PAUP — Phylogenetic Analyses Using Parsimony, Version 4.0b2. Sunderland, MA: Sinauer Associates. Taylor, W. R. & Van Dyke, G. C. (1985). Revised procedures for staining and clearing small fishes and other vertebrates for bone and cartilage study. Cybium, 9, 107 – 119. Tyler, J. C. (1968). A Monograph on Plectognath Fishes of the Superfamily Triacanthoidea. Philadelphia: The Academy of Natural Sciences of Philadelphia. Tyler, J. C. (1980). Osteology, phylogeny, and higher classification of the fishes of the order Plectognathi (Tetraodontiformes). NOAA Technical Report NMFS, 434, 1 – 422. Tyler, J. C. (1997). The Miocene fish Marosichthys, a putative tetrao- 330 dontiform, actually a perciform surgeon fish (Acanthuridae) related to the recent Naso. Beaufortia, 4, 1 – 10. Tyler, J. C. & Santini, F. (2001). A new species of triacanthid fish of the genus Protacanthodes from the Eocene of Monte Bolca, Italy (Tetraodontiformes). Bollettino del Museo Civico di Storia Naturale di Verona, 25, 3 – 9. Tyler, J. C. & Santini, F. (2002). Review and reconstructions of the tetraodontiform fishes from the Eocene of Monte Bolca, Italy, with comments on related Tertiary taxa. Studi e Ricerche Sui Giacimenti Terziari di Bolca Museo Civico di Storia Naturale di Verona. (in press). Tyler, J. C. & Winterbottom, R. (1999). A review of the morphology and relationships of the Oligocene spikefish genera Acanthopleurus Agassiz 1844 and Cryptobalistes Tyler 1968 (Tetraodontiformes: Triacanthidae). Palaontologische Zeitschrift, 73, 351 – 367. Watrous, L. E. & Wheeler, Q. D. (1981). The outgroup comparison method of character analysis. Systematic Zoology, 30, 1 – 11. Wiley, E. O. (1981). Phylogenetics: the Theory and Practice of Phylogenetic Systematics. New York: Wiley-Interscience. Winterbottom, R. (1974). The familial phylogeny of the Tetraodontiformes (Acanthopterygii: Pisces) as evidenced by their comparative myology. Smithsonian Contributions to Zoology, 155, 1 –201. Zelditch, M. L., Swiderski, D. L. & Fink, W. L. (2000). Discovery of phylogenetic characters in morphometric data. In J. J. Wiens (Ed.) Phylogenetic Analysis of Morphological Data (pp. 37 –83). Washington: Smithsonian Institution Press. Zoologica Scripta, 31, 4, September 2002, pp321– 330 • © The Norwegian Academy of Science and Letters