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