Content uploaded by Francesco Santini
Author content
All content in this area was uploaded by Francesco Santini on Feb 10, 2018
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
Author's personal copy
A new multi-locus timescale reveals the evolutionary basis of diversity
patterns in triggerfishes and filefishes (Balistidae, Monacanthidae;
Tetraodontiformes)
Francesco Santini
a,b,
⇑
, Laurie Sorenson
a
, Michael E. Alfaro
a,
⇑
a
University of California Los Angeles, Department of Ecology and Evolutionary Biology, 610 Charles E. Young Drive South, Los Angeles, CA 90095, USA
b
Dipartimento di Scienze Della Terra, Università Degli Studi di Torino, Torino 10125, Italy
article info
Article history:
Received 5 February 2013
Revised 19 May 2013
Accepted 20 May 2013
Available online 30 May 2013
Keywords:
Balistidae
Monacanthidae
Tetraodontiformes
Coral reefs
Fossils
Molecular clock
abstract
Balistoid fishes (triggerfishes, filefishes, leatherjackets) represent one of the most successful radiations of
tetraodontiform fishes across the world’s oceans. Balistids (triggerfishes) are largely circumtropical in
coral reef environments while most monacanthids (filefishes, leatherjackets) are distributed across reef
and non-reef habitats in the Indo-western Pacific. Although members of these clades share a distinctive
mode of locomotion that relies upon coordinated oscillation or undulation of enlarged dorsal and anal
fins, species richness as well as morphologial and ecological diversity are generally considered to be
higher in monacanthids than in triggerfishes. Explicit evolutionary comparisons of diversity patterns
between these sister clades have been hampered by the paucity of systematic studies of filefishes relative
to triggerfishes. Furthermore, a well-sampled molecular timescale for balistoids is lacking, hindering our
understanding of the evolutionary history of these fishes. Here, we produce the largest balistoid molec-
ular dataset to date, based on two mitochondrial and three nuclear loci, for a total of 86 species, and we
time-calibrate it using three tetraodontiform fossils. We show that several of the traditional monacanthid
genera are not monophyletic and that the balistid Xenobalistes tumidipectoris is nested within the genus
Xanthichthys, and suggest that the generic name Xenobalistes be dissolved. Our timetree reveals a Late
Miocene origin of balistids, in accordance with previous studies, but a Late Eocene age for the crown
monacanthids, which experienced significant diversification during the Late Oligocene and Early Mio-
cene. Comparative analyses reveal no significant family-level differences in rates of speciation or body
size evolution, suggesting that the greater diversity of filefishes can be attributed to their more ancient
crown age compared to triggerfishes.
Ó2013 Elsevier Inc. All rights reserved.
1. Introduction
With about 148 extant species distributed across all temperate
and tropical seas (Froese and Pauly, 2012), balistoid fishes (trig-
gerfishes, filefishes, leatherjackets), represent an important radia-
tion of marine percomorph fishes. Currently divided into two
families, the Balistidae (triggerfishes) and the Monacanthidae
(filefishes, leatherjackets), balistoid fishes are characterized by a
number of morphological reductions such as loss of the pelvic
spines; miniaturization of the pelvic fin rays, which become en-
cased within scale plates; and reduction in the number of dorsal
spines and their associated pterygiophores (Tyler, 1980; Santini
and Tyler, 2003). Balistoids are also characterized by a unique
swimming style in which coupled oscillations or undulations of
the soft dorsal and anal fins are used to obtain forward thrust
(Blake, 1978; Dornburg et al., 2011).
In spite of their close evolutionary relationship, triggerfishes
and filefishes exhibit strongly contrasting patterns of species rich-
ness and morphological diversity. Balistids comprise 42 species
while monacanthids include at least 106 species (Froese and Pauly,
2012). Body shape diversity and osteology in balistids appear rela-
tively conserved (e.g., Matsuura, 1979; Tyler, 1980; Dornburg et al.,
2008). In contrast monacanthids, which span from just 2 cm in
body length in Rudarius minutus to over 110 cm in Aluterus scriptus,
range from oblong (e.g., Monacanthus) to almost circular (e.g.,
Brachaluteres) to eel-like (e.g., Anacanthus) in body shape. Sexual
dimorphism in triggerfishes occurs in only a few species (e.g.
Xanthichthys) but occurs in many genera of filefishes (Hutchins,
1992). Distributional patterns also differ signficantly between trig-
gerfishes and filefishes. Whereas almost all triggerfishes are found
circumtropically in association with coral reefs, filefishes are
1055-7903/$ - see front matter Ó2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ympev.2013.05.015
⇑
Corresponding authors. Address: University of California Los Angeles, Depart-
ment of Ecology and Evolutionary Biology, 610 Charles E. Young Drive South, Los
Angeles, CA 90095, USA. Fax: +1 310 206 3987.
E-mail addresses: francesco.santini@alumni.utoronto.ca (F. Santini),
michaelalfaro@ucla.edu (M.E. Alfaro).
Molecular Phylogenetics and Evolution 69 (2013) 165–176
Contents lists available at SciVerse ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Author's personal copy
largely restricted to the Indo-West Pacific, with only a few species
in the Atlantic and Red Sea. However, filefishes occur in a much
greater diversity of habitats across their range, including coral
reefs, rocky reefs, seagrass beds and sandy bottom environments
up to a depth of about 200 m (Froese and Pauly, 2012).
A lack of a robust balistoid-wide phylogeny has hindered quan-
titative study of the underlying causes of differential diversifica-
tion in triggerfishes and filefishes. Filefishes have been especially
understudied, with the pre-cladistic morphological phylogenies
only including one representative per each known genus
(Matsuura, 1979; Tyler, 1980). Although several recently published
molecular phylogenies included some monacanthid species
(Holcroft, 2005; Alfaro et al., 2007; Yamanoue et al., 2009), the
largest study published to date (Yamanoue et al., 2009) included
only 21 species, about 20% of the extant diversity. Furthermore,
these studies have typically included only a single representative
species per genus, thus limiting phylogenetic tests of the current
taxonomy. Testing the monophyly of currently recognized filefish
lineages is important to understanding whether the high morpho-
logical variability found in many of the currently recognized mona-
canthid genera (Hutchins and Swainston, 1985; Hutchins, 1992,
1997) is real or an artifact of errors in taxonomic assignment dri-
ven by the pervasive patterns of sexual dimorphism and cryptic
coloration across the group.
The phylogenetic and evolutionary history of triggerfishes is
better understood due to morphological (Matsuura, 1979; Tyler,
1980) and molecular phylogenies (Yamanoue et al., 2009,Dorn-
burg et al., 2008, 2011), some of which have been time-calibrated
(e.g. Dornburg et al., 2008, 2011) with the help of fossils (Santini
and Tyler, 2003, 2004). Triggerfishes diverged from filefishes in
the Middle Eocene, but the crown radiation did not begin until
the Late Miocene, suggesting that the modern fauna may have
diversified in response to the Late Miocene/Early Pliocene reorga-
nization of coral reef communities (Wood, 1999). Currently avail-
able timescales for monacanthids are extremly limited. Alfaro
et al. (2007) estimated a Late Oligocene origin for crown filefishes
based upon 10% of the extant diversity. In contrast, Yamanoue et al.
estimate a much more ancient divergence in the Mesozoic between
triggerfishes and filefishes based upon mitogenomic data (Yama-
noue et al., 2006).
A more robust phylogenetic framework would permit the first
tests of general hypotheses that could explain differences in balis-
toid species richness, and phenotypic and ecological diversifica-
tion. If crown monacanthids did originate in the Oligocene, the
greater diversity of filefishes could simply reflect the greater length
of time this group has had to diversify relative to triggerfishes,
rather than any intrinsic differences in the rate of speciation or
character evolution between these two clades (Collar et al.,
2005). Alternatively, monacanthid diversity might be due to higher
intrinsic rates of evolution within this group. Habitat may also ex-
plain some aspects of biodiversity. For example, most triggerfishes
and many filefishes are found on coral reefs, which have been
shown to increase diversification rates in other fishes (Alfaro
et al., 2007;Price et al., 2011;Cowman and Bellwood, 2011;Santini
et al., 2013a).
To answer these questions, and to shed further light on the evo-
lutionary history of this group, we assembled the largest molecular
dataset to date for balistoid fishes by sequencing two mitochon-
drial and three nuclear loci for 86 balistoid species (about 60% of
the extant diversity) and three outgroups. We time calibrated the
new molecular phylogeny with three fossil constraints, and used
the resulting timetree to investigate the pattern of lineage diversi-
fication of balistoid fishes, as well as the evolution of body size
among trigger- and filefishes. We also used this new phylogeny
to revise the classification of the balistids by showing that Xenobal-
istes tumidipectoris, known from only a small specimen of about
60 mm found in the stomach of a tuna (Matsuura, 1981), is likely
a juvenile Xanthichthys, and suggest changing its name to Xanthich-
thys tumidipectoris.
2. Materials and methods
2.1. Sampling
We secured tissue samples for 32 species of balistids (76% of ex-
tant species; 11 of 12 described genera), 43 species of monacanth-
ids (41% of extant species; 21 of 31 described genera), plus three
outgroups through loans from university or museum collections,
or purchases through the pet trade (Table 1). We also obtained se-
quence data for the mitochondrial genes used in this study for
Xenobalistes tumidipectoris, the only representative of the one bal-
istid genus missing from our sampling, and an additional 10 spe-
cies of monacanthids from GenBank, bringing our final sampling
to 33 balistids and 53 monacanthids. To verify the identification
of our tissue samples all available voucher specimens were com-
pared to fish identification keys, and all extractions were barcoded
for cytochrome oxidase subunit I; sequences were compared to
those in the barcode of life database, with only those having 99%
or higher similarity to corresponding sequences in the database
being retained. (http://www.barcodinglife.com/index.php/
IDS_OpenIdEngine).
The sister lineage to the balistoids is not known with confidence
due to uncertainty among higher-level tetraodontiform relation-
ships (Winterbottom, 1974; Tyler, 1980; Santini and Tyler, 2003,
2004; Holcroft, 2005; Alfaro et al., 2007; Yamanoue et al., 2008).
We included two molids (Mola mola,Ranzania laevis) and a diodon-
tid (Diodon hystrix) as outgroups. These lineages have been sug-
gested to be closely related to balistoids in other recent
molecular studies (Holcroft, 2005; Alfaro et al., 2007; Yamanoue
et al., 2008).
2.2. DNA extraction, PCR amplification, and sequencing
DNA was extracted from muscle tissue samples or fin clips
stored in 70% ethanol using the Qiagen DNeasy Blood and Tissue
Kit (Qiagen, Valencia, CA, USA) following the protocol suggested
by the manufacturer. Two mitochondrial genes, cytochrome oxi-
dase subunit I (cox1) and cytochrome b (Cytb), and three nuclear
genes, cardiac muscle myosin heavy chain 6 alpha (myh6), recom-
bination activating gene 1 (Rag1), and rhodopsin (Rh), were ampli-
fied using the polymerase chain reaction (PCR). Primers and PCR
conditions are as described in Santini et al. (2013a, 2013b) and
Sorenson et al. (2013). We used ExoSap (Amersham Biosciences)
to remove the excess dNTPs and unincorporated primers from
the PCR products. Purified products were cycle-sequenced using
the BigDye Terminator v.3.1 cycle sequencing kit (1/8th reaction)
(Applied Biosystems) with each gene’s original or additional inter-
nal primers used for amplification. The cycle sequencing protocol
consisted of 25 cycles with a 10-s 94 °C denaturation, 5-s of 50 °C
annealing, and a 4-min 60 °C extension stage. Sequencing was con-
ducted at the Yale University DNA Analysis Facility using an ABI
3730xl DNA Genetic Analyzer (Applied Biosystems).
2.3. Phylogenetic analysis
Chromatograms were visually assessed and assembled into con-
tigs using Geneious 5.4 (Drummond et al., 2011). The consensus se-
quences for each individual gene were aligned in Geneious using
the MUSCLE software (Edgar, 2004), and the alignments subse-
quently checked by eye for accuracy. The sequences were trimmed
to minimize missing characters, and our final data matrix consisted
of 651 bp for cox1, 1107 bp for Cytb, 774 bp for myh6, 1392 bp for
166 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
Author's personal copy
Table 1
List of taxa included in this study with tissue voucher and GenBank numbers. Tissues were from the personal collections of Michael E. Alfaro (MEA), Brian Victor Alfaro (VRA),
Sean Kong (SK) and Francesco Santini (FS) at UCLA, and Peter Wainwright (PW) at UC Davis, as well as from the Australian Museum (EBU), Natural History Museum in Victoria
(NMV), South African Institute of Aquatic Biology (SAIAB), California Academy of Sciences (CAS), Natural History Museum of the University of Kansas (KU), and Scripps Institute of
Oceanography (SIO).
Scientific name Tissue # Cox1 Cytb Myh6 Rag1 Rh
Diodon hystrix FS 05 KF025664 KF025730 KF025800 KF025868 KF025906
Mola mola EBU 38997 KF025665 KF025731 KF025801 KF025869 KF025907
Ranzania laevis SIO 340531 KF025666 KF025732 KF025870 KF025908
Balistidae
Abalistes stellaris PW 1324 KF025667 KF025733 KF025802 KF025871
Abalistes stellatus SAIAB 81914 KF025668 AP009202 KF025803 AY700318 EU108845
Balistapus undulatus MEA 144 KF025669 KF025734 KF025804 EU108869 KF025909
Balistes capriscus MEA 584 KF025670 KF025735 KF025805 AY700308 DQ874818
Balistes polylepis MEA 360 KF025673 KF025738 AY700309 KF025912
Balistes punctatus MEA 283 KF025674 KF025739 KF025808 EU108868 EU108848
Balistes vetula PW 1181 KF025671 KF025736 KF025806 AY700310 KF025910
Balistoides conspicillum FS 2 KF025672 KF025737 KF025807 AY700311 KF025911
Balistoides viridescens KU 4556 KF025675 KF025740 KF025809 AY700320 KF025913
Canthidermis maculata MEA 251 KF025676 KF025741 KF025810 AY700312 EU108851
Canthidermis sufflamen JQ841088
Melichthys indicus KU 7163 KF025677 KF025742 KF025811 KF025872 KF025914
Melichthys niger MEA 562 KF025678 KF025743 AY700313 EU108852
Melichthys vidua MEA 168 KF025679 KF025744 KF025812 EU108870 EU108853
Odonus niger MEA 147 KF025680 AP009208 KF025813 EU108871 EU108854
Pseudobalistes flavimarginatus PW 313 AP009209 AP009209 KF025814 EU108872 EU108855
Pseudobalistes fuscus MEA 158 KF025681 KF025745 KF025815 AY700314 EU108856
Pseudobalistes naufragium MEA 347 KF025682 KF025746 KF025816 GU014460 KF025915
Rhinecanthus aculeatus MEA 194 KF025683 KF025747 KF025817 AY308790
Rhinecanthus assasi MEA 167 KF025684 KF025748 KF025818 AY700315 EU108858
Rhinecanthus lunula MEA 142 KF025685 KF025749 KF025819 EU108873 EU108859
Rhinecanthus rectangulus MEA 110 KF025686 KF025750 KF025820 KF025873 EU108860
Rhinecanthus verrucosus PW 1323 KF025687 KF025751 KF025821 EU108875 EU108861
Sufflamen albicaudatum MEA 100 KF025798 KF025866 KF025904 KF025944
Sufflamen bursa MEA 160 FJ584129 KF025752 KF025822 AY700319 EU108863
Sufflamen chrysopterum PW 1325 JQ350381 KF025753 KF025823 AY700321 EU108864
Sufflamen fraenatum KU 7210 AP004456 KF025754 KF025824 KF025874 KF025916
Sufflamen verres MEA 345 KF025688 KF025755 KF025825 KF025875 KF025917
Xanthichthys auromarginatus MEA 164 KF025689 AP009211 KF025826 AY700316 EU108865
Xanthichthys lineopunctatus SAIAB 74689 KF025690 KF025756 KF025827 KF025876 KF025918
Xanthichthys mento MEA 149 KF025691 KF025757 KF025828 EU108877 EU108866
Xanthichthys ringens MEA 174 KF025692 KF025758 KF025829 EU108878 EU108867
Xenobalistes tumidipectoris AP009182 AP009182
Monacanthidae
Acanthaluteres brownii AP009212 AP009212
Acanthaluteres spilomelanurus NMV Z 10832 KF025693 KF025759 KF025830 KF025877
Acanthaluteres vittiger NMV Z 10906 KF025694 KF025760 KF025878
Acreichthys tormentosus MEA 277 KF025695 KF025761 KF025831 KF025879 KF025919
Aluterus heudeloti KU 5218 KF025696 KF025762 KF025880 KF025920
Aluterus monoceros CAS 123 KF025697 KF025763 KF025832 KF025881 KF025921
Aluterus schoepfi KU 5120 KF025698 KF025764 KF025833 KF025882 KF025922
Aluterus scriptus MEA 170 KF025699 KF025765 KF025834 AY700331 KF025923
Amanses scopas KU 4457 KF025700 KF025766 KF025835 AY308793
Brachaluteres jacksonianus EBU 20482 KF025701 KF025767 KF025836 AY700337 KF025924
Brachaluteres ulvarum AP009216 AP009215
Cantherhines dumerilii MEA 188 EU791285 KF025768 AY700332
Cantherhines macroceros JQ842801
Cantherhines pardalis SIO 340451 KF025702 KF025769 KF025837 KF025883 KF025925
Cantherhines pullus MEA 246 KF025703 KF025770 KF025838 AY700333 KF025926
Cantherhines sandwichiensis MEA 912 JQ411523 KF025771 KF025839 KF025884
Cantherhines verecundus DQ521021
Chaetodermis penicilligerus MEA322 KF025704 KF025772 KF025840 KF025885 KF025927
Eubalichthys bucephalus NMV Z 7896 KF025705 KF025773 KF025841 KF025886
Eubalichthys mosaicus NMV Z 10836 KF025706 KF025774 KF025887
Meuschenia freycineti NMV Z 10819 KF025707 KF025775 KF025842
Meuschenia hippocrepis NMV Z 7894 KF025708 KF025776 KF025843
Meuschenia trachylepis EBU 30878 KF025709 KF025777 KF025844 AY700338 KF025928
Monacanthus chinensis MEA 512 KF025710 KF025778 KF025845 KF025888 KF025929
Monacanthus ciliatus PW 1183 KF025711 KF025779 KF025846 KF025930
Monacanthus tuckeri MEA 254 JQ840165 KF025780 KF025847 KF025889
Nelusetta ayraudi NMV Z 10823 KF025712 KF025781 KF025848 AY700340
Oxymonacanthus longirostris MEA 134 KF025713 KF025782 KF025849 AY700339 KF025931
Paraluteres prionurus PW 1268 KF025714 KF025783 KF025850 AY700336 KF025932
Paramonacanthus choirocephalus VRA 011 KF025715 AP009223 KF025851 KF025890 KF025933
Paramonacanthus filicauda EBU 22126 KF025716 KF025784 KF025852 KF025891 KF025934
Paramonacanthus oblongus EBU 40666 KF025717 KF025785 KF025853
(continued on next page)
F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176 167
Author's personal copy
Rag1, and 741 bp for Rh, for a total of 4665 nucleotides used in the
concatenated analyses. All sequences generated for this study were
deposited in GenBank (Table 1).
We used jModelTest (Posada, 2008) to select the best fitting
model of sequence evolution from the candidate pool of models
that can be utilized in MrBayes 3.2 (Ronquist et al., 2012) using
AICc (Akaike, 1973). We did not include the proportion of invariant
sites parameter in the candidate pool, as this parameter is already
taken into consideration by the gamma parameter (Yang, 2006).
The GTR + G model was selected as the most appropriate for
myh6 and Rag1, while HKY + G was selected as the best model for
cox1,Cytb, and Rh.
Individual gene datasets were subjected to maximum likelihood
analyses using RAxML (Stamatakis, 2006) to test for incongruence
between gene histories of different loci and to identify potentially
contaminated or mislabeled sequences. We assigned a GTR + G
model to each individual gene partition, implementing the RAxML
model closest to the jModelTest results, and ran 100 fast bootstrap
replicates using the GTR + CAT model. The five individual gene
datasets were then concatenated in Mesquite 2.75 (Maddison
and Maddison, 2011), and the full dataset was partitioned by gene
and subjected to maximum likelihood analyses with RAxML (Sta-
matakis, 2006). Each partition was again assigned its own GTR + G
model, and 1000 fast bootstrap replicates were generated using the
GTR + CAT model.
We used MrBayes 3.2 (Ronquist et al., 2012) to perform Bayes-
ian analyses. We partitioned the concatenated dataset by locus and
assigned the model selected by jModelTest. We ran multiple repli-
cates with two analyses of 20 million generations each, with four
chains (one cold, three heated) sampling every 1000 generations.
The trace files were checked in Tracer 1.5 (Drummond and Ram-
baut, 2007) to ensure that the chains had reached convergence,
and the first 25% of trees was discarded as burnin. Post-burnin
trees were combined to obtain a 50% majority rule consensus tree.
The topologies of the different replicates were then compared to
each other to assess support for the results of the analyses.
2.4. Divergence time estimation
The concatenated alignment was analyzed as five unlinked
gene partitions, with each locus being assigned the model se-
lected by jModelTest, with uncorrelated lognormal priors in
BEAST 1.7.2 (Drummond and Rambaut, 2007). A birth-death prior
was assigned to rates of cladogenesis. We ran two analyses of 50
million generations each, with sampling every 5000 generations,
and used Tracer 1.5 (Rambaut and Drummond, 2009) to inspect
the trace files, ensuring that the chains had reached convergence
and the ESS values for all parameters were greater than 200. We
removed the first 20% of the trees from each analysis as burnin
and used LogCombiner to merge the files with the remaining
trees, and TreeAnnotator (Drummond and Rambaut, 2007) to ob-
tain a timetree.
Several fossil calibration points were used to convert the molec-
ular phylogeny into a timetree. The oldest known balistoid fossil is
Gornylistes prodigiosus, a likely stem balistid from the Middle
Eocene of northern Caucasus (Late Lutetian, Early Bartonian,
41–42 Ma) (Bannikov and Tyler, 2008). Several additional stem
balistids are known from Oligocene deposits of Caucasus and
Switzerland (Tyler and Santini, 2002; Santini and Tyler 2003,
2004), while a number of crown balistids are known from Miocene
deposits in both Europe and northern Africa (Santini and Tyler,
2003; Schultz, 2004). Unfortunately, the uncertain phylogenetic
placement of these crown fossils, partially due to their highly
incomplete state (Schultz, 2004) and partly due to their lack of
inclusion in morphological phylogenetic studies, prevents us from
considering them as reliable calibration points. The fossil record of
monacanthids is much less well known. Currently, only two spe-
cies have been described, both from the Middle-Upper Pliocene
and Lower Pleistocene of Italy and Greece (Sorbini and Tyler,
2004). Both species have been assigned to the Aluterus-like extinct
genus Frigocanthus, and appear to be too young to be assigned to
any split in our tree.
We thus used Gornylistes prodigiosus to date the minimum age
for the crown balistoid clade, and the Middle Eocene stem balistoid
Bolcabalistes vari (Tyler and Santini, 2002; Santini and Tyler, 2003,
2004) to put a soft upper boundary (marking the 95% of the prior
probability for that split) of 50 Ma on this node. We used the Early
Miocene Austromola (Gregorova et al., 2009) to provide a minimum
age of 22 Ma for the split between Mola and Ranzania among the
outgroups, and used the Eocene Eomola to set the soft upper bound
for this node (Tyler and Santini, 2002). We also used the Paleocene
Moclaybalistes danekrus, a stem balistoid from the Late Paleogene
of Denmark (58–59 Ma) (Tyler and Santini, 2002), to put a mini-
mum age prior on the root of the tree. The Santonian Protriacanthus
gortani (85 Ma) (Santini and Tyler, 2003) was used to set the soft
upper bound for the root age (crown balistoids: offset = 41,
Table 1 (continued)
Scientific name Tissue # Cox1 Cytb Myh6 Rag1 Rh
Paramonacanthus pusillus SAIAB 82334 KF025718 KF025786 KF025854 KF025892
Paramonacanthus sulcatus EF607471
Pervagor janthinosoma PW 1322 KF025719 KF025787 KF025855 KF025893 KF025935
Pervagor melanocephalus SIO 340458 KF025720 KF025788 KF025856 KF025894 KF025936
Pervagor nigrolineatus MEA 273 KF025721 KF025789 KF025857 KF025895 KF025937
Pseudolutarius nasicornis MEA 196 KF025722 KF025790 KF025858
Pervagor aspricaudatus JQ431991
Pervagor spilosoma DQ521020
Pseudomonacanthus macrurus MEA 503 KF025723 KF025791 KF025859 KF025896 KF025938
Pseudomonacanthus peroni MEA 319 KF025724 KF025792 KF025860 KF025897 KF025939
Rudarius ercodes MEA 501 KF025725 AP009227 KF025861 KF025898 KF025940
Scobinichthys granulatus EBU 40838-032 KF025726 KF025793 KF025899
Stephanolepis auratus SAIAB 82011 KF025727 KF025862 KF025900 KF025941
Stephanolepis cirrhifer KU 8688 KF025728 KF025794 KF025863 KF025901 KF025942
Stephanolepis hispidus PW 1209 KF025729 KF025795 KF025864 AY700335 DQ197910
Stepanolepis setifer MEA 227 JQ840719 KF025796 KF025902 KF025943
Thamnaconus arenaceous JF494680
Thamnaconus fajardoi SAIAB 82037-1 KF025799 KF025867 KF025905
Thamnaconus modestus SK 4 EF607583 KF025797 KF025865 KF025903
Thamnacosus septentrionalis EF607583
Thamnacosus tessellatus JQ681349
168 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
Author's personal copy
mean = 3.0; crown molids: offset = 22, mean = 6.5; root: off-
set = 59, mean = 6.5).
2.5. Comparative analyses
All analyses were performed using the software package R ver-
sion 2.15.2 (R Development Core Team, 2012), using functions in
the packages Geiger (Harmon et al., 2008), Laser (Rabosky, 2006),
and APE (Paradis et al., 2004).
To test whether rates of speciation differed between triggerfish-
es and filefishes, we used the BiSSE model as implemented in
Diversitree (Maddison et al., 2007; FitzJohn, 2012). We compared
a model where speciation rates were allowed to vary by family
to a model where a single speciation rate was held constant across
the tree. We fixed the extinction rates in both models to be 0 since
estimates of extinction from molecular phylogenies are often unre-
liable (Rabosky, 2010).
To identify exceptionally radiating lineages of balistoids we
used MEDUSA, an AIC-based method that can use phylogenetic
and taxonomic richness data to estimate rate shifts on a chrono-
gram with incomplete taxon sampling (Alfaro et al., 2009). We col-
lapsed the timetree to a backbone tree where most of the balistoid
diversity was assigned to 30 lineages. Birth-death models of
increasing complexity were used to estimate the rates of speciation
and extinction, and an improvement in AIC scores of four units or
greater were used as the threshold for retaining rate shifts (Burn-
ham and Anderson, 2002).
To determine the ancestral habitat for monacanthids, and inves-
tigate the number of transitions between coral reefs and non-reef
environments, we used habitat information found in Fishbase
(Froese and Pauly, 2012), as well as published literature (e.g.,
FAO identification sheets) to assign a reef/non-reef status to each
species in our balistoid tree. As we were interested in the evolu-
tionary consequences of the ability to colonize reef habitats, spe-
cies that are found in both reef and non-reef habitats were coded
as reef following the example of Santini et al. (2013a) for habitat
types in pufferfishes. We then used a maximum likelihood ap-
proach to reconstruct ancestral habitats on the timetree using
the MK1 model in Geiger (Harmon et al., 2008).
To examine size diversity in balistoids we obtained size data for
33 species of triggerfish and 53 species of filefish from Fishbase
(Froese and Pauly, 2012). Most records were for total length (TL)
(Table 2). Measures of standard length were converted to total
length using a regression based upon multispecies measurements
(Gaygusuz et al., 2006). We compared log-transformed size diver-
sity between monacanthids and balistids using an Ftest. We recon-
structed the ancestral size at each node on the tree using the MK1
model as implemented in Geiger (Harmon et al., 2008). To compare
rates of size evolution between monacanthids and balistids we
used BrownieLite (O’Meara et al., 2006) as implemented in Phy-
Tools (Revell, 2012). We used a Chi-squared test to determine
whether rates of log-transformed size evolution differed signifi-
cantly between families.
3. Results
3.1. Phylogenetic analyses
Both maximum likelihood and Bayesian analyses of the concat-
enated dataset inferred similar topologies with high support values
present for most inter-specific and generic relationships. The
monophyly of both families is supported with bootstrap propor-
tions (BSP) of 100% in the maximum likelihood phylogeny
(Fig. 1), and posterior probabilities (PP) greater than 0.98 in the
Bayesian phylogeny (Fig. S1).
Table 2
List of taxa with information about total length (in cm) and habitat association (data
from http://www.fishbase.org).
Species Length Habitat
Abalistes stellatus 60 Reef
Abalistes stellaris 60 Non-reef
Balistapus undulatus 30 Reef
Balistes capriscus 60 Reef
Balistes polylepis 76 Reef
Balistes punctatus 60 Non-reef
Balistes vetula 60 Reef
Balistoides conspicillum 50 Reef
Balistoides viridescens 75 Reef
Canthidermis maculata 50 Reef
Canthidermis sufflamen 65 Reef
Melichthys indicus 25 Reef
Melichthys niger 50 Reef
Melichthys vidua 40 Reef
Odonus niger 50 Reef
Pseudobalistes flavimarginatus 60 Reef
Pseudobalistes fuscus 55 Reef
Pseudobalistes naufragium 100 Reef
Rhinecanthus aculeatus 30 Reef
Rhinecanthus assasi 30 Reef
Rhinecanthus lunula 28 Reef
Rhinecanthus rectangulus 30 Reef
Rhinecanthus verrucosus 23 Reef
Sufflamen albicaudatum 22 Reef
Sufflamen bursa 25 Reef
Sufflamen chrysopterum 30 Reef
Sufflamen fraenatum 38 Reef
Sufflamen verres 40 Reef
Xanthichthys auromarginatus 30 Reef
Xanthichthys lineopunctatus 30 Reef
Xanthichthys mento 30 Reef
Xanthichthys ringens 25 Reef
Xenobalistes tumidipectoris 7 Non-reef
Acanthaluteres brownii 55 Reef
Acanthaluteres spilomelanurus 14 Non-reef
Acanthaluteres vittiger 35 Non-reef
Acreichthys tomentosus 12 Reef
Aluterus heudelotii 45 Non-reef
Aluterus monoceros 76 Reef
Aluterus schoepfii 61 Reef
Aluterus scriptus 110 Reef
Amanses scopas 20 Reef
Brachaluteres jacksonianus 10 Reef
Brachaluteres ulvarum 8 Non-reef
Cantherhines dumerilii 38 Reef
Cantherhines macrocerus 46 Reef
Cantherhines pardalis 25 Reef
Cantherhines pullus 20 Reef
Cantherhines sandwichiensis 19 Reef
Cantherhines verecundus 13 Non-reef
Chaetodermis penicilligerus 31 Reef
Eubalichthys bucephalus 50 Non-reef
Eubalichthys mosaicus 60 Non-reef
Meuschenia trachylepis 40 Reef
Meuschenia freycineti 60 Non-reef
Meuschenia hippocrepis 51 Non-reef
Monacanthus ciliatus 20 Reef
Monacanthus chinensis 38 Reef
Monacanthus tuckeri 10 Reef
Nelusetta ayraud 100 Non-reef
Oxymonacanthus longirostris 12 Reef
Paraluteres prionurus 11 Reef
Paramonacanthus choirocephalus 13 Non-reef
Paramonacanthus filicauda 22 Non-reef
Paramonacanthus oblongus 8 Non-reef
Paramonacanthus pusillus 18 Non-reef
Paramonacanthus sulcatus 24 Non-reef
Pervagor janthinosoma 16 Reef
Pervagor melanocephalus 16 Reef
Pervagor nigrolineatus 10 Reef
Pseudalutarius nasicornis 19 Reef
Pervagor aspricaudus 13 Reef
Pervagor spilosoma 18 Reef
(continued on next page)
F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176 169
Author's personal copy
Within balistids, several major clades are identified (Fig. 1): the
first contains a lineage formed by Melichthys,Balistapus and Balisto-
ides conspicillum, which represents the sister group to the remain-
ing balistids. The monotypic Odonus niger appears to be the next
lineage to branch off the balistid tree, followed by a clade formed
by Balistoides viridescens and Pseudobalistes flavimarginatus, plus a
Xanthichthys and Xenobalistes clade (with Xenobalistes nested dee-
ply within Xanthichthys). The last balistid clade is composed of
two main lineages: the first includes Balistes,Pseudobalistes fuscus
and P. naufragium; the second lineage includes Canthidermis,Abal-
istes,Sufflamen and Rhinecanthus. Monophyly of most genera is
strongly supported, although Pseudobalistes and Balistoides are
shown to be polyphyletic.
Within monacanthids, both Bayesian and maximum likelihood
trees reveal the existence of three well-supported clades. The first
contains a lineage, formed by Pseudolutarius +Oxymonacanthus,
which is sister to all other monacanthids. The second clade in-
cludes a subclade formed by Brachaluteres,Paraluteres, and Rudari-
us;aChaetodermis +Acreichthys +Paramonacanthus +Monacanthus
chinensis subclade and a Pervagor +Stephanolepis +Paramonacan-
thus pusillus lineage. The third clade includes all species of Aluterus
in our dataset, as well as two of the three species of Monacanthus
(M. ciliatus and M. tuckeri); a Cantherhines +Amanses scopas group
sister to Pseudomonacanthus; and a group containing Eubalichthys,
Thamnaconus,Nelusetta,Meuschenia,Acanthaluteres and Scobinich-
thys. Unlike the case with balistids, the monophyly of many of
the traditional genera (Acanthaluteres,Aluterus,Cantherhines,Mon-
acanthus,Paramonacanthus,Stephanolepis) was not supported:
Monacanthus ciliatus and tuckeri appear to be nested within
Aluterus, while M. chinensis belongs within Paramonacanthus;
Paramonacanthus pusillus appears to nest within Stephanolepis;
Cantherhines is paraphyletic without the inclusion of Amanses sco-
pas,asisAcanthaluteres without Scobinichthys; and the two species
of Eubalichthys in our dataset never appear as sister taxa.
3.2. Divergence time estimation
The topology of the BEAST analysis (Fig. 2) is largely congruent
with the RAxML and MrBayes trees. Within balistids two subclades
are recovered: the first includes Melichthys,Balistapus and
Balistoides conspicillum, as well as Odonus, and is the sister group
to Balistoides viridescens +Pseudobalistes flavimarginatus plus
Xanthichthys and Xenobalistes. All taxa that are hypothesized in
the BEAST tree that belong to different lineages than in the MrBa-
yes or RAxML analyses are supported by nodes with very weak PP
or BSP (Figs. 1 and S1). The second subclade includes Canthidermis,
Pseudobalistes fuscus and P. naufragium, and Balistes, as well as Abal-
istes,Sufflamen and Rhinecanthus. Relationships among the species
are identical to the RaxML and MrBayes trees. Within the mon-
acanthids only a few minor differences exist between the BEAST
and the MrBayes/RAxML topologies: Monacanthus ciliatus and M.
tuckeri appear to be the sister group to the largest monacanthid
subclade, instead of being nested within Aluterus (itself a member
of this subclade), while Pseudomonacanthus is not the sister taxon
to Cantherhines +Amanses, but is sister to a group containing Can-
therhines +Amanses as well as several additional genera (Eubalich-
thys,Thamnaconus,Nelusetta,Meuschenia,Acanthaluteres and
Scobinichthys).
Our timetree suggests an Early to Middle Eocene age for the
split between monacanthids and balistids, with a mean age of
52 Ma and a 95% confidence interval for the highest posterior den-
sity (HPD) of 42–60 Ma. The crown balistids are 20 Ma (95% HPD:
16–25 Ma), while the two main balistid subclades are about 18 Ma
each (95% HPD: 13–22 Ma and 14–22 Ma, respectively). The diver-
sity of several balistid genera appears to be the product of fairly re-
cent radiations, with ages of 6 Ma (95% HPD: 4–8 Ma 95% HPD) for
Rhinecanthus, 8 Ma (95% HPD: 6–10 Ma) for Sufflamen, 3 Ma (95%
HPD: 2–4 Ma) for Xanthichthys and 1 Ma (95% HPD: 0.6–1.7) for
Melichthys.
The monacanthids appear to have a Late Eocene origin, with a
crown age of 38 Ma (95% HPD: 31–44 Ma). The Pseudolutari-
us +Oxymonacanthus group originated 15 Ma (95% HPD: 10–
21 Ma), while the two other main subclades both originated within
a fairly short interval in the Early Oligocene: 28 Ma (95% HPD: 23–
33 Ma) and 32 Ma (95% HPD: 26–39 Ma), respectively. Most filefish
genera are paraphyletic and also appear to have originated by the
Miocene, with only Thamnaconus having originated within the Pli-
ocene (95% HPD: 4 Ma, 2–5 Ma).
3.3. Comparative analyses
BiSSE results revealed that triggerfishes diversified somewhat
faster than filefishes: k
triggerfishes
= 0.127, k
filefishes
= 0.090. However,
allowing family-dependent diversification rates did not produce a
significant improvement in model fit (Chi-squared P = 0.129).
MEDUSA (Fig. 3) reveals no rate shifts within balistids, while a ma-
jor increase in diversification rate (r
1
= 0.094; r
2
= 0.249) is found
within monacanthids in the lineage leading to the clade formed
by Eubalichthys,Thamnaconus,Nelusetta,Meuschenia,Acanthaluter-
es and Scobinichthys. This clade has a very young age (12 Ma,
10–15 Ma 95% HPD) and accounts for roughly 30% of the filefish
diversity (32 species out of 106).
Maximum likelihood reconstruction of the habitat type on the
molecular timetree (Fig. S2) shows that while both balistids and
monacanthids are ancestrally reef-associated, only a handful of
triggerfish species have colonized non-reef environments. Within
filefishes, at least seven lineages have independently invaded
non-reef environments (this is likely to be an underestimate due
to our sampling only including 50% of living filefish species). In
spite of this multiple invasion of non-reef environments, only
one of these lineages appears to have achieved a significant diver-
sity (the clade formed by Eubalichthys,Thamnaconus,Nelusetta,
Meuschenia,Acanthaluteres and Scobinichthys). Most lineages of
non-reef monacanthids have also subsequently re-invaded reefs
(Fig. S2).
Triggerfishes are significantly larger than filefishes in a non-
phylogenetic comparison (r
2
= 0.115, F= 12.1, p< 0001; Fig. 4)
and show a smaller range of body sizes. However this difference
in size diversity is not statistically significant (F= 0.5961;
p< 0.1205). Brownie-estimated rates of body size evolution be-
tween families were also not statistically significantly different
(Chi-squared P = 0.328). Maximum likelihood reconstruction of
the log transformed TL (Fig. S3) reveals several lineages which
may have experienced directional evolution towards large or small
body size. Within triggerfishes, the Balistoides viridescens +Pseudo-
Table 2 (continued)
Species Length Habitat
Pseudomonacanthus macrurus 25 Reef
Pseudomonacanthus peroni 48 Reef
Rudarius ercodes 8 Non-reef
Scobinichthys granulatus 30 Non-reef
Stephanolepis auratus 28 Non-reef
Stephanolepis cirrhifer 30 Non-reef
Stephanolepis hispidus 28 Reef
Stephanolepis setifer 20 Reef
Thamnaconus arenaceus 23 Non-reef
Thamnaconus fajardoi 22 Non-reef
Thamnaconus modestus 36 Reef
Thamnaconus septentrionalis 26 Non-reef
Thamnaconus tessellatus 25 Non-reef
170 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
Author's personal copy
Fig. 1. Maximum likelihood phylogenetic hypothesis of balistoid relationships based on analysis of the concatenated dataset using RAxML. Bootstrap proportions over 50%
indicated below branches.
F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176 171
Author's personal copy
Fig. 2. Balistoid timetree based on a Bayesian relaxed clock approach implemented in BEAST 1.7.2. Horizontal bars on each node indicate 95% HPD confidence intervals.
Timescale at bottom of figure is in million years before present. Fish images modified under Creative Commons license from original photographs by J.E. Randall (retrieved
from http://www.fishbase.org).
172 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
Author's personal copy
balistes flavimarginatus clade, and Canthidermis +Pseudobal-
istes +Balistes exhibit large body size (Fig. S3). Within
monacanthids, Aluterus,Cantherhines, and a clade containing
Acanthaluteres +Nelusetta +Meuschenia contain mostly large-
bodied species. A clade of miniaturized filefishes (i.e., Rudarius,
Paraluteres,Brachaluteres) is also recovered.
4. Discussion
4.1. Balistoid relationships
Balistoid fishes are not among the tetraodontiform subclades
that have so far been investigated using numerical cladistic meth-
ods (e.g., Klassen, 1995; Santini and Tyler 2002a, 2002b). The only
detailed phylogenetic treatments of balistoid relationships are
found in Matsuura (1979) and the evolutionary taxonomic study
of Tyler (1980). However the amount of phylogenetic signal pres-
ent in these morphological matrices is low, and a Bayesian analysis
of a data set of balistid species produced by combining these two
earlier studies resulted in trees that are almost entirely unresolved
(Dornburg et al., 2008).
This new study supports the monophyly of both Balistidae and
Monacanthidae, corroborating all previous studies of tetraodonti-
form relationships based on both morphological and molecular
datasets (Winterbottom, 1974; Matsuura, 1979; Tyler, 1980; San-
tini and Tyler, 2003, 2004; Holcroft, 2005; Alfaro et al., 2007;
Yamanoue et al., 2008, 2009).
Our topology of balistid relationships differs from some of the
previous molecular studies (Dornburg et al., 2008; Yamanoue
et al., 2009), which identified Balistes as the sister group to the
other balistids split in two major lineages, and agrees with Dorn-
burg et al. (2011) in recovering Balistes nested deeply within trig-
gerfishes. Our study agrees with Dornburg et al. (2008, 2011) in
inferring non-monophyly of both Balistoides and Pseudobalistes
(including a sister group relationship of P. flavimarginatus and B.
viridescens), while the relationships among the other genera
50 40 30 20 10 0
Divergence Time (MYA)
Fig. 3. MEDUSA diversity tree for analyses of lineage diversification in balistoids. Clades from Fig. 2 have been collapsed into 30 representative stem lineages. Each lineage is
identified with the name of the species from that taxon that was retained for the MEDUSA analysis, and the total species diversity that was assigned to that lineage. The red
branch indicates the diversification rate shift (r
2
= 0.0249 vs. r
1
= 0.094). (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
●
Balistidae Monacanthidae
2.0 2.5 3.0 3.5 4.0 4.5
log (TL)
Fig. 4. Plot illustrating range and mean value of body size (expressed as Ln of total
length, TL) in balistid and monacanthid fishes.
F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176 173
Author's personal copy
inferred in our study are highly congruent to these of Yamanoue
et al. (2009).
Within monacanthids, our topology recovers virtually all of the
same subclades found in Yamanoue et al. (2009). The only differ-
ence involves the placement of Chaetodermis and Acreichthys,
which in our topologies appear as sequential sister taxa to the Par-
amonacanthus +Monacanthus chinensis clade, while in Yamanoue
et al. (2009) preferred hypothesis appear as each other’s sister
groups (although Yamanoue et al. (2009) note that their placement
differs in the likelihood and Bayesian analyses).
Even though Matsuura (1979) and Tyler (1980) examined mul-
tiple species for many genera, both of these morphological studies,
as well as Yamanoue et al. (2009), produced generic level phylog-
enies. By including only one representative species per genus,
Yamanoue et al. (2009), Holcroft (2005), and Alfaro et al. (2007)
were unable to test the monophyly of the currently recognized
genera (the only genus for which Holcroft and Alfaro et al. have
more than one species is Monacanthus). Our study is the only one
to date to broadly sample among and within monacanthid genera,
making it difficult to assess if the massive non-monophyly that we
infer is the result of major bias in sequence evolution, poor identi-
fication of the specimens from which the tissue samples were ta-
ken, or highly incorrect existing taxonomy. In order to reduce the
risk of misidentification/contamination biasing the results of this
study, the cox1 fragment sequenced from the tissues that we had
obtained was blasted against the barcode of life database. Multiple
sequences exist in the barcode of life database for many of the spe-
cies included in this study and offered a similarity of sequence
identity greater than 99%, validating the original identification. In
some cases, the additional sequences for the barcode of life suggest
that the non-monophyly of the traditional genera is warranted. For
example, sequences of Monacanthus chinensis are over 99% similar
to one another, but only show 92% similarity to sequences of
other species of Monacanthus, corroborating our finding that M.
chinensis may not be closely related to other Monacanthus species
in spite of a very similar external morphology.
4.2. The timing of triggerfish and filefish evolution
Our timetree suggests an Early to Middle Eocene origin for the
balistoids, with the two families splitting from one another
52 Ma. These results are in close agreement with the ages in-
ferred by Alfaro et al. (2007) in their analyses of tetraodontiform
family relationships. No other timetree exists for balistoids, or for
monacanthids, even though Dornburg et al. (2008, 2011) produced
timetrees for the balistids. Their timetrees suggested a Late Mio-
cene age of 11 Ma for the origin of the crown balistids; our time-
tree suggests older ages (20 Ma), with no overlap of the two 95%
HPD intervals (7–14 Ma in Dornburg et al. (2011), 16–25 Ma in this
study). This age discrepancy could be due to the inclusion of non-
balistoid outgroups in our study, which allowed us to incorporate
additional fossil calibration points and obtain a more reliable esti-
mate for the ages of the balistoid taxa. Both our study and Dorn-
burg et al. (2011), however, suggest that the radiation of
triggerfishes must have been a fairly recent event, with many gen-
era originating towards the end of the Miocene, and most of the
diversity appearing during the Pliocene and Pleistocene. The tempo
of diversification appears to have been rather different for mon-
acanthids. Crown filefishes appear in the Late Eocene, and all major
clades have stem ages that date back to the Late Eocene or Early
Oligocene and crown ages that are no younger than the Early Mio-
cene. Most monacanthid genera appear to be rather old, with many
dating back to the Middle Miocene (15 million years or older).
Unlike balistids, which possess a rich fossil record (Tyler and
Santini, 2002; Bannikov and Tyler, 2008; Schultz, 2004) suggestive
of an intense rate of turnover during the Oligocene and Early Mio-
cene before the origin of the crown taxa, fossil monacanthids are
only known from Pliocene and Pleistocene deposits in Italy and
Greece, and appear to be already very derived crown taxa. This
paucity of fossils might be due to the lack of suitable fish fossilsites
in the Indo-western Pacific, where the bulk of the monacanthid
diversity is currently found, or could be an indication that for much
of their existence, filefishes have been predominantly associated
with environments that make fossilization unlikely, such as rocky
reefs.
4.3. Comparative analyses
Balistids and monacanthids were shown to be among the fastest
diversifying groups of tetraodontiforms (Alfaro et al., 2007); within
balistoid fishes only one lineage of predominantly non-reef
(Fig. S2), large size (Fig. S3) filefishes is shown to have a signifi-
cantly higher rate of net diversification than the balistoids as a
whole (r
2
= 0.249 vs. r
1
= 0.094). Both families are derived from
coral reef-affiliated ancestors (Fig. S2). Within balistids only a
handful of species have subsequently lost reef association, even
though a number of reef species have a habitat range that includes
non-reef environments (e.g. Balistes capriscus,Canthidermis macula-
ta). Within monacanthids, multiple lineages have moved away
from coral reef environments to occupy either rocky reefs in tem-
perate waters (e.g., New Zealand and southern Australia in the
Indo-western Pacific, or the Northwestern Atlantic) or seagrass
beds and sandy bottoms in tropical latitudes (Froese and Pauly,
2012). Interestingly, most of these lineages have subsequently gi-
ven origin to species that have moved back on coral reefs, a pattern
not seen in balistids.
Although filefishes are commonly regarded as being more phe-
notypically diverse than triggerfishes, our comparative analyses re-
veal that it is the older crown age, rather than an exceptional
diversification rate within filesfishes, that underlies this difference
(Collar et al., 2005), at least with regard to body size. Filefishes
encompass a greater diversity of body sizes, but this difference is
not statistically significant (Fig. 4). Furthermore, our evolutionary
analysis of body size reveals that the filefishes and triggerfishes
have evolved at similar rates. Thus the observed difference in the
range of body sizes within filefishes can be explained by the great-
er age of crown monacanthids (and thus a longer amount of evolu-
tionary time for the accumulation of diversity) relative to balistids.
Although we did not find any significant differences in the rate of
evolution between these two families, ancestral state reconstruc-
tion suggests that some filefish lineages may have experienced
directional trends in size evolution (Fig. S3). For example, direc-
tional evolution may explain large body size in Aluterus and in
the clade containing Acanthaluteres +Scobinichthys +Neluset-
ta +Meuschenia. Similarly, directional evolution towards small
body size may explain trait diversity in Monacanthus ciliatus +M.
tuckeri,inAcanthaluteres spilomelanurus +A. vittiger, and in the
clade containing the smallest filefish genera, Rudarius,Paraluteres
and Brachaluteres. Further comparative tests are needed to deter-
mine if evolutionary trends do characterize these and other balis-
toid lineages, and if there are ecological or geographical
correlates to these patterns.
4.4. Taxonomic reclassification
The results of our study suggest that the current taxonomy of
monacanthids is in need of major revisions. Due to the current
incompleteness of our dataset, which only represents half of the
extant 106 described species, we do not feel that we can perform
a taxonomic revision of this group. Our study, however, shows con-
vincingly that the balistid Xenobalistes tumidipectoris belongs with-
in the genus Xanthichthys. The monotypic genus Xenobalistes was
174 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
Author's personal copy
described on the basis of a single, partially digested specimen
found in the stomach of a tuna (Matsuura, 1981). The specimen
was characterized by small size (60 mm standard length) and
the presence of large protuberances, which appear to be lateral
expansions of the coracoid bones just below the pectoral fins
(Fig. 4 in Matsuura, 1981). We believe that the small size of the
Xenobalistes tumidipectoris holotype, as well as the large size of
the eyes compared to other balistids (Matsuura, 1981; Yamanoue
et al., 2009, Fig. 4) indicate that this specimen was likely a juvenile.
We examined postlarval and juvenile balistids in the fish collection
of a number of museums (LA County Natural History Museum;
Museum of Comparative Zoology, Harvard) and observed an
enlargement of the region below the pectoral fins in specimens
of several balistid species, although the enlargement is not as pro-
nounced as in Xenobalistes tumidipectoris. On the basis of both our
phylogeny and this information, we suggest renaming Xenobalistes
tumidipectoris as Xanthichthys tumidipectoris and eliminating the
generic name Xenobalistes.
5. Conclusions
This study, containing 33 of the 42 extant balistid and 53 of the
106 monacanthid species, represents the most comprehensive
molecular phylogeny yet produced for balistoid fishes. Our study
supports the monophyly of both families, as well as the presence
of two robust subclades of balistids and four highly supported
subclades of monacanthids. Our study is largely in agreement with
the only previous molecular study of generic-level relationships
among monacanthids (Yamanoue et al., 2009), while it conflicts
with some of the previous studies of balistid relationships (Dorn-
burg et al., 2008,2011;Yamanoue et al., 2009) in several nodes
of the topology. We show that the only known species of the bal-
istid genus Xenobalistes was likely described on the basis of a juve-
nile Xanthichthys, and suggest renaming the species as Xanthichthys
tumidipectoris. We also show that many of the currently recognized
monacanthid genera are not monophyletic, and that the taxonomy
of this group is in need of major revisions.
The Bayesian chronogram, produced by calibrating the molecu-
lar phylogeny with three tetraodontiform fossils, suggests a Mio-
cene age, and a largely Pliocene/Pleistocene radiation for
balistids, as well as a Late Eocene origin for monacanthids, with
the major filefish lineages having originated between the Late Oli-
gocene and Early Miocene. This pattern is largely similar to that
observed in other tetraodontiform clades (e.g., Santini et al.,
2013a, 2013b).
The maximum likelihood reconstruction of the ancestral habitat
over the molecular chronogram reveals that the common ancestors
of both clades were associated with coral reefs. Balistids have re-
mained largely a reef-associated group, but at least five lineages
of filefishes have radiated into non-coral reef environments, often
with some members of these lineages moving back over scleractin-
ian reefs. Although filefishes exhibit greater species richness and
phenotypic diversity, our analyses reveal no significant family-le-
vel differences in the rate of diversification or body size evolution.
Thus, the greater diversity of filefishes may be simply due to the
longer time this family has had to diversify.
Acknowledgments
Funding for this project was provided by NSF Grant DEB
0842397 ‘‘Systematics and Influence of Coral Reefs on Diversifica-
tion in Tetraodontiform Fishes.’’ to MEA and FS; FS was also sup-
ported by an ISI Lagrange fellowship. This research project was
made possible by the generous loan or gift of tissues from a num-
ber of colleagues and institutions: Victor Brian Alfaro (UCLA), Sean
Kong (UCLA), Peter Wainwright (UC Davis), Tomio Iwamoto (Cali-
fornia Academy of Sciences), Andrew Bentley and Ed Wiley (Uni-
versity of Kansas), H.J. Walker and Phil Hastings (Scripps
Institute of Oceanography), Unathi Lwana (South African Institute
for Aquatic Biodiversity), Dianne Bray (Museum Victoria), and
Mark McGrouther (Australian Museum). We thank Kelly Huynh
and Mai Nguyen for help with the lab work, and Giorgio Carnevale,
Giacomo Bernardi and an anonymous reviewer for helpful com-
ments on an earlier version of this manuscript.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ympev.2013.05.
015.
References
Akaike, H., 1973. Information theory as an extension of the maximum likelihood
principle. In: Petrov, B.N., Caski, F. (Eds.), Second International Symposium on
Information Theory. Akademiai Kiado, Budapest, pp. 267–281.
Alfaro, M.E., Santini, F., Brock, C.D., 2007. Do reefs drive diversification in marine
teleosts? Evidence from the pufferfishes and their allies (order
Tetraodontiformes). Evolution 61, 2104–2126.
Alfaro, M.E., Santini, F., Brock, C.D., Alamillo, H., Dornburg, A., Rabosky, D.L., et al.,
2009. nine exceptional radiations plus high turnover explain species diversity in
jawed vertebrates. Proc. Natl. Acad. Sci. USA 106, 13410–13414.
Bannikov, A.F., Tyler, J.C., 2008. A new genus and species of triggerfish from the
middle Eocene of the northern Caucasus, the earliest member of the Balistidae
(Tetraodontiformes). Paleontol. J. 42, 615–620.
Blake, R.W., 1978. Balistiform locomotion. J. Mar. Biol. Assoc. UK 58, 73–80.
Burnham, K.P., Anderson, D.R., 2002. Model Selection and Multimodel Inference: A
Practical Information-theoretic Approach. Springer, New York.
Collar, D.C., Near, T.J., Wainwright, P.C., 2005. Comparative analysis of
morphological diversity: does disparity accumulate at the same rate in two
lineages of centrarchid fishes? Evolution 59, 1783–1794.
Cowman, P.F., Bellwood, D.R., 2011. Coral reefs as drivers of cladogenesis:
expanding coral reefs, cryptic extinction events, and the development of
biodiversity hotspots. J. Evolution Biol. 24, 2543–2562.
Dornburg, A., Santini, F., Alfaro, M.E., 2008. The influence of model averaging on
clade posteriors: an example using the triggerfishes (Family: Balistidae). Syst.
Biol. 57, 905–919.
Dornburg, A., Sidlauskas, B., Sorenson, L., Santini, F., Near, T., Alfaro, M.E., 2011. The
influence of an innovative locomotor strategy on the phenotypic diversification
of triggerfish (Family: Balistidae). Evolution 65, 1912–1926.
Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evolut. Biol. 7, 214.
Drummond, A.J., Ashton, B., Buxton, S., Cheung, M., Cooper, A., Duran, C., et al., 2011.
Geneious v5.4. <http://www.geneious.com>.
Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and
high throughput. Nucl. Acids Res. 32, 1792–1797.
FitzJohn, R.G., 2012. Diversitree: comparative phylogenetic analyses of
diversification in R. Meth. Ecol. Evolut. 3, 1084–1092.
Froese, R., Pauly, D., 2012. FishBase. <http://www.fishbase.org>.
Gaygusuz, O., Gürsoy, C., Özulu, M., Tarkan, A.S., Acipinar, H., Bilge, G., Filiz, H., 2006.
Conversions of total, fork and standard length measurements based on 42
marine and freshwater fish species (from Turkish Waters). Turk. J. Fisher. Aquat.
Sci. 6, 79–84.
Gregorova, R., Schultz, O., Harzhauser, M., Kroh, A., Coric, S., 2009. A giant early
Miocene sunfish from the North Alpine foreland basin (Austria) and its
implication for molid phylogeny. J. Vert. Paleont. 29, 359–371.
Harmon, L.J., Weir, J.T., Brock, C.D., Glor, R.E., Challenger, W., 2008. GEIGER:
investigating evolutionary radiations. Bioinformatics 24, 129–131.
Holcroft, N.I., 2005. A molecular analysis of the interrelationships of
tetraodontiform fishes (Acanthomorpha: Tetraodontiformes). Mol. Phylogenet.
Evolut. 34, 525–544.
Hutchins, J.B., 1992. Sexual dimorphism in the osteology and myology of
monacanthid fishes. Rec. West. Aust. Mus. 15, 739–747.
Hutchins, J.B., 1997. Review of the monacanthid fish genus Paramonacanthus, with
descriptions of three new species. Rec. West. Aust. Mus. 15, 1–57.
Hutchins, J.B., Swainston, R., 1985. Revision of the monacanthid fish genus
Brachaluteres. Rec. West. Aust. Mus. 12, 57–78.
Klassen, G.J., 1995. Phylogeny and biogeography of the Ostraciinae
(Tetraodontiformes: Ostraciidae). Bull. Mar. Sci. 57, 393–441.
Maddison, W.P., Maddison, D.R., 2011. Mesquite: A Modular System for
Evolutionary Analysis, Version 2.75. <http://www.mesquiteproject.org>.
Maddison, W.P., Midford, P.E., Otto, S.P., 2007. Estimating a binary character’s effect
on speciation and extinction. Syst. Biol. 56, 701–710.
Matsuura, K., 1979. Phylogeny of the superfamily Balistoidea (Pisces:
Tetraodontiformes). Mem. Fac. Fish. Hokkaido Univ. 26, 49–169.
F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176 175
Author's personal copy
Matsuura, K., 1981. Xenobalistes tumidipectoris, a new genus and species of
triggerfish (Tetraodontiformes, Balistidae) from the Marianas. Bull. Nat. Sci.
Mus. Tokyo Ser. A 7, 191–200.
O’Meara, B.C., Ane, C., Sanderson, M.J., Wainwright, P.C., 2006. Testing for different
rates of continuous trait evolution using likelihood. Evolution 60, 922–933.
Paradis, E., Claude, J., Strimmer, K., 2004. APE: analyses of phylogenetics and
evolution in R language. Bioinformatics 20, 289–290.
Posada, D., 2008. JModelTest: phylogenetic model averaging. Mol. Biol. Evolut. 25,
1253–1256.
Price, S.A., Holzman, R., Near, T.J., Wainwright, P.C., 2011. Coral reefs promote the
evolution of morphological diversity and ecological novelty in labrid fishes.
Ecol. Lett. 14, 462–469.
Rabosky, D.L., 2006. LASER: a maximum likelihood toolkit for detecting temporal
shifts in diversification rates from molecular phylogenies. Evolut. Bioinform. 2,
257–260.
Rabosky, D.L., 2010. Extinction rates should not be estimated from molecular
phylogenies. Evolution 64, 1816–1824.
Rambaut, A., Drummond, A.J., 2009. Tracer v1.5. <http://beast.bio.ed.ac.uk/Tracer>.
Revell, L.J., 2012. Phytools: an R package for phylogenetic comparative biology (and
other things). Meth. Ecol. Evolut. 3, 217–223.
Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D.L., Darling, A., Höhna, S., Larget,
B., Liu, L., Suchard, M.A., Huelsenbeck, J.P., 2012. MrBayes 3.2: efficient bayesian
phylogenetic inference and model choice across a large model space. Syst. Biol.
61, 539–542.
Santini, F., Tyler, J.C., 2002a. Phylogeny and biogeography of the extant species of
triplespine fishes (Triacanthidae, Tetraodontiformes). Zool. Scr. 31, 321–330.
Santini, F., Tyler, J.C., 2002b. Phylogeny of the ocean sunfishes (Molidae,
Tetraodontiformes), a highly derived group of teleost fishes. Ital. J. Zool. 69,
37–43.
Santini, F., Tyler, J.C., 2003. A phylogeny of the families of fossil and extant
tetraodontiform fishes (Acanthomorpha, Tetraodontiformes), Upper Cretaceous
to recent. Zool. J. Linn. Soc. 139, 565–617.
Santini, F., Tyler, J.C., 2004. The importance of even highly incomplete fossil taxa in
reconstructing the phylogenetic relationships of the Tetraodontiformes
(Acanthomorpha: Pisces). Integr. Comp. Biol. 44, 349–357.
Santini, F., Nguyen, M., Sorenson, L., Waltzek, T.B., Alfaro, J.W.L., Eastman, J.M.,
Alfaro, M.E., 2013a. Do habitat shifts drive the diversity in teleost fishes? An
example from the pufferfishes (Tetraodontidae). J. Evol. Biol. 66, 1003–
1018.
Santini, F., Sorenson, L., Marcroft, T., Dornburg, A., Alfaro, M.E., 2013b. A multilocus
molecular phylogeny of boxfishes (Aracanidae, Ostraciidae; Tetraodontiformes).
Mol. Phylogenet. Evolut. 66, 153–160.
Schultz, O., 2004. A triggerfish (Osteichthyes: Balistidae: Balistes) from the
Badenian (Middle Miocene) of the Vienna and the Styrian Basin (Central
Paratethys). Annal. Naturhist. Mus. Wien 106A, 345–369.
Sorbini, C., Tyler, J.C., 2004. Review of the fossil file fishes of the family
Monacathidae (Tetraodontiformes), Pliocene and Pleistocene of Europe, with a
new genus, Frigocanthus, and two new species related to the Recent Aluterus.
Boll. Mus. Civ. Stor. Nat. Verona 28, 41–76.
Sorenson, L., Santini, F., Carnevale, G., Alfaro, M.E., 2013. A multi-locus timetree of
surgeonfishes (Acanthuridae, Percomorpha), with revised family taxonomy.
Mol. Phylogenet. Evolut. 68, 150–160.
Stamatakis, A., 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic
analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688–
2690.
Tyler, J.C., 1980. Osteology, phylogeny, and higher classification of the fishes of the
order Plectognathi (Tetraodontiformes). NOAA Tech. Rep. NMFS Circ. 434, 1–
422.
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. Boll. Mus. Civ. Stor. Nat. Verona, Studi Ric. Giac. Terz. Bolca 9, 47–119.
Winterbottom, R., 1974. The familial phylogeny of the Tetraodontiformes
(Acanthopterygii: Pisces) as evidenced by their comparative myology.
Smithson. Contr. Zool. 155, 1–201.
Wood, R., 1999. Reef Evolution. Oxford University Press, New York.
Yamanoue, Y., Miya, M., Inoue, J.G., Matsuura, K., Nishida, M., 2006. The
mitochondrial genome of spotted green pufferfish Tetraodon nigroviridis
(Teleostei: Tetraodontiformes) and divergence time estimation among model
organisms in fishes. Genes Genet. Syst. 81, 29–39.
Yamanoue, Y., Miya, M., Matsuura, K., Katoh, M., Sakai, H., Nishida, M., 2008. A new
perspective on phylogeny and evolution of tetraodontiform fishes (Pisces:
Acanthopterygii) based on whole mitochondrial genome sequences: basal
ecological diversification? BMC Evolut. Biol. 8, 212.
Yamanoue, Y., Miya, M., Matsuura, K., Sakai, H., Katoh, M., Nishida, M., 2009. Unique
patterns of pelvic fin evolution: a case study of balistoid fishes (Pisces:
Tetraodontiforms) based on whole mitochondrial genome sequences. Mol.
Phylogenet. Evol. 50, 179–189.
Yang, Z., 2006. Computational Molecular Evolution. Oxford University Press, Oxford.
176 F. Santini et al. / Molecular Phylogenetics and Evolution 69 (2013) 165–176
