MOLECULAR
PHYLOGENETICS
AND
EVOLUTION
Molecular Phylogenetics and Evolution 26 (2003) 215–221
www.elsevier.com/locate/ympev
Molecular phylogenetic evidence refuting the hypothesis of
Batoidea (rays and skates) as derived sharks
Christophe J. Douady,a Mine Dosay,b,1 Mahmood S. Shivji,c and Michael J. Stanhopeb,*,2
a
c
Department of Biochemistry and Molecular Biology, Dalhousie University, 5859 University Avenue, Halifax, Nova Scotia, Canada B3H 4H7
b
Biology & Biochemistry, The QueenÕs University of Belfast, 97 Lisburn Road, Belfast BT9 7BL, UK
Guy Harvey Research Institute, Oceanographic Center, Nova Southeastern University, 8000 North Ocean Drive, Dania Beach, FL 33004, USA
Received 1 March 2002; revised 30 July 2002
Abstract
Early morphological studies regarding the evolutionary history of elasmobranchs suggested sharks and batoids (skates and rays)
were respectively monophyletic. More modern morphological cladistic studies, however, have tended to suggest that batoids are
derived sharks, closely related to sawsharks and angelsharks, a phylogenetic arrangement known as the Hypnosqualea hypothesis.
Very few molecular studies addressing interordinal relationships of elasmobranchs have been published; the few that do exist, are
very limited in terms of both taxon representation and/or aligned sequence positions, and are insufficient to answer the question of
whether batoids are derived sharks. The purpose of this study was to address this issue with more complete taxon representation,
concomitant with a reasonable number of aligned sequence positions. The data set included a 2.4-kb segment of the mitochondrial
12S rRNA—tRNA valine—16S rRNA locus, and in terms of taxa, representatives of two orders of Batoidea, at least one representative of all orders of sharks, and as an outgroup, the widely recognized sister group to elasmobranchs—Holocephali. The results
provide the first convincing molecular evidence for shark monophyly and the rejection of the Hypnosqualea hypothesis. Our
phylogenetic placement of batoids as a basal elasmobranch lineage means that much of the current thinking regarding the evolution
of morphological and life history characteristics in elasmobranchs needs to be re-evaluated.
Ó 2002 Elsevier Science (USA). All rights reserved.
1. Introduction
Compagno (1973, 1977) and more recently de Carvalho (1996) divide elasmobranch fishes into the
following superorders and orders: Galeomorphii:
Orectolobiformes (Carpet sharks), Heterodontiformes
(Bullhead sharks), Carcharhiniformes (Ground sharks),
and Lamniformes (Mackerel sharks); Squatinomorphii:
Squatiniformes (Angelsharks); Squalomorphii: Hexanchiformes (Frilled and cow sharks), Squaliformes
(Dogfish shark), Pristiophoriformes (Sawsharks); Batoidea: Myliobatiformes (Stingrays), Pristiformes
(Sawfishes), Torpediniformes (Electric rays), Rajiformes
*
Corresponding author. Fax: +610-917-7901.
E-mail address: [email protected] (M.J. Stanhope).
1
Present address: Department of Zootechny, Agriculture Faculty,
University of Canakkale, Canakkale, Turkey.
2
Present address: Bioinformatics, GlaxoSmithKline, 1250 South
Collegeville Road, UP1345, Collegeville, PA 19426-0989, USA.
(Skates), and Rhinobatiformes (Guitarfishes). The relative relationships between these orders and superorders,
and indeed the ordinal composition of the superorders,
remains uncertain. Early morphological work suggested
that there was a fundamental split between batoids and
sharks (e.g., Bigelow and Schroeder, 1948, 1953; also see
review by Seret, 1986). However, the more modern
morphological cladistic studies have tended to suggest
that batoids are derived sharks, closely related to sawsharks and angelsharks (Squatinomorphii (Pristiophoriformes, Batoidea)) (Compagno, 1977; de Carvalho,
1996; de Carvalho and Maisey, 1996; Deets, 1994;
Dingerkus, cited in Seret, 1986; Shirai, 1992a,b, 1996),
proposed as superorder Hypnosqualea by de Carvalho
and Maisey (1996). From the morphological perspective, evidence for the hypnosqualean group was sufficient in ShiraiÕs (1996) opinion for the issue of batoidÕs
phylogenetic position to now be settled.
However, very few molecular studies addressing interordinal relationships of sharks have been published.
1055-7903/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S 1 0 5 5 - 7 9 0 3 ( 0 2 ) 0 0 3 3 3 - 0
216
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
The three relevant studies concern Dunn and Morrissey
(1995), Kitamura et al. (1996), and Arnason et al. (2001).
The Kitamura et al. study involved a 732-bp sequence of
the mitochondrial cytochrome b gene from 8 of the 13
elasmobranch orders; no support values were indicated
in their study and the resulting trees were rooted with a
very distant outgroup (lamprey), instead of a chimaera
(Holocephali), which are widely recognized to be the
sister group to elasmobranchs (Bigelow and Schroeder,
1948; Compagno, 1990). The Dunn and Morrissey study
involved a 303-bp fragment of the mitochondrial 12S
rRNA gene, included only five orders, and did not include the necessary taxa to address the Hypnosqualea
hypothesis. Neither study included a skate as one of
their batoid representatives. The Kitamura et al. study
did support a possible distinction of batoids from
Hypnosqualea, however, it did not include sufficient
ordinal diversity to evaluate whether sharks were
monophyletic and/or where the batoid lineage arose in
the diversification of elasmobranchs. Furthermore, the
fact chimaeras were not used as an outgroup and the fact
cytochrome b is now known to be a highly questionable
marker for deep evolutionary splitting events (Springer
et al., 2001) suggest the results are difficult to interpret.
The Arnason et al. study included complete mitochondrial genome sequences of three orders of sharks (Carcharhiniformes, Squaliformes, and Heterodontiformes)
and a skate, with ratfish as an outgroup, but did not
include the necessary taxa to address the Hypnosqualea
hypothesis. Their study does, however, suggest rejection
of the concept of Squalea (Shirai, 1992b), which has the
six and seven gill sharks as the sister group to Squaliformes/Hypnosqualea, because their trees do not have
the squaliform representative (spiny dogfish) on the
same branch as the skate. These earlier studies, encompassing either those with poorly represented elasmobranch taxonomic diversity, or those with few sequence
positions, concomitant with their basic disagreement
with modern morphological studies, regarding the possibility that batoids are derived sharks related to sawsharks and angelsharks, indicates that a more thorough
molecular perspective is certainly warranted. Modern
molecular studies of interordinal relationships of placental mammals, for example, have gradually expanded
the data sets to include 16.4 kb of sequence data for
several representatives of all orders, before a resolution
of many of the questions regarding branching arrangement were eventually reached (e.g., Murphy et al., 2001).
The primary purpose of this study was to provide a
molecular phylogenetic perspective on the question of
whether sharks are paraphyletic, due to a derived position for Batoidea, or whether sharks are monophyletic,
to the exclusion of Batoidea. A corollary aspect to the
study, arising from the inclusion of representatives of all
orders of sharks, two different ordinal representatives of
Batoidea, concomitant with a chimaera as outgroup,
was examination of several additional questions pertaining to elasmobranch interordinal affinities.
2. Materials and methods
2.1. Taxa and phylogenetic loci
The elasmobranch taxa included in this study encompass all orders of sharks and two of the possible five
orders of Batoidea (Table 1). The monophyly of Batoidea has never been in question, and thus we argue
that missing two or three (i.e., depending on the classification scheme) of the batoid orders (guitarfish, sawfish,
and electric rays) should have minimal affects on our
ability to address the issue of whether elasmobranchs
have a diphyletic history, with sharks and batoids respectively monophyletic, or whether sharks are paraphyletic, due to a derived position of the Batoidea (cf.
below). All phylogenetic analyses included at least one
chimaera sequence as one of the outgroups.
The phylogenetic locus was the mitochondrial 12S–
16S and intervening tRNA valine, chosen because it has
been shown to be a powerful phylogenetic marker for
other groups of vertebrates (Springer et al., 2001) and
because preliminary sequence data indicated there was
little or no transition saturation, even in comparisons
involving the chimaera sequence(s) (data not shown).
Various primer combinations were used to amplify this
locus including three forward and two reverse primers.
Forward primer sequences are: 12SF16690: 50 CTCGA
AAAACCCCTAAAACGAGGGCC 30 ; 12SF22: 50 GC
ATGGCACTGAAGATGCTAAGATGA 30 ; and 12S
F204: 50 GCCCAAGACACCTTGCTAAGCCAC 30 .
Reverse primer sequences are: 16SR2600: 50 AATCGTT
GAACAAACGAACCCTTAATAGC 30 and 16SR2
633: 50 CTCCGGTCTGAACTCAGATCACGT 30 . The
numbers in the primer names refer to the 50 position of
each of these oligonucleotides in the complete mitochondrial genome sequence of the thorny skate
(AF106038). Because different primer pairs were used to
amplify different species, the analyses were conducted on
the smallest resulting fragment homologous to all taxa,
which was the 12SF204/16SR2600 region. The PCR
fragments were direct sequenced using dye terminator
chemistry on an ABI 373 or 377 automated sequencer.
The sequence alignments were constructed according
to the procedure suggested by Cassens et al. (2000) in
which alignments without ambiguous positions were
obtained using the software SOAP (L€
oytynoja and
Milinkovitch, 2001), under 25 different settings and
employing weighted matrix, gap penalties from 11 to 19,
by steps of 2, extension penalties from 3 to 11, also by
steps of 2, and 100% conservation. This technique was
inspired by Gatesy et al. (1993) and has the effect of
removing all the positions that are sensitive to variation
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
in alignment position. All alignments are available upon
request from MJS ([email protected]).
Rooting phylogenetic trees is arguably the most
problematic step in phylogenetic reconstruction (Swofford et al., 1996). Outgroups generally increase the noise
in molecular sequence data (Burck, 1999) but are required to polarise the trees. Furthermore, single outgroups are more subject to the affects of long branch
attraction than are multiple paraphyletic taxa. In an
attempt to circumvent these dilemmas, we used several
different outgroup permutations. The more outgroups
utilized, the fewer positions were included in the resulting alignment, because fewer positions were identified as conserved by SOAP. To identify these different
data sets we use numbers that refer to the number of
positions in the resulting alignments. The 2010 data set
used only one outgroup, the chimaera, Hydrolagus colliei (spotted ratfish). The 1963 data set used two outgroups (Silver eye and chimaera; see Table 1 for
217
scientific names). The 1880 data set used four outgroups
(Sea lamprey, Silver eye, Lesser siren, and chimaera;
Table 1). Since chimaeras are widely recognized to be
the sister group to elasmobranchs, we also took advantage of the existence of four partial 16S sequences
available on GenBank for additional chimaera representatives (Rabbit fish, Pacific spookfish, Silver chimaera, and Ninespot chimaera; Table 1). These
sequences were 538–539 bp in length and correspond
with position 2053–2598 of the complete skate mitochondrial genome sequence. We added our ratfish sequence to these four (data set 440). Data set 437 and 420
were obtained by adding ray-finned fish or the ray-finned fish plus lamprey plus tetrapod to the five chimaera
outgroup, respectively. This permutation of outgroups
provided an additional perspective on the batoid/shark
diphyletic hypothesis, with the benefit of a highly subdivided sister group branch, albeit with a relatively
limited number of sequence positions.
Table 1
Accession numbers and taxa common names
Order
Family
Scientific name
Common name #
Accession No.
Ingroup
Squaliformes
Squalidae
Pristiophoriformes
Squatiniformes
Hexanchiformes
Pristiophoridae
Squatinidae
Hexanchidae
Heterodontiformes
Orectolobiformes
Lamniformes
Heterodontidae
Ginglymostomatidae
Odontaspididae
Alopiidae
Lamnidae
Carcharhiniformes
Rajiformes
Myliobatiformes
Triakidae
Scyliorhinidae
Carcharhinidae
Rajidae
Dasyatidae
Squalus acanthias
Centrophorus granulosus
Pristiophorus nudipinnis
Squatina californica
Hexanchus griseus
Heptranchias perlo
Heterodontus francisci
Ginglymostoma cirratum
Carcharias taurus
Alopias vulpinus
Carcharodon carcharias
Isurus oxyrinchus
Isurus paucus
Lamna nasus
Mustelus manazo
Scyliorhinus canicula
Carcharhinus porosus
Raja radiata
Urobatis jamaicensis
Piked dogfish
Gulper shark
Shortnose sawshark
Pacific angelshark
Bluntnose sixgill shark
Sharpnose sevengill shark
Horn shark
Nurse shark
Sand tiger shark
Thresher
Great white shark
Shortfin mako
Longfin mako
Porbeagle
Starspotted smooth-hound
Small-spotted catshark
Smalltail shark
Thorny skate
Yellow stingray
Y18134a
AY147884
AY147885
AY147886
AY147887
AY147888
AY147889
AY147890
AY147891
AY147892
AY147893
AY147894
AY147895
AY147896
AB015962b
Y16067c
AY147897
AF106038d
AY147898
Outgroup
Chimaeriformes
Chimaeridae
Polymixiiformes
Sirenoidea
Petromyzontiformes
Rhinochimaeridae
Polymixiidae
Sirenidae
Petromyzontidae
Hydrolagus colliei
Hydrolagus barbouri
Chimaera phantasma
Chimaera monstrosa
Rhinochimaera pacifica
Polymixia japonica
Siren intermedia
Petromyzon marinus
Spotted ratfish
Ninespot chimaera
Silver chimaera
Rabbit fish
Pacific spookfish
Silver eye (ray-finned fish)
Lesser siren (tetrapods)
Sea lamprey (lamprey)
AY147899
AF288202e;f
AF288201e;f
AF288200e;f
AF288203e;f
AB034826g
Y10946
U11880
Note. # Following FishBase (http://www.fishbase.org) for the ingroup and NCBI (http://www.ncbi.nlm.nih.gov) for the outgroup.
*
Sequences new to this study.
a
Rasmussen and Arnason (1999a).
b
Cao et al. (1998).
c
Delarbre et al. (1998).
d
Rasmussen and Arnason (1999b).
e
Spolsky and Didier, unpub.
f
Partial sequences (16S).
g
Miya and Nishida (2000).
218
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
2.2. Phylogenetic reconstruction
Modeltest software (Posada and Crandall, 1998) was
used to objectively determine the best suited model of
sequence evolution and the accompanying parameter
values for these data (base frequency, instantaneous
rates for each substitution type, shape of the distribution
used to accommodate the among-site rate variation,
proportion of invariant sites). For maximum likelihood
(ML) the TN + G + I (i.e., Tamura-Nei model of sequence evolution with an allowance for invariant sites
and a gamma shape for among-site rate variation) was
used for the 420, 437, 440 bp data sets that encompassed
all five chimaera species. The 1880, 1963, and 2010 data
sets all used GTR + G + I (general time reversible with
gamma, plus invariant sites) model. All ML analyses
used tree-bisection-reconnection (TBR) as the branch
swapping algorithm. ML bootstrapping was performed
on 500 replicates of the data sets. Starting trees were
obtained by stepwise addition, and AsIs sequence addition was used.
Maximum parsimony (MP) used TBR, and the
starting tree was obtained via stepwise addition, with
addition sequences 10X randomized. Minimum evolution (ME) was conducted using both ML and LogDet
distances. Parameters for ML distance calculations were
the same as those for the ML search. Starting trees were
obtained by stepwise addition with sequence addition set
AsIs. ME and MP bootstrap analyses both involved 500
replicates. MP, ME, and ML analyses were all conducted using PAUP* 4b8 (Swofford, 1998).
Bayesian phylogenetic analyses were performed using
the ‘‘metropolis-coupled Markov chain Monte Carlo’’
(MCMCMC) algorithm implemented in Mr.Bayes 2.01
(Huelsenbeck and Ronquist, 2001). The tree space was
explored using four chains: one cold chain and three
incrementally heated ones [heat being set as
1=ð1 þ ði 1ÞT ), where i is the chain number (i.e., 1–4)
and T was set to 0.20 (N.B., for i ¼ 1, heat ¼ 1 ¼ cold
chain)]. We employed a GTR model of sequence evolution allowing a gamma shape of among-site rate variation. Posterior probability distributions were obtained
for the phylogeny and the parameters of the model of
sequence evolution, with the following priors: Branch
length, uniform (0.0, 10.0); Instantaneous rate matrix,
uniform (0.0, 100.0); Base frequencies, Dirichlet (4.0);
gamma shape, uniform (0.0, 10.0). Proposal mechanisms
for the Markov chains included attempted changes to
the rate matrix (2.22%), base frequencies (2.22%),
gamma shape (2.22%), stochastic nearest neighbor interchanges (88.89%) and one worm change to the tree
(4.44%). A random tree was used as seed. We explored
the tree space using 1,000,000 generations, sampled every 100 generations. The number of generations to obtain convergence of the likelihood value for each data
set, and thus the level at which ‘‘burn in’’ was set, were
as follows: data set 2010, 27,100 generations; data set
1963, 39,300 generations, data set 1880, 16,400 generations, data set 440, 7600 generations, data set 437,
13,500 generations, data set 420, 30,400 generations.
Statistical significance, under a likelihood model, of
several phylogenetic hypotheses, was assessed using the
KH test, as implemented in PAUP 4b8. Even though the
validity of this test has recently been debated (Goldman
et al., 2000; Shimodaira and Hasegawa, 1999), our
specific usage, as an a priori hypothesis test, is still regarded as valid (Goldman et al., 2000).
3. Results and discussion
None of our phylogenetic reconstructions support
batoids as a derived shark lineage and all trees support a
much more ancient split between the two groups, with
batoids as the sister group to a clade consisting of all
shark orders (Fig. 1 and Supplementary material may be
found on the journal homepage). In considering the
2010, 1963, and 1880 data sets, which includes 15 different phylogenetic analyses, all of them support this
basal split between batoids and sharks. Bootstrap support for shark monophyly, in reconstructions involving
these three data sets, ranged from 69% to 99%, with 7 of
12 analyses in excess of 90%, and only two less than
75%. Similarly, Bayesian analysis also strongly supported shark monophyly with posterior probability
scores of 0.99 or 1.0. Statistical tests in support of shark
monophyly were not judged to be significant, however,
topologies in which batoids were the sister group to
sawsharks, which is part of the branching arrangement
proposed in the Hypnosqualea hypothesis (Shirai,
1992a,b), were significantly rejected for all three of these
data sets (2010 data set: P ¼ 0:0129; 1963 data set:
P ¼ 0:0177; 1880 data set P ¼ 0:0020). The complete
Hypnosqualea hypothesis, in which a sawshark/batoid
clade is joined by angelsharks was judged even more
unlikely by the KH test (2010 data set: P < 0:0001; 1963
data set, P < 0:0001; 1880 data set, P ¼ 0:0003). Similarly, an angelshark/batoid clade (although we acknowledge that this is not a commonly proposal
phylogenetic hypothesisa) was also significantly rejected
for all three of these data sets (1880, P ¼ 0:0103; 1963,
P ¼ 0:0333; and 2110, P ¼ 0:0311).
Despite the fact our data set lacks representatives of
the putative deepest branching batoid orders (e.g.,
Torpediniformes), it is very unlikely that long branch
attraction has generated the basal position of Batoidea.
First, all tree reconstruction methods recover this position, not only methods such as maximum parsimony
known to be sensitive to this reconstruction artefact.
Second, the batoid-shark branch is far from being the
longest, with the branch leading to nurse shark more
than twice longer (0.1617 vs. 0.0752) and the Hexanch-
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
219
Fig. 1. Maximum likelihood tree obtained using the 2010 bp data set and a single outgroup. Branches are drawn to scale, with the exception of the
branch leading to the Spotted ratfish (reduced almost two times to fit the figure). Frame displays bootstrap values from top to bottom for ML, MP,
ME ml, ME LogDet, and Bayesian probability, respectively; - - in the frame, in place of a bootstrap value, indicates that node did not reflect the
majority consensus branching arrangement after bootstrap analysis with that method.
iformes branch also longer (0.09413). Third, when we
unroot the tree, the branching arrangement remains
unchanged and batoids do not join sawshark or sawshark + angelshark.
In addition to shark monophyly, these three data sets
also tended to support batoid monophyly, although
with somewhat lower bootstrap support; Bayesian support, however, was high and ranged from 0.90 to 1.0.
Although we know of no analysis that challenges batoid
monophyly, these data represent the first time a skate
and ray are included in the same molecular phylogenetic
analysis, and thus represent a further, alternative substantiation of that already accepted view.
The reduced data sets of 440, 437, and 420 bp also
supported the same distinction between batoids and
sharks, with bootstrap support around 60% (range 35–
80, median ¼ 60%) for shark monophyly, and the
Bayesian posterior probability score at 0.99–1.0. These
data sets also exhibited remarkably high support (i.e.,
given the short nature of the sequence) for the monophyly of Holocephali, with bootstrap support ranging
from 88% to 100%, and the Bayesian posterior probability score at 1.0. Within the Holocephali, the surprising result was a strongly supported rabbit fish/Pacific
spookfish clade joined by silver chimaera, then ninespot
chimaera, and lastly the spotted ratfish. This rabbit fish/
220
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
spookfish clade received bootstrap support ranging from
77% to 100%. If these results were corroborated by additional sequences, this would indicate that the family
Chimaeridae is paraphyletic. These reduced data sets
also supported elasmobranch monophyly, chondrichthyan monophyly, and batoid monophyly.
Several superordinal associations within sharks were
strongly supported. Most evident among these was the
association of Squaliformes (Piked dogfish and gulper
shark) with an angelshark/sawshark clade. This is reminiscent of some of the earlier morphological hypotheses
put forward by Shirai (1992b) and de Carvalho (1996), in
which they propose Squaliformes as the sister group to
Hypnosqualea. If we eliminate Batoidea from Hypnosqualea, we are left with an assortment of taxa that would
agree with this arrangement Shirai (1992b, 1996) and de
Carvalho (1996) also went one step further and suggested
the concept of Squalea, which has the six and seven gill
sharks as the sister group to Squaliformes/Hypnosqualea. Again, as long as we disregard the Batoidea from
such a grouping, our trees yield some support for this
association, although not unanimous between methods
and data sets. Interestingly, the sole mode of reproduction for all orders within this modified Squalea (i.e., the
association of Hexanchiformes, Squaliformes, Pristiophoriformes, and Squatiniformes) is yolksac viviparous
(Compagno, 1990). Within sharks, this type of reproduction is also found in Orectolobiformes, and Carcharhiniformes, but not solely. Within batoids it is the
sole mode of reproduction in the sawfishes, guitarfishes,
and electric rays, with Myliobatiformes and Rajiformes
exhibiting uterine and extended oviparous respectively.
Myliobatiformes and Rajiformes are proposed by some
investigators (e.g., McEachran et al., 1996; Nishida,
1990) as the two more derived orders within Batoidea.
Compagno (1990) suggests that the phylogenetic sequence of life history development within elasmobranchs
is: extended oviparous ) retained oviparous ) yolksac
viviparous ) cannibal, placental or uterine viviparous.
In contrast, our diphyletic phylogeny for elasmobranchs,
concomitant with the proposed interordinal phylogenies
for Batoidea, would suggest that yolksac viviparous reproduction is plesiomorphic for elasmobranchs, and that
its existence in Squalea is either symplesiomorphic or
convergent. Resolution of this symplesiomorphic vs.
synapomorphic question regarding characteristics for
Squalea will depend largely on the ultimate phylogenetic
placement for Heterodontiformes (extended oviparous
reproduction) and Orectolobiformes (mostly extended
oviparous, but with some yolksac viviparous), in addition to the obvious necessity for greater phylogenetic
confidence in the proposed superorder Squalea, and the
branching arrangement within Batoidea.
Another superordinal concept put forward by
Compagno (1973, 1977) and revisited by Shirai (1996)
and de Carvalho (1996) is Galeomorphii, which has a
Lamniformes/Carcharhiniformes clade joined by Orectolobiformes (carpet sharks including nurse shark), and
then Heterodontiformes (bullhead or horn sharks). The
only portion of this hypothesis that is clearly evident
from our analyses is the lamniform/carcharhiniform
association. The nurse shark and horn shark placements
are relatively unstable in all our reconstructions and it
seems likely that these two orders will be the most difficult to accurately place. Certainly our data are insufficient to either accept or reject Shirai and de CarvalhoÕs
Galeomorphii arrangement (2110 data set, not statistically rejected, P ¼ 0:6020; 1963 data set, not statistically
rejected, P ¼ 0:6966; 1880 data set, not statistically
supported P ¼ 0:3848), and in our opinion, phylogenetic
placement of Orectolobiformes and Heterodontiformes
will have to await increased taxon representation for
both these orders, concomitant with a good deal more
sequence positions.
Our study provides the first convincing molecular case
for rejection of the Hypnosqualea hypothesis and as
such, suggests that the seven or more (i.e., depending on
the author), putative morphological synapomorphies
used to define this superorder are either symplesiomorphic or the consequence of convergent evolution in the
ancestor to sawsharks/angelsharks and batoids (depending on the resolution of Galeomorphii). Our molecular sequence data strongly suggest that a much more
likely evolutionary scenario for the diversification of
sharks and batoids is an early fundamental split into two
respectively monophyletic lineages. A basal vs. derived
position for Batoidea will have significant implications in
the polarization of morphological and life history characteristics for elasmobranchs and means that much of the
current thinking in this regard needs to be re-evaluated.
Acknowledgments
This work was made possible by an infrastructure
operating grant to M.J.S., from The QueenÕs University
of Belfast and via a Florida Sea Grant and Hai Stiftung
Foundation grant to M.S.S. M.D. was supported by a
studentship award from the Turkish Ministry of Education and Canakkale 18 Mart University. C.J.D. was
supported by a Killam Postdoctoral Fellowship. We
thank the following individuals for their assistance in
the collection of samples: S. Anderson, J. Morrisey, and
L. Natanson. C.J.D. is particularly grateful to W. Ford
Doolittle for his generous provision of laboratory
facilities.
References
Arnason, U., Gulberg, A., Janke, A., 2001. Molecular phylogenetics of
gnathostomous (jawed) fishes: old bone, new cartilage. Zool. Scr.
30, 249–255.
C.J. Douady et al. / Molecular Phylogenetics and Evolution 26 (2003) 215–221
Bigelow, H.W., Schroeder, W.C., 1948. Sharks. In: Tee-Van, J.,
Breder, C.M., Hildebrand, S.F., Parr, A.E., Schroeder, W.C.
(Eds.), Fishes of Western North Atlantic, Part 1. Yale University,
New Haven, CT, pp. 59–576.
Bigelow, H.W., Schroeder, W.C., 1953. Sawfishes, guitarfishes, skates,
and rays. In: Tee-Van, J., Breder, C.M., Hildebrand, S.F., Parr,
A.E., Schroeder, W.C. (Eds.), Fishes of Western North Atlantic,
Part 2. Yale University, New Haven, CT, pp. 1–514.
Burck, A., 1999. A chronicle of kangaroo evolution: phylogenetic
relationships among Macropodoidea based on mitochondrial
rRNA genes with implications for the evolution of morphological
characters. Ph.D. Dissertation, University of California, Riverside.
Cao, Y., Waddell, P.J., Okada, N., Hasegawa, M., 1998. The complete
mitochondrial DNA sequence of the shark Mustelus manazo:
evaluating rooting contradictions to living bony vertebrates. Mol.
Biol. Evol. 15, 1637–1646.
Cassens, I., Vicario, S., Waddell, V.G., Balchowsky, H., Van Belle, D.,
Ding, W., Fan, C., Mohan, R.S., Simoes-Lopes, P.C., Bastida, R.,
Meyer, A., Stanhope, M.J., Milinkovitch, M.C., 2000. Independent
adaptation to riverine habitats allowed survival of ancient cetacean
lineages. Proc. Natl. Acad. Sci. USA 97, 11343–11347.
Compagno, L.J.V., 1973. Interrelationships of living elasmobranchs.
Zool. J. Linn. Soc. 53, 15–61.
Compagno, L.J.V., 1977. Phylogenetic relationships of living sharks
and rays. Am. Zool. 17, 303–322.
Compagno, L.J.V., 1990. Alternative life-history styles of cartilaginous
fishes in time and space. Environ. Biol. Fish 28, 33–75.
de Carvalho, M.R., 1996. Higher-level elasmobranch phylogeny, basal
squaleans, and paraphyly. In: Stassny, M.L.J., Parenti, L.R.,
Johnson, G.D. (Eds.), Interrelationships of Fishes. Academic
Press, San Diego, pp. 35–62.
de Carvalho, M.R., Maisey, J.G., 1996. Phylogenetic relationships of
the Late Jurassic shark Protospinax Woodward 1919 (Chondrichthyes: Elasmobranchii). In: Arratia, G., Viohl, G. (Eds.), Mesozoic
Fishes: Systematics and Paleoecology. Verlag Dr Friedrich Pfeil,
Munich, pp. 9–46.
Delarbre, C., Spruyt, N., Delmarre, C., Gallut, C., Barriel, V., Janvier,
P., Laudet, V., Gachelin, G., 1998. The complete nucleotide
sequence of the mitochondrial DNA of the dogfish, Scyliorhinus
canicula. Genetics 150, 331–344.
Deets, G.B., 1994. Copepod-chondricthyan coevolution: a cladistic
consideration. Ph.D. Dissertation, University of British Columbia,
Vancouver.
Dunn, K.A., Morrissey, J.F., 1995. Molecular phylogeny of elasmobranchs. Copeia 3, 526–531.
Gatesy, J., DeSalle, R., Wheeler, W., 1993. Alignment-ambiguous
nucleotide sites and the exclusion of systematic data. Mol.
Phylogenet. Evol. 2, 152–157.
Goldman, N., Anderson, J.P., Rodrigo, A.G., 2000. Likelihood-based
tests of topologies in phylogenetics. Syst. Biol. 49, 652–670.
Huelsenbeck, J.P., Ronquist, F.R., 2001. MRBAYES: Bayesian
inference of phylogenetic trees. Bioinformatics 17, 754–755.
Kitamura, T., Takemura, A., Watabe, S., Taniuchi, T., Shimizu, M.,
1996. Molecular phylogeny of the sharks and rays of superorder
221
Squalea based on mitochondrial cytochrome b gene. Fisheries Sci.
62, 340–343.
L€
oytynoja, A., Milinkovitch, M.C., 2001. SOAP, cleaning
multiple alignments from unstable blocks. Bioinformatics 17,
573–574.
McEachran, J.D., Dunn, K.A., Miyake, T., 1996. Interrelationships
within the batoid fishes (Chondrichthyes: Batoidea). In: Stiassny,
M.L.J., Parenti, L.R., Johnson, G.D. (Eds.), Interrelationships of
Fishes. Academic Press, San Diego, pp. 63–84.
Miya, M., Nishida, M., 2000. Use of mitogenomic information in
teleostean molecular phylogenetics: a tree-based exploration under
the maximum-parsimony optimality criterion. Mol. Phylogenet.
Evol. 17, 437–455.
Murphy, W.J., Eizirik, E., OÕBrien, S.J., Madsen, O., Scally, M.,
Douady, C.J., Teeling, E., Ryder, O.A., Stanhope, M.J., de Jong,
W.W., Springer, M.S., 2001. Resolution of the early placental
mammal radiation using Bayesian phylogenetics. Science 294,
2348–2351.
Nishida, K., 1990. Phylogeny of the suborder Myliobatoidei. Mem.
Fac. Fish. Hokkaido Univ. 37, 259–298.
Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of
DNA substitution. Bioinformatics 14, 817–818.
Rasmussen, A.S., Arnason, U., 1999a. Phylogenetic studies of
complete mitochondrial DNA molecules place cartilaginous
fishes within the tree of bony fishes. J. Mol. Evol. 48, 118–
123.
Rasmussen, A.S., Arnason, U., 1999b. Molecular studies suggest that
cartilaginous fishes have a terminal position in the piscine tree.
Proc. Natl. Acad. Sci. USA 96, 2177–2182.
Seret, B., 1986. Classification et phylogenese des chondrichthyens.
Oceanis 12, 161–180.
Shimodaira, H., Hasegawa, M., 1999. Multiple comparisons of loglikelihoods with applications to phylogenetic inference. Mol. Biol.
Evol. 16, 1114–1116.
Shirai, S., 1992a. Phylogenetic relationships of the angel sharks, with
comments on elasmobranch phylogeny (Chondrichthyes: Squatinidae). Copeia 2, 505–518.
Shirai, S., 1992b. Squalean Phylogeny: A New Framework of
ÔSqualoidÕ Sharks and Related Taxa. Hokkaido University Press,
Sapporo.
Shirai, S., 1996. Phylogenetic interrelationships of Neoselachians
(Chondrichthyes: Euselachii). In: Stiassny, M.L.J., Parenti, L.R.,
Johnson, G.D. (Eds.), Interrelationships of Fishes. Academic
Press, San Diego, pp. 9–34.
Springer, M.S., DeBry, R.W., Douady, C., Amrine, H.M., Madsen,
O., de Jong, W.W., Stanhope, M.J., 2001. Mitochondrial versus
nuclear gene sequences in deep-level mammalian phylogeny
reconstruction. Mol. Biol. Evol. 18, 132–143.
Swofford, D.L., 1998. Paup*: Phylogenetic Analysis using Parsimony
(*and other Methods). Sinauer, Sunderland, MA.
Swofford, D.L., Olsen, G.P., Waddell, P.J., Hillis, D.M., 1996.
Phylogenetic inference. In: Hillis, D.M., Moritz, C., Mable, B.K.
(Eds.), Molecular Systematics. Sinauer Associates, Sunderland, pp.
407–514.