M83, U
Accepted 31 May 2013
Available online xxxx
Keywords:
Mellita
Leodia
Phylogeny
Mel
of the genus we sequenced parts of the mitochondrial cytochrome oxidase I and of 16S rRNA as well
as part of the nuclear 28S rRNA gene from a total of 185 specimens of all ten described morphospecies
shallow water regular sea urchins, heart urchins and sand dollars,
specimen from the midst of the range of M. longi?ssa Michelin,
1858. Mayr?s (1954) model of allopatric speciation in echinoids
has been supported by molecular phylogenies of regular sea urch-
ins (Lessios et al., 1999, 2001, 2003, 2012; McCartney et al., 2000;
Zigler and Lessios, 2004; Palumbi and Lessios, 2005), but to-date
this hypothesis has not been tested with a molecular phylogeny
of a sand dollar genus.
Atlantic and ?ve from the eastern Paci?c. Harold and Telford
e genus and con-
hese have
rold and T
0 (type lo
Beaufort, North Carolina, USA) is distributed along the eas
of North America, from Nantucket, Massachusetts to Fort L
dale, Florida, and M. tenuis Clark, 1940 (type locality: Sanibel Is-
land, Florida, USA) from Louisiana to west Florida in the Gulf of
Mexico. Harold and Telford (1990) synonymised M. lata Clark,
1940 (type locality: Puerto Lim?n, Costa Rica) and M. latiambulacra
Clark, 1940 (type locality: Cuman?, Venezuela) with M. quinquies-
perforata (Leske, 1778) (type locality: Veracruz, Mexico), and gave
M. quinquiesperforata?s geographic range as Louisiana to Brazil,
including Central America and the Greater Antilles.
? Corresponding author. Fax: +507 212 8790.
Molecular Phylogenetics and Evolution xxx (2013) xxx?xxx
Contents lists available at
ne
.e lE-mail address: coppardse@gmail.com (S.E. Coppard).including those of the genus Mellita L. Agassiz, 1841. However, one
species of Mellita, M. grantii Mortensen, 1948, proved problematic
for Mayr (1954) and he chose to ignore it in his analysis, stating
that Mortensen (1948) had described M. grantii based on a single
(1990) conducted a morphological analysis of th
cluded that seven species were valid. Three of t
non-overlapping distributions in the Atlantic (Ha
1990). Mellita isometra Harold and Telford, 1991055-7903/$ - see front matter  2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ympev.2013.05.028
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the sand dollar genus Mellita: Cryptic speciation along the coasts of the Am
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028largely
elford,
cality:
t coast
auder-The interplay of factors that result in speciation, habitat special-
isation and geographic distributions of marine organisms with
planktonic larvae remains poorly understood. Mayr (1954)
strongly advocated that speciation in shallow water echinoids oc-
curs allopatrically. This conclusion was based on the geographic
distributions of morphospecies from sixteen genera of tropical
cord. This record dates from the Early Pliocene, and is helpful in
dating splitting events between lineages (Mooi and Peterson,
2000). The genus is geographically restricted to the Americas (Mooi
et al., 2000), which avoids complications that can arise from spe-
cies invasions out of other geographic regions.
Ten morphospecies of Mellita have been described, ?ve from theDivergence
Speciation
1. Introductionfrom 31 localities. Our analyses revealed the presence of eleven species, including six cryptic species.
Sequences of ?ve morphospecies do not constitute monophyletic molecular units and thus probably rep-
resent ecophenotypic variants. The fossil-calibrated phylogeny showed that the ancestor of Mellita
diverged into a Paci?c lineage and an Atlantic + Paci?c lineage close to the Miocene/Pliocene boundary.
Atlantic M. tenuis, M. quinquiesperforata and two undescribed species of Mellita have non-overlapping dis-
tributions. Paci?c Mellita consist of two highly divergent lineages that became established at different
times, resulting in sympatric M. longi?ssa and M. notabilis. Judged by modern day ranges, not all diver-
gence in this genus conforms to an allopatric speciation model. Only the separation of M. quinquiesperfo-
rata from M. notabilis is clearly due to vicariance as the result of the completion of the Isthmus of Panama.
The molecular phylogeny calibrated on fossil evidence estimated this event as having occurred 3 Ma,
thus providing evidence that, contrary to a recent proposal, the central American Isthmus was not com-
pleted until this date.
 2013 Elsevier Inc. All rights reserved.
Sand dollars in the genus Mellita are an ideal group in which to
trace phylogeographic phenomena, as they have a rich fossil re-Received 1 February 2013
Revised 20 May 2013
tropical and subtropical regions on the two coasts of the Americas. To reconstruct the phylogeographyPhylogeography of the sand dollar genus
along the coasts of the Americas
Simon E. Coppard a,?, Kirk S. Zigler a,b, H.A. Lessios a
a Smithsonian Tropical Research Institute, Box 0843-03092, Balboa, Panama
bDepartment of Biology, Sewanee: The University of the South, Sewanee, Tennessee 373
a r t i c l e i n f o
Article history:
a b s t r a c t
Sand dollars of the genus
Molecular Phyloge
journal homepage: wwwellita: Cryptic speciation
nited States
lita are members of the sandy shallow-water fauna. The genus ranges in
SciVerse ScienceDirect
tics and Evolution
sevier .com/ locate /ympevericas.
In the Paci?c, Harold and Telford (1990) recognised four spe-
cies: M. longi?ssa (type locality: unknown), M. notabilis Clark,
1947 (type locality: unknown); M. grantii (type locality: San Felipe,
Mexico); M. kanakof? Durham, 1961 (type locality: Upper Pleisto-
cene, Newport Beach, California, USA). They synonymised M. edu-
ardobarrosoi Caso, 1980 (type locality: Acapulco, Mexico) with M.
notabilis, ?nding them to be morphologically identical. Paci?c spe-
cies have extensively overlapping distributions, from the west
coast of Baja California, Mexico to Panama; the range of M. longif-
issa extends to the Galapagos Islands (Isla Santa Maria (Charles Is-
land)). According to Mortensen (1948), M. grantii was only known
from the Gulf of California, but Harold and Telford (1990) extended
its range to Panama. Lessios (2005) suggested that specimens from
the Gulf of San Miguel, Panama, belonged to this species. Mellita
grantii is morphologically very similar to juvenile M. longi?ssa,
which has led to considerable confusion in identi?cation.
In Harold and Telford?s (1990) cladistic analysis based on mor-
phological characters the Atlantic species of Mellita do not form a
monophyletic group. Mellita isometra and M. tenuis are sister to
each other, whereas M. quinquiesperforata is sister to a group con-
taining all the Paci?c species. From their analysis Harold and Tel-
ford (1990) stated that they were unable to determine whether
speciation in the Paci?c clade had occurred allopatrically.
Experimental evidence suggests that the larvae of M. quinquies-
perforata can settle in as little as seven days if they encounter
favourable conditions, but are also able to remain in the plankton
for up to four weeks (Caldwell, 1972). As Mellita specialise in living
in terrigenous (siliceous) sands (Telford and Mooi, 1986), this ?ex-
ibility in timing of settlement is vital for successful recruitment
and is likely to be mediated by a chemosensitive response to either
suitable terrigenous sands or adult conspeci?cs.
In this study we combine mitochondrial and nuclear gene se-
quences to reconstruct the phylogeny of Mellita. We attempt to an-
swer the following questions: (1) Are species as recognised by
morphology valid? (2) When did speciation events occur? (3) Does
the dating of the phylogeny from fossils concur with the dating
from vicariant events? (4) What were the physical barriers that re-
sulted in speciation? (5) Does speciation in Mellita conform to the
allopatric model?
2. Materials and methods
Specimens representing all described extant species of Mellita
were collected throughout the range of the genus (Fig. 1). Collec-
tion sites included the type localities of M. eduardobarrosoi and
M. isometra. Samples included members of the morphological vari-
ants of M. quinquiesperforata. Specimens of Leodia sexiesperforata
Leske, 1778 and of Encope grandis L. Agassiz, 1841 were also col-
lected for use as outgroups. The Aristotle?s lantern was extracted
from each specimen and preserved in 95% Ethyl Alcohol or high
salt Dimethyl Sulfoxide buffer. Fossil Mellita, including representa-
tives of both extant and extinct species, were examined from the
collections of the Florida Museum of Natural History (FLMNH),
Humbolt State University?s Natural History Museum (HSU), and
the Natural History Museum of Los Angeles County (LACM).
2.1. DNA extraction, sequencing and alignment
Genomic DNA was extracted from the lantern muscle of 185
specimens of Mellita, 9 specimens of Leodia and 3 specimens of
Encope using a DNeasy tissue kit (Qiagen). We ampli?ed three
r to
79W
uita
N, 7
: Be
P:
798
(8.
am
2 S.E. Coppard et al. /Molecular Phylogenetics and Evolution xxx (2013) xxx?xxxFig. 1. Collection localities of Mellita, Leodia and Encope used in this study. Letters refe
Carolina, USA (34.6934N, 76.6981W); B: Fort Pierce, Florida, USA (27.4380N, 80.27
(27.8339N, 97.0610W); E: Playa Lim?n, Costa Rica (9.9957N, 83.0280W); F: Playa Cah
H: Palmas Bellas, Colon, Panama (9.2322N, 80.0878W); I: Portobelo, Panama (9.5509
Colombia (11.2592N, 74.2050W); L: Cocos Bay, Trinidad (10.4152N, 61.0230W); M
(28.9492N, 113.5576W); O: Bahia de Kino, Sonora, Mexico (28.8189N, 111.9392W);
Mazatlan, Mexico (23.1807N, 106.3926W); R: Playa Azul, Michoacan, Mexico (17.9
Puerto Armuelles, Panama (8.2223N, 82.8588W); U: Playa Las Lajas, Chiriqui, Panama
Peninsula, Panama (7.3753N, 80.2702W); X: Playa Venado, Azuero Peninsula, Pan
80.9445W); Z: Punta Mala, Azuero Peninsula, Panama (7.4679N, 80.0007W); C: Punt
(8.5516N, 79.8654W); H: Punta Chame, Panama (8.6446N, 79.7006W); K: Bahia Sola
78.2727W); R: Play?n Chico, San Blas, Panama (9.3012N, 78.233W).
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028localities, numbers to sample size, and colours to morphospecies. A: Beaufort, North
); C: Mullet Key, Florida, USA (27.6216N, 82.7385W); D: Port Aransas, Texas, USA
, Costa Rica (9.7356N, 82.8346W); G: Bocas del Toro, Panama (9.3458N, 82.2519W);
9.6675W); J: Ustupo Island, San Blas, Panama (9.1374N, 77.9288W); K: Santa Marta,
ssa Beach, Paraiba, Brazil (7.0656N, 34.8249W); N: Bahia de Los Angeles, Mexico
Malcomb, Baja California Sur, Mexico (26.7126N, 113.2670W); Q: Isla de la Piedra,
N, 102.3497W); S: Playa Encantada, Acapulco, Mexico (16.6982N, 99.6652W); T:
1618N, 81.8598W); V: Chiriqui, Panama (7.9422N, 81.6503W); W: Isla Ca?a, Azuero
a (7.5146N, 80.1280W); Y: Playa Cambutal, Azuero Peninsula, Panama (7.0067N,
a Blanca, Santa Elena, Ecuador (2.1517N, 80.7905W); D: Playa Gorgona, Panama
no, Colombia (6.2304N, 77.4029W); P: Gulfo de San Miguel, Panama (8.3996N,
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
netiregions of mitochondrial DNA (mtDNA) and one of a nuclear gene.
The 50 region of cytochorome c oxidase subunit I (COI) was ampli-
?ed using the HCOI and LCOI primers of Folmer et al. (1994), the 30
region of COI using the primers COIa and COIf of Lessios et al.
(2001), and the 16S rRNA using the 16Sar and 16Sbr primers of
Kessing et al. (1989). Each reaction contained 0.2?0.5 ll of ex-
tracted genomic DNA (approximately 10?15 ng), 12.0 ll of nucle-
ase free H20 (adjusted to 12.3 ll when using less genomic DNA),
5.0 ll GoTaq Flexi Buffer (5), 2.5 ll MgCl2 (25 mM), 2.5 ll dNTPs
(8 mM), 1.25 ll (10 lM) of each forward and reverse primer, and
0.6 units of Flexi-GoTaq polymerase (Promega). These were
ampli?ed using the following protocol: (1) 96 C for 10 s, (2)
94 C for 30 s, (3) 50 C for 45 s, (4) 72 C for 1 min for 39 cycles,
and a ?nal extension at 72 C for 5 min. The 50 end of nuclear
28S rRNA was ampli?ed using a HotStartTaq PCR ampli?cation
kit (Qiagen) with the primers and protocol of Littlewood and
Smith (1995). PCR products of 28S were cloned using Promega
pGEM-T Easy kits to avoid sequences with heterozygous single-
nucleotide polymorphisms (double peaks on chromatograms).
One clone was cycle-sequenced for each specimen using Promega
M13 and M13R primers, before being sequenced in one direction
using the M13 primer.
After puri?cation in Sephadex columns, ampli?cation with the
same primers, and labelling with Applied Biosystems (ABI) Prism
BigDye Terminators, nucleotides were sequenced in both direc-
tions using an ABI 3130 XL automatic sequencer. Pairwise se-
quence alignments were performed in MacClade (Maddison and
Maddison, 2005). The two sequenced COI regions overlapped by
22 base pairs (bp) and after the removal of the primer regions pro-
duced a contiguous sequence of 1236 bp with no indels, covering
approximately 80% of the complete coding sequence of the gene.
571 bp of 16S rRNA and 1137 bp of 28S were sequenced including
several 1 bp indels. All sequences have been deposited in GenBank
with the Accession numbers KF204670?KF204860 for COI,
KF204861?KF205051 for 16S and KF205052?KF205242 for 28S.
2.2. Phylogenetic reconstruction
jMODELTEST v. 2.1.1 (Darriba et al., 2012) was used to determine
the best model of molecular evolution for each gene based on
the AIC criterion (Akaike, 1974). For COI the general time-revers-
ible model (Tavare, 1986) was selected, with a gamma distribution
shape parameter (a) of 0.1310 (GTR + G). The Hasegawa, Kishino
and Yano model (Hasegawa et al., 1985) was selected for 16S
(HKY + I + G; I = 0.4520 and a = 0.2350), and a transition/transver-
sion model TIM1 + I was suggested for 28S, where I = 0.9350. As
different models were selected for each gene, partitions were used
in the phylogenetic analysis of the concatenated data. Saturation
tests for each gene and for the concatenated data were conducted
using the software package DAMBE (V.5.3.00) (Xia et al., 2003; Xia
and Lemey, 2009). For COI, the index of substitution saturation ISS
was calculated for all sites using all codon positions, and separately
using just the third position, which is prone to saturation. For the
ribosomal genes, sites with indels were not included in the analy-
sis, because they can reduce the sensitivity of this method (Xia and
Lemey, 2009). In all tests ISS was signi?cantly smaller than the crit-
ical index of substitution saturation (ISS.c) under the assumption of
either a symmetrical (ISS.cSym) or extremely asymmetrical (ISS.cAsym)
tree (p = 0 for COI, 16S, 28S and the concatenated data, p = 0.02 for
the third codon position in COI).
Bayesian phylogenetic analysis was carried out on the concate-
nated data using MRBAYES v. 3.2.1. (Ronquist and Huelsenbeck,
2003). Each gene was analysed as a separate partition with the
S.E. Coppard et al. /Molecular Phylogemodel suggested by jMODELTEST with parameters unlinked across
partitions. A haplotype of Encope grandis was randomly selected
as an outgroup (only one outgroup is permitted in MRBAYES). The
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028heating parameter T of the runs was 0.15. The analysis was started
with Dirichlet priors for rates and nucleotide frequencies and was
run for 60 million generations, sampling every 1000th tree from
two runs. Convergence was assessed according to the average stan-
dard deviation of split frequencies <0.01 and potential scale reduc-
tion factor (Gelman and Rubin, 1992) reaching 1.00 for all
parameters. The runs were also visually checked in Tracer v1.5
(Rambaut and Drummond, 2007). The ?rst 25% of trees were dis-
carded from each run as burn-in, and a 50% majority rule tree
was constructed from 90,002 trees. Clades with less than 85% sup-
port were collapsed.
Partitioned maximum likelihood analysis was also carried out
in GARLI V.2.0 (Zwickl, 2006) using the model suggested by jMODEL-
TEST for each gene. Five replicate runs, each of two million itera-
tions, were conducted. Node support values were calculated in
GARLI based on 400 bootstrap replicates, and the bootstrapped
consensus tree was calculated in PAUP* (Swofford, 2002).
Genetic distances between clades, based on the appropriate
models, were estimated for each gene by calculating the mean of
all pairwise comparisons between species. When a clade contained
more than one subclade, its mean distance from its sister clade was
calculated as the average distance between clades in each group.
FST values were calculated between members of described mor-
phospecies when the phylogeny indicated that they did not belong
to separate clades in order to determinewhether there was reduced
gene ?ow between them. For this analysis, all samples of each
morphospecies were pooled. FST values were calculated using the
concatenated 2944 bp sequences and Tamura and Nei (1993) dis-
tances in Arlequin v. 3.1.5.3 (Excof?er and Lischer, 2010). The prob-
ability that the FST values could be due to chance was estimated
with 10,000 reshuf?ings of sequences between morphospecies.
2.3. Timing of divergence
The COI and combined set of data were analysed in BEAST
(v1.7.1) (Drummond et al., 2012). To determine time of most re-
cent common ancestor (TMRCA), we analysed COI and the com-
bined set of data independently using an uncorrelated lognormal
relaxed clock (where the rate of substitutions per site per unit time
is estimated) with the substitution models selected by jMODELTEST
and the Yule speciation process. All analyses were run for 150 mil-
lion generations with parameters logged every 1500 iterations and
convergence assessed using Tracer. The ?rst 15000 samples (10%)
were discarded as burn-in before trees were viewed in FigTree
v1.3.1. Molecular clocks were calibrated in two independent runs,
one using the 3.1?2.8 million year ago (Ma) date range for the ?nal
closure of the isthmus (Coates and Obando, 1996; Coates et al.,
2003), another using dates of fossils to constrain estimates of node
age. An additional run used both the Isthmus of Panama and the
fossils as calibration points. The oldest fossil assigned to Mellita,
M. caroliniana (Ravenel, 1841) (FLMNH UF80503) from the York-
town Formation of the middle Pliocene (Krantz, 1991), is problem-
atic, as it is morphologically unlike all other Mellita; it may belong
to either an ancestor of Mellita or to a different genus. Mellita aclin-
ensis Kier, 1963 (FLMNH UF40428) from the Late Pliocene Piacenz-
ian (3.60?2.59 Ma) Tamiami Formation (Lyons, 1991; Jones et al.,
1991; Mooi and Peterson, 2000) is morphologically and geograph-
ically (Florida) very close to extant M. tenuis (some populations of
this extant species also have six lunules (Cerame-Vivas and Gray,
1964)). We therefore constrained the minimum estimate of node
age for the split of M. tenuis from the Atlantic + Paci?c lineage of
Mellita to 3.60?2. 59 Ma. This calibration point was used in con-
junction with minimum node ages of extant Mellita that have well
dated fossils. These were fossils of M. grantii (HSU, NHM943) and
cs and Evolution xxx (2013) xxx?xxx 3M. tenuis (FLMNH UF14478) from the Gelasian stage of the Pleisto-
cene (2.59?1.8 Ma), and of M. kanakof? (LACM 1121, 1122) from
the Tarantian stage of the Pleistocene (0.13?0.01 Ma).
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
3. Results
3.1. Phylogeny
Bayesian inference (BI) and Maximum likelihood (ML) analyses
of the 132 unique haplotypes of Mellita, 9 of Leodia and 1 of
E. grandis produced congruent phylogenies for all well supported
clades and subclades (support >85%), with only slight differences
in weakly supported subclades within species (Figs. 2a and 2b).
In the phylogeny based on the concatenated data rooted on
Encope, Leodia was sister to Mellita with a genetic distance from
d w
4 S.E. Coppard et al. /Molecular Phylogenetics and Evolution xxx (2013) xxx?xxxFig. 2a. Phylogeny of Mellita using concatenated COI, 16S and 28S data reconstructe
Bayesian (?rst number next to node) and Maximum Likelihood (second number) re
indistinguishable haplotypes, scale bar re?ects number of changes per site. Names next to
the pictures are our interpretation as to species af?liation according to the molecular p
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028ith MRBAYES and rooted on Encope grandis. Clade credibility values >85% in both the
construction are shown. Numbers after locality names indicate individuals with
terminal branches indicate the morphology of the specimens, names to the right of
hylogeny.
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
netiS.E. Coppard et al. /Molecular Phylogethe latter genus of 47.22% in COI. The ?rst split within Mellita sepa-
rated a Paci?c clade consisting ofM. grantii andM. longi?ssa from all
other species, with 39.76% divergence in COI from all other extant
Mellita (Table 1). Within this clade, M. grantii, from both sides of
the Baja California peninsula formed a group, which had 10.48%
divergence in COI from a clade containing three subclades of
M. longi?ssa. Mellita longi?ssa from the Paci?c side of Colombia
(M. sp. 3) was sister to the two other clades of M. longi?ssa. Individ-
uals of this molecular clade had no obvious morphological differ-
ences from M. longi?ssa but its COI sequences were different by
5.60%. Specimens of M. longi?ssa from Panama and from the mouth
of the Gulf of California formed a separate clade. Its sister clade,
M. sp. 4, was 3.97% different in COI and was composed exclusively
of sequences from individuals collected in Panama. Mellita sp. 4
Fig. 2b. Phylogeny of Mellita
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028cs and Evolution xxx (2013) xxx?xxx 5contained two well supported subclades, each of which included
specimens that weremorphologically typical ofM. longi?ssa, aswell
as specimens fromPanama identi?ed by Lessios (2005) asM. grantii.
The sister clade to M. longi?ssa-M. grantii incorporated M. isom-
etra from the Atlantic Coast of the United States and M. tenuis from
the Gulf of Mexico. This clade included a well-supported subclade
of M. tenuis from Fort Aransas, Texas. However, other specimens of
M. tenuis were intermixed with M. isometra and did not form sep-
arate clades as they would if they were separate species.
Sister to the M. tenuis-isometra clade (18.70% divergent from it in
COI) was a polytomy that included Atlantic and Paci?c lineages, the
separation of which corresponds with the most recent transisthmian
divergence in the genus.Within this polytomy, mean genetic distance
in COI between Atlantic and Paci?c species was also 18.70%.
continued from Fig. 2a.
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
usin
ess.
16S
(%)
10.6
10.0
10.9
9.1
7.7
8.4
3.1
netiTable 1
Mean difference and divergence times between clades and species of Mellita calculated
COI and the concatenated set of genes with a relaxed clock and the Yule speciation proc
of the Panama Isthmus, and the combined fossil/isthmus calibration.
Divergence COI
(%)
Leodia from all Mellita 47.22
Leodia from M. grant. + M. long. + M. sp. 3 and 4 45.56
Leodia from M. tenuis + M. quinq. + M. notab. + M. sp. 1, 2, 5 and 6 42.29
M. grant. + M. long. + M. sp. 3 and 4 from M. tenuis + M. quinq. + M.
notab. + M. sp. 1, 2, 5 and 6
39.76
M. tenuis from M. quinq. + M. notab. + M. sp. 1, 2, 5 and 6 18.70
M. quinq. + M. sp. 1 and 2 from M. notab. + M. sp. 5 and 6 18.70
M. sp. 2 from M. quinq. + M. sp. 1 12.23
6 S.E. Coppard et al. /Molecular PhylogeThe Atlantic lineages were formed of specimens with a range of
morphologies that have previously been included in M. quinquies-
perforata. Mellita sp. 2 from Trinidad and Brazil formed a well-sup-
ported clade with 12.23% divergence in COI from a lineage
containing M. quinquiesperforata from Costa Rica and Panama as
well as M. sp. 1 from Santa Marta, Colombia. Between the Colom-
bian and the Central American subclades there was 8.64% diver-
gence in COI. Specimens from Costa Rica and Panama had tests
that were particularly broad relative to their length, and are there-
fore representative of M. quinquiesperforata as originally described
(Klein, 1734 (Pre-Linnean) and Leske, 1778).
Within the Paci?c lineage, three distinct subclades formed a
polytomy. Mellita notabilis included adult specimens from Mexico
and Panama with typical M. notabilis morphology, as well as M.
eduardobarrosoi from Mexico and M. kanakof? from Panama. This
M. grant. from M. long. + M. sp. 3 nd 4 10.48 2.6
M. quinq. from M. sp. 1 8.64 3.1
M. notab. from M. sp. 5 13.12 2.7
M. notab. from M. sp. 6 8.64 2.6
M. sp. 5 from M. sp. 6 11.33 3.5
M. sp. 3 from M. lon. + M. sp. 4 5.60 1.5
M. long. from M. sp. 4 3.97 0.7
Fig. 3. Timing of cladogenesis based on concatenated COI, 16S, and 28S data, as derived
Panama Isthmus. Ages of stages and epoch series are based on International Commission o
light blue. Colours of clades indicate geographic range (red = Atlantic and Caribbean, blu
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028g models suggested by jMODELTEST. Timing of divergence was calculated in BEAST using
Separate molecular clock calibrations were based on the fossil record, the ?nal closure
28S
(%)
COI
fossil
(Ma)
COI
isthmus +
fossil (Ma)
COI + 16S + 28S,
isthmus (Ma)
COI + 16S + 28S,
fossil (Ma)
COI + 16S + 28S
isthmus + fossil
(Ma)
5 2.05 5.96 6.02 5.95 6.14 6.01
9 2.13 5.96 6.02 5.95 6.14 6.01
7 2.01 5.96 6.02 5.95 6.14 6.01
8 0.64 5.23 5.61 5.30 5.46 5.42
9 0.29 3.68 3.66 3.69 3.83 3.76
6 0.17 2.92 3.03 3.09 3.21 3.18
5 0.17 2.92 3.03 3.09 3.21 3.18
cs and Evolution xxx (2013) xxx?xxxclade had 13.12% divergence in COI from M. sp. 5 from Ecuador,
and 8.64% divergence in COI from M. sp. 6, which consisted of
two specimens from Mazatlan, Mexico. The M. sp. 5 clade included
members with some test characters of both M. notabilis and M.
kanakof?i, whereas the M. sp. 6 was morphologically closer to M.
kanakof?i. Genetic distance in COI between M. sp. 5 and M. sp. 6
was 11.33%.
The molecular phylogeny revealed which morphospecies corre-
sponded with monophyletic molecular clades, but also suggested
the existence of cryptic species.
An FST value of 0.09 between M. tenuis and M. isometra,
(p > 0.05), indicated that the two putative species on either side
of the Florida Peninsula interbreed freely. FST of 0.15 between M.
notabilis (including M. eduardobarrosoi) and M. kanakof? (p < 0.01)
suggests that there may be some barriers to gene ?ow between
the two sympatric morphospecies in the eastern Paci?c.
4 0.13 2.34 2.40 3.03 3.12 3.06
2 0.09 1.78 1.81 2.09 2.23 2.12
4 0.02 1.59 1.63 1.82 1.88 2.03
3 0.01 1.59 1.63 1.82 1.88 2.03
2 0.21 1.59 1.63 1.82 1.88 2.03
0 0.05 1.47 1.50 2.32 2.38 2.34
6 0.09 1.15 1.18 1.81 1.87 1.84
from analysis on BEAST calibrated using the fossil record and the ?nal closure of the
n Stratigraphy stratigraphic chart 2012. Error bars estimated by BEAST are shown in
e = eastern Paci?c, green = Gulf of California).
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
that the completion of the Panama Isthmus 3 Ma separated M.
netiOur analyses revealed that Leodia and the ancestor of Mellita di-
verged in the Late Miocene (6.0 Ma). Separation between Leodia
and Mellita was accompanied by niche partitioning, with extant
Leodia living only in biogenic, carbonate sands and Mellita special-
ising in terrigenous siliceous sands. Such specialisation by Leodia
may have evolved in concert with increased carbonate deposition
in the Caribbean in the Late Miocene/Early Pliocene as the result
of the post-Miocene proliferation of coral reefs in the Caribbean
(Johnson et al., 2007, 2008; O?Dea et al., 2007; Smith and Jackson,
2009).
The ancestor of Mellita split into a Paci?c lineage and an Atlantic
lineage around the Miocene/Pliocene boundary. Origination ratesquinquiesperforata from its sister lineage in the Paci?c produced a
mean molecular divergence rate of 6.23% per million years
(my1) in COI, of 2.82% my1 in 16S and of 0.06% my1 in the nu-
clear gene 28S. This resulted in a per-lineage substitution rate of
3.12% my1 in COI, 1.41% my1 in 16S and 0.03% my1 in 28S.
When divergence was calibrated with fossils, it produced similar
estimates of the timing of splitting as those obtained from the cal-
ibration based on the ?nal closure of the Panama Isthmus (Table 1).
The combined fossil-isthmus calibration using either just COI or
COI + 16S + 28S also produced similar divergence dates, except in
the M. longi?ssa-grantii lineage, where COI appeared to underesti-
mate divergence times relative to those obtained by the other
methods. The timing of cladogenic events using COI + 16S + 28S
and a calibration based on both fossil ages and the closure of the
isthmus is shown in Fig. 3. Leodia diverged from Mellita during
the Messinian stage of the Miocene, and Mellita split into two ma-
jor lineages near to the Miocene/Pliocene boundary. In the Paci?c
lineage, M. grantii diverged from M. longi?ssa in the Piacenzian
stage of the Late Pliocene; splits within M. longi?ssa occurred in
the Gelasian stage of the Pleistocene. In the Atlantic + Paci?c line-
age, M. tenuis diverged from the ancestor of M. quinquiesperforata
in the Zanclean stage of the Pliocene. The ?nal closure of the Pan-
ama Isthmus in the Piacenzian resulted in a split of the ancestor of
the clades morphologically assigned to M. quinquiesperforata from
the ancestor of the lineages assigned by morphology to M. notabilis.
Further splits occurred between M. quinquiesperforata and M. sp. 1,
and between M. notabilis, M. sp. 5 and M. sp. 6 during the Gelasian
stage of the Pleistocene.
4. Discussion
4.1. Timing and possible causes of divergenceGenetic distance in COI and 16S between the Paci?c M. longi?s-
sa-grantii lineage and all other Mellita is similar to that between
both these Mellita lineages and L. sexiesperforata, (Table 1), suggest-
ing that their differentiation is equivalent to that found between
genera of the Mellitidae. Nevertheless, variation in COI among all
species of Mellita consisted only of silent substitutions. Only three
?xed amino acid differences differentiated Leodia from all Mellita in
COI. Eleven single-bp indels were found in 16S, four of which dis-
tinguish Leodia from Mellita, but none of which differentiate the Pa-
ci?c M. longi?ssa-grantii lineage from the Atlantic + Paci?c lineage
of Mellita. Two indels occurred in 28S, one insertion was unique
to Leodia.
3.2. Timing of divergence
The calibration of divergence values based on the assumption
S.E. Coppard et al. /Molecular Phylogeof many marine taxa are reported to have peaked at this time, in
response to increasing habitat heterogeneity in shallow-water
marine environments (Jackson et al., 1993; Cheetham and Jackson,
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.0281996; Collins, 1996; Knowlton and Weigt, 1998; Budd and John-
son, 1999;Marko, 2002; Smith and Jackson, 2009; Jagadeeshan
and O?Dea, 2011).
Through most of the Neogene, tropical and subtropical America
was biogeographically divided into two distinct regions, the Caloo-
sahatchian province, from North Carolina through the North of the
Gulf of Mexico, and the Gatunian province comprised the rest of
the tropical Atlantic and the modern day tropical eastern Paci?c
(Petuch, 1982; Vermeij, 2005; see Fig. 4). The provinces contained
unique assemblages of species, with limited overlap (Vermeij,
2005). The divergence of M. tenuis from the Atlantic + Paci?c line-
age at the end of the Zanclean stage of the Pliocene (3.8 Ma) re-
sulted in M. tenuis becoming restricted to the Caloosahatchian
province, with all other Recent Mellita limited to the Gatunian
province. These distributions are re?ected in the fossil record and
their present day species ranges. As the FST statistics indicate, the
?ne silt sediments and biogenic (carbonate) sands off the tip of
the Florida Peninsula do not act as a barrier to dispersal between
Atlantic and Gulf Coast M. tenuis as had been suggested by Harold
and Telford (1990).
In the Late Pliocene, the ancestor of M. quinquiesperforata di-
verged from the ancestor of M. notabilis following the ?nal closure
of the Panama Isthmus. Levels of divergence between these gemi-
nate species of Mellita are similar to those of other echinoid species
separated by the ?nal closure of the Panama Isthmus (see Lessios,
2008). Timing of divergence of geminate species in Mellita using a
relaxed molecular clock and the fossil record (but independent of
the age of the Isthmus) produces a date of 3 my for the ?nal split
of Atlantic and Paci?c lineages, consistent with the hypothesis that
it was due to the closure of the Panama Isthmus. This is congruent
with the hypothesis that sea water connections between the Atlan-
tic and the Paci?c existed until the late Pliocene, as a large accumu-
lation of palaeoceanographic (Keigwin, 1982; Collins, 1996; Haug
and Tiedemann, 1998; O?Dea et al., 2007), palaeontological (Webb,
1976; Coates and Obando, 1996), and genetic (Lessios, 2008) data
have indicated. Our data do not agree with the proposal of Montes
et al. (2012) that the isthmus has been an uninterrupted chain
above sea level since the Eocene. As no extant or fossil species of
Mellita has ever been found outside the Americas, the possibility
of post-isthmian genetic connections by stepping-stones around
the world is extremely remote. The existence of dry land separat-
ing the two oceans since the Eocene, 30 my before the Pliocene,
is entirely incompatible with the molecular phylogeny of Mellita.
Molecular phylogenies of regular echinoids (Lessios et al.,
1999, 2001, 2003; McCartney et al., 2000; Zigler and Lessios,
2004; Palumbi and Lessios, 2005) have agreed with Mayr?s
(1954) model of allopatric speciation. However, our data re-
vealed that not all divergence highlighted in this study ?ts an
allopatric model.
Divergence between Leodia and Mellita occurred 6 Ma when
there was no obvious geographic isolation, but rather at a time
when an increase in the interchange of transisthmian waters
has been reported, following the deepening of the canal basin
(Collins et al., 1996). Divergence between the Paci?c M. longi?s-
sa-grantii lineage and the Atlantic + Paci?c lineage appears to
have occurred prior to the Pliocene shoaling event (which began
4.7 Ma, Coates et al., 2003), without a clear barrier to gene
?ow.
Only the separation of M. quinquiesperforata from M. notabilis
(and its related Paci?c M. sp. 5 and M. sp.6) is clearly due to vicar-
iance due to the completion of the Isthmus of Panama 3 Ma.
4.2. How many species of Mellita are there?
cs and Evolution xxx (2013) xxx?xxx 7Genetic distance in COI between echinoderm congeners is typ-
ically no larger than 15.33% (Ward et al., 2008). Our COI data
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
als
netishow that the genetic distance between Leodia and the Paci?c
M. longi?ssa-grantii lineage was 45.56%, and between L. sexiesper-
forata and the Atlantic + Paci?c lineage 42.29%. L. Agassiz (1841)
regarded L. sexiesperforata as synonymous with Mellita, whereas
Mortensen (1948) preferred to consider Leodia only as a subgenus
of Mellita. The large genetic distances we have found between
L. sexiesperforata and Mellita justify Gray?s (1851) transfer of the
former species to a separate genus. Further subdivision of Mellita
may well be appropriate, because the deepest split between
Fig. 4. Geographic distributions of extant species of Mellita. The ?gure
8 S.E. Coppard et al. /Molecular Phylogeclades within the genus, that between the M. longi?ssa + M. grantii
lineage and the Atlantic + Paci?c clade, in COI (39.67%) is sugges-
tive of genus-level differentiation. These lineages are morpholog-
ically differentiated by the width of the ambulacral regions
between the food grooves on the oral surface that surround the
ambulacral lunules. Such regions are narrow in members of
the M. longi?ssa + M. grantii lineage, and broad in members of
the Atlantic + Paci?c lineage.
Which of the clades we have observed within a morphospe-
cies should be considered separate biological species is a dif?cult
question, because there are no data on reproductive isolation.
One approach to answering this question is the one typically
used in COI barcoding. Intraspeci?c variation of COI in animals
(except the Cnidaria) is rarely more than 2% and more typically
less than 1% (Avise, 2000); similar values have also been found
in echinoderms (Ward et al., 2008). Reproductive barriers be-
tween echinoid species have been found to arise in species
separated for only 250,000 years (COI divergence of 0.9%), as
they did between Echinometra oblonga (Blainville, 1825) in the
central Paci?c and an unnamed species from the western Paci?c
(Landry et al., 2003).
Mellita longi?ssa in the Paci?c was split into three mtDNA lin-
eages in the Pleistocene. Two of these lineages (M. longi?ssa and
M. sp. 4) today have distributions that overlap in the Gulf of Pan-
ama. However, with 3.97% divergence in COI it is likely that they
have become reproductively isolated. This is also true for M. sp. 3
from Paci?c Colombia, which has a mean divergence of 5.67% in
COI from M. longi?ssa + M. sp. 4. The three lineages that diverged
from the ancestor of M. notabilis have more than 8.6% divergence
in COI from one another suggesting that they are also separate
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028species. This is also true for M. quinquiesperforata and M. sp. 1, in
which divergence in COI between these two clades is also 8.6%.
Short branch lengths and intermixing of morphospecies within
the Gulf of Mexico and Atlantic clade suggest that M. isometra is
not a separate species from M. tenuis. Seven of the eight specimens
of M. tenuis sampled from Texas did form a geographically exclu-
sive subclade. This is likely the result of restricted larval dispersal
between east and west coast populations across the freshwater
plume of the Mississippi Delta.
o shows the Neogene biogeographic provinces sensu Vermeij (2005).
cs and Evolution xxx (2013) xxx?xxxWe therefore propose that there are four extant species of Mel-
lita in the Atlantic (M. tenuis, M. quinquiesperforata, M. sp. 1 and M.
sp. 2) and three extant species in the Paci?c (M. notabilis, M. sp. 5
and M. sp. 6). A further four Paci?c species (M. grantii, M. longi?ssa,
M. sp. 3, M. sp. 4) are probably best placed in a new genus. We pro-
pose that M. isometra should be synonymised with M. tenuis, and
suggest that M. kanakof? and M. eduardobarrosoi should be consid-
ered junior synonyms of M. notabilis.
Considerable plasticity in test structures was encountered in
the M. longi?ssa-grantii clade, particularly between those living in
sheltered bays in relation to those living on exposed beaches. This
was particularly evident in M. sp. 4 from the sheltered Gulfo de San
Miguel in Panama. Specimens from this location had almost circu-
lar tests and a posterior interambulacral lunule that projected only
halfway between the posterior petals, similar to M. grantii from the
Gulf of California. Other members of the M. sp. 4 clade from ex-
posed beaches outside of the Gulfo de San Miguel had pentagonal
test outlines and interambulacral lunules that projected to the
periproct, being more typical of M. longi?ssa. A similar pattern in
posterior lunule development occurred in M. grantii. Members of
this species from the Paci?c coast of the Baja Peninsula have a
longer posterior interambulacral lunule than those from within
the Gulf of California.
Members of M. notabilis also exhibited morphological variation
within subclades. Some members had very hummocky lunule mar-
gins, deep pressure drainage channels and a sub-rectangular test
margin, while others of similar size had smooth lunule margins,
shallow drainage channels and a sub-circular shaped test. Popula-
tions on exposed beaches at Acapulco and Michoacan in Mexico
had deep pressure drainage channels and hummocky lunule
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
(Fig. 4). Mellita tenuis occurs in the northern region of the Gulf of
levels of morphological plasticity. The molecular phylogeny re-
21?56.
netivealed the presence of eleven probable species in Mellita, including
six cryptic species, whereas ?ve described morphospecies do not
include monophyletic molecular units and thus represent ecophe-
notypic variation within species. Leodia sexiesperforata diverged
from the ancestor of Mellita in the Late Miocene. The ancestor of
Mellita diverged into a Paci?c lineage and an Atlantic + Paci?c line-
age close to the Miocene/Pliocene boundary. High levels of genetic
differentiation occur between these lineages suggesting genus le-
vel differentiation. Paci?c Mellita, therefore consist of two highly
divergent lineages that became established in the Paci?c at differ-
ent times, resulting in sympatric distributions. Atlantic species, on
the other hand, are alloparic with respect to one another. Judged by
modern day ranges, not all divergence in this genus conforms to an
allopatric model. Fossil calibration of some of the nodes of the
molecular phylogeny dated the separation of M. quinquiesperforata
and M. notabilis event at 3 Ma in the Late Pliocene, a date consis-
tent with a great deal of evidence regarding the ?nal closure of theMexico and off the US Atlantic coast. Mellita quinquiesperforata
(name designation based on the morphology of the holotype) is
distributed along the Atlantic coasts of Costa Rica and Panama.
Specimens from the type locality of Veracruz, Gulf of Mexico need
to be sequenced to establish whether they belong to the same
clade as those from Costa Rica and Panama. Mellita sp. 1 occurs
off Santa Marta, Colombia. This species may also occur throughout
the upwelling Guajira region (Andrade and Barton, 2005), where
endemic species are not uncommon (Petuch, 2004). Mellita sp. 2
is distributed from the Lesser Antilles through tropical Brazil.
In the Paci?c, M. grantii occurs in the Gulf of California and
along the adjacent Paci?c coast of the Baja California Peninsula.
Mellita sp. 6 occurs in the mouth of the Gulf of California, while
M. notabilis and M. longi?ssa are sympatric from Northern Mexico
to Panama. Mellita sp. 5 was only recorded from Paci?c Colombia
and M. sp. 5 from Ecuador.
5. Conclusions
Our molecular analyses indicate that the species designations of
Mellita according to morphology were often erroneous due to highmargins, while populations in bays at Playa Venado and Punta
Mala Panama had shallow pressure drainage channels and smooth
lunule margins. FST statistics indicated only a moderate degree of
gene ?ow between these ecophenotypes, suggesting that the mor-
phological differences may, in fact be due to divergence between
entities that do not interbreed freely.
A similar range of test characters occurred in M. sp. 5 from Santa
Elena, Ecuador. However, in contrast to M. notabilis, the extremes
of morphological variation in M. sp. 5 occurred in an identical hap-
lotype (see Fig. 2b). Variation in colour pattern was also observed
in this population with approximately half its members having
spots down the ambulacra and interambulacra aborally (Fig. 2b),
the other members being uniformly green. These colour patterns
were mixed among morphotypes and both colour patterns were
present in identical haplotypes. Such colour pattern variation has
been observed in other populations from Ecuador (Sonneholzner
Varas, pers. com.).
4.3. Species distributions
Based on our ?ndings regarding molecular clades, we can pro-
pose ranges for each of the presumed species we have discovered
S.E. Coppard et al. /Molecular PhylogePanamanian Isthmus, but at odds with a recent suggestion (Montes
et al., 2012) that uninterrupted land was present as early as the
Eocene.
Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028Coates, A.G., Aubry, M.P., Berggren, W.A., Collins, L.S., Kunk, M., 2003. Early Neogene
history of the Central American arc from Bocas del Toro, western Panama.
Geological Society of America Bulletin 115, 271?287.
Collins, L.S., 1996. Environmental changes in Caribbean shallow waters relative to
the closing tropical American seaway. In: Jackson, J.B.C., Coates, A.G., Budd, A.
(Eds.), Evolution and Environment in Tropical America. University of Chicago
Press, Chicago, IL, pp. 130?167.
Collins, L.S., Coates, A.G., Berggen, W.A., Aubry, M.P., Zhang, J., 1996. The late
Miocene Panama isthmian strait. Geology 24 (8), 687?690.
Darriba, D., Taboada, G.L., Doallo, R., Posada, D., 2012. JModelTest 2: more models,
new heuristics and parallel computing. Nature Methods 9, 772.
Drummond, A.J., Suchard, M.A., Xie, D., Rambaut, A., 2012. Bayesian phylogenetics
with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29, 1969?
1973.
Durham, J.W., 1961. The Echinoid Mellita in the Paci?c Coast Cenozoic.
Contributions in Science Los Angeles County Museum 48, 3?12.
Excof?er, L., Lischer, H.E.L., 2010. Arlequin suite ver 3.5: A new series of programs to
perform population genetics analyses under Linux and Windows. Molecular
Ecology Resources 10, 564?567.
Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for
ampli?cation of mitochondrial cytochrome c oxidase subunit I from diverse
metazoan invertebrates. Molecular Marine Biology and Biotechnology 3, 294?Acknowledgments
We thank A. Calderon, L. Calderon, L.B. Geyer, D. Gonz?lez, S.
Lockhart and A.H. Magallon for their assistance in the laboratory.
We are grateful to R. Collin, S.W.C. Coppard, T. Duda, A. Hiller, J.
Lawrence, S. Lockhart, D. McClay, A. O?Dea, D. Pope, R.W. Portell,
L. Rocha, R. Riosmena-Rodriguez, F. Rodriguez, T.I. Sancho-Mejias,
A.N. Su?rez Castillo, and A. Sweeney for help with collecting sam-
ples. We thank R. Mooi for helpful discussion and A. O?Dea for his
comments on the manuscript. This work was supported by a fel-
lowship to SEC from SENACYT (La Secretar?a Nacional de Educaci?n
Superior Ciencia Tecnolog?a e Innovaci?n) (Panama) as part of pro-
ject COL08-002, and fellowships to KSZ from the Smithsonian Insti-
tution, Friday Harbor Laboratories, the National Science
Foundation (#0202773), and a Grant-in-Aid of Research from Sig-
ma Xi.
References
Agassiz, L., 1841. Des Scutelles. Monographies des Echinoderms Vivans et Fossiles.
Monograph 2, 1?151.
Akaike, H., 1974. A new look at the statistical model identi?cation. IEEE
Transactions on Automatic Control 19, 716?723.
Andrade, C.A., Barton, E.D., 2005. The Guajira upwelling system. Continental Shelf
Research 25 (9), 1003?1022.
Avise, J.C., 2000. Phylogeography: The History and Formation of Species. Harvard
University Press, Cambridge, MA, 447 pp.
Blainville, H.M.D. de, 1825. Oursins. In: Dictionnaire des Sciences Naturelles 37. p.
95. F.G. Levrault, Strasbourg, Paris. Budd A.F. and Johnson K.G., Origination
preceding extinction during late Cenozoic turnover of Caribbean reefs,
Paleobiology 25, 1999, 188?200.
Budd, A.F., Johnson, K.G., 1999. Origination preceding extinction during late
Cenozoic turnover of Caribbean reefs. Paleobiology 25, 188?200.
Caldwell, J.W., 1972. Development, Metamorphosis and Substrate Selection of the
Sand Dollar Mellita quinquiesperforata (Leske, 1778). MSc Thesis, University of
Florida. 163 pp.
Caso, M.E., 1980. Contribuci?n al estudio de los Echinozoa de M?xico. La Familia
Mellitidae Stefanini. Descripti?n de una nueva especie del g?nero Mellita.
Mellita eduardobarrosoi sp. nov. Anales del Centro de Ciencias del Mar y
Limnologia Universidad Nacional Aut?noma de M?xico 7, 141?180.
Cerame-Vivas, M.J., Gray, I.E., 1964. The Presence of a sixth lunule in the sand dollar,
Mellita quinquiesperforata. Bulletin of Marine Science 14, 303?305.
Cheetham, A.H., Jackson, J.B.C., 1996. Speciation, extinction, and the decline of
arborescent growth in Neogene and Quaternary cheilostome Bryozoa of tropical
America. In: Jackson, J.B.C., Coates, A.G., Budd, A. (Eds.), Evolution and
Environment in Tropical America. University of Chicago Press, Chicago, IL, pp.
205?233.
Clark, H.L., 1940. Revision of the keyhole urchins (Mellita). Proceedings of the United
States National Museum 89, 435?444.
Clark, H.L., 1947. A new and remarkable keyhole urchin, Mellita notabilis n. sp.
Bulletin of the Southern California Academy of Sciences 46, 77?78.
Coates, A.G., Obando, J.A., 1996. The geologic evolution of the Central American
Isthmus. In: Jackson, J.B.C., Coates, A.G., Budd, A. (Eds.), Evolution and
Environment in Tropical America. University of Chicago Press, Chicago, IL, pp.
cs and Evolution xxx (2013) xxx?xxx 9297.
Gelman, A., Rubin, D.B., 1992. Inference from iterative simulation using multiple
sequences. Statistical Science 7, 457?511.
d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.
Gray, J.E., 1851. Description of two new genera and new species of Scutellidae and
Echinolampidae in the collection of the British Museum. Proceeding of the
Zoological Society of London 19, 34?38.
Harold, A.S., Telford, M., 1990. Systematics, phylogeny and biogeography of the genus
Mellita (Echinoidea: Clypeasteroida). Journal of Natural History 24, 987?1026.
Haug, G.H., Tiedemann, R., 1998. Effect of the formation of the Isthmus of Panama
on Atlantic Ocean thermohaline circulation. Nature 393, 673?676.
Jagadeeshan, S., O?Dea, A., 2011. Integrating fossils and molecules to study
cupuladriid evolution in an emerging isthmus. Evolutionary Ecology 26 (2),
337?355.
Jackson, J.B.C., Jung, P., Coates, A.G., Collins, L.S., 1993. Diversity and extinction of
tropical American molluscs and emergence of the Isthmus of Panama. Science
Mayr, E., 1954. Geographic speciation in tropical echinoids. Evolution 8, 1?18.
McCartney, M.A., Keller, G., Lessios, H.A., 2000. Dispersal barriers in tropical oceans
and speciation of Atlantic and eastern Paci?c Echinometra sea urchins.
Molecular Ecology 9, 1391?1400.
Michelin, H., 1858. Revue des especes du genre Mellita. Revue et Magasin de
Zoologie Pure et Appliqu?e 1858, p. 2. PI. VIII, I.
Montes, C., Cardona, A., MacFadden, R., Moron, S.E., Silva, C.A., Restrepo-Moreno, S.,
Ramirez, D.A., Wilson, J., Farris, D., Bayona, G.A., Jaramillo, C., Valencia, V., Flores,
J.A., 2012. Evidence for middle Eocene and younger emergence in Central
Panama: implications for Isthmus closure. Geological Society of America
Bulletin 120, 780?799.
10 S.E. Coppard et al. /Molecular Phylogenetics and Evolution xxx (2013) xxx?xxx260, 1624?1626.
Johnson, K.G., Todd, J.A., Jackson, J.B.C., 2007. Coral reef development drives
molluscan diversity increase at local and regional scales in the late Neogene and
Quaternary of the southwestern Caribbean. Paleobiology 33, 24?52.
Johnson, K.G., Jackson, J.B.C., Budd, A.F., 2008. Caribbean reef development was
independent of coral diversity over 28 million years. Science 319, 1521?1523.
Jones, D.S., MacFadden, B.J., Webb, S.D., Mueller, P.A., Hodell, D.A., Cronin, T.M., 1991.
Integrated geochronology of a classic Pliocene fossil site in Florida: linking marine
and terrestrial biochronologies. The Journal of Geology, 637?648.
Keigwin, L.D., 1982. Isotopic paleoceanography of the Caribbean and east Paci?c:
role of Panama uplift in Late Neogene time. Science 217, 350?353.
Kessing, B.H., Croom, H., Martin, A., McIntosh, C., McMillan, O.W., Palumbi, S., 1989.
The Simple Fools Guide To PCR. University Of Hawaii, Honolulu, 45 pp.
Kier, P.M., 1963. Tertiary echinoids from the Caloosahatchee and Tamiami
formations of Florida: Smithsonian Miscellaneous. Collections 145, 1?63.
Klein, J.T., 1734. Naturalis dispositio echinodermatum. Accessit lucubratiuncula de
aculeis echinorum marinorum, cum spicilegio de belemnitis. 1?78. Gedani,
Schreiber.
Knowlton, N., Weigt, L.A., 1998. New dates and new rates for divergence across the
Isthmus of Panama. Proceedings of the Royal Society Biological Sciences 265,
2257?2263.
Krantz, D.E., 1991. A chronology of Pliocene sea-level ?uctuations, US Atlantic
Coastal Plain. Quaternary Science Reviews 10, 163?174.
Landry, C., Geyer, L.B., Arakaki, Y., Uehara, T., Palumbi, S.R., 2003. Recent speciation
in the Indo-West Paci?c: rapid evolution of gamete recognition and sperm
morphology in cryptic species of sea urchin. Proceedings of the Royal Society
Biological Sciences 270, 1839?1847.
Leske, N.G., 1778. Iacobi Theodori Klein natvralis dispositio echinodermatvm.
Accesservnt lvcvbrativncvla de acvleis echinorvm marinorvm et spicilegivm de
belemnitis. 278 pp. Lipsiae, Gleditsch.
Lessios, H.A., 2005. Echinoids of the Paci?c waters of Panama: Status of knowledge
and new records. Revista de Biologia Tropical 53, 147?170.
Lessios, H.A., 2008. The Great American Schism: divergence of marine organisms
after the rise of the Central American Isthmus. Annual Reviews of Ecology
Evolution and Systematics 39, 63?91.
Lessios, H.A., Kessing, B.D., Robertson, D.R., Paulay, G., 1999. Phylogeography of the
pantropical sea urchin Eucidaris in relation to land barriers and ocean currents.
Evolution 53, 806?817.
Lessios, H.A., Kessing, B.D., Pearse, J.S., 2001. Population structure and speciation in
tropical seas: global phylogeography of the sea urchin Diadema. Evolution 555,
955?975.
Lessios, H.A., Kane, J., Robertson, D.R., 2003. Phylogeography of the pantropical sea
urchin Tripneustes: contrasting patterns of population structure between
oceans. Evolution 57, 2026?2036.
Lessios, H.A., Lockhart, S., Collin, R., et al., 2012. Phylogeography and bindin
evolution in Arbacia, a sea urchin genus with an unusual distribution. Molecular
Ecology 21, 130?144.
Littlewood, D.T.J., Smith, A.B., 1995. A combined morphological and molecular
phylogeny for sea urchins (Echinoidea: Echinodermata). Philosophical
Transactions of the Royal Society of London 347B, 213?234.
Lyons, W.G., 1991. Post-Miocene species of Latirus Monfort, 1810 (Mollusca:
Faciolariidae) of southern Florida, with a review of the regional marine
biostratigraphy. Bulletin of the Florida Museum of Natural History 35, 131?208.
Maddison, D.R., Maddison, W.P., 2005. MacClade 4: Analysis of Phylogeny and
Character Evolution. Version 4.08a. <http://macclade.org>.
Marko, P.B., 2002. Fossil calibration of molecular clocks and the divergence times of
geminate species pairs separated by the Isthmus of Panama. Molecular Biology
and Evolution 19, 2005?2021.Please cite this article in press as: Coppard, S.E., et al. Phylogeography of the san
Mol. Phylogenet. Evol. (2013), http://dx.doi.org/10.1016/j.ympev.2013.05.028Mooi, R., Mart?nez, S., Parma, S.G., 2000. Phylogenetic systematics of Tertiary
monophorasterid sand dollars (Clypeasteroida: Echinoidea) from South
America. Journal of Paleontology 74, 263?281.
Mooi, R., Peterson, D., 2000. A new species of Leodia (Clypeasteroida: Echinoidea)
from the Neogene of Venezuela and its importance in the phylogeny of mellitid
sand dollars. Journal of Paleontology 74, 1083?1092.
Mortensen, T., 1948. A Monograph of the Echinoidea. IV, 2. Clypeasteroida.
Clypeasterid?, Arachnoid?, Fibulariid?, Laganid? and Scutellid?. CA Reitzel;
Copenhagen. 471 pp.
O?Dea, A., Jackson, J.B.C., Fortunato, H., Smith, J.T., D?Croz, L., Johnson, K.G., Todd, J.A.,
2007. Environmental change preceded Caribbean extinction by 2 million years.
Proceeding of the National Academy of Science of the United States of America
104, 5501?5506.
Palumbi, S.R., Lessios, H.A., 2005. Evolutionary animation: How do molecular
phylogenies compare to Mayr?s reconstruction of speciation patterns in the sea?
Proceedings of the National Academy of Sciences of the United States of
America 102 (Suppl. 1), 6566?6572.
Petuch, E.J., 1982. Geographical heterochrony: contemporaneous coexistence of
Neogene and Recent molluscan faunas in the Americas. Palaeogeography,
Palaeoclimatology Palaeoecology 37, 277?312.
Petuch, E.J., 2004. Cenozoic Seas: the View From North America. CRC Press, Boca
Raton, Fl, 308 pp.
Rambaut, A., Drummond, A.J., 2007. Tracer v1.4. <http://beast.bio.ed.ac.uk/Tracer>.
Ravenel, E., 1841. Description of two new Scutella from South Carolina. Proceedings
of the Academy of Natural Sciences of Philadelphia 1, 81?82.
Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes version 3.0: Bayesian phylogenetic
inference under mixed models. Bioinformatics 19, 1572?1574.
Smith, J.T., Jackson, J.B.C., 2009. Ecology of extreme faunal turnover of tropical
American scallops. Paleobiology 35, 77?93.
Swofford, D.L., 2002.PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods). Sinauer Associates, Sunderland, Massachusetts.
Tamura, K., Nei, M., 1993. Estimation of the number of nucleotide substitutions in
the control region of mitochondrial DNA in humans and chimpanzees.
Molecular Biology and Evolution 10, 512?526.
Tavare, S., 1986. Some probabilistic and statistical problems in the analysis of DNA
sequences. Lectures on Mathematics in the Life Sciences (American
Mathematical Society) 17, 57?86.
Telford, M., Mooi, R., 1986. Resource partitioning by sand dollars in carbonate and
siliceous sediments: evidence from podial and particle dimensions. Biological
Bulletin 171, 197?207.
Vermeij, G.J., 2005. One-way traf?c in the western Atlantic: causes and
consequences of Miocene to early Pliocene molluscan invasions in Florida and
the Caribbean. Paleobiology 31, 624?642.
Ward, R.D., Holmes, B.H., O?Hara, T.D., 2008. DNA barcoding discriminates
echinoderm species. Molecular Ecology Resources 8, 1202?1211.
Webb, S.D., 1976. Mammalian faunal dynamics of the great American interchange.
Paleobiology 2, 220?234.
Xia, X., Xie, Z., Salemi, M., Chen, L., Wang, Y., 2003. An index of substitution
saturation and its application. Molecular Phylogenetics and Evolution 26, 1?7.
Xia, X., Lemey, P., 2009. Assessing substitution saturation with DAMBE. In: Lemey,
P., Salemi, M., Vandamme, A.M. (Eds.), The Phylogenetic Handbook: A Practical
Approach to DNA and Protein Phylogeny, second ed. Cambridge University
Press, pp. 615?630.
Zigler, K.S., Lessios, H.A., 2004. Speciation on the coasts of the new world:
phylogeography and the evolution of bindin in the sea urchin genus Lytechinus.
Evolution 58, 1225?1241.
Zwickl, D.J., 2006. Genetic Algorithm Approaches for the Phylogenetic Analysis of
Large Biological Sequence Datasets Under the Maximum Likelihood Criterion.
Ph.D. Dissertation, The University of Texas at Austin.d dollar genus Mellita: Cryptic speciation along the coasts of the Americas.