OPEN 0 ACCESS Freely available online

A Novel Widespread Cryptic Species and Phylogeographic Patterns within Several Giant Clam Species (Cardiidae: Tridacna) from the Indo-Pacific OceanThomas H uelsken19, Jude K eyse19, Libby Liggins1, Shane Penny2, Eric A. Treml1'3, Cynthia Riginos1*1 The University of Queensland, School of Biological Sciences, St Lucia, Australia, 2 Charles Darwin University, Research Institute for Environment and Livelihoods, Casuarina, Australia, 3 University of Melbourne, Department of Zoology, Melbourne, Australia

AbstractGiant d a m s (genus Tridacna) a re iconic coral reef animals of th e Indian and Pacific Oceans, easily recognizable by their massive shells an d vibrantly colored m antle tissue. Most Tridacna species are listed by CITES a n d th e IUCN Redlist, as their p opulat ions have been extensively harvested and d e p le te d in m any regions. Here, w e survey Tridacna crocea and Tridacna maxima from th e eas te rn Indian an d w este rn Pacific Oceans for m itochondria l (COI and 765) a n d nuclear {ITS) s e q u e n ce variation a n d conso lida te th e se d a ta with previous published results using phy logenetic analyses. We find d e e p intraspecific differentiation within bo th T. crocea and T. maxima. In T. crocea w e describe a previously u n d o c u m e n te d phy logeograph ic division to th e eas t of Cenderawasih Bay (northw es t New Guinea), w hereas for T. maxima t h e previously described, dist inctive lineage o f Cenderawasih Bay can be se e n to also typify w este rn Pacific populat ions. Furthermore, w e find an undescrlbed, m onophy le t lc g ro u p th a t Is e v o lu t io n a r y distinct from n am ed Tridacna species a t b o th m itochondria l and nuclear loci. This cryptic taxon Is geographical ly w idesp read with a range ex te n t th a t minimally Includes m uch o f th e central Indo-Pacific region. Our results reinforce th e e m erg in g pa rad igm th a t cryptic species are c o m m o n a m o n g marine invertebrates , even for co nsp icuous an d culturally significant taxa. Additionally, o u r results ad d to identified locations of g en e t ic differentiation across th e central Indo-Pacific a n d highlight h ow p h y logeograph ic pa tte rns may differ even b e tw ee n closely related a n d co-dis tr ibuted species.

C i t a t i o n : Huelsken T, Keyse J, Liggins L, Penny S, Treml EA, e t al. (2013) A Novel Widespread Cryptic Species and Phylogeographic Patterns within Several Giant Clam Species (Cardiidae: Tridacna) from the Indo-Pacific Ocean. PLoS ONE 8(11): e80858. doi:10.1371/journal.pone.0080858

E d i to r : Mikhail V. Matz, University of Texas, United States of America

R e c e iv e d July 30, 2013; A c c e p t e d October 14, 2013; P u b l i s h e d November 20, 2013

C o p y r i g h t : © 2013 Huelsken e t al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

F u n d i n g : Funding for this work was provided by the Australian Research Council (www.arc.gov.au, DP0878306 to CR), the German Research Foundation (www. dfg.de/en, DFG, Hu 1806/1-1, Hu 1806/2-1 to TH), the World Wildlife Fund (worldwildlife.org/initiatives/fuller-science-for-nature-fund, Kathryn Fuller Post-doctoral Research Fellowship to EAT), the Malacological Society of Australasia (www.malsocaus.org, to TH), and the Joyce Vickery Fund (linneansocietynsw.org.au/ grants.html, to JK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

C o m p e t i n g I n t e r e s t s : The authors have declared that no competing interests exist.

* E-mail: c.riginos@uq.edu.au

9 These authors contributed equally to this work.

Introduction

G iant clams o f the genus Tridacna are am ong the most conspicuous m arine invertebrates on coral reefs due to their large size and brilliantly colored m antle that contains photosynthesizing symbionts. G iant clanis have traditionally provided raw m aterial for tools, containers, an d ornam ents [1], and m any populations are harvested for m eat, shells, an d the ornam ental aquarium trade [2,3]. Despite local m anagem ent efforts, including m ariculture [3], wild stocks o f giant clanis are depleted and some species are locally extinct in m any areas o f Southeast Asia and the South Pacific [3— 5]. C onsequently, m ost Tridacna species are listed by C IT E S (Appendix II) [6] and the IETCN Redlist [7].

T h ere are currently eight [8] described species w ithin the genus Tridacna (T. crocea Lam arck, 1819, T. derasa (Röding 1798), T. gigas (Linnaeus 1758), T. maxima (Röding 1798), T. mbalavuana Ladd, 1934, T. rosewateri Sirenko and Scarlata 1991, T. squamosa Lam arck 1819, an d T. squamosina S turany 1899), differentiated by m orphology an d hab ita t preference [9-12]. Tridacna squamosina, T. rosewateri, and T. mbalavuana have restricted distributions (Red Sea, M auritius, and Fiji to Tonga, respectively), whereas T. derasa,

I . gigas, T. crocea, T. squamosa an d T. maxima are widely distributed in the Ind ian an d Pacific O ceans, w ith the latter two extending their distribution into the R ed Sea [8,9], M olecular phylogenetic investigations support m onophyly of the described species [13-15], albeit w ith some disagreem ent am ong species relationships. An unpublished M aster’s thesis [16] also reports a m orphologically distinct clam from T aiw an and uses m tD N A loci to show that this clam is highly divergent from sym patric T. maxima, potentially indicative o f an additional unnam ed species.

T h e junctu re betw een the Ind ian and Pacific O ceans (Fig. 1), w here several species o f Tridacna are sym patric [8], i s a well-known epicenter o f tropical m arine biodiversity [17,18], Genetic surveys in this region have revealed cryptic species, even am ong conspicuous an d well-studied m arine invertebrates [19,20], M any species show substantial intraspecific genetic division betw een the ocean basins (reviewed by [21]), w ith the Sunda Shelf, M olucca and Flores Seas, M akassar Strait, and B ird’s Plead region of northw est New G uinea em erging as locations o f genetic discon­tinuities [21,22]. T hese locations span the archipelago com m only referred to as W allacea, w hich falls betw een the Sunda (southeast Asia) an d Sahul (Australia an d New Guinea) continental shelves

PLOS ONE I www.plosone.org 1 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

Sumatra 4

Red Sea wm.

4 k

Sunda Shelf

Bird’s Head Peninsula

Cenderawasih Bay

Wallacea New Guinea

ru Basin

PACIFIC OCEAN

Solomon IslandsSahul Shelf

INDIAN OCEAN

t1000 km

Torres Strait

VIGreat Barrier Reef

\

Figure 1. Study region. T he ligh t g re y o u tlin e re p re sen ts th e lo w est P le is to cen e sea level (120 m d e p th co n to u r). do i:10 .1371/jo u rn a l.pone.0080858 .g001

and was the only po in t o f p e rm anen t oceanic connection betw een the Ind ian and Pacific O ceans th roughout the Pleistocene [23].

Phylogeographic and population genetic surveys have intensely sam pled T. maxima an d T. crocea th roughout W allacea using m itochondrial (mtDNA) m arkers [24-26], allozymes [27,28], and m icrosatellites [29]. Both T. crocea and T. maxima have been shown to contain distinct m tD N A clades associated with Sum atra (Sunda), W allacea, and northw est New G uinea (Sahul, particularly in C enderaw asih Bay) [24—26], T hese lineages are sym patric in some populations, for instance T. maxima from n o rthern Ja v a has bo th S um atran and W allacean mitotypes, and similarly T. crocea populations from H alm ahera eastw ard th rough C enderaw asih Bay contain b o th W allacean and northw est New G uinean lineages [26,29]. M icrosatellite genotyping o f T. crocea corroborates the distinctiveness o f S um atran and C enderaw asih populations, with evidence for m ixing in W allacea o f local genotypes with Cenderawasih-like genotypes [29]. Thus, substantial genetic differentiation typifies at least two Tridacna species in this region.

In the Pacific O cean, T. derasa, T. gigas, T. maxima an d T. crocea have been genetically surveyed, prim arily w ith allozyme m arkers [27,28,30-35], bu t also w ith m tD N A [36], These studies show genetic divisions betw een w estern an d central Pacific populations bu t with some indication that eastern A ustralian populations show greater affinities w ith Philippine populations than they do with o ther w estern Pacific populations [30,33]. G reat B arrier R eef populations (eastern Sahul) form a cluster distinct from , bu t closely

related to, Philippine populations for T. maxima an d T. derasa bu t with low sam pling in the Philippines (two and one populations, respectively) and no sam pling in W allacea or Sunda regions. Thus, it is unknow n w hether substantial genetic divergence reflects the geographic distance separating the Philippines and eastern Sahul or is indicative o f distinct regional groupings.

H ere, we exam ine D N A sequence diversity o f T. crocea and T. maxima whose sam pled distributions include the eastern Indian O cean, W allacea, and western Pacific O ceans. D a ta from new samples, p redom inantly from the w estern Pacific, are m erged with d a ta from previous studies, especially from W allacea (e.g. [24,25,26]), to present a unified sum m ary o f phylogeographic patterns an d a point o f contrast to earlier broadscale studies based on allozymes [30,32,33,35]. W e use phylogenetic analyses to assess evolutionary relationships am ong species an d also gauge regional geographical divisions w ithin species.

Materials and M ethods

S am p lin g a n d p e rm itsSmall m antle biopsies were non-lethally collected from anim als

with m orphology characteristic o f Tridacna maxima and T. crocea at 0 -2 0 m dep th from the Solom on Islands, and in A ustralia from Ningaloo Reef, H ero n Island, L izard Island, the T orres Strait and L ihou Reef. All sam pling an d tissue transport was in accordance with local an d international regulations. Perm it details are as

PLOS ONE I www.plosone.org 2 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

follows: L ihou Reef, Australia: D epartm en t o f Sustainability, E nvironm ent, W ater, Population & Com m unities (Access to Biological Resources in a C om m onw ealth A rea for N on- C om m ercial Purposes perm it num ber: A U -C O M 2008042); Liz­ard Island an d H eron Island, Australia: G reat B arrier R eef M arine Park A uthority an d Q ueensland Parks and W ildlife (M arine Parks Perm its: G 08 /28114 .1 , G 09 /31678 .1 , G 1 0 / 33597.1, G 1 1/34640.1); N ingaloo Reef, Australia: W estern A ustralia D epartm en t o f E nvironm ent and Conservation (License to take F auna for Scientific Purposes: SF007126, SF006619, SF008861; A uthority to E nter C alm L a n d /o r W aters: C E002227, CE002627, D epartm en t o f Fisheries, W estern A ustralia Exem p­tion 2046); Q ueensland: Q ueensland G overnm ent D epartm en t o f Prim ary Industries (General Fisheries Permits: 118636, 150981); T orres S trait Islands, Australia: C om m onw ealth o f A ustralia T orres S trait Fisheries Act 1984 and A ustralian Fisheries M anagem ent A uthority (Perm it for Scientific Purposes: 8562); Solom on Islands: Solom on Islands G overnm ent M inistry o f Education and H u m an R esource D evelopm ent and M inistry o f Fisheries and M arine Resources (research perm it: to S Albert, expiry 3 1 /1 0 /2 0 1 1 ); Solom on Islands G overnm ent M inistry o f E nvironm ent, Conservation and M eteorology (Convention on In ternational T rad e in E ndangered Species o f W ild F auna and F lora export perm it: E X 2010/102); A ustralian G overnm ent D epartm en t o f the E nvironm ent, W ater, H eritage and the Arts (C onvention on In ternational T rad e in E ndangered Species o f W ild F auna and Flora im port perm it: 2010-AU-616020); A ustra­lian Q u aran tine Inspection Service (Perm it to Im port Q uaran tine M aterial: IP10017966).

DNA s e q u e n c e sD N A was extracted using a m odification o f the Q iagen DN easy

protocol [37]. Prim ers th a t targeted m itochondrial cytochrom e oxidase 1 (COI) [24,26,38] an d ribosom al 16S [39] w ere used to amplify 390 and 417 basepair segments o f the respective gene regions. A subset o f samples were am plified for the partial nuclear 18S an d ITS1 region (referred to as IT S in text) to provide independent estim ates o f phylogenetic relationships using prim ers from [13,40]. P C R products were purified following a standard Exo-Sap protocol (New England Biolabs) and were sequenced by M acrogen (Korea). T race files were edited in C odonC ode Aligner (ver. 4.0.3). In addition, the N C B I repository o f nucleotide sequences was searched for all published Tridacna COI a n d 1 6S sequences (August 2012) representing bo th intraspecific [24— 26,41] and interspecific [9,15,16] surveys. T hese sequences were m anually aligned [42] against our new sequences and against outgroups (Hippopus hippopus, Hippopus porcellanus, Cerastoderma glaucum, Fragum sueziense, and Corculum cardissa) and trim m ed to a com m on length. For IT S there were several insertions/deletions th a t could no t be reconciled, so these areas o f low overlap were m asked an d no t used for phylogenetic analyses.

P h y lo g en e t ic ana lysesPrevious m tD N A surveys have used either 16S [9,15,26] o r COI

[24-26,41] gene regions. T o unify these sources o f da ta and address interspecific relationships, we initially took representative sequences across studies and linked them by ou r samples for which bo th gene regions had been sequenced in a concatenated search. For samples w ith only a single gene region (that is, sequences acquired from NCBI), inform ation from the m issing gene region was treated as m issing data. U p to four individuals pe r species were retained representing the diversity o f their species clade and prioritizing individuals w ith bo th 16S and COI sequenced. U sing StarB EA ST v. 1.6.2 [43] each m tD N A gene region was treated as

a separate partition. A general tim e reversible m odel w ith gam m a distributed an d invariant sites (G TR +G +I) was applied to each gene, with additional partition ing by codon position (1+2, 3) for COI. A relaxed m olecular clock with an uncorrelated lognorm al m utation ra te was used for each gene. T h e COI and 16S gene trees were linked, as m tD N A is a single linked locus (i.e. concatenated gene regions). Priors were set for nodes defining species as a log norm al date (m ean = 0, SD = 1) w ith an offset representing the m ost recent estim ate o f the earliest fossil (71 crocea: 1.8, 71 maxima: 5.3, and 71 squamosa: 1.8 m illion years). T h e roo t o f the T ridacn inae was set as norm al with m ean date o f 14 an d SD of 2.5 m illion years. All fossil dates were based on [15,44], Spéciation was m odeled bo th as b irth -death and Yule processes in independent runs o f 250 m illion steps, w ith a burn-in o f 25% , and yielded similar results.

A dditional genealogical searches were perform ed using M rBayes ver. 3.1.2 [45] and R A xM L (Random ized Axelerated M axim um Likelihood, Blackbox interface) [46]. U sing the concatenated file o f the same m tD N A sequences as above, searches w ere partitioned such th a t 16S form ed one partition, and COI form ed a second partition with th ird codon positions partitioned separately from first and second (1+2, 3) for COI. In MrBayes, a G T R + G + I (nst = 6, invgam m a) m odel for all three partitions was used, with a search length o f 10 m illion steps, sam pling every 10,000 steps, an d a bu rn -in o f 25% (2.5 mill steps). Similarly, the G T R + G + I m odels were applied to these partitions in R A xM L in a m axim um likelihood search with 100 bootstrap replicates.

Locus-specific genealogies w ere also inferred for COI, 16S, and IT S using bo th M rBayes an d RAxM L. T otal da ta sets for each locus w ere assem bled from all available sequences and then simplified by rem oving any identical haplotypes. Searches were perform ed under the same conditions previously described for 1 6S (no partitions) and for COI (1+2, 3) w ith four separate searches o f 10 m illion steps and the final 25% percen t o f trees retained (effectively a burn-in o f 7.5 m illion steps). Search conditions for the partial nuclear IT S sequences were as above with indels treated as missing da ta and no partitioning.

T h e software Figtree (R am baut: h ttp ://tre e .b io .ed .ac .u k / softw are/fig tree/) was used to assist w ith tree visualization and graphics p reparation .

P h y lo g e o g ra p h ic p a t t e rn sIntraspecific phylogeographic patterns were assessed by exam ­

ining all available COI and 16S sequences for T. crocea, T. maxima, and the distinct clade (Tridacna sp.) identified in the previous analyses. For each species-locus com bination, a heuristic m axi­m um parsim ony search was conducted in PAUP* [47]. Because frequencies o f published haplotypes are no t consistently available, it was no t possible to conduct standard population genetic analyses such as m easures o f diversity and differentiation. For intraspecific parsim ony searches, the m axim um num ber o f trees was set to 1000 in PA U P*[47].

Results

DNA s e q u e n c e sNew D N A sequence da ta was generated for individuals from

five locations (including 55 COI, 65 16S, and 50 IT S sequences: G enbank Ace. Nos. JX 974838-JX 975007). C om bining these new sequence da ta with previously published da ta yielded aggregations o f 405 COI, 132 16S, and 50 IT S sequences for Tridacna species, w ith 335 unique haplotypes for C O /an d 54 unique haplotypes for 16S. In the new data generated for this study nearly all included

PLOS ONE I www.plosone.org 3 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

individuals were sequenced for b o th C O I an d 1 6S allowing us to link results from these two loci and provide a com m on context for the aggregated sequences from previous studies. Similarly, I T S sequences were obtained from an overlapping subset o f individuals sequenced for C O I an d 1 6S. Nexus files have been deposited in T reebase (h ttp :// p u rl.o rg /p h y lo /treeb ase /p h y lo w s/s tu d y /TB2:S13501).

P h y lo g en e t ic ana lysesPhylogenetic analyses resulted in well-resolved topologies

defining several clades w ithin Tridacna. T ree topologies for the concatenated and single gene datasets w ere similar (Figs. 2—T), providing evidence for a robust an d consistent phylogenetic signal. T h e concatenated analyses o f m itochondrial C O I and 1 6 S loci (Fig. 2) strongly support m onophyly o f T. squamosa, T. crocea, and a previously undescribed clade (but reported in [16]) form ed well- supported term inal taxa, w ith m ore m odest support for the m onophyly o f T. maxima. This undescribed clade (which we refer to as Tridacna sp.) was also well supported in single gene analyses o f C O I and 1 6 S (Fig. 3) an d I T S (Fig. 4). T. sp. sequences were evolutionarily distinct from other species; the average pairwise C O I sequence divergence betw een T. sp and T. crocea was 14.4% and

was 12.6% betw een T. sp. and T. squamosa, as com pared to 9.5% betw een T. crocea and T. squamosa (uncorrected pairw ise distances).

G ene trees for C O I and 1 6 S show concordan t relationships am ong species (Fig. 3), confirm ing that independent research groups have sam pled similar genotypes. T h e notable exception to the consistency across studies was the 1 6S T. derasa sequence from [15] w hich did no t cluster consistently with our 1 6 S T. derasa sequence (specimen ET358) even though our C O I sequence from this same individual clustered w ith o ther T. derasa sequences including G Q 1 66591 from [48], For this reason, the T. derasa sequence from [15] was re ta ined in the 1 6 S tree, bu t excluded from the jo in t C O I and 1 6 S searches. All m tD N A -based genealogies supported T. squamosa and T. crocea as sister species (Figs. 2 and 3) w hereas I T S based analyses gave m odest support for T. sp. an d T. crocea as sister species (Fig. 4). W ith in the m tD NA- based analyses, T. derasa, T. gigas, an d T. mbalavuana appear consistently as basal lineages w ithin Tridacna (Figs. 2 and 3). (No I T S sequences w ere available for these taxa.)

P h y lo g e o g ra p h ic p a t t e rn sW ithin T. crocea and T. maxima, there was broadscale

phylogeographic concordance o f m tD N A gene trees (as shown in Fig. 5). T. crocea and T. maxima haplotypes from the Solom on

100, 96

100. 100

100. 100

98. 99

100, 100

99, 56

81. 5363

98. 100100

Tridacna squamosa (EU 961: JX974924, JX974840) Tridacna squamosa (ET358: JX974925, JX974841)

— Tridacna squamosa (EU346363)Tridacna squamosa (AM909753)

I— Tridacna crocea (EU341346)1— Tridacna crocea (EU341354)

Tridacna crocea (ET544: JX974956, JX974872) Tridacna crocea (ET1879: JX974945, JX974861)

r Tridacna sp. (DQ168140, DQ119339)Tridacna sp. (ET920: JX974910, JX974879)Tridacna sp. (ET918: JX974908, JX974878)Tridacna sp. (ET2339: JX974919, JX974895) Tridacna maxima (ET2386: JX974932, JX974849) Tridacna maxima (ET303: JX974936, JX974855) Tridacna maxima (AM909743)Tridacna maxima (DQ155301, DQ115320)

100, 58 r Adacna squamosina (AM909732)Tridacna squamosina (AM909726)Tridacna derasa (ET356: JX974957, JX974902) Tridacna derasa (GQ166591)Tridacna gigas (AF122975)Tridacna mbalavuana (AF122977)Hippopus porcellanus (AF122974)Hippopus hippopus (AF122973)

25 20 15 10

Millions of years before present

Figure 2. Species relationships within Tridacna based on concaten ated m itochondrial DNA (COI and 755) seq u en ces. The to p o lo g y sh o w n is a tim e ca lib ra ted m ax im um c lad e credib ility tre e in ferred w ith StarBEAST u n d e r a b ir th -d e a th m o d e l. Bayesian p o s te r io r p ro b ab ilities from StarBEAST a n d M rBayes a re a b o v e b ra n c h es an d RAxML b o o ts tra p s u p p o r t p e rc e n ta g e s a re b e lo w b ran ch es . Individuals w ith tw o access ion n u m b e rs inc lude b o th COI a n d 7öS se q u e n c e s . Individuals th a t a re u n d e rlin e d a lso a p p e a r in Fig. 4. do i:10 .1371/jo u rn a l.p o n e .0080858 .g002

PLOS ONE I www.plosone.org 4 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

5 5 \COI 16S

Tridacnacrocea 100

10098

100

100

Tridacnasquamosa

77100100

100 .100 Tridacna sp.

9 9 /100

100i

100100100

Tridacnamaxima

100

10066

54

Tridacnasquamosina' 100Tridacna derasa-

Tridacna derasa -

\100/ /--------------------MOO"Hippopus porcellanus-----------

Hippopus hippopus

100

Tridacna gigasTridacna mbalavuanaTridacna derasa

Cerastoderma glaucumFragum sueziense

Corculum cardissa

2.0 0.3

PLOS ONE I www.plosone.org 5 November 2013 | Volume 8 | Issue 11 I e80858

Giant Clams

Figure 3. Bayesian ph ylogen etic trees for m itochondrial COI and 16S. M rBayesian co n se n su s tre e s c o n s tru c te d fo r e ac h g e n e reg io n using all availab le d a ta . A lthough d iffe ren t sp ec ies a n d reg io n s h ave d ifferential re p re sen ta tio n , th e tw o g e n e tre e s a re c o n co rd an t, as is e x p e c te d fo r linked loci. Thus, overall p a tte rn s a re c o n s is te n t a m o n g resea rch g ro u p s . Branch co lo rs c o rre sp o n d to d is tin c t lin eag es w h o se g e o g ra p h ic d is trib u tio n s a re d e sc rib ed in Fig. 5.do i:10 .1371/jo u rn a l.p o n e .0080858 .g003

Islands, the T orres S trait and L izard Island (and additionally w estern N ew G uinea/C enderw asih Bay, L ihou R eef an d H eron Island for T. maxima) form ed a distinct m onophyletic ‘Pacific’ group (colored blue in Fig. 5). Sequences from the Sunda Shelf form ed a second m onophyletic group (colored orange in Fig. 5) as described in the original publications [24-26], although the location or the genetic b reak differed slightly for each species. Finally, sequences from Indonesia, Singapore, w estern New G uinea/C enderw asih Bay an d T aiw an form ed a th ird group (black in Fig. 5). M ost sequences published in G enbank are not georeferenced. W e were, however, able to deduce the distinct clades typifying m ajor regions from previously published surveys by recreating previously published analyses; T. crocea (yellow haplotypes o f [24], grey clade o f [26]) are shown in green and orange respectively, and T. maxima (yellow haplotypes o f [25]) are shown in blue and orange respectively in Fig. 5.

For T. maxima, the northw est New G uinea clade form ed a cluster w ith the Pacific clade, although no haplotypes were shared betw een the two locations. For T. crocea, however, haplotypes from northw est N ew G uinea and the w estern Pacific were m em bers o f two distinct m onophyletic groups: the Pacific (blue) and the W allacea (black) groups (Fig. 5). T he T. crocea and T. maxima 1 6 S .sequences from [15], described as having been obtained from individuals sourced from aquarium stores, bo th fell w ithin Pacific haplotype groups, suggesting that these purchased specimens had a Pacific origin.

Despite the reduced sam pling for T. sp., a ‘Pacific’ lineage was similarly positioned in the Solom on Islands, and a distinct lineage, com prising samples from w estern A ustralia and T aiw an, geo­graphically overlapped with the W allacea (black) lineage portrayed in T. crocea and T. maxima. Sim ilar phylogeographic patterns were evident for C O I and 1 6 S for each species despite only partially overlapping sets o f individuals form ing the basis for each tree.

96

100100

52

7089

97

-E_ Tridacna squamosa 9 0 ,— c

"ggi_F Tridacna sp. Tridacna crocea

9887

Er Tridacna maxima

Tridacna derasa

0.2

Figure 4. Species relationships within Tridacna based an ITS MrBayes consen su s tree. U nalignab le reg io n s h av e b e e n ex c lu d ed . Bayesian p o s te r io r p ro b a b ilitie s a re a b o v e b ra n c h e s a n d RAxML b o o ts tra p s u p p o r t p e rc e n ta g e s a re b e lo w b ran c h es . do i:10 .1371/jo u rn a l.pone .00 8 0 8 5 8 .g 0 0 4

Discussion

Despite their distinct shell m orphology an d longstanding cultural and com m ercial significance, our da ta reveal cryptic diversity w ithin giant clanis. H ere, we find a previously undescribed clade o f Tridacna (Tridacna sp.). T his clade is supported by bo th m tD N A and nuclear gene regions (Figs. 2-4), which identify it as a unique, evolutionarily significant unit [49] with reference to previously described species. O u r m olecular phylo­genetic analyses place T. sp. as a sister clade to T. squamosa a n d /o r T. crocea, bu t in no instance was a close relationship betw een T. sp. and T. maxima suggested in our gene trees. Thus, m olecular data do no t support T. sp. being a variety o f T. maxima as was suggested by T an g [16], C lanis with T. sp. m itotypes were found bo th at N ingaloo R eef in w estern A ustralia an d in the Solom on Islands. A lthough only T. sp. and T. squamosa were identified am ong our clam samples from N ingaloo, it is likely th a t T. maxima also occur at N ingaloo (Penny unpub., [50]), and we found T. sp. sym patric with T. maxima and T. crocea in the Solomons.

T h e T. sp. clade includes the single haplotype (C O I and a 16S) described from T aiw an [16]. T an g et al. (2005) suggested that there are m orphological differences betw een T. sp. and T. maxima, including m antle pa tte rn , shell lip shape, posterior adductor weight an d the position o f the incurren t aperture. Qualitative exam ination o f an individual from Ningaloo R eef w ith T. sp. m tD N A shows shell characters typical o f T. maxima', asym m etry of the valve with posterior elongation and dense rows o f scales on folds (Fig. 6). T. maxima is well know n for its m orphological variability [51] and thus it is possible that previous m orphological exam inations o f T. sp m ay have been identified it as T. maxima. (Additional m orphological samples are no t presently available as m ost collecting perm its only allow non-lethal sam pling of giant clanis.) O u r findings, therefore, lend support to T an g ’s conclusion that T. sp. is an undescribed species bu t we show that, ra th e r than being a narrow -range endem ic (such as Tridacna rosewateri from M auritius [10]), T. sp. is widely distributed. A lthough it is not possible a t present to delineate the distribution o f T. sp., it seems probable that T. sp. occurs at locations in betw een Australia, T aiw an an d the Solom on Islands. T. sp. individuals from the w estern Pacific were reciprocally m onophyletic from the individ­uals from N ingaloo (Indian O cean) an d the single sequence from T aiw an (Fig. 5).

M tD N A genealogies place T. sp as sister species to T. crocea and T. squamosa, w ith strong support for m onophyly o f this group of three species (Figs. 2 and 3). Tridacna maxima an d T. squamosina form ed a second clade, bu t w ith less support across phylogenetic analyses (Fig. 2) p robably because only 1 6 S sequences were available for T. squamosina. M onophyly o f T. crocea and T. squamosa was reported in previous m tD N A based phylogenetic analyses [9,15], bu t no t in allozyme analyses [14] w here T. squamosa was sister to T. crocea and T. maxima. M onophyly o f the Chametrachea subgenus (including T. squamosina, T. crocea, T. maxima, T. sp. and T. squamosa) [15,44] was supported in individual gene analyses and the concatenated StarB EA ST searches (Fig. 2). M onophyly o f the Tridacna subgenus (including T. derasa, T. mbalavuana, an d T. gigas) was no t well supported in any of ou r m tD N A analyses, w ith these taxa appearing basal to the Chametrachea, bu t m issing an d non­overlapping da ta m ay have con tribu ted to the low resolution.

PLOS ONE I www.plosone.org 6 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

Tridacna crocea

100

COI 100

n = 206

16S

n = 28

Tridacna sp.

J J - 100

\n = 22

>n = 22

Tridacna maxima

100

100

4.0n = 147

n = 33

Figure 5. Unrooted parsim ony trees and sam pling locations for Tridacna crocea, Tridacna sp., and Tridacna maxima. M ajor lineages on n e tw o rk s a re co lo red a n d th e g e o g ra p h ic e x te n t o f e ac h lin eag e is in d ica ted o n th e m ap . Relative freq u en c ie s o f e ac h h a p lo ty p e a re n o t d e p ic ted ; e ac h h a p lo ty p e is sh o w n in eq u a l size (see tex t). D ots o n m ap s in d ica te sam p lin g loca tions a n d loca tions w ith tw o d is tin c t sym patric lineages a re sh o w n as b isec ted circles (n o t ind ica tive o f re la tive frequenc ies). S u p p o r t for m o n o p h y ly o f m ajo r c lad es a m o n g COI t re e s is b a se d o n 100 p e rc en t c o n s is ten cy o f e ac h b ran ch a m o n g all equally p a rs im o n io u s tre e s (a ran d o m ly c h o se n tre e is d ep ic ted ) . A m ong 76S tre e s , th e sing le m ost p a rs im o n io u s tre e for T. sp. a n d T. maxima a re sh o w n , a n d for T. crocea b o th g re e n a n d b lue lineages w e re p re s e n t in all six eq u ally pars im o n io u s tre e s . Colors in d ica te g e o g ra p h ic loca tions o f h a p lo ty p e s a n d in te rna l b ra n ch es a re in gray. d o i:1 0 .1371 /jou rna l.pone.0080858 .g005

Previous phylogeographic studies o f T. crocea [24,26,29] and T. maxima [25] from Indonesia show geographic restriction o f several clades. T he m tD N A gene trees w ithin these papers delineate clusters com prising haplotypes from w estern Sum atra (Sunda), W allacea, and northw est N ew G uinea (Sahul) [24—26,29] with some m ixing betw een clades particularly in the B ird’s H ead Peninsula o f northw est N ew G uinea [26], O u r samples showed an additional and deeper evolutionary break for T. crocea to the east o f C enderaw asih Bay, w hereby individuals from the Solom on Islands, T orres Strait, and G reat B arrier R eef form a m onophy­letic group and do not share any m tD N A haplotypes with northw est N ew G uinea or locations in W allacea (Fig. 5). Therefore, it appears that the distinct clade o f T. crocea haplotypes from northw est N ew G uinea (with some spillover westw ard into W allacea [26]) is regionally endem ic and does not extend into the west Pacific. These patterns are not due to differences in D N A sequencing in terpreta tion betw een research groups, as samples (from [24,26,41]) are m utually consistent and a single T. crocea (from [15]) falls w ithin the larger Pacific T. crocea clade. Based on present sampling, we can place this newly discovered genetic discontinuity betw een C enderaw asih Bay and the Solom on Islands in the n o rth and betw een the A ra Basin and T orres S trait in the south. F o r T. maxima, in contrast, the distinct haplotypes from northw est N ew G uinea fall in the same clade as west Pacific

haplotypes. T hus the northw est N ew G uinea clade o f T. maxima can now be viewed as a w estward extension o f Pacific variants, albeit w ith no shared haplotypes betw een locations.

W ith only two species to com pare, we can only speculate as to why the m tD N A patterns differ betw een species, although greater overall population genetic structure in T. crocea com pared to T. maxima is consistent w ith previously co-sam pled regions (for instance, [24] in com parison to [25]). Because o f the diffuse sam pling for T. crocea, we cannot p inpoint a specific location of geographic differentiation east o f C enderaw asih Bay, yet at a m acroscale this observation is consistent w ith m tD N A patterns in a butterflyfish [52], a reef fish [53], and a sea star (Crandall pers. comm.) and m ay be associated w ith a long stretch (> 7 0 0 km) of coastline east o f C enderaw asih Bay w ith sparse reef hab ita t [54],

In T. maxima, we found th at Solom on Islands haplotypes cluster with haplotypes from the G reat B arrier Reef; this affinity contrasts with allozyme results that show substantial divergence betw een Solom on Islands and G reat B arrier R eef populations [33], T he nature o f these differing patterns cannot be explored further as allozyme results are not directly com parable across research groups.

T he broadscale geographic and m ultispecies phylogenetic results o f this study, consolidated w ith those o f previous investigations, reveal new aspects o f regional patterns and

PLOS ONE I www.plosone.org 7 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

An terio r Posterio r

Figure 6. An individual with Tridacna sp. mtDNA dem onstrating valve m orphology consisten t with Tridacna maxima. A) Tridacna maxima from H ibernia Reef, WA, Australia. A ccession N o # P.52722 (M useum Art Gallery N orthern T erritory (MAGNT)), orig inal iden tifica tion b a se d on m o rp h o lo g y , B) Tridacna sp. from Five F inger Reef, S o u th o f Coral Bay, N ingaloo M arine Park, WA, Australia. A ccession N o # . P.51911 (M useum Art Gallery N orthern T erritory (MAGNT)), C) Tridacna maxima, from n o rth w e s te rn WA, A ustralia, u n re g is te re d (M useum Art Gallery N orthern Territory (MAGNT)). P h o to c red it: S h an e Penny. do i:10 .1371/jo u rn a l.pone .00 8 0 8 5 8 .g 0 0 6

highlight key uncertainties in the curren t knowledge o f Tridacna. A com m on result am ong population genetic studies o f Tridacna species to date is that there is substantial population structure. Such genetic differentiation m ay be due in pa rt to the relatively short planktonic larval duration o f approxim ately 9 days [12] that is likely to restrict dispersal distances. T h e discovery o f an undescribed species adds to o ther recent species discoveries in

Tridacna [9—11], bu t the b ro ad distribution of T. sp. illustrates that cryptic species can rem ain undetected even in such conspicuous groups as giant clanis.

Both the discovery o f a new species and the observation o f substantial geographic differentiation a re relevant to m onitoring o f local stocks an d hu m an transport o f clanis. First, the presence o f a cryptic sym patric species w ould result in overestim ates o f species

PLOS ONE I www.plosone.org November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

abundance w here clam populations are censused. Second, hum an- aided m ovem ents could cause species to be in troduced to regions outside their natural range and, similarly, are likely to introduce foreign genetic m aterial into local populations. Tridacna maxima, T. squamosa, T. derasa, T. mbalavuana and T. gigas were frequendy translocated during the 1980’s and 1990’s (some hum an assisted m ovem ents continuing into this century) by governm ental, com m ercial and conservation organizations to com bat local depletion and facilitate the live culture trade [55]. T h ird , depleted populations are unlikely to receive im m igrants from geographically d istant locations via planktonic dispersal and, therefore, recovery m ay be slow or negligible even w hen local harvesting has ceased. Results from g iant clams underscore two im portan t them es em erging from genetic investigations o f m arine organisms: cryptic species are com m on [19,20,56,57], an d m any species are genetically heterogeneous across their geographic range [58].

Supporting Information

Document SI Genbank accession numbers for all included sequences.(XLS)

References1. H eslin g a G (1996) C lam s to cash: H o w to m ak e a n d sell g ian t c lam shell

p ro d u cts . H aw aii, U S A : C e n te r fo r T ro p ic a l a n d S u b tro p ica l A q u acu ltu re . P u b lica tio n N u m b e r 125: 47.

2. W a b n itz C , T a y lo r M , G re e n E , R a z a k T (2003) F ro m o c e an to a q u a riu m : T h e g lobal t ra d e in m arin e o rn a m e n ta l species. C am b rid g e , U K : U N E P W o rld C o n se rv a tio n M o n ito rin g C e n tre . 6 4 p.

3. T isdell C , M e n z K M (1992) G ia n t c lam fa rm in g a n d susta in ab le d eve lopm ent: A n overview . In : T isd ell C , ed ito r. G ia n t c lam s in th e susta inab le d e v e lo p m en t o f th e S o u th Pacific: S o c ioeconom ic issues in m aricu ltu re a n d conserva tion . C a n b e r ra , A ustralia : A C IA R p p . 3 -1 4 .

4. T o m asc ik T , M a h A J, N o n tji A , M o o sa M K (1997) T h e ecology o f the In d o n es ia n seas: P a r t one. H o n g K o n g : P erip lus Ed itions. 656 p.

5. W ells S M (1997) G ia n t clam s: S tatus, t ra d e a n d m aricu ltu re , a n d th e ro le o f C IT E S in m an a g e m en t. G la n d , S w itzerlan d a n d C a m b rid g e , U K : IU C N -T h e W o rld C o n se rv a tio n U n io n . 77 p.

6 . U N E P -W C M C (2013) C o n v e n tio n o n in te rn a tio n a l t ra d e in e n d a n g e re d species o f w ild fau n a a n d flora . A vailable: w w w .cites.o rg . A ccessed 2013 O c to b e r 20.

7. IU C N {2013) IU C N re d list o f th re a te n e d species. A vailable: w w w .IU C N red lis t. org. A ccessed 2013 O c to b e r 20.

8 . b in O th m a n A S , G o h G H S , T o d d P A (2010) T h e d istr ib u tio n a n d s ta tus o f g ian t c lam s {family T r id a c n id ae ) - a sh o rt review . R affles Bull Z oo l 58: 103—111.

9. R ic h te r C , R o a -Q u ia o it H , J a n tz e n C , A l-Z ib d a h M , K o ch z iu s M {2008) C o llap se o f a n ew liv ing species o f g ian t c lam in th e R e d Sea. C u r r B iol 18: 1 3 4 9 -1 3 5 4 .

10. S irenko B I, S carla to O A {1991) Tridacna rosewateri sp. n ., a n e w species o f g ian t c lam fro m th e In d ia n O c e a n . L a C o n ch ig lia 22: 4—9.

11. L u c a sJS , L ed u a E , B raley R D {1991) Tridacna tevoroa L ucas, L e d u a a n d Braley: A recen d y -d esc rib ed species o f g ian t c lam {Bivalvia; T r id acn id ae ) f ro m Fiji a n d T o n g a . N au tilu s 105: 9 2 -1 0 3 .

12. L ucas J S {1988) G ia n t clam s: D esc rip tio n , d istr ib u tio n a n d life h istory. In: C o p la n d J W , LucasJS , ed ito r. G ia n t c lam s in A sia a n d th e Pacific, p p . 21—32.

13. M a ru y a m a T , Ish ik u ra M , Y am azak i S, K a n a i S {1998) M o lecu lar phy lo g en y o f z o o x an th e lla te bivalves. B iol Bull 195: 70—77.

14. B e n z ie J A H , W illiam s S T {1998) P hy logenetic rela tio n sh ip s a m o n g g ian t c lam species {M ollusca: T r id a c n id ae ) d e te rm in e d b y p ro te in e lectrophoresis . M a r Biol 132: 1 23-133 .

15. S c h n e id e r J A , O T o ig h il D {1999) P h y logeny o f g ian t c lam s {C ardiidae: T r id a c n in ae ) b a sed o n p a rtia l m ito c h o n d ria l 16S rD N A gene sequences. M ol P hy l E vol 13: 5 9 -6 6 .

16. T a n g Y C {2005) T h e system atic sta tus o f Tridacna maxima {Bivalvia: T rid acn id ae ) b a sed o n m o rp h o lo g ica l a n d m o lecu la r ev idence. M a s te r’s thesis. K ee lu n g , T a iw a n , R e p u b lic o f C h in a : N a tio n a l T a iw a n O c e a n U niversity . 80 p.

17. H o ek sem a B {2007) D e lin ea tio n o f th e In d o -M a la y a n cen tre o f m ax im u m m arin e biodiversity : T h e C o ra l T rian g le . In: R e n e m a W , ed ito r. B iogeography , tim e, a n d place: D istrib u tio n s, b a rrie rs , a n d islands, p p . 117—178.

18. E k m a n S {1953) Z o o g eo g rap h y o f th e sea. L o n d o n : S idgw ick a n d Ja ck so n . 417 p.19. V o g ler C , B enzie J , Lessios H , B a rb e r P, W ö rh e id e G {2008) A th re a t to co ra l

reefs m ultip lied? F o u r species o f cro w n -o f-th o rn s starfish. B iol L e tt 4: 69 6 —699.20. D u d a T F , K o h n AJ, M a th e n y A M {2009) C ry p tic species d iffe ren tia ted in Conus

ebraeus, a w id esp read tro p ica l m arin e gastro p o d . Biol. Bull. 217 :292—305

A cknow ledgm entsSampling in the Coral Sea was supported by the M arine Division of the Australian Governm ent D epartm ent o f Sustainability, Environment, W ater, Population & Communities. We are grateful to the staff of the Australian Museum’s Lizard Island Research Station and the H eron Island Research Station for their facilities and support. Sampling in the Torres Strait Islands was assisted by the staff and students o f Tagai State College, Thursday Island Primary and the Torres Strait Regional Authority. Sampling in the Solomon Islands was made possible via the Pacific Strategy Assistance Program within the Australian Government D epart­m ent o f Climate Change and Energy Efficiency and with the assistance of the Roviana Conservation Foundation. We especially thank JD Aguirre, S Albert, A Denzin, N Gemmell, M Jim uru, F M acGregor, V M cG rath (Senior Com m unity Liaison Officer, Land and Sea M anagem ent Unit, Torres Strait Regional Authority), A Mirams, R Pearce, Stephen, Lavud and Takenda for their logistical support and field assistance. JS Lucas, LG Cook, A Toon, L Pope and JM Pandolfi provided helpful comments and suggestions, as did several anonymous reviewers.

Author ContributionsConceived and designed the experiments: T H JK LL SP EAT CR. Performed the experiments: T H J K LL SP EAT CR. Analyzed the data: T H JK SP CR. Contributed reagents/m aterials/analysis tools: T H J K SP EAT CR. W rote the paper: C R JK T H LL.

21. C a rp e n te r K E , B a rb e r P H , C ra n d a ll E D , A b la n -L ag m a n M C A , A m b ariy an to , e t a l. {2010) C o m p a ra t iv e p h y lo g e o g ra p h y o f th e C o ra l T r ia n g le a n d im plica tions fo r m ar in e m an a g e m en t. J M a r B iol 2011: 1—14.

22. B a rb e r P H , E rd m a n n M V , P a lu m b i S R {2006) C o m p a ra tiv e p h y lo g eo g rap h y o f th re e co d is trib u ted s to m a to p o d s: O rig in s a n d tim in g o f reg io n a l lin eag e d iversification in th e C o ra l T rian g le . E vo lu tio n 60: 1825—1839.

23. V o ris H K {2000) M ap s o f P le istocene sea levels in so u th e a s t Asia: Shorelines, r iver system s a n d tim e d u ra tio n s. J B iogeogr 27: 1153—1167.

24. K o ch z iu s M , N u ry a n to A {2008) S tro n g g en etic p o p u la tio n s tru c tu re in the b o r in g g ia n t c lam , Tridacna crocea, a c ro ss th e In d o -M a la y a rc h ip e la g o : Im p lica tio n s re la te d to ev o lu tio n ary processes a n d connectiv ity . M o l E co l 17: 3 7 7 5 -3 7 8 7 .

25. N u ry a n to A , K o ch z iu s M {2009) H ig h ly res tric te d g ene flow a n d d eep e v o lu tio n ary lineages in th e g ian t c lam Tridacna maxima. C o ra l R eefs 28: 60 7 — 619.

26. D eB o er T S , S ub ia M D , A m b a riy a n to , E rd m a n n M V , K ovitv o n g sa K , e t al. {2008) P h y lo g eo g rap h y a n d lim ited genetic connectiv ity in th e en d a n g e re d b o r in g g ian t c lam across th e C o ra l T rian g le . C o n s Biol 22: 1255—1266.

27. R av ag o -G o tan c o R G , M ag sin o R M , J u in io -M e ñ e z M A {2007) In flu en ce o f the N o r th E q u a to ria l C u r re n t o n th e p o p u la tio n genetic s tru c tu re o f Tridacna crocea {M ollusca: T r id acn id ae ) a lo n g th e ea ste rn P h ilip p in e seab o ard . M a r E co l P rog S er 336: 1 6 1 -1 6 8 .

28. J u in io -M e ñ e z M A , M ag sin o R M , R av ag o -G o tan c o R G , Y u E T {2003) G en ed c s tru c tu re o f Linckia laevigata a n d Tridacna crocea p o p u la tio n s in th e P a law a n shelf a n d shoal reefs. M a r Biol 142: 7 1 7 -7 2 6 .

29. D eB o er T S , N a g u it M R , E rd m a n n M V , A b la n -L ag m a n M C A , A m b a riy a n to , e t al. {2014) C o n c o rd a n t ph y lo g en e tic p a tte rn s in fe rred f ro m m ito c h o n d ria l a n d m icrosatellite D N A in th e g ian t c lam Tridacna crocea. Bull M a r Sei in press.

30. M a ca ra n a s J M , A b lan G A , P an te M JR , B enzie J A H , W illiam s S T {1992) G en e tic s tru c tu re o f g ian t c lam {Tridacna derasa) p o p u la tio n s f ro m reefs in the Indo-P acific . M a r Biol 113: 2 3 1 -2 3 8 .

31. L a u ra n t V , P lanes S, S a lv at B {2002) H ig h v ariab ility o f g en e tic p a tte rn in g ian t c lam {Tridacna maxima) p o p u la tio n s w ith in F re n ch Polynesia. B io lJ L in n S oc 77: 2 2 1 -2 3 1 .

32. B e n z ie J A H , W illiam s S T {1995) G e n e flow a m o n g g ian t c lam {Tridacna gigas) p o p u la tio n s in the Pacific does n o t p a ra lle l o c e an c ircu la tio n . M a r Biol 123: 781—787.

33. B enzie J A H , W illiam s S T {1997) G en e tic s tru c tu re o f g ian t c lam {Tridacna maxima) p o p u la tio n s in th e w est Pacific is n o t co n sis ten t w ith d ispersal b y p resen t- d ay o c e an cu rren ts . E v o lu tio n 51: 768—783.

34. B enzie J A H , W illiam s S T {1992) N o genetic d iffe ren tia tio n o f g ian t c lam {'Tridacna gigas) p o p u la tio n s in th e G re a t B a rrie r R eef, A ustra lia . M a r Biol 113: 3 7 3 -3 7 7 .

35. B enzie J A H , W illiam s S T {1992) G en e tic s tru c tu re o f g ian t c lam {Tridacna maxima) p o p u la tio n s fro m reefs in th e w estern C o ra l Sea. C o ra l R eefs 11: 135— 141.

36. G a rd n e r J P A , B oesche G, M ey er J M , W o o d A R {2012) A nalyses o f D N A o b ta in e d f ro m shells a n d b r in e -p re se rv e d m e a t o f the g ian t c lam Tridacna maxima f ro m th e c e n tra l Pacific O c e a n . M a r E co l P ro g S er 453: 297—301.

37. H u e lsk en T , S ch re ib e r S, H o llm a n n M {2011) C O I am p lificatio n success fro m m u cu s-rich m arin e g astro p o d s {G astropoda: N a tic id ae) d e p en d s o n D N A

PLOS ONE I www.plosone.org 9 November 2013 | Volume 8 | Issue 11 | e80858

Giant Clams

e x tra c tio n m e th o d a n d p rese rv in g agent. M itte ilu n g en d e r D eu tsch e n M alak o - 48.zoo log ischen G esellschaft 85: 17—26.

38. F o lm er O , Black M , H o e h W , L u tz R , V rijen h o ek R (1994) D N A p r im e rs foram p lificatio n o f m ito c h o n d ria l cy to ch ro m e c oxidase su b u n it I fro m diverse 49.m e ta z o an in v erteb ra te s . M o l M a r Biol B io tech 3: 2 9 4 -2 9 9 .

39. P a lu m b i S (1996) N u cle ic acids ii: T h e p o ly m erase c h a in reac tio n . In : H illis D MM o ritz C . M ab le B K . ed ito rs. M o lecu lar system atics. S u n d e rla n d , M assach u - 50.setts: S in au er & A ssociates Inc. p p . 205—247.

40. Y u E T , J u in io -M e n e z M A , M o n je V D (2000) S eq u en ce v a ria tio n in the rib o so m al D N A in te rn a l tra n sc rib e d sp acer o f Tridacna crocea. M a r B io tech n o l 2:5 1 1 -5 1 6 . 5 L

41. N e o M L , T o d d P A (2012) P o p u la tio n d ensity a n d genetic s tru c tu re o f th e g ian t c lam s Tridacna crocea a n d T. squamosa o n S in g a p o re ’s reefs. A q u a tic Biology 14:2 6 5 -2 7 5 .

42. R a m b a u t A E (1996) Se-Al: S eq u en ce a lig n m en t ed ito r v 2 .0 a l l . A vailable: ^h t tp : / / t r e e .b io .e d .a c .u k /s o f tw a re / . A ccessed 2013 O c to b e r 20.

43. H e le d J , D ru m m o n d AJ (2010) B ayesian in fe ren ce o f species trees fro mm ultilocus d a ta M o l Biol E vol 27: 5 7 0 -5 8 0 . c „54.

44. H a rz h a u se r M , M a n d ie O , P iller W E , R e u te r M , K ro h A (2008) T ra c in g b ack th e o rig in o f the In d o -P ac ific m ollusc fauna : B asal T r id a c n in a e f ro m the O lig o cen e a n d M io cen e o f th e S u lta n a te o f O m a n . P a laeo n to lo g y 51: 199—213.

45. R o n q u is t F , H u e lsen b eck J P (2003) M rB ayes 3: B ayesian p h y lo g en e tic in ference u n d e r m ix ed m odels. B io in fo rm atics 19: 1572—1574.

46. S tam atak is A, H o o v e r P , R o u g e m o n t J (2008) A rap id b o o ts tra p a lg o rith m fo r 5 5

th e R A x M L w eb-servers. Syst Biol 57: 7 5 8 -7 7 1 .47. Sw offord D L (1998) PA U P*, phy lo g en etic analysis us in g p a rs im o n y {*and o th e r 5 7 .

m ethods). V e rs io n 4. S u n d e rla n d , M assachusetts: S in au er A ssociates. 58.

P lazzi F , C e reg a to A , T a v ian i M , P assam o n ti M {2011) A m o lecu la r p h y logeny o f b iv alv e m ollusks: A n c ie n t rad ia tio n s a n d d iv erg en ces as rev e a led b y m ito c h o n d ria l genes. P L o S O n e 6 : e27147.M o ritz C , D o w lin g T E , B ro w n W M (1987) E vo lu tio n o f a n im a l m ito ch o n d ria l D N A : R e lev an ce fo r p o p u la tio n b io logy a n d system atics. A n n R e v E co l Syst 18: 2 6 9 -2 9 2 .B lack R , J o h n s o n M S , P rin ce J , B rearley A , B o n d T (2011) E v id en ce o f large , lo ca l v a ria tio n s in re c ru itm e n t a n d m o rta lity in the sm all g ian t c lam , Tridacna maxima, a t N in g alo o M a rin e P ark , W e ste rn A ustra lia . M a r F W R es 62: 1318— 1326.R o sew a te r T (1965) T h e fam ily T r id a c n id a e in th e In do-Pacific . M ollusc R es 1:3 4 7 -3 9 6 .M cM illan W O , W eig t L A , P a lu m b i S R (1999) C o lo r p a tte rn ev o lu d o n , a sso rtative m atin g , a n d gene tic d iffe ren tia tio n in b r ig h d y co lo red butterflyfishes (C h aeto d o n tid ae). E v o lu tio n 53: 2 4 7 -2 6 0 .Sim s C A , R ig inos C , B lo m b erg SP, H u e lsk en T , D r e w J , e t al. (2014) C lean in g u p th e b io g eo g rap h y o f Labroides dimidiatus u s in g p h y logenetics a n d m o rp h o m e t­ries. C o ra l Reefs: in press.S p a ld in g M D , F o x H E , A llen G E , D a v id so n N , F e rd a ñ a ZA , e t al. (2007) M a rin e ecoregions o f th e w orld : A b io reg io n aliza tio n o f c o asta l a n d sh e lf a reas.B ioscience 57: 5 7 3 -5 8 3 .K in c h J , T e ite lb a u m A (2009) P ro ceed in g s o f th e reg io n a l w o rk sh o p o n the m a n a g e m e n t o f susta in ab le fisheries for g ian t c lam s (T rid acn id ae ) a n d C IT E S c ap ac ity b u ild ing . N a d i, Fiji: S e c re ta ria t o f th e P acific C o m m u n ity . 5 4 p. B ickford D , L o h m a n D J, S o d h i N S , N g P K , M eie r R , e t al. (2006) C ry p tic species as a w in d o w o n diversity a n d co nserva tion . T r E co l E vol 22: 148—155. K n o w lto n N (1993) S ib ling species in th e sea. A n n R ev E co l Syst 24: 1 8 9 -2 1 6 . H e llb e rg M E (2009) G e n e flow a n d iso la tion a m o n g p o p u la tio n s o f m arin e an im als. A n n R e v E co l Syst 40: 2 9 1 -3 1 0 .

PLOS ONE I www.plosone.org 10 November 2013 | Volume 8 | Issue 11 | e80858