Biodiversity and phylogeography of Northeast Atlantic and Mediterranean sponges - Chapter 4: Cryptic diversity in a "cosmopolitan" excavating sponge : the Cliona aff. celata complex (Porifera, Clionaidae)

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Biodiversity and phylogeography of Northeast Atlantic and Mediterranean

sponges

Bogalho Teixeira Xavier, J.R.

Publication date

2009

Link to publication

Citation for published version (APA):

Bogalho Teixeira Xavier, J. R. (2009). Biodiversity and phylogeography of Northeast Atlantic

and Mediterranean sponges.

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

C

“

”

SPONGE

:

CLIONA

.

(P

, C

)

Joana R. Xavier, Paola G. Rachello-Dolmen, Fernando Parra-Velandia,

Christine H.L. Schönberg, Johannes A.J. Breeuwer & Rob W.M. van Soest

Accepted for publication in Molecular Phylogenetics and Evolution

Over the past several decades molecular tools have shown an enormous potential to aid in the clarification of species boundaries in the marine realm, particularly in morphologically simple groups. In this paper we report a case of cryptic speciation in an allegedly cosmopolitan and ecologically important species – the excavating sponge Cliona celata (Clionaidae, Hadromerida). In the northeast Atlantic and Mediterranean C. celata displays a discontinuous distribution of its putative growth stages (boring, encrusting, and massive) leading us to investigate its specific status. Phylogenetic reconstructions of mitochondrial (COI, Atp8) and nuclear (28S) gene fragments revealed levels of genetic diversity and divergence compatible with interspecific relationships. We therefore demonstrate Cliona

celata as constituting a species complex comprised of at least four morphologically indistinct species, each showing a far more restricted distribution: two species on the Atlantic European coasts and two on the Mediterranean and adjacent Atlantic coasts (Macaronesian Islands). Our results provide further confirmation that the different morphotypes do indeed constitute either growth stages or ecologically adapted phenotypes as boring and massive forms were found in two of the four uncovered species. We additionally provide an overview of the cases of cryptic speciation which have been reported to date within the Porifera, and highlight how taxonomic crypsis may confound scientific interpretation and hamper biotechnological advancement. Our work together with previous studies suggests that overconservative systematic traditions but also morphological stasis have led to genetic complexity going undetected and that a DNA-assisted taxonomy may play a key role in uncovering the hidden diversity in this taxonomic group.

KEYWORDS: cryptic species, phylogeny, Northeast Atlantic, Mediterranean,

CLIONAAFF. CELATASPECIES COMPLEX

INTRODUCTION

The ability of taxonomists to recognize and delineate species boundaries has relied largely on a morphologically-based classification system. However, speciation is not always necessarily paired with morphological differentiation (Avise, 1994). Indeed, since the incorporation of molecular techniques into the field of systematics, the discovery of morphologically-cryptic but genetically-divergent species has increased exponentially over the past two decades (Knowlton, 1993; Bickford et al., 2006).

Cryptic species are especially prevalent in the marine realm and more so in taxonomic groups with few and highly plastic morphological characters. Our failure to detect such species complexes may have profound implications on management and conservation strategies, particularly for commercially-exploited, invasive, endangered or biotechnologically-promising species, and it also hinders our understanding of speciation and biogeographical patterns (Bickford, 2006; Knowlton, 2000).

Sponges constitute one of the dominant invertebrate groups in hard-bottom benthic communities, and play important roles in ecosystem functioning (e.g. Bell, 2008). As prolific producers of bioactive compounds, sponges increasingly attract the interest of the pharmaceutical industry. Yet, they are one of those groups for which conservative systematic traditions allied to the paucity, simplicity and plasticity of morphological characters confounds species boundaries that may lead to an underestimation of the phylum’s biodiversity, but most importantly to ambiguous results in bioassays. Indeed, in recent years an increasing number of studies have revealed cases of cryptic speciation in what had previously been considered widespread and even cosmopolitan species (Table 1).

The family Clionaidae represents a particularly important group of sponges, given their ability to bioerode calcium carbonate substrates such as limestone, coral, mollusk shells and coralline algae (e.g. Rützler, 2002; Schönberg, 2008). They play a key role in reef dynamics (e.g. Goreau & Hartman, 1963; Bell, 2008), and can adversely affect shellfish cultures (e.g. Rosell et al., 1999; Fromont et al., 2005). Some clionaid sponges are presumed to pass through three putative growth stages: boring, encrusting and massive stages, also known as alpha, beta and gamma (Topsent, 1888; Rützler, 2002). However, these forms were never demonstrated to be developmental stages, and they could alternatively reflect the

influence of environmental variation or represent morphologies of different species (see Hartman 1958; Schönberg, 2008).

The present study addresses the species status of an ecologically/ biotechnologically- important and allegedly cosmopolitan excavating ‘species’ - Cliona celata Grant, 1826, which has been widely reported to occur in all three of the above-mentioned growth forms (van Soest et al., 2000; Rützler, 2002) (Fig. 1). Due to its conspicuous habitus (especially in the massive stage) and its effects on major shellfish cultures (in the boring stage), C. celata has been one of the most widely cited species in the sponge literature since its description. Its presence has been extensively documented in the Atlantic, Indian, and Pacific Oceans, as well as in the Mediterranean and other enclosed seas (e.g. Topsent, 1900; Volz, 1939; Vacelet & Vasseur, 1971; Schönberg, 2000; Calcinai et al., 2000). With a skeleton composed of a single type of spicule (tylostyles 140-360 μm in length; Rosell & Uriz, 2002) this species lacks other distinctive morphological characters to aid in taxonomical assignment (but see Rosell &

Uriz, 2002; Schönberg et al., 2006). C. celata is regarded as a eurytopic species, able to withstand and adapt to varying environmental conditions such as temperature, hydrodynamism and sedimentation load (Nicol & Reisman, 1976; Carballo et al., 1994). Furthermore, it has been proposed that both macro- (growth form) and micro-morphology (shape and size of the tylostyles) in C. celata constitute an adaptation to changing environmental conditions such as current extremes (Bell et al., 2002).

(a)

(b)

(c) Figure 1. Cliona celata growth forms: (a) - boring or alpha form (© P. Wirtz); (b) - encrusting or beta form (© C. Goodwin); and (c) - massive or gamma form (© B. Picton).

CLIONAAFF. CELATASPECIES COMPLEX Table 1 . Repo rt ed ca ses o f crypt ic specia ti o n in spo nges enz – e n z y me e lec tro p ho re si s; a ll o – a ll o z y mes ; is o – i so z ymes ; m tDNA – m ito cho ndr ia l DN A; C O I – cy to chr o me o x id a se s u b u n it I ; A tp8 – AT P sy ntha se su bu ni t 8 ; r R NA – r ibo so mal R N A; ITS-1 /2 – i nter n al tr a n sc ri bed spa cer s 1 a n d 2 ; ATP Sb -i II – ATP -s yntheta se be ta s u b u n it i n tr on II . Ta x o no m ic grou p Mole cul a r m a rk e r Re fe re nc e Order Fa m ily Putative con sp e cie s en z mt D N A rRNA othe r Ho m o sc le ro p h o rid a P lak in id ae Cort ic ium can d el abrum al lo S o lé -C av a et al ., 19 92 Osc arel la lobu lar is a ll o Bo u ry-Esna u lt et a l., 19 92 P lak in a t ril op ha al lo M u ricy e t al ., 19 96 Had ro m er id a Cl io n aid ae Cli ona v iri d is a ll o B a rb ie ri e t al ., 199 5 C lio na cel at a a ll o B a vestrel lo et a l., 1 99 6 28S B a ru cc a et a l., 2 00 7 C O I, Atp 8 2 8 S th is s tudy Pl a cospon g ii d a e Plac os po n gia s p . ITS's N icho ls & B a rnes , 2 0 05 Sp ir a strel li da e Sp ir ast rell a cf . m ol lis a ll o Bo u ry-Esna u lt et a l., 19 99 Su ber it id a e Suberites f icus is o So lé -C a v a & Tho rp e, 19 86 Tethyi d a e Teth ya r obust a al lo Sar à et a l. , 19 93 C h o ndr o si d a C h o ndr il li d a e Ch on dr osi a ren if or mis al lo L azo sk i et al ., 20 01 Ch on dr il la cf . nucula al lo K lau ta u e t al ., 19 99 CO I Du ra n & Rü tz ler , 2 0 06 P o eci lo sc le ri d a Tedan ii d ae Ted an ia i g nis CO I Wu lf f, 2 00 6 La tru n cu li id a e Latrunculia brev is al lo M iller et al ., 20 01 Ha li cho ndr id a Axi nel li da e A x inel la d amic ornis al lo S o lé -C av a et al ., 19 91 b Di ctyo nel li da e S co pa lin a l op hy ro po da CO I 28S , 18S B lan qu er & U riz , 20 07 Ag ela si d a As tro scl er id a e A st rosclera w il ley ana ITS's Wö rheide et al ., 2 00 0 Ha p lo scl er id a P etr o si d a e Petrosi a f ici form is a ll o B a vestrel lo & S a rà , 1 99 2 Di ctyo cer a ti d a Tho recti d a e Ph y llos p on g ia l amel los a a ll o Ben z ie et a l., 1 99 4 C la thr in id a C la thr in id a e Clat hrin a c lat hru s is o S o lé -C av a et al ., 19 91 a Clat h rin a cerebrum is o S o lé -C av a et al ., 19 91 a Clat hrin a prim or dia lis al lo K lau ta u e t al ., 19 94 Clat hrin a br asi lien sis al lo K lau ta u e t al ., 19 94 Leu cett id a e Leucet ta ch ag osen si s ITS's, 28 S ATP Sb-i II Wö rheide et al ., 2 00 8

However, in the Northeast Atlantic and Mediterranean areas the different growth forms show a discontinuous distribution: the massive form is common off the west coasts of Britain, France, and the Iberian Peninsula but does not occur in the North Sea or the Mediterranean, whereas around the Macaronesian islands only the boring form can be found (J.R. Xavier, pers. obs.). These combined observations prompted us to investigate the phylogenetic relationships among the boring and massive forms in order to assess whether these constitute real growth stages, ecologically-adapted phenotypes of the same species or conversely different species previously undetected by classical taxonomy. For this purpose we sequenced two mitochondrial (COI, Atp8) and one nuclear ribosomal (28S) gene fragments of boring and massive specimens collected throughout the Northeast Atlantic and Mediterranean.

MATERIAL AND METHODS

Sampling

Specimens of Cliona celata Grant, 1826 were collected by scuba diving at nine locations throughout the Northeast Atlantic and the Mediterranean Sea, covering part of the species distribution range (Fig. 2). Morphological analysis of the skeletal features (spicules) showed that specimens conformed to the typological species Cliona celata (see Rützler, 2002 for a full description). Specimens of the congener Cliona viridis (Schmidt, 1862) were collected in areas where both putative species co-occurred and used as an outgroup for the phylogenetic analysis.

Voucher specimens were preserved in 96% ethanol and deposited in the Porifera collection of the Zoological Museum of Amsterdam (ZMAPOR). Sample fragments taken for genetic analyses were preserved in absolute ethanol and kept at -10 °C until further processing. The investigated specimens are listed in Appendix A.

DNA extraction, amplification and sequencing

Total DNA was extracted from the specimens using a DNeasy® tissue kit (QIAGEN), following the instructions of the manufacturer. Partitions of two mitochondrial genes, cytochrome oxidase subunit I (COI) and ATP synthase subunit 8 (Atp8), and one nuclear ribosomal (28S) gene were amplified and

75

CLIONAAFF.CELATASPECIESCOMPLEX sequenced. The COI fragment was amplified using the spongespecific primers PorCOI2fwd(5’–AATATGNGGGCNCCNGGNATNAC–3’)andPorCOI2rev(5’ –ACTGCCCCCATNGATAAAACAT– 3’) developed by the first author. These primers amplify an alternative partition of the COI gene that overlaps approximately 60 base pairs with Folmer’s 3’ partition (Folmer et al., 1994) and includesErpenbeck’s‘I3M11’(Erpenbecketal.,2006).Thispartitionwasshown to be more suitable than Folmer’s for addressing phylogenetic relationships at interspecific or population levels (Erpenbeck et al., 2006; LopézLegentil & Pawlik,2008).AfragmentcomprisingtheATPsynthasesubunit8wasamplified usingtheprimersdscox2f1(5’TGGNGCAAATCATTCNTTTATGC–3’)andds atp6r1 (5’ – CTACATTAAATTGATCAAAATANGC – 3’) (Lavrov D., pers. comm.), and the nuclear ribosomal 28S gene with the primers C1 (5’ – ACCCGCTGAATTTAAGCAT – 3’) and Ep3 (ATKCGYTTCCCTCCYAACGG – 3’)(Borchiellinietal.,2004).

Figure2.SamplinglocationsintheNEAtlanticandMediterraneanSea.Symbolsrefer to the specimens growth forms: O  boring (or alpha) andǻ  massive (or gamma) forms.FilledsymbolscorrespondtoClionaaff.celataandopensymbolsarespecimens of Cliona cf. viridis (used as outgroup). Colors correspond to the clades found in the

BER AZO CAN nIRL GAL BLA TAR nSEA WAL

Amplifications were carried out in 25 μl volume reactions containing 2.5 μl of 10x buffer (Sphaero Q), 4 μl dNTPs (1 mM), 1.6 μl BSA (10 mg/ml), 1.6 μl MgCl2 (25 mM), 0.3 μl (5 U/μl) of Taq polymerase (Sphaero Q), 0.8 μl of each

primer (10 μM), and 1.5 μl of DNA. The amplification profile for COI and 28S was as follows: initial denaturing step of 95 o_{C for 3 min, 36 cycles (94} o_{C for 30 s, 57} o_C

for 45 s, and 70 o_{C for 90 s), and a final extension of 72} o_{C for 10 min. For the Atp8}

the amplification profile was: initial denaturing step of 95 o_{C for 3 min, 36 cycles}

(96 o_{C for 20 s, 45} o_{C for 20 s, and 72} o_{C for 30 s), and a final extension of 72} o_{C for}

10 min.

Amplified products were excised from 1% TAE gels and purified with QIAquick Gel Extraction kit (QIAGEN) following the manufacturer’s instructions. The same primers were used for the sequencing reaction with the ABI-Big-Dye Ready-Reaction, and purified products were sequenced in both directions on an ABI 3700 automated sequencer at the Amsterdam Academic Medical Centre.

Data analysis

The poriferan origin of the obtained sequences was verified through BLAST searches (http://www.ncbi.nlm.nih.gov/BLAST/). Sequences were aligned and edited in the BioEdit Sequence Alignment Editor v. 7.0.0 (Hall 1999). Identical sequences were collapsed into genotypes in DnaSP v. 4.0 (Rozas et al. 2003) (see Appendix A). Phylogenetic reconstructions were performed under maximum likelihood (ML) and Bayesian inference (BI) criteria on the nucleotide data sets, implemented in PAUP* v. 4.0b10 (Swofford, 2002) and MrBayes v. 3.04b (Huelsenbeck & Ronquist, 2001), respectively. The likelihood settings for ML analysis were derived from Modeltest 3.06 (Posada & Crandall, 1998) under the Akaike Information Criterion (AIC) as the best-fitting models for each of the nuclear and mitochondrial genes, as well as for the concatenated dataset. Topological confidence of the trees was assessed by heuristic search with 200 bootstrap replicates of stepwise addition with tree bisection and reconnection branch swapping (TBR). Bayesian inference reconstructions were obtained under one million generations and four Metropolis-coupled Markov chains, sampled every hundredth generation and with a burnin of 0.2, and the best-fitting model as estimated by Modeltest 3.06 (Posada & Crandall, 1998). The software MEGA (Kumar et al., 2004) was used to calculate mean intra- and inter-clade p-distances.

CLIONAAFF. CELATASPECIES COMPLEX

RESULTS

In total 1375 bp (564 bp of COI, 474 bp of Atp8, and 337 bp of 28S) of 46 ingroup and 17 outgroup specimens were generated. The Atp8 primers failed to amplify this partition in the C. viridis samples. We obtained 15 COI (C1-C15), 12 Atp8 (A1-A12) and six 28S (S1-S6) ingroup genotypes. Once concatenated, the data resulted in 30 distinct genotypes (22 for the ingroup and 8 for the outgroup). All mitochondrial and nuclear genotypes were deposited in GenBank and accession numbers are listed in Appendix A. Sequence data and further morphological information of the examined specimens is also available in the Sponge Barcoding Project homepage (http://www.spongebarcoding.org/).

P h y l o g e n e t i c reconstructions for each of the genes and concatenated data consistently recovered four well-supported clades (bootstraps values above 95% and Bayesian posterior probabilities above 0.99). Clade I comprises all massive specimens from the British Isles, Berlengas and Galicia, along with a single boring specimen from Galicia, while Clade II includes all the boring specimens from the northernmost locations (North Sea, Northern Ireland and Galicia) together with three massive specimens from Galicia. All boring specimens from the southernmost locations grouped in Clade III with an A t l a n t o - M e d i t e r r a n e a n distribution (Taranto Gulf, Canary Islands and Azores). Clade IV, which also contained only boring specimens, is apparently restricted to the

N h Hd π COI Clade I 16 3 0.342 0.00099 Clade II 12 6 0.818 0.00368 Clade III 11 4 0.691 0.00464 Clade IV 7 2 0.476 0.00084 Overall ingroup 46 15 0.885 0.05703 Overall outgroup 17 8 0.853 0.00681 Atp8 Clade I 16 3 0.592 0.00146 Clade II 12 4 0.803 0.00465 Clade III 11 4 0.691 0.00298 Clade IV 7 1 0.000 0.00000 Overall ingroup 46 11 0.901 0.08256 Overall outgroup --- --- --- --- 28S Clade I 16 2 0.125 0.00037 Clade II 12 1 0.000 0.00000 Clade III 11 2 0.545 0.00162 Clade IV 7 1 0.000 0.00000 Overall ingroup 46 6 0.790 0.02973 Overall outgroup 17 1 0.000 0.00000 Table 2. Diversity measures for the COI, Atp8, and 28S datasets of Cliona aff. celata (ingroup) and Cliona viridis (outgroup). Sample size (N), number of haplotypes (h), haplotype diversity (Hd), and nucleotide diversity (π) are indicated

Mediterranean Sea (Taranto Gulf and Blanes). The relationships among the clades are not clear when regarding the individual gene trees. In the concatenated dataset, Clades I-III form a monophyletic group with a sister-group relationship to Clade IV (Fig. 3). Specimens from clades I/II and III/IV were found to occur in sympatry at Galicia/Northern Ireland and at Taranto Gulf, respectively (Fig. 2).

Overall ingroup genetic diversities (π) varied from 0.0297 in 28S to 0.0825 in Atp8, with values several orders of magnitude lower for individual clades (Table 2).

Mean COI intraclade divergence varied between 0.1 and 0.5%, whereas mean interclade divergence varied between 6.2 and 8.4%. A divergence of over 14% between ingroup and outgroup specimens was found. Similar intraclade values were detected when looking at the Atp8 gene (0-0.5%), while interclade divergences showed a wider distribution of values ranging from 7.9 to 16%. Comparisons among ingroup and outgroup specimens for this gene were not possible given the lack of outgroup Atp8 sequences. Divergence values found in the nuclear ribosomal 28S gene were lower than those found for the mitochondrial genes: mostly 0% within each clade versus 2.8-5.4% between clades (Fig. 4).

DISCUSSION

Cliona celata as a cryptic species complex

The consistent detection of four well-supported clades, with the various gene fragments, the amount of sequence variation and the clear gap between intra and interclade divergences found in our study, strongly support the existence of a species complex in the studied area hereafter called the Cliona aff. celata complex. The overall sequence variation found in our study for C. celata COI (π = 0.057) was approximately 100-fold higher than the values which have been found at the intraspecific level in other sponge species: Crambe crambe (π = 0.0006; Duran et al., 2004), Astrosclera willeyana (π = 0.00049; Wörheide, 2006), and Xestospongia muta (π = 0.00058; López-Legentil & Pawlik, 2008). However, the sequence variation found within clades I and IV of C. celata is of the same magnitude as the previously mentioned values, suggesting conspecific relationships among specimens of each clade. Clades II and III exhibit intermediate values of sequence variation (π ranging from 0.00368 and 0.00464), which may indicate more structured and partially isolated populations within these clades.

CLIONAAFF. CELATASPECIES COMPLEX

Figure 3. Phylogenetic reconstructions of Cliona aff. celata. Strict-consensus trees are shown with Maximum Likelihood bootstrap values and Bayesian posterior probabilities (ML/BI) for each of the gene fragments and concatenated dataset. Sequence types and geographical labels refer to Appendix A.

Clade I Clade II Clade III Clade IV Hap 1 Hap 2 Hap 4 Hap 5 Hap 3 Hap 6 Hap 7 Hap 8 Hap 12 Hap 13 Hap 14 Hap 9 Hap 10 Hap 11 Hap 15 Hap 16 Hap 20 Hap 19 Hap 18 Hap 17 Hap 21 Hap 22 Hap 23 Hap 24 Hap 25 Hap 27 Hap 26 Hap 28 Hap 29 Hap 30 0.005 substitutions/site 100/1.00 100/1.00 100/1.00 100/1.00 100/---88/1.00 56/0.64 Outgroup Concatenated data

BER; GAL; WAL; nIRL GAL

GAL; nIRL; nSEA GAL

TAR; CAN; AZO

TAR; BLA C1 C2 C3 C4 C6 C5 C7 C8 C9 C10 C11 C13 C12 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 0.01 substitutions/site 99/1.00 99/1.00 97/1.00 96/0.99 100/---COI _A1 A2 A3 A4 A7 A5 A6 A8 A10 A11 A9 A12 0.01 substitutions/site 100/1.00 100/1.00 100/1.00 Atp8 S1 S2 S3 S4 S5 S6 S7 0.005 substitutions/site 100/1.00 ---97/1.00 92/1.00 28S

Mean COI genetic distances found in C. aff. celata clades are in accordance with values previously found for other sponge species: the mean intraspecific distances obtained in these sponges ranged from 0.16% in Tedania ignis to 0.41% in Astrosclera willeyana (Wullf, 2006; Wörheide, 2006). The genetic divergence found in our study for COI and 28S was smaller but of the same order of magnitude as that found in a study of cryptic species of the genus Scopalina across the same geographical area (Blanquer & Uriz, 2007).

Moreover, the high levels of genetic divergence found between clades occurring in sympatry provide strong evidence that members of t h e s e c l a d e s a r e reproductively isolated. Our results further show that the boring and massive growth forms truly constitute either different growth stages or e c o l o g i c a l l y - a d a p t e d phenotypes within the same species, as these consistently group in Clades I and II. The encrusting form of C. celata (rarely reported in the literature) was not found during sampling, which suggests that this form may be either seasonal or adapted to very specific environmental conditions.

Similar to previous reports in Porifera, we have provided conclusive evidence that Cliona celata, which until now has been considered to be a cosmopolitan species, in fact constitutes a species-complex comprising at least four morphologically-similar species, each showing a far more restricted distribution: two species on the Atlantic European coasts occurring in sympatry on the west coasts of the British Isles and Galicia, and two species in the Mediterranean and adjacent Atlantic with a sympatric zone in the Taranto Gulf (Ionian Sea). 0 2 4 6 8 10 12 14 16 18 COI 28S % of sequenc e divergence ( p -dis tance s) Atp8

Figure 4. Schematic representation of intra- (o) and inter-clade (i) divergences as mean p-distances (%) for each of the analyzed genes.

CLIONAAFF. CELATASPECIES COMPLEX

Our results confirm the findings of Bavestrello and colleagues who detected by allozyme electrophoresis the presence of two C. aff. celata species inhabiting the same limestone veins in the Ionian Sea (Bavestrello et al., 1996). More recently, Barucca and colleagues also suggested the presence of more than one C. aff. celata species in the Mediterranean (Barucca et al., 2007). These authors, however, based their study on a small sample set (N=3) and did not include Atlantic specimens, from where C. celata was originally described. Grant (1826), in his original description of C. celata, refers to the boring (alpha) form as the “yellow

fleshy substance [occupying] the perforated shells of the living oyster” in the Firth of Forth (east coast of Scotland). The fact that all boring specimens from the North Sea analyzed in our study group in Clade II suggests that these specimens are probably conspecific with the type material, hence representing C. celata sensu

stricto. Consequently, the specimens occurring in the Mediterranean Sea and the adjacent Macaronesian islands in the Atlantic do not belong to C. celata but constitute different species which await formal taxonomic description. The extensive evidence of cryptic speciation here reported for the Northeast Atlantic and Mediterranean area suggests that records of C. celata from distant locations such as Vietnam (Calcinai et al., 2006), Madagascar (Vacelet & Vasseur, 1971) and the Great Barrier Reef (Schönberg, 2000) will probably be found to constitute different species within the overall species-complex. However, the unclear relationships among the species uncovered in this study suggest that some may have been introduced from distant areas along with the shellfish into which they bore (e.g. the Pacific oyster). The evolutionary history of this morphologically conservative, but genetically diverse and globally-distributed species complex, provides an interesting and challenging topic for future research.

It seems that morphological stasis, i.e., genetic divergence unaccompanied by phenotypic modification, explains why the C. aff. celata complex has remained undetected until now. Our findings for C. celata add to a number of clionaid sponges which have traditionally been regarded as prime examples of cosmopolitanism but that have recently been proposed to represent species complexes (see Tab. 1 in Rosell & Uriz, 1997). Within the C. aff. celata complex, classical morphological characters seem insufficient to permit confident identification of the species; in these cases the study of other characteristics such as: burrowing patterns (e.g. Rosell & Uriz, 2002), papillar morphometric analysis or ecological studies on niche differentiation (e.g. Hartman, 1957) may prove to be useful.

Our study additionally highlights the potential of molecular data such as DNA barcodes to uncover morphologically cryptic species, in a promising field of DNA-assisted taxonomy (Wörheide et al., 2007; Wörheide & Erpenbeck, 2007).

Cause and consequence: cryptic species in Porifera

In a previous study, Klautau and colleagues suggested cosmopolitan distributions in Porifera to result from overconservative systematic traditions (Klautau et al., 1999). From our review of all cases of cryptic speciation in sponges reported in the literature (Table 1) some distinct trends were observed: (i) a clear incidence of crypsis uncovered within the most morphologically-conservative orders (e.g. 32% in Hadromerida; 17% in Homosclerophorida); (ii) a bias towards taxa that constitute the type species for their genera (40%) and taxa described prior to the 1900’s (over 85%). These studies translated into an average ratio of three genetically-detected species to each putative species based only on morphological characters (N=23), and resulted in a significant reduction of the geographical range of the investigated species. These findings are in accordance with the low dispersal potential generally attributed to Porifera. Sponges, as sessile organisms, disperse only by means of lecitotrophic larvae which have a short life span between a few days and two weeks (Maldonado, 2006). Furthermore, several studies have shown sponge larvae to display philopatric behavior, thus remaining in the immediate vicinity of the parental location (Mariani et al., 2005, 2006; Uriz et al., 2008). Therefore, a scenario of genetic connectivity across large geographical areas is unlikely within this taxonomic group. As such, cosmopolitan distributions of sponges should be regarded with skepticism and investigated further.

In some cases, the status of newly-uncovered cryptic species was subsequently confirmed by examining additional morphological characters previously overlooked or regarded as normal intraspecific phenotypic variation (e.g. Blanquer & Uriz, 2008). In other cases, different sources of information such as biochemical (e.g. Loukaci et al., 2004) or ecological data (e.g. Duran & Rützler, 2006; Wulff, 2006) complemented morphology to delineate species (e.g. Rützler et al., 2007; Blanquer & Uriz, 2008). Yet, many of the species discovered using molecular methods have remained in taxonomic crypsis, as no morphologically-diagnostic characters have been found. The morphological stasis displayed by these species suggests that molecular and morphological evolution may often proceed independently within this phylum.

CLIONAAFF. CELATASPECIES COMPLEX

Overall, conservative systematic traditions but also morphological stasis seems to be the main causes of sponge cryptic species going undetected. According to the World Porifera Database there are over 8,200 currently accepted nominal sponge species (van Soest et al., 2008), but several authors estimate that this number probably represents only half of the true total number of extant species (e.g. Hooper & Lévi, 1994). Undetected cryptic species contribute to an overall underestimation of the species-level diversity within this taxonomic group and consequently of its role in ecosystem functioning.

Furthermore, sponges, which often possess antiviral, antibiotic, or analgesic properties, are among the most prolific sources of natural products, and are thus of great pharmacological interest (Munro et al., 1999). In this field of research, if cryptic species remain undetected, results of bioassays may be confounded and biotechnological advancement seriously compromised. This problem is further magnified when the cryptic species occur in sympatry (as is the case for C. aff. celata). An example is the use of ‘Cliona celata’, collected from various parts of the world, in research areas such as immunity or neurodegeneration (e.g. Keyzers et al., 2008); cancer biology (e.g. Castellanos et al., 2006); biomonitoring (e.g. Marques et al., 2006, 2007), as well as the acknowledged impact of this taxon on natural and commercial populations of various shellfish species (Rosell et al., 1999; Wesche et al., 1997). We therefore suggest that a taxonomic validation of specimens, through DNA barcodes, should become a standard procedure in these applied areas of research.

ACKNOWLEDGEMENTS

We are grateful to several colleagues for sending us samples from their study locations: James Bell, Kate Lock, Caterina Longo, Steve Cook, Johanna Frotscher and Claire Goodwin. We also thank Mansour Aliabadian and Annika Hillers for their helpful suggestions about data analyses; Nuno Curado for the construction of Fig. 2; and Elly Beglinger, Peter Kuperus, Betsie Voetdijk, and Lin Dong for technical assistance. We are very grateful to Dennis Lavrov for sharing his unpublished Atp8 primers with us. We further thank Steph Menken, Dirk Erpenbeck, Cleve Hicks and two anonymous reviewers for insightful comments and suggestions on an earlier version of this manuscript. Fieldwork was financially supported by grants from the Schure-Beijerinck Poping Fonds (KNAW, The Netherlands) to J. R. Xavier, the AUV-Fonds (University of

Amsterdam) and Sociedade Afonso Chaves (Azores, Portugal) to P. G. Rachello-Dolmen. This study was funded by Fundação para a Ciência e Tecnologia – FCT, Portugal (grant no. SFRH/BD/16024/2004) as part of J. R. Xavier’s PhD project on the phylogeography and taxonomy of marine sponges from the Azores and North Atlantic Seamounts.

CLIONAAFF. CELATASPECIES COMPLEX A PPE N D IX I. Info rma ti o n o n the i n v es ti g ated speci m en s Spe cime n G rowth form Sa mplin g sta ti o n L o ca tion Code H a plo type s (field code) CO I 2 n d Atp8 2 8 S C lio na a ff. c el at a Clad e I B .05.09.15 m assive (g am m a) B erleng as Island s W coast Portug al B E R C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) B .05.09.77 m assive (g am m a) B erleng as Island s W coast Portug al B E R C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) B .05.09.78 m assive (g am m a) B erleng as Island s W coast Portug al B E R C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) B .05.09.109 m assive (g am m a) B erleng as Island s W coast Portug al B E R C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) B .05.09.110 m assive (g am m a) B erleng as Island s W coast Portug al B E R C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) GA L .05. 05. 04 mas si ve (gam ma) Galicia NW coas t of S pain GA L C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) GA L .05. 05. 05 mas si ve (gam ma) Galicia NW coas t of S pain GA L C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) GA L .05. 05. 07 mas si ve (gam ma) Galicia NW coas t of S pain GA L C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) GA L .05. 05. 16 mas si ve (gam ma) Galicia NW coas t of S pain GA L C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) GA L .05. 05. 51 boring (alpha) G alicia NW coas t of S pain GA L C2 (FJ 670572 ) A3 (GQ 169 270) S 1 (FJ8286 64) CcBRI 1 mas si ve (gam ma) S k omer I sland W coas t Wales WAL C3 (FJ 670573 ) A1 (GQ 169 268) S 1 (FJ8286 64) CcBRI 2 mas si ve (gam ma) S k omer I sland W coas t Wales WAL C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) CcBRI 4 mas si ve (gam ma) S k omer I sland W coas t Wales WAL C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) CcBRI 8 mas si ve (gam ma) S k omer I sland W coas t Wales WAL C3 (FJ 670573 ) A1 (GQ 169 268) S 2 (FJ8286 65) Mc 4067 mas si ve (gam ma) S trangford L o u g h E coas t of N I reland nI RL C1 (FJ 670571 ) A1 (GQ 169 268) S 1 (FJ8286 64) Mc 4038 mas si ve (gam ma) S trangford L o u g h E coas t of N I reland nI RL C1 (FJ 670571 ) A2 (GQ 169 269) S 1 (FJ8286 64) Clad e I I GA L .05. 05. 01 mas si ve (gam ma) Galicia NW coas t of S pain GA L C4 (FJ 670574 ) A4 (GQ 169 271) S 3 (FJ8286 66) GA L .05. 05. 02 mas si ve (gam ma) Galicia NW coas t of S pain GA L C4 (FJ 670574 ) A4 (GQ 169 271) S 3 (FJ8286 66) GA L .05. 05. 06 mas si ve (gam ma) Galicia NW coas t of S pain GA L C5 (FJ 670575 ) A4 (GQ 169 271) S 3 (FJ8286 66)

A PPE N D IX I (cont .). Info rma ti o n o n the i n v esti ga ted spe ci mens Spe cime n G rowth form Sa mplin g sta ti o n L o ca tion Code H a plo type s (field code) CO I 2 n d Atp8 2 8 S Clad e I I (con t. ) GA L .05. 05. 50 boring (alpha) G alicia NW coas t of S pain GA L C6 (FJ 670576 ) A4 (GQ 169 271) S 3 (FJ8286 66) GA L .05. 05. 52 boring (alpha) G alicia NW coas t of S pain GA L C4 (FJ 670574 ) A5 (GQ 169 272) S 3 (FJ8286 66) NT H14 boring (alpha) Zeeland S W coas t of T he Netherland s nS E A C4 (FJ 670574 ) A5 (GQ 169 272) S 3 (FJ8286 66) HL G01a boring (alpha) Helgoland arc hipela go S E part North S ea nS E A C7 (FJ 670577 ) A6 (GQ 169 273) S 3 (FJ8286 66) HL G02a boring (alpha) Helgoland arc hipela go S E part North S ea nS E A C7 (FJ 670577 ) A6 (GQ 169 273) S 3 (FJ8286 66) HL G03a boring (alpha) Helgoland arc hipela go S E part North S ea nS E A C7 (FJ 670577 ) A5 (GQ 169 272) S 3 (FJ8286 66) HL G04a boring (alpha) Helgoland arc hipela go S E part North S ea nS E A C7 (FJ 670577 ) A7 (GQ 169 274) S 3 (FJ8286 66) Mc 4018 boring (alpha) S trangford L o u g h E coas t of N I reland nI RL C8 (FJ 670578 ) A7 (GQ 169 274) S 3 (FJ8286 66) Mc 4054 boring (alpha) S trangford L o u g h E coas t of N I reland nI RL C9 (FJ 670579 ) A7 (GQ 169 274) S 3 (FJ8286 66) Clad e I II APU. 06. 02 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C10 (F J67058 0) A8 (GQ 169 275) S 4 (FJ8286 67) APU. 06. 04 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C10 (F J67058 0) A8 (GQ 169 275) S 4 (FJ8286 67) APU. 06. 06 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C11 (F J67058 1) A9 (GQ 169 276) S 4 (FJ8286 67) APU. 06. 09 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C11 (F J67058 1) A9 (GQ 169 276) S 4 (FJ8286 67) CAN. 07. 09 boring (alpha) T enerif e Canary I sland s CAN C10 ( F J67058 0) A8 ( G Q169 275) S 5 ( F J8286 68) CAN. 07. 28 boring (alpha) T enerif e Canary I sland s CAN C10 ( F J67058 0) A8 ( G Q169 275) S 5 ( F J8286 68) CAN. 07. 35 boring (alpha) T enerif e Canary I sland s CAN C10 ( F J67058 0) A8 ( G Q169 275) S 5 ( F J8286 68) CAP. 05. 09. 23 boring (alpha) Faial Azores Archipelago AZO C10 (F J67058 0) A8 (GQ 169 275) S 5 (FJ8286 68) RI B .05. 09. 02 boring (alpha) Faial Azores Archipelago AZO C12 (F J67058 2) A10 (GQ 16 9277 ) S 5 (FJ8286 68) RI B .05. 09. 08 boring (alpha) Faial Azores Archipelago AZO C12 (F J67058 2) A10 (GQ 16 9277 ) S 4 (FJ8286 67) RI B .05. 09. 09 boring (alpha) Faial Azores Archipelago AZO C13 (F J67058 3) A11 (GQ 16 9278 ) S 4 (FJ8286 67)

CLIONAAFF. CELATASPECIES COMPLEX Clad e I V APU. 06. 01 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C14 (F J67058 4) A12 (GQ 16 9279 ) S 6 (FJ8286 69) APU. 06. 03 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C14 (F J67058 4) A12 (GQ 16 9279 ) S 6 (FJ8286 69) APU. 06. 05 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C14 (F J67058 4) A12 (GQ 16 9279 ) S 6 (FJ8286 69) APU. 06. 07 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C14 (F J67058 4) A12 (GQ 16 9279 ) S 6 (FJ8286 69) APU. 06. 08 boring (alpha) Porto Ces areo S coas t I taly, T aranto Gulf T A R C14 (F J67058 4) A12 (GQ 16 9279 ) S 6 (FJ8286 69) CcJF B la1 boring (alpha) Port L ligat NE coas t of S pain B L A C15 (F J67058 5) A12 (GQ 16 9279 ) S 6 (FJ8286 69) CcJF B la2 boring (alpha) Port L ligat NE coas t of S pain B L A C15 (F J67058 5) A12 (GQ 16 9279 ) S 6 (FJ8286 69) C li o n a c f vi ri di s B .05. 09. 148 boring ( alpha) B erleng as Is land s W coast Portug al B E R C16 (F J67058 6) --- S 7 (FJ8286 70) (o utgr o u p) B .05. 09. 149 boring ( alpha) B erleng as Is land s W coast Portug al B E R C16 (F J67058 6) --- S 7 (FJ8286 70) B .05. 09. 150 boring ( alpha) B erleng as Is land s W coast Portug al B E R C16 (F J67058 6) --- S 7 (FJ8286 70) B .05. 09. 151 boring ( alpha) B erleng as Is land s W coast Portug al B E R C16 (F J67058 6) --- S 7 (FJ8286 70) PI X. 05. 09. 11 boring (alpha ) Pico Azores Archipelago AZO C17 (F J67058 7) --- S 7 (FJ8286 70) FUR. 05. 09. 01 boring (alpha) Faial Azores Archipelago AZO C17 (F J67058 7) --- S 7 (FJ8286 70) FUR. 05. 09. 08 boring (alpha) Faial Azores Archipelago AZO C17 (F J67058 7) --- S 7 (FJ8286 70) PRB .05. 08. 25 mas si ve (gam ma) S .M iguel Azores Archipelago AZO C18 (F J67058 8) --- S 7 (FJ8286 70) PCP. 05. 08. 41 mas si ve (gam ma) S . Miguel Azores Archipelago AZO C19 (F J67058 9) --- S 7 (FJ8286 70) PCP. 05. 08. 43 mas si ve (gam ma) S . Miguel Azores Archipelago AZO C20 (F J67059 0) --- S 7 (FJ8286 70) B L A.06. 23 boring (alpha) B lan es NE coas t of S pain B L A C21 (F J67059 1) --- S 7 (FJ8286 70) B L A.06. 24 boring (alpha) B lan es NE coas t of S pain B L A C16 (F J67058 6) --- S 7 (FJ8286 70) B L A.06. 25 boring (alpha) B lan es NE coas t of S pain B L A C21 (F J67059 1) --- S 7 (FJ8286 70) B L A.06. 26 boring (alpha) B lan es NE coas t of S pain B L A C16 (F J67058 6) --- S 7 (FJ8286 70) MAD. 05. 02. 36 boring (alpha) Mad eira Madeira archipelago MAD C22 (F J67059 2) --- S 7 (FJ8286 70) MAD. 05. 02. 14 boring (alpha) Mad eira Madeira archipelago MAD C22 (F J67059 2) --- S 7 (FJ8286 70) MAD. 04. 02. 15 boring (alpha) Madeira Madeira archipelago MAD C23 (F J67059 3) --- S 7 (FJ8286 70) A PPE N D IX I (cont .). Info rma ti o n o n the i n v esti ga ted spe ci mens