UvA-DARE (Digital Academic Repository)

On the Phylogeny of Halichondrid Demosponges

Erpenbeck, D.J.G.

Publication date 2004

Link to publication

Citation for published version (APA): Erpenbeck, D. J. G. (2004). On the Phylogeny of Halichondrid Demosponges. Universiteit van Amsterdam.

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.

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

Download date:30 Sep 2021 Chapterr 6

Conflictingg phylogenetic signal between nuclear ribosomal andd mitochondrial data in halichondrid demosponges

D.. Erpenbeck, J.A.J. Breeuwer and R.W.M. van Soest

Abstract t

Wee sequenced a fragment of the mitochondrial cytochrome oxidase subunit 1 (COl) for a selected sett of halichondrid demosponges. The topology of the COl is in parts significantly different from molecularr phylogenies obtained from nuclear ribosomal genes. We find no support for a close relationshipp between Halichondriidae and Suberitidae (Hadromerida), as suggested by various 28SrDNAA gene trees. However, the COl gene tree does support previous 28SrDNA results and analysess of biochemical data in finding a close relationship of the order Agelasida to the halichondrid taxa.. The combination of these results therefore suggests that the order Halichondrida should be mergedd with Agelasida.

91 1 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges

Introduction n

"Secondd opinions" are becoming the standard in molecular systematics. In the previous decades molecularr phylogenies were entirely based on single genes from which the species tree was extrapolated.. Phylogenetic hypotheses based on morphological characters were often questioned, if onee single gene tree differed in the certain aspects from the expectations. However, different genes frequentlyy give rise to different phylogenies, particularly if sequences with different evolutionary historiess are regarded. A typical example of such a conflict is the higher phylogeny of sponges, inn which the gene trees reconstructed from Hsp70, cPKC, 28SrDNA, 18SrDNA genes and their combinationss find the three demosponge classes of sponges either monophyletic in almost all possible constellationss (Borchiellini et al. 1998; Kim et al. 1999; Medina et al. 2001), or paraphyletic at the basee of the metazoan tree (Zrzavy et al. 1998), or partially in a sister group relationship with other diploblastss (Lafay et al. 1992; CavalierSmith et al. 1996; Collins 1998). Severall factors can be responsible for the divergent topologies of the reconstructed gene trees. Amongg these may be different evolutionary rates, different DNA repair mechanisms, furthermore thee bacterial origin of mitochondria with its clonal maternal inheritance (but see also Smith and Smithh 2002), which is in contrast to recombination and paralogy of nuclear genes. Not every gene iss a suitable marker for a second opinion. Mitochondrial genes display a relatively homogeneous internall substitution behavior (Simon et al. 1994) similar to nuclear gene clusters like cytoplasmatic ribosomalribosomal genes, which are influenced by mechanisms such as concerted evolution (Hillis and Dixon 1991).. Therefore, genes of the same gene cluster appear unsuitable as secondary opinions because theirr evolution cannot be regarded as independent. Thee vast majority of molecular sponge phylogenies are based on the large and small nuclear ribosomall genes (28S and 18SrDNA) and therefore from the same gene cluster (Kelly-Borges et al. 1991;; Lafay et al. 1992; Kelly-Borges and Pomponi 1994; Chombard et al. 1997; Chombard et al. 1998;; Chombard and Boury-Esnault 1999; Alvarez et al. 2000; McCormack et al. 2002; McCormack andd Kelly 2002). Inn sponge phylogeny second opinions from multiple independent molecular markers are stilll uncommon, although they have successfully been recruited in macro-evolutionary studies inn cnidarians (an overview in Shearer et al, 2002). The first mitochondrial sponge fragment was amplifiedd by Watkins and Beckenbach (1999) in order to compare gene arrangement patterns with otherr diploblasts. The first mitochondrial data set of Wórheide et al. (2000) based on the cytochrome oxidasee subunit 2 revealed low evolutionary rates that were also observed in other diploblast mitochondriaa (Shearer et al. 2002). Erpenbeck et al. (2002, + in press) and Duran et al. (2002) analysedd fragments of the cytochrome oxidase subunit 1 (COl). Schroder et al. (2003), published a phylogeneticc study based on COl and the nuclear tubulin-intron. Here,, we present a COl gene tree of the demosponge order Halichondrida and compare it withh previously established nuclear phylogenies of Erpenbeck et al. (chapter 3). The phylogeny of Halichondridaa is of particular interest in sponge systematics, as their changing systematic positions reflectt recent major changes in demosponge systematics. During the history of sponge systematics demospongess have frequently been separated into two major clades firstly based on morphological criteriaa (see Levi 1957) or by presumed reproductive features (Levi 1953). Levi (1957) regarded the Halichondridaa as morphological intermediates between these two groups and divided their families by theirr assumed oviparous or viviparous reproductive mode. Van Soest (1987,1991) and Hooper (1990) pointedd out the paraphyly of Levi's classification and re-united the families in the order Halichondrida (vann Soest et al. 1990), which consists to date of the families Axinellidae, Dictyonellidae, Bubaridae, Desmoxyidaee and Halichondriidae (van Soest and Hooper 2002). However, the monophyly of the 92 2 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges Halichondridaa and its families still has to be verified because morphological apomorphies are dubious orr non-existent. Erpenbeck et al. (chapter 2 and 3) presented a nuclear gene tree, which implied that thee order Halichondrida is not monophyletic (Fig. 1 right). This was consistent with previous reports thatt indicated a close relationship of Halicondrida families to other demosponge orders, such as Axinellidaee (order Halichondrida) to Agelasidae (order Agelasida) (Lafay et al. 1992; Alvarez et al. 2000)) or Halichondriidae (order Halichondrida) to Suberitidae (order Hadromerida) (Chombard and Boury-Esnaultt 1999; McCormack and Kelly 2002). Based on the latter results Chombard and Boury- Esnaultt (1999) suggested to abandon the taxon Halichondrida in favour of a taxon "Suberitina" which comprisedd the families Halichondriidae (Halichondrida) and Suberitidae (Hadromerida). However, alll these reconstructions are based on the same gene (28SrDNA), although sequences of partially differentt domains were used. These analyses are clearly non-independent and there is a strong need too reconstruct halichondrid gene trees on new, independent markers to get a better insight in the phylogeneticc relationships of this order. Here,, we reconstruct a phylogenetic tree based on a fragment of the mitochondrial cytochrome oxidasee subunit 1 for selected halichondrid species and compare this tree with our nuclear 28S rDNA phylogeny.. We assess the suitability of COl for demosponge phylogeny and specifically test if the taxonomiee relationships found in the 28S analysis of Erpenbeck et al. (submitted, chapter 2 and 3), aree supported by the independent mitochondrial gene tree.

Materiall and Methods

Thee list of species studied and their sample location is given in tab. 1. Samples were either freshly collectedd by SCUBA diving, or taken from collection material of the Zoological Museum Amsterdam (ZMA),, where all vouchers for this investigation are kept. Total DNA was extracted from the choanosomee to reduce the chance of amplifying DNA templates from outside the sponge. We used standardd protocols (see Erpenbeck et al. 2002) or used a DNA extraction kit (Quiamp DNA Mini Kit, Quiagen)) and followed the manufacturer's protocol. DNAA fragments were amplified in a two-step nested PCR. A first PCR product was obtained withh universal PCR primers under temperature regimes that were used in previous studies (Erpenbeck ett al. 2002). 1 ftl of the PCR product was taken as DNA template in a consecutive step in order too specifically re-amplify the sponge fragments. For this step the following specific primers were designed:: COlporFl: 5' CCN CAN TTN KCN GMN AAA AAA CA 3' and COlporRl 5' AAN TGNN TGN GGR AAR AAN G 3' and used under a temperature regime of 3 min 94°C, followed by 355 cycles of (30s 94°C,, 30s 45°C, lmin 72°C) and a final elongation time of 10 min 72°C. All PCR productss were excised from a 2% TAE gel, re-extracted with glassmilk in a 65°C water bath, washed threee times with 80% ethanol, dried and resolved in lOpil H20 followed by ligation in a pGEM T-easy vectorr (Promega) and cloned in E. coli according to the manufacturer's protocol. Plasmid DNA was extractedd in alkaline lysis method (Sambrook 1989) and cycle sequenced (Amersham) with labelled M133 primers. Both strands of the template were sequenced on a LiCor automated sequencer. The sequencess of several specimens were double-checked by re- sequencing of either the PCR products orr the cloned plasmids by direct sequencing with an automated (ABI) sequencer using the BigDye Terminatorr vl.1 according to the manufacturer's protocol. Positions in contradiction between two sequencess were coded with question marks. The sequence alignment was unambiguous and done by eye. . Spongess are filter feeders without a true, sealing epithelium and may therefore contain microbiall symbionts or otherwise ingested DNA templates. To prevent the incorporation of non- spongee sequences in our data set we verified the poriferan origin of all sequences with phenetic BLASTT searches (Altschul et al. 1990) and cladistic tree-reconstructions as described in Erpenbeck

93 3 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges

Order r Family y Species s ZMAA No. origin Agelasida a Agelasidae e AgelasAgelas nakamurai PORR 17662 Sulawesi Agelasida a Agelasidae e AgelasAgelas oroides POR144355 Blanes/E Hadromerida a Suberitidae e AaptosAaptos suberitoides PORR 17498 Sulawesi Hadromerida a Suberitidae e SuberitesSuberites massa POR83277 Dinard/F Hadromerida a Suberitidae e SuberitesSuberites suberia POR97266 North Sea Hadromerida a Suberitidae e SuberitesSuberites virgultosa PORR 12969 North Sea Halichondrida a Axinellidae e AxinetlaAxinetla damicornis PORR 14097 Roscoff/F Halichondrida a Axinellidae e AxinellaAxinella polypoides PORR 14093 Roscoff/F Halichondrida a Axinellidae e AxinellaAxinella verrucosa PORR 14590 Blanes/E Halichondrida a Axinellidae e CymbastelaCymbastela vespertina PORR 11006 Darwin, Austral. Halichondrida a Axinellidae e PtilocaulisPtilocaulis fusiformis PORR 14505 Sulawesi Halichondrida a Desmoxyidae e DidiscusDidiscus oxeata PORR 14326 Curacao Halichondrida a Desmoxyidae e HigginsiaHigginsia mixtal PORR 17677 Sulawesi Halichondrida a Desmoxyidae e HigginsiaHigginsia mixtal PORR 17678 Sulawesi Halichondrida a Desmoxyidae e MyrmekiodermaMyrmekioderma granulatal PORR 17680 Sulawesi Halichondrida a Desmoxyidae e MyrmekiodermaMyrmekioderma granulatal PORR 14530 Sulawesi Halichondrida a Dictyonellidae e AcanthellaAcanthella acuta PORR 14589 Spain Halichondrida a Dictyonellidae e DictyonellaDictyonella sp.I POR14614POR14614 Oman Halichondrida a Dictyonellidae e DictyonellaDictyonella sp.1 PORR 6094 Marseille/F Halichondrida a Dictyonellidae e LiosinaLiosina paradoxa I PORR 14478 Sulawesi Halichondrida a Dictyonellidae e LiosinaLiosina paradoxa 1 PORR 14665 Sulawesi Halichondrida a Dictyonellidae e ScopalinaScopalina lophyropoda PORR 14434 Blanes/E Halichondrida a Dictyonellidae e StylissaStylissa carteri 1 PORR 17666 Sulawesi Halichondrida a Dictyonellidae e StylissaStylissa carteri 1 PORR 17667 Sulawesi Halichondrida a Dictyonellidae e StylissaStylissa flabelliformis PORR 17668 Sulawesi Halichondrida a Dictyonellidae e StylissaStylissa massa 1 PORR 17669 Sulawesi Halichondrida a Dictyonellidae e StylissaStylissa massa 1 PORR 17670 Sulawesi Halichondrida a Dictyonellidae e SvenzeaSvenzea devoogdae I POR176711 Sulawesi Halichondrida a Dictyonellidae e SvenzeaSvenzea devoogdae 1 POR176811 Sulawesi Halichondrida a Halichondriidae e AmorphinopsisAmorphinopsis excavans POR146277 Oman Halichondrida a Hatichondriidae e AmorphinopsisAmorphinopsis siamensis PORR 17672 GulfofSiam Halichondrida a Halichondriidae e AxinyssaAxinyssa ambrosia PORR 14311 Curacao Halichondrida a Halichondriidae e AxinyssaAxinyssa sp. POR145011 Sulawesi Halichondrida a Halichondriidae e AxinyssaAxinyssa topsenti PORR 17682 South Africa Halichondrida a Halichondriidae e CiocalyptaCiocalypta penicillus POR141111 Roscoff/Fr Halichondrida a Halichondriidae e HalichondriaHalichondria (Halichondria) bowerbankiPORR 1768 3 Netherlands Halichondrida a Halichondriidae e HalichondriaHalichondria (Halichondria) paniceaPORR 14125 Roscoff/Fr Halichondrida a Halichondriidae e BymeniacidonBymeniacidon perlevis 1 PORR 14140 Roscoff/Fr Halichondrida a Halichondriidae e HymeniacidonHymeniacidon perlevis 1 PORR 17676 Netherlands Halichondrida a Halichondriidae e TopsentiaTopsentia halichondroides PORR 8450 Sri Lanka Halichondrida a Halichondriidae e TopsentiaTopsentia ophiraphidites PORR 14312 Curacao Haplosclerida a Callyspongiidae e CallyspongiaCallyspongia (Cladochalina) aerizusaPORR 17684 Sulawesi Haplosclerida a Chalinidae e HaliclonaHaliclona (Haliclona) oculata PORR 17501 Netherlands Haplosclerida a Chalinidae e HaliclonaHaliclona (Soestella) xena PORR 17504 Netherlands Haplosclerida a "'Jiphatidae e AmphimedonAmphimedon paraviridis PORR 17685 Sulawesi Haplosclerida a >Jiphatidae e GelliodesGelliodes fibulata PORR 17686 Sulawesi Poecilosclerida a Mycalidae e MycaleMycale (Naviculina) flagellifera PORR 17503 Sulawesi

Tab.1:: Samples used in this analysis with sampling locations.

94 4 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges ett al. (2002). Thee COl sequence of Sarcophyton glaucum (Cnidaria:Anthozoa, Genbank accession number AP064823)) was used as outgroup sequence for all analyses. Phylogenetic reconstructions were performedd under maximum-likelihood and bayesian inference criteria. PAUP*4.10b (Swofford 2002)) was used to for the maximum likelihood analyses for which the relatively best-fitting model wass estimated with the likelihood-ratio test as implemented in MODELTEST 3.06 (Posada and Crandalll 1998). Bayesian inference reconstructions were obtained with MrBayes 3.04b (Huelsenbeck andd Ronqvist 2001) under 1,000,000 generations and four Metropolis-coupled Markov chains. We performedd two different bayesian approaches, (a) with the unpartitioned data set and (b) with a data sett partitioned by the codon positions under the relatively best-fitting substitution model for each partitionn as calculated by MrModeltest (Nylander). The resulting clades were compared with the resultss from the maximum-likelihood analysis. Alternative tree configurations were tested with a heuristicc T-PTPtest (Faith 1991) as implemented in PAUP*4.10b under 1000 permutation steps and a treee buffer limited to 1000 per step. To compare the COl gene tree statistically with a 28SrDNA tree, wee tested trees of identical taxa sets with TreeMap 1.1b (Page 1994) under heuristic searches and both "Yulee (Markowian)" and "Proportional to distinguishable" models under 10.000 random repetitions.

Results: :

Thee amplified fragment had a length of 459 bp, and could be unambiguously aligned using their proteinn coding nature. We calculated the base frequencies of the different codon partitions and observedd a bias in thee first and third position against cytosine in the coding strand: 1st pos: A: 28%; C:: 8%; G: 34%; T: 30% (N=48), 2nd pos: A: 18%; C: 20%; G: 18%; T: 44% (N=48), 3rd pos: (A: 31%;; C: 9%;G: 17%; T: 43% (N=48). MrModeltestt estimated the GTR+G+I model (Rodriguez et al. 1990) as the relatively best fittingfitting for the unpartitioned data set, the first and third codon position, and GTR+G for the second. Thee three data sets were analysed in MrBayes in a combined analysis under their relatively best- fittingfitting model. Additionally, amino acid and first and second codon position were analysed separately whichh lead to a congruent result. Ass bayesian posterior probabilities tend to have higher values than bootstrap probabilities (Huelsenbeckk et al. 2002), we restricted ourselves to clades supported by a posterior probability of 800 or higher. The resulting tree is therefore less resolved but provides a more robust estimation of the phylogeneticc relationships. Figuree 1 left shows the combined reconstructed COl tree, which was not significantly congruentt (rejected randomisation test in TreeMap) with the 28SrDNAtree (fig. 1 right).Amin o acid dataa and analysis restricted to (a) the second codon position only and (b) first+secondcodo n position onlyy resulted in similar topologies (unpublished results). AxinellaAxinella damicornis (Halichondrida: Axinellidae) is in the COl tree sister taxon to Stylissa (Halichondrida:: Dictyonellidae) and forms a well-supported cluster with the two species of Agelas off the order Agelasida. This constellation is in congruence with the 28SrDNAtree (fig. 1 cluster A), withh the exception of two other Axinella species, A. polypoides and A. verrucosa which group with DictyonellaDictyonella and Liosina (both Halichondrida: Dictyonellidae). Inn the COl tree all members of the family Halichondriidae form a single clade, but cluster withh Scopalina lophyropoda, Acanthella acuta (both Halichondrida: Dictyonellidae) and Cymbastela vespertinavespertina (Halichondrida: Axinellidae). The 28SrDNA tree also groups the Halichondriidae togetherr (fig. 1 cluster C), with the exception of Axinyssa spp., which clusters with Acanthella acuta andd Cymbastela vespertina more basal and distant from the other Halichondriidae (fig. 1 cluster D).

95 5 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges

5 5

fe fe

s s

o o 0 10 0 CO O 0 10 10 SI I o o fe fe 0 s s 5 5 8 8 §§ 8 10 0 0 11

c c 1 1 ITT Ti n II I IIII II 000 11

"22 <

ISS 11 .. s s i i i « ; « < II 3 s I I ! i's I _ 6.. Q Q Q SQ-4 j jj j S 5 ï s a 3 liliiisllillill l II s i i 33!!!!!! !! St! 88i8Siïas838 8

O O ü ü

96 6 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in hatichondrid demosponges Previouss page: Fig. 1. Left: Reconstructed phylogenetic tree of the cytochrome oxidase subunit 1 fragment basedd on the unpartitioned bayesian analysis with the posterior probabilities above the branches and the posteriorr probabilities of the partitioned analysis below the branches, the values of the unpartitioned analysis, cladess with posterior probabilities lower than 80% are collapsed. Right: Phylogenetic tree based on 28SrDNA dataa from Erpenbeck et al. (chapter 3, this thesis). The different taxon compositions have no influence on the concludingg topologies of the two analyses (unpublished results). The numbers on the branches are posterior probabilites.. See Erpenbeck et al. (chapter 3, Üüs thesis) for details. The polygons highlight clusters of both treess which are discussed in the text. Orders: AGE = Agelasida, CHO = Chondrosida, HAD = Hadromerida, HALL = Halichondrida, HAP = Haplosclerida, LIT = "Lithistida", POE = Poecilosclerida. Families: AGL == Agelasidae, ASC = Astroscleridae, AXI = Axinellidae, CHA = Chalinidae, CHN = Chondrillidae, CLI = Clionaidae,, COR = Corallistidae, DCT = Dictyonellidae, DES = Desmoxyidae, HLC = Halichondriidae, MYCC = Mycalidae, NIP = Niphatidae, PET = Petrosiidae, SUB = Suberitidae, THE = Theonellidae, VET = Vetulinidae. .

HigginsiaHigginsia and Myrmekioderma (both Halichondrida: Desmoxyidae) form a monophyletic cladee but the phylogenetic position of a third desmoxyid species, Didiscus oxeata, cannot be assessed unambiguouslyy (fig. 1 cluster E). Thee 28SrDNA gene trees support a close relationship between Halichondriidae (Halichondrida) andd the Suberitidae (Hadromerida), resulting in a non-monophyletic assemblage of the Halichondriid families.. The topology of the COl tree, does not support this outcome (fig. 1 cluster B). Differences betweenn these trees are independent from the different taxon sets (unpublished results). To obtain statisticall support we tested in the T-PTP test the probability of an alternative topology as resulting treee of our COl data set. This alternative topology had Suberitidae among the Halichondriidae as "backbone"" topology and represented this "Suberitina" topology as suggested by the 28SrDNA data. Thee alternative tree was significantly not supported by our data set (T-PTP test, p < 0.01). Thee haplosclerid taxa Haliclona (Haplosclerida: Chalinidae) and Acanthostrongylophora (Haplosclerida:: Petrosiidae) form a paraphyletic assemblage with Suberites (Hadromerida: Suberitidae).. Two other Haplosclerida species, Gelliodes fibulata and Amphimedon paraviridis, bothh from the family Niphatidae, form a sister group relationship but their phylogenetic position in thiss gene tree is uncertain. Also the phylogenetic affinities of Callyspongia aerizusa (Haplosclerida: Callyspongiidae),, Mycale flagellifera (Poecilosclerida: Mycalidae) and the two sequences of Svenzea devoogdaedevoogdae (Halichondrida: Dictyonellidae) are unclear.

Discussion n

MitochondrialMitochondrial base frequency bias in sponges

Wee have obtained mitochondrial sequences from a large number of sponge species using a nested- PCRR approach. The base composition of sponge COl sequences is skewed towards adenine and thymine.. This appears to be a characteristic of mitochondrial sequences of several protostome invertebratess (Simon et al. 1994, Frati et al. 1997) and may be related to differences in metabolic rate (Jermiinn and Crozier 1994). This frequently results in a bias against codons ending with G or C as observedd in springtails (Frati et al. 1997) and other metazoans (Jermiin and Crozier 1994) and may bee due to selection against tRNAs with CNN anticodons during an assumed process of genome size economisationn (Osawa et 1992).

DifferentDifferent topologies of the cytochrome oxidase gene tree

Thee cytochrome oxidase subunit 1 gene tree is in major parts different from a 28S rDNA tree as previouslyy reconstructed by us (Erpenbeck et al, submitted, chapter 2 and 3, this thesis) and in parts fromm previous authors (Chombard and Boury-Esnault 1999; McCormack and Kelly 2002). 28SrDNA

97 7 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges reconstructionss of demosponge phylogenies repeatedly lead to topologies largely inconsistent with morphologicall classification (e.g., McCormack et al. 2002; McCormack and Kelly 2002), but this is noo objective criterion to prefer one tree above another. Moore (1995) discusses mitochondrial gene treess as more likely to be congruent with the species tree than nuclear gene trees, because population geneticss of the mitochondrial genome is determined by effective population size. The smaller the effectivee population size, the higher the probability of congruence between the gene tree and the speciess tree. In the mitochondrial genome this effective population size is one fourth as large as that of aa nuclear genome because the mitochondrial genome is effectively haploid and maternally inherited. Onn the other hand a marker locus with a lower effective population size, such as mitochondria, is oftenn more strongly affected by historical events such as founder effects or bottlenecks than nuclear geness such as 28SrDNA (Arnaud-Haond et al. 2003). Inn addition, bilaterian mitochondrial genes are in many aspects different from diploblast mitochondriall genes and hardly explored so far. Among these are structural differences as in certain cnidarians,, in which unlike most other Metazoa, mitochondrial DNA strands are linear instead of circularr (Bridge et al. 1992). Furthermore, mtDNA of Anthozoa, (and probably of other diploblasts ass well) encodes only 1 to 2 tRNA genes instead of 22 as usual in Metazoa (Wolstenholme 1992), whichh requires import of tRNAs and causes different evolutionary constraints. Additionally, the evolutionaryy rate is discussed to be up to 10-20 times lower in diploblasts (Shearer et al. 2002), which iss potentially suitable to resolve older clades or deeper phylogenies than in Bilateria. However, the radiationn time of halichondrid demosponges is largely unknown because the fossil record is scarce andd correct taxonomie assignments of sponge fossils is ambiguous (Van Kempen 1977). Shearer et al.. (2002) describe cases for diploblasts in which the mitochondrial gene clearly failed to reconstruct speciess trees: Genetic variation among species also can be disproportionate to the taxonomie classificationn with a divergence rate too variable among clades. (Romano and Palumbi (1996, 1997) reportt six identical mitochondrial 16S strands from taxa of three different scleractinian families in whichh Shearer et al. (2002) assume additional reason not to overestimate the phylogenetic inference off mitochondrial gene trees. The need of secondary or even more independent opinions in (sponge) plylogeneticss is therefore highly evident.

PhylogeneticPhylogenetic implications

Thee analysis of the COl gene tree in combination with the 28SSrDNA data provides insight in the phylogeneticc classification of the Halichondrida. Despite the incongruent topologies of the two (independent)) gene trees there is agreement in several points, which should be taken into account Evidencee for a placement of the order Agelasida in Halichondrida is well-supported. Both in thee 28SrDNA and in the COl tree Agelasida are in a sister group relationship to Axinella damicornis andd Stylissa spp. The DNA data also agree with biochemical data (Braekman et al. 1992; Van Soest andd Braekman 1999; Costantino et al. 1996). Halichondrida is therefore non-monophyletic, unless Halichondridaa and Agelasida are merged. "Suberitina",, however, i.e., the predicted relationship between the families Halichondriidae (order Halichondrida)) and Suberitidae (order Hadromerida), as suggested by Chombard and Boury-Esnault (1999),, are not supported by the COl gene tree and rejected in a T-PTP test. While the rDNA tree clearlyy favours a close relationship between halichondriid and suberitid demosponges, COl provides noo indication to merge these two families. Thee family Dictyonellidae Van Soest, Diaz and Pomponi, is not supported as a monophyletic groupp by this data set and the 28S gene tree. Scopalina, Stylissa, Acanthella, Dictyonella and Liosina aree positioned in various parts of the tree indicating that the family Dictyonellidae is polyphyletic. Dictyonellidaee is based on the lack of both an ectosomal skeleton and typically halichondroid

98 8 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges choanosomee skeleton, differentiating it from Halichondriidae and Desmoxyidae. Furthermore, they lackk an Axinellidae-type of surface structure. The fleshy appearance of the ectosomal region (Soest et al.. 2002) is a character of ambiguous quality making Dictyonellidae an assemblage of genera related too different higher taxa, sharing the absence of their taxonomie relevant characters. Thee close relationship of the members of the family Halichondriidae is supported by the COll data set and is in congruence to 28SrDNA data. However, the placement of Acanthella, (Halichondrida:: Dictyonellidae) and Cymbastela (Halichondrida: Axinellidae) with Axinyssa spp. (Halichondrida:: Halichondriidae) in both data sets suggests a strong phylogenetic signal for such constellation.. Axinyssa, the only Halichondriidae in this data set lacking a clear specialized ectosomal skeleton,, clusters in the COl gene tree with the other Halichondriidae, while 28S data suggests its placementt outside the Halichondriidae. Additional information from an independent molecular data sett is necessary to determine the position of Axinyssa, Acanthella and Cymbastela. Thee desmoxyid taxa Higginsia and Myrmekioderma display a close relationship, in opposition too the 28S data, but the placement of Didiscus oxeata in the Desmoxyidae is not supported by the COl genee tree. The close relationship of Scopalina and Svenzea, suggested by (non skeletal) morphological dataa (Rützler et al. 2003) and previous 28SrDNA data is not retrieved in this analysis.

MorphologicalMorphological cladistics in halichondrids and the need of independent markers

Ourr results raises questions on the use of morphological data in sponge systematics. Morphological or biochemicall synapomorphies between Suberitidae and Halichondriidae are not obvious, and several homoplasiess must be accepted to postulate a closer relationship between these two groups using morphologicall characters. For this reason the "Suberitina" hypothesis has never been discussed in detaill and the gene trees of Chombard and Boury-Esnault (1999) and McCormack and Kelly (2002) weree probably widely regarded as artefacts of their phylogenetic markers. The same, however, would applyy for the agelasid / axinellid relationship: From a morphological point of view the resemblance off agelasids with Stylissa or Axinella is low: All groups possess styles, one of the most abundant spiculee types in sponges. However, styles in Agelasida are morphologically and functionally different ass they are echinating and possess a "verticillid" morphology, i.e. with unique rings of spines, while thosee in Stylissa are smooth and plumoreticulated. Historically this was reason enough to place themm in different demosponge orders. Now we can present multiple, independent evidence for a closee relationship of these taxa, which is not immediately obvious from morphological systematics. Consequentlyy we have to ask, in how many other cases morphology is misleading and masking true evolutionaryy histories in sponges. Nevertheless, the relationship of Agelasida to halichondrid clades slowlyy becomes accepted, but we have to regard this cytochrome oxidase 1 gene tree as only one additionall piece of evidence out of many. Therefore, the validation or rejection of "Suberitina" should bee based on more additional markers, as its entity is supported by 28S data but neither by morphology, norr by chemical compounds or COl-DNA. Wee clearly need more independent markers especially in (demo)sponge systematics. Thus far, severall molecular results appear to contradict each other depending on the gene regarded and only the recruitmentt of additional independent genes will provide more objective insight in the phylogenetic historyy of demosponges.

Acknowledgements s

Wee would like to thank Martin R. Genner, Ann N. Knowlton, Peter Kuperus, Frederick R. Schram, Jann J. Vermeulen, Betsie Voetdijk and Nicole J de Voogd for contributions in various ways to this paper.. Part of the material was freshly collected during the S YMBIOSPONGE project (EU-MAS3CT

99 9 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges 97-0144).97-0144). This work was financed by the Dutch Organisation for Scientific Research (NWO) Nr: ALWW 809.34.003. All experiments comply with the current laws of the Netherlands and the European Union. .

References s

Altschull SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Moll Biol 215: 403-410 Alvarezz B, Crisp MD, Driver F, Hooper JNA, Soest RWM van (2000) Phylogenetic relationships of thee family Axinellidae (Porifera: Demospongiae) using morphological and molecular data. Zoologicaa Scripta 29: 169-198 Arnaud-Haondd S, Bonhomme F, Blanc F (2003) Large discrepancies in differentiation of allozymes, nuclearr and mitochondrial DNA loci in recently founded Pacific populations of the pearl oysterr Pinctada margaritifera. Journal of Evolutionary Biology 16: 388-398 Axx P (1995) Das System der Metazoa I. Gustav Fischer, Stuttgart Borchiellinii C, Boury-Esnault N, Vacelet J, Le Parco Y (1998) Phylogenetic analysis of the Hsp70 sequencess reveals the monophyly of Metazoa and specific phylogenetic relationships between animalss and fungi. Molecular Biology and Evolution 15: 647-655 Braekmann JC, Daloze D, Stoller C, Soest RWM van (1992) Chemotaxonomy of Agelas (Porifera, Demospongiae).. Biochemical Systematics and Ecology 20: 417-431 Bridgee D, Cunningham CW, Schierwater B, Desalle R, Buss LW (1992) Class-Level relationships inn the phylum Cnidaria - evidence from mitochondrial genome structure. Proceedings of the Nationall Academy of Sciences of the United States of America 89: 8750-8753 CavalierSmithh T, Chao EE, BouryEsnault N, Vacelet J (1996) Sponge phylogeny,, animal monophyly, andd the origin of the nervous system: 18S rRNA evidence. Canadian Journal of Zoology- Revuee Canadienne De Zoölogie 74: 2031-2045 Chombardd C, Boury-Esnault N (1999) Good Congruence between morphology and molecular phylogenyy of Hadromerida, or how to bother sponge taxonomists. Memoirs of the Queensland Museumm 44: 100 Chombardd C, Boury-Esnault N, Tillier A, Vacelet J (1997) Polyphyly of "sclerosponges" (Porifera, Demospongiae)) supported by 28S ribosomal sequences. Biological Bulletin 193: 359-367 Chombardd C, Boury-Esnault N, Tillier S (1998) Reassessment of homology of morphological characterss in tetractinellid sponges based on molecular data. Systematic Biology 47: 351- 366 6 Collinss AG (1998) Evaluating multiple alternative hypotheses for the origin of Bilateria: An analysis off 18S rRNA molecular evidence. Proceedings of the National Academy of Sciences of the Unitedd States of America 95: 15458-15463 Costantinoo V, Fattorusso E, Mangoni A, DiRosa M, Ianaro A, Maffia P (1996) Glycolipids from spongess .4. Immunomodulating glycosyl ceramides from the marine sponge Agelas dispar. Tetrahedronn 52: 1573-1578 Durann S, Pascual M, Estroup A, Turon X (2002) Genetic variation in populations of the sponge CrambeCrambe crambe (Poecilosclerida) assessed using polymorphic microsattelite markers and mtDNAA sequence data. Boll. Mus. 1st. bio. Univ. Genova 66-67: 60 Erpenbeckk D, Breeuwer J AJ, van der Velde HC, Soest RWM van (2002) Unravelling host and symbiont phylogeniess of halichondrid sponges (Demospongiae, Porifera) using a mitochondrial marker. Marinee Biology 141: 377-386 Erpenbeckk D, Knowlton AL, Talbot SL, Highsmith RC, Soest RWM van. A molecular comparison of Alaskann and North East Atlantic Halichondrid panicea (Pallas 1766) (Porifera: Demospongiae)

100 0 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges Populations.. Boll. Mus. 1st. bio. Univ. Genova. in press Faithh DP (1991) Cladistic Permutation Tests for Monophy ly and Nonmonophyly. Systematic Zoology 40:: 366-375 Fratii F, Simon C, Sullivan J, Swofford DL (1997) Evolution of the mitochondrial cytochrome oxidase III gene in Collembola. Journal of Molecular Evolution 44: 145-158 Hilliss DM, Dixon MT (1991) Ribosomal DNA - Molecular Evolution and Phylogenetic Inference. Quarterlyy Review of Biology 66: 410-453 Hooperr JNA (1990) Character Stability, Systematics, and Affinities between Microcionidae (Poecilosclerida)) and Axinellida. In: Rützler K (ed) New perspectives in Sponge Biology. Smithoniann Institution Press, Washington D.C., pp 284-294 Huelsenbeckk JP, Larget B, Miller RE, Ronquist F (2002) Potential applications and pitfalls of Bayesiann inference of phylogeny. Systematic Biology 51: 673-688 Huelsenbeckk JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17:: 754-755 Jermiinn LS, Crozier RH (1994) The cytochrome-B region in the mitochondrial-DNA of the ant TetraponeraTetraponera rufoniger - Sequence divergence in Hymenoptera may be associated with nucleotidee content. Journal of Molecular Evolution 38: 282-294 Kelly-Borgess M, Bergquist PR, Bergquist PL (1991) Phylogenetic-Relationships within the Order Hadromeridaa (Porifera, Demospongiae, Tetractinomorpha) as Indicated by Ribosomal-RNA Sequencee Comparisons. Biochemical Systematics and Ecology 19: 117-125 Kelly-Borgess M, Pomponi SA (1994) Phylogeny and classification of lithistid sponges (Porifera: Demospongiae):: a preliminary assessment using ribosomal DNA sequence comparisons. Molecularr Marine Biology and Biotechnology 3: 87-103 Kempenn TMG van (1977) Roepella solanansis, a new sponge from the mid-tertiary of S.E. Spain. First fossill record of a hymeniacidonoid, with remarks on other fossilized sponges with monaxid megascleres.. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, Amsterdam,, series B 80: 106-131 Kimm JH, Kim W, Cunningham CW (1999) A new perspective on lower metazoan relationships from 18SS rDNA sequences. Molecular Biology and Evolution 16: 423-427 Lafayy B, Boury-Esnault N, Vacelet J, Christen R (1992) An analysis of partial 28S ribosomal RNA sequencess suggests early radiation of sponges. BioSystems 28: 139-151 Levii C (1953) Sur une nouvelle classification des Démosponges. Comptes rendus des séances de 1'Académiee des Sciences 236: 853-855 Levii C (1957) Ontogeny and Systematics in Sponges. Syst. Zool. 6: 1-4 McCormackk GP, Erpenbeck D, Soest RWM van (2002) Major discrepancy between phylogenetic hypothesess based on molecular and morphological criteria within the Order Haplosclerida (Phylumm Porifera: Class Demospongiae). Journal of Zoological Systematics and Evolutionary Researchh 40: 237-240 McCormackk GP, Kelly M (2002) New indications of the phylogenetic affinity of Spongosorites suberitoidessuberitoides Diaz et al., 1993 (Porifera, Demospongiae) as revealed by 28S ribosomal DNA. Journall of Natural History 36: 1009-1021 Medinaa M, Collins AG, Silberman JD, Sogin ML (2001) Evaluating hypotheses of basal animal phylogenyy using complete sequences of large and small subunit rRNA. Proceedings of the Nationall Academy of Sciences of the United States of America 98: 9707-9712 Mooree WS (1995) Inferring Phylogenies from mtDNA Variation - Mitochondrial-Gene Trees Versus Nuclear-Genee Trees. Evolution 49: 718-726 Nylanderr J, MrModeltest. http://www.ebc.uu.se/systzoo/staff/nylander.html Osawaa S, Jukes TH, Watanabe K, Muto A (1992) Recent-Evidence for Evolution of the Genetic-

101 1 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges Code.. Microbiological Reviews 56: 229-264 Pagee RDM (1994) Parallel Phylogenies - Reconstructing the History of Host- Parasite Assemblages. Cladistics-thee International Journal of the Willi Hennig Society 10: 155-173 Posadaa D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14:817-818 8 Rodriguezz F, Oliver JF, Mann A, Medina JR (1990) The general stochastic model of nucleotide substitution.. J Theor Biol 142: 485-501 Romanoo SL, Palumbii SR (1996) Evolution of scleractinian corals inferred from molecular systematics.. Science 271: 640-642 Romanoo SL, Palumbi SR (1997) Molecular evolution of a portion of the mitochondrial 16S ribosomal genee region in scleractinian corals. Journal of Molecular Evolution 45: 397-411 Rützlerr K, Soest RWM van, Alvarez B (2003) Svenzea zeai, a Caribbean reef sponge with a giant larva,, and Scopalina ruetzleri: a comparative fine-structural approach to classification (Demospongiae,, Halichondrida, Dictyonellidae). Invertebrate Biology 122: 203-222 Sambrookk E, Fritsch F, Maniatis T (1989) Molecular cloning. Cold Spring Harbor Press, Cold Spring Harbor,, New York Schroderr HC, Efremova SM, Itskovich VB, Belikov S, Masuda Y, Krasko A, Muller IM, Muller WEG,, Medina M, Weil E, Szmant AM (2003) Molecular phylogeny of the freshwater sponges inn Lake Baikal. Journal of Zoological Systematics and Evolutionary Research 41: 80-86 Shearerr TL, Van Oppen MJH, Romano SL, Wörheide G (2002) Slow mitochondrial DNA sequence evolutionn in thee Anthozoa (Cnidaria). Molecular Ecology 11: 2475-2487 Simonn C, Frati F, Beckenbach A, Crespi B, Liu H, Rook P (1994) Evolution, weighting, and phylogeneticc utility of mitochondrial gene-sequences and a compilation of conserved polymerasee chain- reaction primers. Annals of the Entomological Society of America 87: 651-701 1 Smithh JM, Smith NH (2002) Recombination in animal mitochondrial DNA. Molecular Biology and Evolutionn 19: 2330-2332 Soestt RWM van (1987) Phylogenetic Exercises with Monophyletic Groups of Sponges. In: Vacelet J, Boury-Esnaultt N (eds) NATO ASI Series. Springer, Berlin Heidelberg, pp 227-241 Soestt RWM van (1991) Demosponge Higher Taxa Classification Re-Examined. In: Reitner J, Kreupp HH (eds) Fossil and recent sponges. Springer, Berlin, pp 54-71 Soestt RWM van, Erpenbeck D, Alvarez B (2002) Family Dictyonellidae Van Soest, Diaz & Pomponi 1990.. In: Hooper JNA, Soest RWM van (eds) Systema Porifera, A guide to the classification off Sponges. KluwerAcademic/Plenum Publishers, New York, pp 773-786 Soestt RWM van, Braekman JC (1999) Chemosystematics of Porifera: A review. Memoirs of the Queenslandd Museum 44: 569-589 Soestt RWM van, Diaz MC, Pomponi SA (1990) Phylogenetic classification of the Halichondrids (Porifera,, Demospongiae). Beaufortia 40: 15-62 Soestt RWM van, Hooper JNA (2002) Order Halichondrida Gray, 1867. In: Hooper JNA, Soest RWM vann (eds) Systema Porifera, A guide to the classification of Sponges. KluwerAcademic/ Plenumm Publishers, New York, pp 721-723 Swoffordd DL (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Sinauer Associates,, Sunderland, Massachusetts Watkinss RF, Beckenbach AT (1999) Partial sequence of a sponge mitochondrial genome reveals sequencee similarity to Cnidaria in cytochrome oxidase subunit II and the large ribosomal RNAA subunit. Journal of Molecular Evolution 48: 542-554 Wolstenholmee DR (1992) Animal mitochondrial DNA: structure and evolution. In: Wolstenholme DR,, Jeon KW (eds) International Review of Cytology. Academic Press, New York, pp 173-

102 2 6:: Conflicting phylogenetic signal between nuclear ribosomal and mitochondrial data in halichondrid demosponges 216 6 Worheidee G, Degnan BM, Hooper JNA (2000) Population phylogenetics of the common coral reefspongess Leucetta spp. and Pericharax spp. (Porifera: Calcarea) from the Great Barrier Reeff and Vanuatu. In: Hopley D, Hopley P, Tamelander J, Done T (eds) Ninth International Corall Reef Symposium, Vol Abstracts. Bali, Indonesia, pp 21 Zrzavyy J, Mihulka S, Kepka P, Bezdek A, Tietz D (1998) Phylogeny of the Metazoa based on morphologicall and 18S ribosomal DNA evidence. Cladistics-the International Journal of the Willii Hennig Society 14: 249-285

103 3 104 4