RECONSTRUCTION OF THE PHYLOGENY OF THE

OPISTHOBRANCHIA (MOLLUSCA: GASTROPODA) BY MEANS

OF 18S AND 28S rRNA GENE SEQUENCES
VERENA VONNEMANN 1 , MICHAEL SCHRÖDL 2 , ANNETTE KLUSSMANN-KOLB 3
AND HEIKE WÄGELE 1
1
Lehrstuhl für Spezielle Zoologie, Ruhr-Universität Bochum, 44780 Bochum, Germany;
2
Zoologische Staatssammlung München, 81247 München, Germany;
3
Zoologisches Institut der J. W. Goethe Universität, 60054 Frankfurt am Main, Germany
(Received 20 November 2003; accepted 25 August 2004)

ABSTRACT
The complete 18S (SSU) rRNA gene and the partial 28S (LSU) rRNA gene of 49 taxa, including
representatives of the Opisthobranchia, Pulmonata and Allogastropoda, were sequenced to infer the
phylogeny of the Opisthobranchia and their major subordinate taxa Nudibranchia, Pleurobranchoidea,
Tylodinoidea, Sacoglossa, Cephalaspidea s.l. and Anaspidea. For the first time the enigmatic taxon
Acochlidiacea was included. Several tests were applied to the datasets to investigate relative evolutionary
rates of the taxa, saturation of substitution, base composition and the amount of phylogenetic
information. Maximum parsimony, maximum likelihood and distance methods of tree construction were
used. The opisthobranch subgroups, including the Acochlidiacea, were confirmed as monophyletic in
most analyses, whereas monophyly vs paraphyly of Opisthobranchia could not be resolved. All trees
showed a sister-group relationship between Nudibranchia and Pleurobranchoidea (Nudipleura). In most
trees, sister-group relationships between Cephalaspidea s.s. and Anaspidea, and between Nudipleura and
Acteonoidea (Pupa solidula, Hydatina physis), were supported. The latter sister-group relationship supports
an affinity that has not been mentioned for nearly 40 years (Ghiselin, 1966, Malacologia, 3: 327– 378). It
was ‘rediscovered’ in the molecular analyses by Grande et al. (2004, Molecular Biology and Evolution, 21:
303– 313) and represents a valid and interesting opisthobranch clade.

INTRODUCTION phylogenetic signal, additional characters (domain D3) have

been included.
The Opisthobranchia are an assemblage of morphologically very
The aims of the present paper are to reconstruct opisthobranch
diverse snails and slugs occurring in marine habitats all over the phylogeny and to assess the suitability of the 28S rRNA gene for
world. Their appearance ranges from inconspicuous caenogas- phylogenetic questions in comparison with the well-studied 18S
tropod-like animals burrowing in sand to extravagantly coloured rRNA gene.
and highly derived swimming organisms. Their smallest repre-
sentatives (Microhedylidae, Acochlidiacea) measure little more
than 1 mm (Arnaud, Poizat & Salvini-Plawen, 1986), whereas MATERIALS AND METHODS
the largest species (Aplysia vaccaria Winkler, 1955, Anaspidea) can
reach nearly 1m in length (Behrens, 1980). The earliest fossil Biological material
opisthobranchs are from the Lower Carboniferous (ca 310 Ma), Forty-three species of opisthobranchs and six outgroup species
resembling members of the extant Scaphandridae and Retusidae (Pulmonata and Allogastropoda) were collected at various sites
(Cephalaspidea; Gosliner, 1981). The number of species of all over the world (Table 1). Within the scope of the present study
Opisthobranchia is of the order of 4500 – 5000 (Wägele, the partial 28S rRNA gene of 48 taxa and the complete 18S
unpublished) and is very different among the opisthobranch rRNA gene of 41 taxa were sequenced, and the remaining
subgroups (e.g. Tylodinoidea about 15 species, Nudibranchia sequences were taken from GenBank.
nearly 3000 species).
Despite numerous publications dealing with the Opisthobran- DNA extraction, amplification and sequencing
chia, their phylogeny is mainly unresolved. This is due to the lack
of comprehensive morphological investigations and the contra- Genomic DNA was extracted from alcohol-preserved specimens
dictory results produced by molecular studies (Thollesson, 1999; by means of the Blood and Tissue Kit (Qiagen), guided by the
Dayrat et al., 2001; Wägele, Vonnemann & Wägele, 2003). The enclosed protocol. The amplification of the 18S rRNA gene
present study is the first extensive phylogenetic examination of region by PCR (Saiki et al., 1988) was performed with primers
developed by Trisha Spears (personal communication): forward
opisthobranch 28S rRNA gene sequences and also the first
(18A1), 50 -CCT A(CT)C TGG TTG ATC CTG CCA GT-30
considering the domains D1, D2 and D3. Dayrat et al. (2001)
and reverse (1800), 50 -GAT CCT TCC GCA GGT TCA CCT
investigated the D1 and D2 domains of 31 euthyneuran species
ACG-30 . The PCR was carried out in the thermal cycler Progene
and found that some groups were well supported, but resolution (Techne) under the following conditions: 95 8C for 4 min,
was poor for other major clades. In order to obtain an increase of followed by 38 cycles of 30 s at 94 8C, 30 s at 52.5 8C, 2.5 min at
72 8C and a final extension at 72 8C for 10 min. The domains D1–
Correspondence:V.Vonnemann; D3 of the 28S rRNA gene were amplified with forward primer
e-mail: [email protected] C10 (Dayrat et al., 2001): 50 -ACC CGC TGA ATT TAA GCA

Journal of Molluscan Studies (2005) 71: 113–125 doi: 10.1093/mollus/eyi014

q The Author 2005. Published by Oxford University Studies on behalf of The Malacological Society of London, all rights reserved.
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

Table 1. List of specimens examined in this study, sampling localities, accession numbers and lengths of complete 18S and partial 28Sr DNA genes.

Length of Length of partial

Species Family Collecting locality Accession numbers* complete 18S 28S (D1 –D3)

NUDIBRANCHIA
Bathydoris clavigera Bathydorididae Antarctica AY165754 2007 1383
Thiele, 1912 AY427444
Chromodoris krohni Chromodorididae Spain, Atlantic Ocean AJ224774 1826 1064
(Verany, 1846) AY427445
Hypselodoris villafranca Chromodorididae Spain, Atlantic Ocean AJ224780 1832 1071
(Risso, 1818) AY427446
Dendrodoris fumata Dendrodorididae Great Barrier Reef, Australia AF249216 1842 1062
(Rüppell & Leuckart, 1830) AY427447
Cuthona nana Tergipedidae Helgoland, Germany AY165760 1881 1027
(Alder & Hancock, 1842) AY427448
Flabellina babai Flabellinidae Spain, Mediterranean Sea AY165768 1870 1045
(Schmekel, 1972) AY427449
Dendronotus dalli Dendronotidae USA, Atlantic Ocean AY165757 1947 1063
Bergh, 1879 AY427450
PLEUROBRANCHOIDEA
Euselenops luniceps Pleurobranchidae Great Barrier Reef, Australia AF249218 2047 1136
(Cuvier, 1817) AY427451
Tomthompsonia antarctica Pleurobranchidae Antarctica AY427492 2105 1156
(Thiele, 1912) AY427452
Bathyberthella antarctica Pleurobranchidae Weddel Sea, Antarctica AF249219 2024 1123
Willan & Bertsch, 1987 AY427453
Berthellina citrina Pleurobranchidae Elba, Italy AY427493 2028 1090
(Rüppell & Leuckart, 1828) AY427454
Pleurobranchus peroni Pleurobranchidae Wollongong, NSW, Australia, AY427494 2022 1120
Cuvier, 1804 AY427455
Berthella stellata Pleurobranchidae Elba, Italy AY427495 2001 1166
(Risso, 1826) AY427456
TYLODINOIDEA
Umbraculum mediterraneum Umbraculidae Meteor Bank, Atlantic Ocean AY165753 1786 1069
(Lamarck, 1829) AY427457
Tylodina perversa Tylodinidae Spain, Mediterranean Sea AY427496 1796 1067
(Gmelin, 1791) AY427458
SACOGLOSSA
Placobranchus ocellatus Plakobranchidae Magnetic Island, QLD, Australia AY427497 1790 1053
van Hasselt, 1824 AY427459
Bosellia mimetica Boselliidae Elba, Italy AY427498 1793 1053
Trinchese, 1891 AY427460
Thuridilla bayeri Elysiidae Great Barrier Reef, Australia AF249220 1786 1053
Marcus, 1965 AY427461
Elysia viridis Elysiidae Mediterranean Sea AY427499 1789 1052
(Montagu, 1804) AY427462
Cyerce nigricans Caliphyllidae Eagle Island, QLD, Australia AY427500 1790 1063
(Pease, 1866) AY427463
Calliopaea bellula Limapontiidae Elba, Italy AY427501 1787 1044
d’Orbigny, 1837 AY427464
Limapontia capitata Limapontiidae North Sea AJ224920 1786 1047
(O.F. Müller, 1773) AY427465
ANASPIDEA
Akera bullata Akeridae Kattegat, Denmark AY427502 1789 1047
Müller, 1776 AY427466
Aplysia californica Aplysiidae Genbank AY039804 1791 1049
Cooper, 1863 AY026366
Dolabella auricularia Aplysiidae Okinawa, Japan AY427503 1785 1042
(Lightfoot, 1786) AY427467

114
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

Table 1. (continued)

Length of Length of partial

Species Family Collecting locality Accession numbers* complete 18S 28S (D1 – D3)

CEPHALASPIDEA s.s.
Haminoea hydatis Haminoeidae Elba, Italy AY427504 1790 1037
(Linné, 1758) AY427468
Phanerophthalmus smaragdinus Haminoeidae Lizard Island, QLD, Australia AY427505 1787 1036
(Rüppell & Leuckart, 1831) AY427469
Atys semistriatus Haminoeidae Dingo Beach, QLD, Australia AY427506 1787 1034
Pease, 1860 AY427470
Odontoglaja guamensis Aglajidae Orpheus Island, QLD, Australia AY427507 1789 1039
Rudman, 1978 AY427471
Navanax inermis Aglajidae La Paz, Gulf of California, Mexico AY427508 1790 1034
(Cooper, 1863) AY427472
Chelidonura inornata Aglajidae Lizard Island, QLD, Australia AY165752 1796 1039
Baba, 1949 AY427473
Philinopsis pilsbryi Aglajidae Magnetic Island, QLD, Australia AY427509 1788 1034
(Eliot, 1900) AY427474
Philinoglossa praelongata Philinoglossidae Rovinj, Croatia AY427510 1784 1041
Salvini-Plawen, 1973 AY427475
cf. Retusa sp. Retusidae Elba, Italy AY427511 1788 1035
AY427476
Bulla cf. striata Bullidae Elba, Italy AY427512 1787 1036
Bruguière, 1792 AY427477
Sagaminopteron psychedelicum Gastropteridae Magnetic Island, QLD, Australia AY427513 1792 1042
Carlson & Hoff, 1974 AY427478
Runcina adriatica Runcinidae Elba, Italy AY427514 1783 1044
Thompson, 1980 AY427479
ACTEONOIDEA
Hydatina physis Hydatinidae Hastings Point, NSW, Australia AY427515 1804 1087
(Linné, 1758) AY427480
Pupa solidula Acteonidae Dingo Beach, QLD, Australia AY427516 1842 1065
(Linné, 1758) AY427481
ACOCHLIDIACEA
Unela glandulifera Microhedylidae Rovinj, Croatia AY427517 1787 1045
(Kowalewsky, 1901) AY427482
Parahedyle cryptophthalma Microhedylidae Elba, Italy AY427518 1787 1045
(Westheide & Wawra, 1974) AY427483
Pontohedyle milatchevitchi Microhedylidae Elba, Italy AY427519 1788 1043
(Kowalewsky, 1901) AY427484
Hedylopsis spiculifera Hedylopsidae Rovinj, Croatia AY427520 1786 1035
(Kowalewsky, 1901) AY427485
PULMONATA
Onchidella floridana Onchidiidae Flatts Village, Bermuda AY427521 1790 1049
(Dall, 1885) AY427486
Onchidium cf. verruculatum Onchidiidae Cockle Bay, QLD, Australia AY427522 1791 1050
(Cuvier, 1830) AY427487
Siphonaria alternata Siphonariidae Flatts Village, Bermuda AY427523 1786 1046
(Say, 1826) AY427488
Cepaea hortensis Helicidae Grietherbusch, Rhine, Germany AY427524 1788 1058
(OF Müller 1774) AY427489
Lymnaea stagnalis Lymnaeidae Grietherbusch, Rhine, Germany AY427525 1800 1042
(Linnaeus, 1758) AY427490
ALLOGASTROPODA
Odostomia sp. Pyramidellidae Banyuls sur Mer, France, AY427526 1791 1044
Mediterranean Sea AY427491

*Accession number for complete 18S rRNA gene sequence above, accession number for partial 28S rRNA gene below.
Specimens have been deposited at either the Zoologische Staatssammlung München (ZSM) or at the Lehrstuhl Spezielle Zoologie, Ruhr-Universität Bochum.

115
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

T-30 and reverse primer D3: 50 -GAC GAT CGA TTT GCA CGT opening 10, gap extension 0.2), but subsequent manual correc-
CA-30 under the same conditions as the 18S rRNA gene. Each tion was necessary. This was based on different and partly
PCR reaction mix (50 ml) contained 5 ml of 10 £ PCR buffer subjective criteria: elimination of obvious errors like displaced
(Qiagen), 10 ml of Q-Solution (Qiagen), 5 ml of dNTP-mix sequence parts, optimization of the homologies among groups
(2 mM per dNTP), 0.5 ml of each primer (18A1 and 1800 or C1 that are well supported morphologically and maximization of
and D3), 0.3 ml of Taq polymerase (Qiagen), 0.25 to 3.0 ml of correspondences by insertion of gaps. This manual treatment was
genomic DNA and 21.55 – 25.7 ml H2O. executed by means of the program GENEDOC (Nicholas &
After purification of the PCR products by means of the Nicholas, 1997).
QIAquick PCR purification kit (Qiagen), the 18S and 28S rRNA
genes were sequenced directly by the chain-termination method
(Sanger, Miklen & Coulson, 1977) using the Thermo-sequenase Phylogenetic signal
fluorescent-labelled primer cycle sequencing kit (Amersham) on The software PHYSID (Phylogenetic Signal Detection, Wägele
the automated sequencers 4000 and 4000IR2 (Licor). Both & Rödding, 1998) allows an a priori estimation of the information
strands of the 18S and 28S rRNA genes were sequenced using the content of aligned sequences without reference to a tree topology
primers from the PCR and several internal primers (Table 2). or to a model of sequence evolution. The method consists of a
search for splits in an alignment supported by putative
apomorphies. If the most frequent partitions are compatible a
Sequence alignment strong phylogenetic signal is assumed (Wägele, 1996; Wägele &
Rödding, 1998). An a posteriori evaluation can follow by
Sequences were aligned with CLUSTAL X (Thompson et al., comparing the splits with the dendrograms obtained by
1997) using the default parameters (pairwise parameters: gap maximum parsimony, maximum likelihood or distance methods
opening 10, gap extension 0.1; alignment parameters: gap and examinating the fit of the splits on the trees.
Table 2. Primers used for sequencing the complete 18S and partial 28S
Another (a posteriori) approach for evaluating the phylogenetic
rRNA genes. signal of the 18S and 28S rRNA gene alignments is the
permutation-tail probability (PTP) test (Archie, 1989). This
Name Sequence (50 – 30 ) Reference test is based on comparison of the length of the most parsimonious
tree with the lengths of trees obtained after randomly permutat-
18A1seq CTG GTT GAT CCT GCC According to PCR primer ing the character state assignments within each character. The
AGT CAT ATG C permutation-tail probability is then defined as the estimate of the
400F ACG GGT AAC GGG GAA Wollscheid, Dreyer & Englisch† proportion of times that a tree can be found that is as short
TCA GGG
or shorter than the original tree (Faith & Cranston, 1991). The
PTP test was conducted by means of PAUP 4.0 beta 3a
470F CAG CAG GCA CGC AAA Wollscheid, Dreyer & Englisch†
(Swofford, 1999).
TTA CCC
700F GTC TGG TGC CAG CAG Wollscheid, Dreyer & Englisch†
CCG CG
Statistical tests and phylogenetic reconstruction
†
1155F CTG AAA CTT AAA GGA Wollscheid, Dreyer & Englisch Base composition and saturation of substitution were investigated
ATT GAC GG by means of the software PAUP 4.0 beta 3a (Swofford, 1999) and
1600F CGT CCC TGC CCT TTG Wollscheid, Dreyer & Englisch† rates of evolution by means of the software k2WuLi (Wu & Li,
TAC ACA CC 1985).
1800seq GAT CCT TCC GCA GGT According to PCR primer
All tree-building methods were conducted using PAUP 4.0 beta
3a (Swofford, 1999). Parameters of maximum-parsimony analyses
TCA CCT ACG
were: ACCTRAN, gaps counted as fifth base; heuristic search
1500R CAT CTA GGG CAT CAC Wollscheid, Dreyer & Englisch†
options: stepwise addition ¼ random, branch-swapping option ¼
AGA CC TBR. The option ‘gaps are missing data’ was not used, since
1155R CCG TCA ATT CCT TTA Wollscheid, Dreyer & Englisch† elongations in sequences are separate characters, and should be
AGT TTC AG coded as such. The parameters for distance and maximum-
700R CGC GGC TGC TGG CAC Wollscheid, Dreyer & Englisch† likelihood methods were determined with likelihood ratio tests
CAG AC using Modeltest Version 3.06 (Posada & Crandall, 1998).
400R CCC TGA TTC CCC GTT Wollscheid, Dreyer & Englisch† For testing incongruence in combined analyses, the incon-
ACC CGT gruence length difference (ILD) test (Farris et al., 1994) was
28SC1 ACC CGC TGA ATT TAA According to PCR primer
applied. With this test, it can be inferred whether combining the
data improves phylogenetic accuracy (P . 0.01) or whether
GCA T
accuracy of the combined data suffers relative to the individual
28SC2F (C20 )* GAA AAG AAC TTT GAA Dayrat et al. 2001, modified
partitions (P , 0.001, Cunningham, 1997).
GAG AGA GT
28SD2F CCC GTC TTG AAA CAC Vonnemann et al., present study
RESULTS
GGA CCA AGG
28SD3 GAC GAT CGA TTT GCA According to PCR primer
Sequence alignment
CGT CA
28SD2R CCT TGG TCC GTG TTT Vonnemann et al., present study The alignment of the 18S rRNA gene (28S rRNA gene)
CAA GAC GGG
sequences comprises 2211 (1149) positions, of which 1124 (441)
positions are constant, 293 (149) are variable but parsimony-
28SC2R (C2)* ACT CTC TCT TCA AAG Dayrat et al., 2001, modified
uninformative and 794 (559) are parsimony-informative.
TTC TTT TC
Parts with uncertainties of homology were removed from the
* Names of unmodified primers in brackets (Dayrat et al., 2001). alignments (unpublished data of secondary structure analyses of
†
Molecular Phylogeny Working group, Department of Spezielle Zoologie, senior author). In the 18S rRNA gene the Pleurobranchoidea
Ruhr-Universität Bochum. possess long non-alignable insertions (positions 865– 1004 in

116
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

structure E23_3) in region V4 (Wuyts, van de Peer & Pleurobranchoidea the very high rates of Berthella stellata are
De Wachter, 2001). In structure 43 (region V7: positions striking (18S, 7.3 – 14.7; 28S, 2.3– 9.3). Its Z-scores are much
1892 – 1974) the sequence lengths of Nudibranchia, and higher than the values of the other Pleurobranchoidea and
especially of Bathydoris clavigera, are much higher than of the even higher than those of the Anthobranchia. Among the other
other opisthobranch taxa and show a high divergence. These two groups the substitution rates are comparatively homogeneous
regions were excluded from the analyses. In the 28S rRNA gene, (18S, 0 –5.0; 28S, 0 – 4.7), except for Pupa solidula and Hydatina
length variation and extensive sequence divergence occur in the physis, which are distinguished by slightly higher Z-scores in the
domains D2 and D3. Therefore, the alignment positions 491– 539 18S gene (0.1 – 5.9)
(D2), 667– 729 (D2) a nd 1113– 1398 (D3) were not considered
in the phylogenetic analyses. Maximum-parsimony analysis
A maximum-parsimony analysis of the complete 18S rRNA gene
Phylogenetic signal with Odostomia sp. (Pyramidellidae) as outgroup produced two
The application of PHYSID (Wägele & Rödding, 1998) reveals shortest topologies (length 2903 steps, CI ¼ 0.63, RI ¼ 0.78,
that the signal-to-noise relationship is adverse in both alignments. RC ¼ 0.49). The strict consensus tree is shown in Figure 1. All
Most opisthobranch subgroups are supported by as many or opisthobranch subgroups appear monophyletic, with the Nudi-
fewer putative apomorphies as nonsense groups, i.e. the pleura (Nudibranchia and Pleurobranchoidea), Acteonoidea
phylogenetic signals for these clades are not higher than (Pupa solidula and Hydatina physis) and Sacoglossa forming one
background noise. The most frequent partition separates Cuthona major clade and Cephalaspidea s.s., Anaspidea, Tylodinoidea
and Flabellina (Nudibranchia) from the remaining taxa, sup- and Acochlidiacea forming the other. The Acteonoidea are the
ported by 217/102 (18S dataset/28S dataset) putative apomor- sister group of the Nudipleura and the members of the Pulmonata
phies. Concerning the major opisthobranch clades the are basal in the tree. The bootstrap analysis (1000 replicates)
Pleurobranchoidea show the best signal (70/42 putative apo- reveals low support for the position of the major opisthobranch
morphies), followed by Nudibranchia (50/34), Nudipleura (31/ groups. Monophyletic groups with high bootstrap support are,
24), Acteonoidea (31/21), Sacoglossa (10/14), Cephalaspidea s.s. for example, Nudipleura (100%), Pleurobranchoidea (100%),
(12/4), Tylodinoidea (6/10), Anaspidea þ Cephalaspidea s.s. Nudibranchia (90%), Acteonoidea (100%), Nudipleura plus
(7/4), Anaspidea (2/8) and Acochlidiacea (1/1). Acteonoidea (98%), Tylodinoidea (97%), Sacoglossa (100%)
The permutation-tail probability (PTP) test on the other hand and Cephalaspidea s.s. except Runcina adriatica (100%). The
reveals significant phylogenetic information in the 18S as well as sister-group relationship of Cephalaspidea s.s. and Anaspidea is
in the 28S dataset. After generation of 100 random data sets, supported by a bootstrap value of only 56%; the sister-group
computed tree lengths were compared with the MP trees (50% relationship between these two groups and the Tylodinoidea by
majority rule consensus; Figs 1, 2) of the original data. In the 18S only 59%. The positions of the Sacoglossa, Acochlidiacea and
rRNA gene dataset, the shortest tree length of the unpermutated Pulmonata are not resolved at all and the acochlidiacean species
data is 2903 and the shortest random set length is 5996, resulting Hedylopsis spiculifera does not form a monophyletic group with the
in the minimum possible PTP of 0.01. In the 28S the PTP is also other Acochlidiacea.
0.01 (actual length ¼ 3413, shortest random set length ¼ 5678). The result of the maximum-parsimony analysis of the 28S
rRNA gene (Fig. 2; 16 shortest trees, length 3413 steps,
CI ¼ 0.45, RI ¼ 0.63, RC ¼ 0.28) shows several correspon-
Statistical tests
dences with the results based on the 18S rRNA gene alignment.
Execution of the chi-squared test shows a significant hetero- Subordinate taxa of the Opisthobranchia are monophyletic
geneous base composition in the 18S (P ¼ 0) as well as the 28S except the Acochlidiacea (as above). The sister-group relation-
(P ¼ 0.02) rRNA gene data, caused primarily by the sequences of ships between Nudipleura and Acteonoidea as well as between
the Nudibranchia. After exclusion of this taxon, P values of 0.89 Cephalaspidea s.s. and Anaspidea are confirmed, and a sister-
(18S) and 0.92 (28S) are obtained. Additional exclusion of the group relationship between Sacoglossa and Tylodinoidea is
Pleurobranchoidea leads to a homogeneous composition of bases proposed. The bootstrap tree does not differ significantly from
(18S, P ¼ 1.0; 28S, P ¼ 0.99). Regarding the AT and GC- the maximum-parsimony tree, but resolution is lower. Mono-
content of the 18S rRNA gene sequences a progressive increase of phyletic groups with high bootstrap support are, for example,
AT is observed from Cladobranchia (Nudibranchia, 38.4%) to Nudibranchia (97%), Pleurobranchoidea (99%), Nudipleura
Anthobranchia (Nudibranchia, 42.4%), Pleurobranchoidea (100%), Acteonoidea (100%), Cephalaspidea s.s. (88%),
(43.7%), Acteonoidea (46.8%) and the remaining taxa Anaspidea (100%), Sacoglossa (100%), Acochlidiacea without
(48.1%). This trend is not as distinct in the 28S rRNA gene Hedylopsis spiculifera (99%) and Tylodinoidea (91%). The sister-
(Cladobranchia, 31.7%; Anthobranchia, 32.5%; Pleurobran- group relationship between Nudipleura and Acteonoidea has a
choidea, 34.2%; Acteonoidea, 32.9%; others, 36.8%). Investi- bootstrap value of only 54%, and that between Cephalaspidea
gation of the nucleotide substitutions by plotting the P-distances s.s. and Anaspidea has a bootstrap value of 76%.
against the absolute number of transitions and transversions (not The incongruence length-difference test (Farris et al., 1994)
shown) indicated that the 18S and 28S rRNA gene sequences are shows a P value of 0.003 (1000 replicates), indicating a significant
not saturated. degree of incongruence. Cunningham (1997) showed that
The relative-rate test (Sarich & Wilson, 1973) reveals that combining genes generally improved phylogenetic accuracy,
evolutionary rates are significantly different in both genes. The even in cases where incongruence was detected. Only when
Cladobranchia (Nudibranchia) show the highest rates in both P values were lower than 0.001 did the combined data suffer
genes, with a maximum Z-score of 15.9 between Dendronotus dalli relative to the individual partitions. Therefore, the 18S and 28S
and Siphonaria alternata (Pulmonata) in the 18S rRNA gene and rRNA gene data sets have also been analysed combined.
10.9 between Dendronotus dalli and Dolabella auricularia The combined analysis of the two genes (Fig. 3) shows a
(Anaspidea) in the 28S rRNA gene. The Anthobranchia topology (8 shortest trees, length 6352 steps, CI ¼ 0.53,
(Nudibranchia) show higher rates than the other taxa except RI ¼ 0.7, RC ¼ 0.37) that is better resolved than the 28S tree,
Cladobranchia (18S, 5.9 – 10.1; 28S, 2.6 –7.4), and the Pleuro- but not as good as the 18S tree. The Pulmonata are presented as
branchoidea evolve faster than the remaining taxa except the sister group of the Opisthobranchia, the remaining taxa form
Nudibranchia (18S, 5.0 – 9.8; 28S, 1.7 – 7.0). Within the three major clades, of which the relationships are unresolved.

117
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

Figure 1. Strict consensus tree of two shortest maximum-parsimony trees based on complete 18S rRNA gene (length 2903 steps, CI ¼ 0.63, RI ¼ 0.78,
RC ¼ 0.49). Bootstrap indices (50% majority rule) are given.

One clade consists of Nudipleura þ Acteonoidea þ Sacoglossa, gamma distribution shape parameters of a ¼ 0.31 (18S) and
the second of Cephalaspidea s.s. þ Anaspidea þ Tylodinoidea a ¼ 0.52 (28S). The proportions of invariable sites are 0.27
and the third of the Acochlidiacea. The bootstrap analysis shows (18S) and 0.23 (28S). The phylogram of the 18S dataset is
higher resolution compared with the bootstrap trees of the single shown in Figure 4, revealing that branch lengths are
genes. All opisthobranch subgroups are supported by high extremely short in all groups except Nudibranchia, Pleuro-
bootstrap values, as well as Nudibranchia plus Pleurobranchoi- branchoidea and Acteonoidea. The results of the distance trees
dea (100%), Nudipleura plus Acteonoidea (98%) and Cepha- and bootstrap values (18S and 28S gene) are presented in
laspidea s.s. plus Anaspidea (91%). Figure 5A, B. For better clarity and easier comparison of
opisthobranch intrarelationships we prefer to present the
distance and maximum-likelihood trees as cladograms and
Distance methods not as phylograms, and tree presentation is limited to the
For choosing the most appropriate model of sequence relationships of the major opisthobranch groups. If they are
evolution, hierarchical likelihood-ratio tests (Goldman, 1993) not monophyletic, the excluded taxa are specified separately.
are conducted by means of the program Modeltest Version In Figure 5A, B it can be seen that several taxa have varying
3.06 (Posada & Crandall, 1998). For both datasets the best-fit positions within the cladograms (Sacoglossa, Anaspidea,
model is the Tamura-Nei model (Tamura & Nei, 1993) with Acochlidiacea, Pulmonata), and even the monophyly of

118
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

Figure 2. Strict consensus tree of 16 shortest maximum-parsimony trees based on partial 28S rRNA gene (length 3413 steps, CI ¼ 0.45, RI ¼ 0.63,
RC ¼ 0.28). Bootstrap indices (50% majority rule) are given.

some major taxa is not supported (Pulmonata by exclusion of Maximum-likelihood analysis

Cepaea hortensis and/or Lymnaea stagnalis, Cephalaspidea s.s. by
exclusion of Runcina adriatica). Bootstrap values in the 18S tree The phylogenetic reconstructions by means of the maximum-
are high for e.g. Cladobranchia (100%), Nudipleura (96%), likelihood method were also conducted with the settings obtained
Acteonoidea (100%), the latter plus Nudipleura (86%), from the likelihood-ratio tests (see above).
Sacoglossa (92%), Cephalaspidea s.s. except Runcina adriatica The trees based on 18S and 28S rRNA genes (Fig 5C, D) again
(99%), Anaspidea (84%) and Tylodinoidea (99%). For 28S, show differences. Monophyly is shown for the major opistho-
they are high, for example, Anthobranchia (98%), Clado- branch groups except for the Acochlidiacea (because of Hedylopsis
branchia (100%), Nudibranchia (98%), Pleurobranchoidea spiculifera). Moreover Runcina adriatica is excluded again from the
(99%), Nudipleura (99%), Acteonoidea (100%), Anaspidea Cephalaspidea s.s. in the 18S tree. Bootstrap analyses could not
(99%), Cephalaspidea s.s. except Runcina adriatica (90%), be carried out because of the computation time required. The
Sacoglossa (100%) and Tylodinoidea (97%). two datasets were not analysed simultaneously under the
A distance analysis of the combined datasets is not carried out maximum-likelihood criterion (see above).
because the ILD test is presently only implemented with Comparing all four analyses (Fig. 5A– D), the sister-taxa
parsimony. relationship of Nudipleura and Acteonoidea occurs three times,

119
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

Figure 3. Strict consensus tree of 8 shortest maximum-parsimony trees based on the combined dataset of 18S and 28S rRNA genes (length 6352 steps,
CI ¼ 0.53, RI ¼ 0.7, RC ¼ 0.37). Bootstrap indices (50% majority rule) are given.

and the sister-taxa relationship of Cephalaspidea s.s. and Guerdoux & Tillier, 1994; Tillier, Masselot & Tillier, 1996;
Anaspidea twice. Ponder & Lindberg, 1997; Thollesson, 1999; Dayrat et al., 2001).
A third phylogenetic hypothesis for the Euthyneura has
recently been presented by Grande et al. (2003), based on the
analysis of eight mitochondrial genes. They recovered the
DISCUSSION pulmonates as a polyphyletic group with the Stylommatophora
To date, the phylogeny of the Opisthobranchia is unresolved, as an early split and the basommatophoran pulmonate Siphonaria
both in respect of the origin of the whole group as well as of the placed within the opisthobranchs.
interrelationships of its major clades. Two possible evolutionary Because of the well-known rampant parallelism in opistho-
scenarios have been discussed: Opisthobranchia and Pulmonata branch gastropods, there are very few ‘uniquely derived
evolved from a common ancestor and are therefore sister-groups attributes’ in the Opisthobranchia (Gosliner & Ghiselin, 1984).
(e.g. Gosliner, 1981, 1994; Gosliner & Ghiselin, 1984; Wade & Mikkelsen (2002) mentioned the following synapomorphies:
Mordan, 2000; Mikkelsen, 2002; Wägele et al., 2003); Pulmonata loss of the adult operculum, presence of parapodia (with
originated within the Opisthobranchia, rendering the three reversals) and loss of the posterior stomach chamber.
latter paraphyletic (e.g. Haszprunar, 1985; Tillier, Masselot, Many phylogenetic investigations concerning the various

120
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

Figure 4. Distance phylogram based on the complete 18S rRNA gene. Model of sequence evolution: Tamura-Nei model (Tamura & Nei, 1993), gamma
distribution shape parameter ¼ 0.31, proportions of invariable sites ¼ 0.27.

opisthobranch subgroups have been carried out and have (e.g. Mikkelsen, 1996). The ‘lower heterobranchs’ comprise
produced some good results. The Cephalaspidea s.l. are several families, for example Rissoellidae, Architectonicidae and
commonly accepted as a paraphyletic group consisting of the Pyramidellidae, that in the past have been variously aligned with
monophyletic group Cephalaspidea s.s. and a cluster of primitive ‘prosobranchs’ or opisthobranchs.
forms, the Acteonoidea (Acteonidae, Hydatinidae and Ring- The Anaspidea (including Akera) are well supported as a
iculidae). Apomorphies of Cephalaspidea s.s. are the ciliated monophyletic group by several synapomorphies from the
strips flexed at the mantle margin and the presence of three reproductive system, defensive glands, radula, gizzard and
gizzard plates (Mikkelsen, 1996). The Acteonoidea are either nervous system (Morton & Holme, 1955; Ghiselin, 1966;
a basal and probably paraphyletic opisthobranch group Mikkelsen, 1996; Medina & Walsh, 2000). The Sacoglossa,
(e.g. Gosliner, 1994) or have even been excluded from the with reallocation of the genus Cylindrobulla, are confirmed as
Opisthobranchia, belonging to the ‘lower heterobranchs’ monophyletic by, for example, the uniserial radula and the ascus

121
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

Figure 5. Distance (branch lengths not considered) and maximum-likelihood trees (branch lengths not considered) based on complete 18S rRNA and
partial 28S rRNA genes. For better understanding and easier comparison, the terminal taxa are condensed into monophyletic major groups. Bootstrap
indices (50% majority rule) are given. Models of sequence evolution were determined by means of Modeltest Version 3.06 (Posada & Crandall, 1998).
A. Distance tree based on the complete 18S rRNA gene. B. Distance tree based on the partial 28S rRNA gene. C. Maximum-likelihood tree based on the
complete 18S rRNA gene. D. Maximum-likelihood tree based on the partial 28S rRNA gene.

(Mikkelsen, 1998, 2002). Monophyly of the Notaspidea The pelagic Gymnosomata and Thecosomata have not yet
(Pleurobranchoidea plus Tylodinoidea) has been rejected by been studied phylogenetically and are not considered in the
several workers (Minichev & Starobogatov, 1978; Schmekel, present paper.
1985; Salvini-Plawen, 1991; Salvini-Plawen & Steiner, 1996; Our knowledge about the phylogeny of the subgroups within
Wägele & Willan, 2000). Characters presented as apomorphies the opisthobranchs is fragmentary because extensive morpho-
by Willan (1987), for example a bipinnate gill on the right side of logical analyses based on cladistic principles are lacking and
the body and longitudinally slit rhinophores, have since been molecular analyses have produced contradictory or poorly
shown to be plesiomorphies. Monophyly of the Tylodinoidea is resolved trees (e.g. compare the molecular studies by Thollesson,
supported by a cuticularized thickened ring just inside the mouth 1999, and Dayrat et al., 2001 with the morphological analysis by
(Willan, 1987) and the concentration of the nervous system with Wägele & Willan, 2000). Mikkelsen (1996), in her thorough
paired cerebral, pleural-intestinal and pedal ganglia (Schmekel, morphological analyses of Cephalaspidea s.l., found a sister-
1985). The Pleurobranchoidea are united by the internal, group relationship of Cephalaspidea s.s. and Anaspidea, together
rectangular shell and the presence of a pedal gland (Willan, representing the sister-group of the Sacoglossa. In 2002 she
1987). Schmekel (1985) was the first to propose a sister-group conducted a combined analysis of characters from four published
relationship of Pleurobranchoidea and Nudibranchia, confirmed phylogenies based on morphological characters (Willan, 1987;
by Salvini-Plawen (1990, 1991) and Wägele & Willan (2000). Mikkelsen, 1996; Jensen, 1996, 1997; Wägele & Willan, 2000),
The latter authors named the new taxon Nudipleura for although the resulting cladogram was not well resolved. The
Pleurobranchoidea plus Nudibranchia, characterized by the most derived clade consisted of Tylodinoidea and Nudipleura,
possession of a blood gland, an androdiaulic reproductive system forming a monophyletic group with Anaspidea and Cephalaspi-
and the loss of the osphradium. dea s.s. All groups together were the sister taxon of the
A polyphyletic origin has often been assumed for the Sacoglossa. A sister-group relationship of Anaspidea and
Nudibranchia (e.g. Minichev, 1970) but recent analyses Cephalaspidea s.s. was also evident in molecular analyses of the
support their monophyly (Wägele & Willan, 2000). Apomor- 16S rRNA gene in the study by Thollesson (1999) and of the 18S
phies are, for example, the presence of a vacuolated rRNA gene, 16S rRNA gene and COI gene by Wägele et al.
epithelium and a longitudinally orientated pericardial complex (2003).
(Wägele & Willan, 2000). Practically nothing can be said about the position of
The monophyly of the interstitial Acochlidiacea is supported Acochlidiacea, Thecosomata and Gymnosomata, because only
by their unique external organization. They possess a distinct very few data are available. Gosliner & Ghiselin (1984) presented
visceral hump separated from the rest of the body, in which the the Acochlidiacea as the sister-group of the Sacoglossa based
head-foot complex can be partially or completely retracted on morphological characters. Dayrat et al. (2001) found a
(Schrödl, personal communication; Rankin, 1979). monophyletic group consisting of Thecosomata, Gymnosomata

122
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

and Anaspidea, based on analyses of the 28S rRNA gene, whereas partly appear as basal taxa. Although this is the second
Gymnosomata was not part of the Euthyneura but grouped with study to show this sister-group relationship, the introduction
the caenogastropods in the study of Thollesson (1999). of a name for the clade Nudipleura þ Acteonoidea is
An important contribution to the solution of opisthobranch postponed until further support is available.
phylogeny was made by Grande et al. (2004) by means of their † In the present study Tylodinoidea does not form a clade
molecular study of the Euthyneura, based on eight mitochondrial with Nudipleura but shows different positions in the
genes. They recovered the Sacoglossa as the most basal group and different analyses due to the lack of phylogenetic infor-
a well-supported clade consisting of Anaspidea, Cephalaspidea mation in both genes. A relationship with Cephalaspidea
and Tylodinoidea. The Nudibranchia formed a paraphyletic and Anaspidea, as suggested by Grande et al. (2003), is
group, because the Pleurobranchoidea were located within them. shown in the maximum-parsimony trees of the18S rRNA
Certainly the most striking affinity was the sister-group (Fig. 1), the combined dataset (Fig. 3) and the maximum-
relationship between Acteonoidea and Nudibranchia plus likelihood tree of the 18S rRNA gene (Fig. 5C).
Pleurobranchoidea (discussed below). † The position of Sacoglossa is also variable. In the
The results of the present study, based on the 18S and 28S phylogenetic reconstructions of the 18S rRNA gene and
rRNA genes and reconstructed by means of maximum parsi- the combined datasets, this taxon shows an affinity with
mony, maximum likelihood and distance methods, can be Nudipleura and Acteonoidea, but in the trees based on the
summarized as follows. 28S rRNA gene no such relationship can be recognized. A
basal position, as lin Grande et al. (2004), is not supported
† Opisthobranchia is mainly presented as a monophyletic group at all.
if Pupa solidula and Hydatina physis are considered opistho- † Acochlidiacea is not recognized as a monophyletic group in
branchs. Opisthobranchia can be either para- or monophy- several analyses, because Hedylopsis spiculifera, a member of
letic, because Pulmonata is arranged among Opisthobranchia the family Hedylopsidae, does not group with the three
in the maximum-likelihood and distance trees, or as a basal members of the Microhedylidae (Unela, Pontohedyle, Para-
paraphyletic group in the maximum-parsimony analyses. hedyle). Except for the 28S maximum-likelihood tree,
† Monophyly of the major opisthobranch clades is supported in Acochlidiacea shows a more or less basal position, mostly
most analyses, usually with high bootstrap values. However, in close proximity to the Pulmonata.
Acochlidiacea forms a monophyletic group only in the † The amount of phylogenetic information in the 18S and
maximum-parsimony (18S and combined tree) and distance 28S rRNA gene sequences at the family and genus levels is
analyses, and a high bootstrap value is found only in the rather different concerning the several opisthobranch
maximum-parsimony tree of the combined dataset. Further- subgroups. In all analyses the phylogeny of Nudibranchia
more Runcina adriatica is excluded from the Cephalaspidea s.s. is in accordance with Wägele & Willan (2000). The
in the maximum-likelihood and distance tree of the 18S rRNA position of the pleurobranchoidean genera, on the other
gene. In addition to the topological support, the monophyly of hand, is extremely variable, dependent on which gene and
the pleurobranchoidean taxa is corroborated by the corre- method is used. The arrangement of Cephalaspidea s.s. is
sponding insertion (structure E23_3) in region V4 of the 18S comparatively consistent. The Aglajidae (Chelidonura, Odon-
rRNA gene. toglaja, Navanax, Philinopsis) is recovered as a monophyletic
† The arrangement of the major opisthobranch clades shows group in each analysis, and the Haminoeidae (Phaner-
variability caused by the lack of phylogenetic signal in both ophthalmus, Atys, Haminoea) in each except in the 18S
alignments, with exceptions of the following three points. distance tree. Runcina, if included, occupies the most basal
† The sister-group relationship of Nudibranchia and Pleuro- position within the Cephalaspidea s.s. and Retusa represents
branchoidea (Nudipleura) is confirmed in all analyses with the sister-group of the Haminoeidae except in the 18S
high bootstrap values (96 – 100%). distance tree where it groups within the Haminoeidae. The
† The monophyly of Cephalaspidea s.s. plus Anaspidea is also positions of Philinoglossa, Sagaminopteron and Bulla are
confirmed (except in distance trees), although bootstrap variable. The basal position of Runcina adriatica is compa-
support is lower than for Nudipleura (56 –91%). tible with morphological observations, because Runcina
† A sister-group relationship of Nudipleura and Acteonoidea possesses four gizzard plates, whereas all other members
(28/21 putative apomorphies detected by PHYSID) is of the Cephalaspidea s.s. have three (Mikkelsen, 1996),
shown in all trees (except the maximum-likelihood tree of implying that four plates represent the plesiomorphic
the 18S rRNA gene). This is remarkable because the
character state. Concerning Sacoglossa, Limapontiidae
Acteonoidea are usually considered the most basal opistho-
(Limapontia, Calliopaea) is recovered in each tree and a
branchs or are even excluded from this taxon (Mikkelsen
close affinity between Thuridilla, Elysia, Placobranchus and
1996, 2002), whereas the Nudipleura are usually regarded
Bosellia (Placobranchoidea) is always shown except in the
as highly derived (e.g. Wägele et al., 2003). At first sight this
28S distance tree. The position of Cyerce is variable.
relationship therefore seems to be due to a long-branch
Microhedylidae, containing Unela, Parahedyle and Pontohe-
problem, because Pupa, Hydatina and especially the Nudi-
dyle, is always monophyletic.
pleura shows higher evolutionary rates than the other taxa.
On the other hand, it is possible that there are real sister
taxa, because there are some similar morphological char- Nucleotide composition bias and rate heterogeneity, found in
acters. Ghiselin, who united Acoela (Nudipleura and both genes, are estimated to have no affect on the accuracy of the
Tylodinoidea) with Hydatinidae and Acteonidae on the reconstructed phylogenies. The detected differences in the AT-
basis of their reproductive system, mentioned that ‘Eliot content reflect phylogenetic relationships already established on
(1910) notes a number of […] similarities which imply that the basis of morphological investigations, i.e. Cladobranchia,
there is an affinity between the Acoela and the Hydatinidae Anthobranchia and Pleurobranchoidea (Wägele & Willan,
and Acteonidae’ (Ghiselin, 1966: 371). In contrast to the 2000). The relative-rate test supports the same groups by
phylogeny proposed by Ghiselin (1966), Nudipleura and different classes of substitution rates (Cladobranchia .
Acteonoidea represent a very derived clade in the maxi- Anthobranchia . Pleurobranchoidea . remaining taxa). The
mum-parsimony and distance trees of the present study. In only case where higher substitution rates can potentially mislead
the maximum-likelihood analyses on the other hand they phylogeny estimation is that of the Acteonoidea, because within

123
 V. VONNEMANN ET AL.V. VONNEMANN ET AL.

the ‘remaining taxa’ the Acteonoidea show the highest evol- HASZPRUNAR, G. 1985. The Heterobranchia—a new concept of the
utionary rates. phylogeny of the higher Gastropoda. Zeitschrift für Zoologische System-
The suitability of 18S and partial 28S rRNA genes for atik und Evolutionsforschung, 23: 15 –37.
investigating the phylogeny of the Opisthobranchia must be JENSEN, K.R. 1996. The Diaphanidae as a possible sister-group of the
assessed critically. Although monophyly of the individual Sacoglossa (Gastropoda, Opisthobranchia). In: Origin and evolutionary
subgroups is supported, the relationships of several major radiation of the Mollusca (J.D. Taylor, ed.), 231–247. Oxford University
Press, Oxford.
clades remain uncertain. The analyses with PHYSID reveal a
relatively low content of phylogenetic information in both JENSEN, K.R. 1997. Sacoglossernes systematik, fylogeni og evolution (Mollusca,
genes. This is probably due to the many highly conserved Opisthobranchia). Vestjydsk Forlag, Copenhagen.
positions and the loss of phylogenetic signal in the variable MEDINA, M. & WALSH, P.J. 2000. Molecular systematics of the order
positions by multiple hits. The content of phylogenetic Anaspidea based on mitochondrial DNA sequence (12S, 16S, and
COI). Molecular Phylogenetics and Evolution, 15: 41–58.
information is higher in the 18S than in the (investigated
part) of the 28S rRNA gene, probably attributed to the MIKKELSEN, P.M. 1996. The evolutionary relationships of Cephalas-
pidea s.l (Gastropoda: Opisthobranchia): A phylogenetic analysis.
greater sequence length of the 18S rRNA gene. Malacologia, 37: 375–442.
MIKKELSEN, P.M. 2002. Shelled opisthobranchs. Advances in Marine
Biology, 42: 67– 136.
ACKNOWLEDGEMENTS MINICHEV, Y.S. 1970. On the origin and system of nudibranchiate
molluscs (Gastropoda, Opisthobranchia). Monitore Zoologico Italiano, 4:
This project was supported by several grants from the German 169 –182.
Science Foundation to the authors: Wa 618/5 (HW), Wa
MINICHEV, Y.S. & STAROBOGATOV, Y.I. 1978. The subclasses of
618/7 (HW, VV, MS) and KL 1303/1 (AKK). We wish to
Gastropoda and their phylogenetic relations. Zoologischeskii Zhurnal,
express our sincere thanks to some colleagues who provided 58: 293–305.
material for this study: J. Brodie (Townsville), B. Bolland
MORTON, J.E. & HOLME, N.A. 1955. The occurrence at Plymouth of
(Okinawa), J. Bohn (Munich), P. Durán (Santiago de Chile), the opisthobranch Acera bullata, with notes on its habits and
A. Fahrner (Munich), G. Haszprunar (Munich), N. Michiels relationships. Journal of the Marine Biological Association of the United
(Münster), M. Raupach (Bochum) and G. Steiner (Vienna). Kingdom, 34: 101–112.
Thanks also to two unknown reviewers and M. Thollesson for NICHOLAS K.B. & NICHOLAS, H.B. 1997. GeneDoc: analysis and
valuable criticism. visualization of genetic variation, available from http://www.psc.edu/
biomed/genedoc/
PONDER, W.F. & LINDBERG, D.R. 1997. Towards a phylogeny of
REFERENCES gastropod molluscs: an analysis using morphological characters.
Journal of the Linnean Society, 119: 83–265.
ARCHIE, J.W. 1989. A randomization test for phylogenetic information POSADA, D. & CRANDALL, K.A. 1998. Modeltest: testing the model
in systematic data. Systematic Zoology, 38: 219–252. of DNA substitution. Bioinformatics, 14: 817–818.
ARNAUD, P.M., POIZAT, C. & VON SALVINI-PLAWEN, L. 1986. RANKIN, J.J. 1979. A freshwater shell-less mollusc from the Carribean:
Marine-interstitial Gastropoda (including one freshwater interstitial structure, biotics and contribution to a new understanding of the
species). In: Stygofauna Mundi (L. Botosaneanu, ed.), 153–176. Brill, Acochlidiacea. Life Sciences Contributions of the Royal Ontario Museum,
Leiden. 116: 1–123.
BEHRENS, D.W. 1980. Pacific coast nudibranchs. Sea Challengers Natural SAIKI, R.K., GELFAND, D.H., STOFFEL, S., SCHARF, S.J.,
History Books, Danville. HIGUCHI, R., HORN, G.T., MULLIS, K.B. & ERLICH, H.A.
CUNNINGHAM, C.W. 1997. Can three incongruence tests predict 1988. Primer-directed enzymatic amplification of DNA with a
when data should be combined? Molecular Biology and Evolution, 14: thermostable DNA polymerase. Science, 239: 487– 491.
733–740. SALVINI-PLAWEN, L. VON 1990. Origin, phylogeny and classifi-
DAYRAT, B., TILLIER, A., LECOINTRE, G. & TILLIER, S. 2001. cation of the phylum Mollusca. Iberus, 9: 1 –33.
New clades of euthyneuran gastropods (Mollusca) from 28S rRNA SALVINI-PLAWEN, L. VON 1991. The status of the Rhodopidae
sequences. Molecular Phylogenetics and Evolution, 19: 225–235. (Gastropoda: Euthyneura). Malacologia, 32: 301 –311.
ELIOT, C. 1910. A monograph of the British nudibranchiate Mollusca, Part 8 SALVINI-PLAWEN, L. VON & STEINER, G. 1996. Synapomorphies
(Supplement). Ray Society, London. and plesiomorphies in higher classification of Mollusca. In: Origin and
FAITH, D.P. & CRANSTON, P.S. 1991. Could a cladogram this short evolutionary radiation of the Mollusca (J.D. Taylor, ed.), 29–51. Oxford
have arisen by chance alone? On permutation tests for cladistic University Press, Oxford.
structure. Cladistics, 7: 1 –28. SANGER, F., MIKLEN, S. & COULSON, A.R. 1977. DNA sequencing
FARRIS, J.S., KÄLLERSJÖ, M., KLUGE, A.G. & BULT, C. 1994. with chain-termination inhibitors. Proceedings of the National Academy of
Testing significance of incongruence. Cladistics, 10: 315 –319. Sciences of the USA, 74: 5463–5467.
GHISELIN, M.T. 1966. Reproductive function and the phylogeny of SARICH, V.M. & WILSON, A.C. 1973. Immunological time scale for
opisthobranch gastropods. Malacologia, 3: 327–378. hominoid evolution. Science, 158: 1200–1203.
GOLDMAN, N. 1993. Simple diagnostic statistical test of models for SCHMEKEL, L. 1985. Aspects of the evolution within opisthobranchs.
DNA substitution. Journal of Molecular Evolution, 37: 650 –661. In: The Mollusca (E.R. Trueman & M.R. Clarke, eds), Evolution,
GOSLINER, T.M. 1981. Origins and relationships of primitive members Vol. 10. 221–267. Academic Press, Orlando.
of the Opisthobranchia (Mollusca: Gastropoda). Biological Journal of SWOFFORD D.L. 1999. PAUP: Phylogenetic analysis using
the Linnean Society, 16: 197–225. parsimony. Version 4.0 beta 3a. Sinauer Associates, Sunderland,
GOSLINER, T.M. 1994. Gastropoda: Opisthobranchia. In: Microscopic Massachusetts.
anatomy of invertebrates (F.W. Harrison & A.J. Kohn, eds), Mollusca I, TAMURA, K. & NEI, M. 1993. Estimation of the number of nucleotide
Vol. 5. 253 –355. Wiley Liss, New York. substitutions in the control region of mitochondrial DNA in
GOSLINER, T.M. & GHISELIN, M.T. 1984. Parallel evolution in humans and chimpanzees. Molecular Biology and Evolution, 10:
opisthobranch gastropods and its implications for phylogenetic 512 –526.
methodology. Systematic Zoology, 33: 255– 274. THOLLESSON, M. 1999. Phylogenetic analysis of Euthyneura
GRANDE, C., TEMPLADO, J., CERVERA, J.L. & ZARDOYA, R. (Gastropoda) by means of the 16S rRNA gene: Use of a ‘fast’ gene
2004. Molecular phylogeny of Euthyneura (Mollusca: Gastropoda). for ‘higher-level’ phylogenies. Proceedings of the Royal Society of London,
Molecular Biology and Evolution, 21: 303 –313. Series B, 266: 75–83.

124
 PHYLOGENY OF OPISTHOBRANCHIAPHYLOGENY OF OPISTHOBRANCHIA

THOMPSON, J.D., GIBSON, T.J., PLEWNIAK, F., JEANMOUGIN, WÄGELE, J.-W. & RÖDDING, F. 1998. A priori estimation of
F. & HIGGINS, D.G. 1997. The Clustal X windows interface: phylogenetic information conserved in aligned sequences. Molecular
Flexible strategies for multiple sequence alignment aided by quality Phylogenetics and Evolution, 9: 358–365.
analysis tools. Nucleic Acids Research, 25: 4876–4882. WÄGELE, H. & WILLAN, R.C. 2000. On the phylogeny of
TILLIER, S., MASSELOT, M., GUERDOUX, J. & TILLIER, the Nudibranchia. Zoological Journal of the Linnean Society, 130: 83–181.
A. 1994. Monophyly of major gastropod taxa tested from partial WÄGELE, H., VONNEMANN, V. & WÄGELE, J.-W. 2003. Towards
28S rRNA sequences, with emphasis on Euthyneura and hot-vent
a phylogeny of the Opisthobranchia. In: Molecular systematics and
limpets Peltospiroidea. Nautilus, 2(Suppl.): 122–140.
phylogeography of mollusks (C. Lydeard & D. Lindberg, eds), 185–228.
TILLIER, S., MASSELOT, M. & TILLIER, A. 1996. Phylogenetic Smithsonian Institution Press, Washington, DC.
relationships of the pulmonate gastropods from rRNA sequences, and
tempo and age of the stylommatophoran radiation. In: Origin and WILLAN, R.C. 1987. Phylogenetic systematics of the Notaspidea
evolutionary radiation of the Mollusca (J.D. Taylor, ed.), 267 –284. Oxford (Opisthobranchia) with reappraisal of families and genera. American
University Press, Oxford. Malacological Bulletin, 5: 215– 241.
WADE, C.M. & MORDAN, P.B. 2000. Evolution within the WU, C.I. & LI, W.H. 1985. Evidence for higher rates of nucleotide
gastropod molluscs; using the ribosomal RNA gene-cluster as an substitution in rodents than in man. Proceedings of the National Academy
indicator of phylogenetic relationships. Journal of Molluscan Studies, of Sciences of the USA, 82: 1741–1745.
66: 565–570. WUYTS, J., VAN DE PEER, Y. & DE WACHTER, R.D. 2001.
WÄGELE, J.-W. 1996. Identification of apomorphies and the role of Distribution of substitution rates and location of insertion sites in the
groundpatterns in mlecular systematics. Journal of Zoological Systematics tertiary structure of ribosomal RNA. Nucleic Acids Research, 29:
and Evolutionary Research, 34: 31– 39. 5017–5028.

125