Read Schulze2003.pdf text version

Phylogeny of Vestimentifera (Siboglinidae, Annelida) inferred from morphology

Blackwell Publishing Ltd.

ANJA SCHULZE

Accepted: 9 August 2002

Schulze, A. (2003). Phylogeny of Vestimentifera (Siboglinidae, Annelida) inferred from morphology. -- Zoologica Scripta, 32, 321­ 342. Vestimentifera, formerly considered a phylum, are here included in the annelid clade Siboglinidae which also encompasses Frenulata and Sclerolinum. All Siboglinidae inhabit reducing habitats, mostly in the deep sea. Vestimentifera are known from hydrothermal vents and cold seeps. Cladistic analyses of vestimentiferan relationships are performed on three levels: (1) among the vestimentiferan species, (2) among the reconstructed ancestral vestimentiferan and other siboglinids and (3) on the level of the families included in the annelidan clade Sabellida. The monophyly of vestimentiferans is confirmed in all analyses. A group of exclusively ventinhabiting species forms a derived monophyletic clade. The sister group to the vent clade is the Escarpia complex. Lamellibrachia appears to be paraphyletic. Except for the paraphyly of Lamellibrachia, the reconstructed pattern agrees with the molecular phylogeny based on cytochrome c oxidase subunit I. Ancient ridge systems can be invoked to explain modern day geographical distributions. The Pacific Kula Ridge that spanned the Pacific in an east­west direction during the Early Tertiary, may have been a pathway for the ancestor of the vent clade to reach the eastern Pacific. The biogeography is consistent with the recent divergence of Vestimentifera as inferred from molecular data. The reconstructed phylogeny of the Siboglinidae supports the monophyly of the Frenulata and within those, the Thecanephria and Athecanephria. In contrast to molecular and other morphological analyses, Sclerolinum appears as the sister group to the Frenulata. The family level analysis supports the sister group relationship of the Siboglinidae to a clade formed by Sabellariidae, Sabellidae and Serpulidae. Hypothesized homologies of the vestimentiferan obturaculum and vestimentum to structures in related taxa need further investigation. Anja Schulze, Department of Invertebrate Zoology, MCZ, 26 Oxford Street, Harvard University, Cambridge, MA 02138, USA. E-mail: [email protected]

Introduction

Vestimentifera attracted attention in the scientific community and the general public alike when they were first found in biological communities around hydrothermal vents in 1977. Researchers retrieved large specimens of a tube-dwelling worm from a site where elevated temperatures up to 23 °C and sulphide concentrations of up to 160 µM were measured (Jones 1981). The species, Riftia pachyptila, was new to science, but a related species, Lamellibrachia barhami, from cold-water seeps off the coast of Oregon, had been described earlier (Webb 1969). Fifteen species of vestimentiferans, representing 10 genera, have been described. The literature mentions 10 more species, making a total of 25 known species ( Table 1). As adults, Vestimentifera lack a functional digestive system. They derive their nutrition from chemoautotrophic microbial symbionts hosted in the trophosome, a specialized tissue located in their elongated trunk region. A long trunk

region with chemoautotrophic symbionts and the absence of a gut are shared by two other groups of tube worms that have long been considered related to Vestimentifera. One of them is the Frenulata (= Perviata), a taxon comprising over 130 species (Southward 2000). The name Frenulata was suggested by Webb (1969) and delimits the same group as Jones's (1981) Perviata. The second group is represented by the little known genus Sclerolinum, with only seven species (Webb 1964; Southward 1972; Smirnov 2000). Opinions about the phylogenetic affinities and taxonomic ranks of Vestimentifera, Frenulata and Sclerolinum have shifted remarkably throughout the history of scientific investigations on these taxa. For a detailed account see Rouse (2001). Today, Vestimentifera, Frenulata and Sclerolinum are almost unanimously regarded as a clade of derived annelids (but see Christofferson & Araújo-de-Almeida 1994; Salvini-Plawen 2000). Rouse & Fauchald (1997) proposed grouping

321

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

Table 1 Vestimentiferan species described or mentioned in the literature to date, their inclusion in the present study, general geographical area,

specific sites, depth ranges and type of habitat. General area: WP, west Pacific; EP, east Pacific; WA, west Atlantic; EA, east Atlantic; specific sites: Cle, San Clemente Fault; Edi, Edison Seamount, Papua New Guinea; Fij, North Fiji Basin; Flo, Florida Escarpment, Gulf of Mexico; Gal, Galapagos Ridge; Gua, Guaymas Basin; Guy, Guyana continental margin; Jav, Java Trench; Kag, Kagoshima Bay, Japan; Kan, Kanasuno-se, Japan; Lau, Lau Basin; Lou-I, Louisiana, lower continental slope; Lou-u, Louisiana, upper continental slope; Man, Manus Basin; Med, Mediterranean off Turkey; Mex, Mid-American Trench off Mexico; Mid, Middle Valley, Juan de Fuca Ridge; Mon, Monterey Bay; Nan-T13, Nankai Trough, sites of increasing depth; Oki, Okinawa Trough; Ore, Oregon Prism; Sag, Sagami Bay; Uru, Uruguay continental margin; Vig, shipwreck off Vigo, Spain; 9­13°N, 21°N, geographical latitude on east Pacific rise; habitat type: bv, basaltic vent; se, seep; sv, sedimented vent; wh, whale bones; wr, shipwreck.

Included in present analysis Yes No No No Yes Yes Yes No No Yes

Species

General area WP WP WP WP WP WA EP WP WP EP

Site(s) Lau Sag Oki Man Man, Fij Flo, Lou-I Mex, Gua WP Nan-T1 Nan-T2, Oki, Man Mid, Ore, Mon, Cle

Depth range 1750­1890 m 800­1450 m 1400 m 1700­1900 m 2190 m 1000(?)-3500 m 2000­4000 m 300 m 1200­1400 m 600­2400 m

Habitat type bv se ? ? bv se se/sv/wh se se/sv? se/sv

Reference(s) Southward (1991) Kojima et al. (2000) Kojima et al. (2000) Kojima et al. (2000) Southward & Galkin (1997); Southward et al. ( 2002) Jones (1985); Sibuet & Olu (1998) Jones (1985); Sibuet & Olu (1998) Kojima et al. (1997) Kojima et al. (1997, 2000) Webb (1969); Jones (1985); Southward et al. (1996); Sibuet & Olu (1998) Southward (1991) Van der Land & Nørrevang (1975, 1977) Fisher, pers. comm.; Sibuet & Olu (1998) Kojima et al. (1997); Miura et al. (1997) Mañé-Garzón & Montero (1986) Kojima et al. (1997) Kojima et al. (1997) Dando et al. (1992) Woodside (1997) Jones (1985); Van Dover (2000) Southward et al. (2002) Jones (1985); Tunnicliffe et al. (1998); Van Dover (2000) Jones (1981, 1985); Tunnicliffe et al. (1998); Van Dover (2000) Gardiner et al. (2001) Jones (1985); Tunnicliffe et al. (1998); Van Dover (2000)

Alaysia spiralis Alaysia? sp. 1 Alaysia? sp. 2 Alaysia? sp. 3 Arcovestia ivanovi Escarpia laminata Escarpia spicata Escarpia sp. 1 Escarpia sp. 2 Lamellibrachia barhami

Lamellibrachia columna Lamellibrachia luymesi Lamellibrachia cf. luymesi Lamellibrachia satsuma Lamellibrachia victori Lamellibrachia sp. 1 Lamellibrachia sp. 2 Lamellibrachia sp. 3 Lamellibrachia sp. 4 Oasisia alvinae Paraescarpia echinospica Ridgeia piscesae Riftia pachyptila Seepiophila jonesi Tevnia jerichonana

Yes Yes No Yes Yes No No No No Yes Yes Yes Yes Yes Yes

WP WA WA WP WA WP WP EA EA EP WP EP EP WA EP

Lau Guy Lou-u Kag, Kan Uru Sag, Kan, Nan-T2 Nan-T3 Vig Med 21°N, 9­13°N Edi, Jav JdF Gua, 21°N, 9­13°N, Gal Lou-u 9­13°N

1830­1890 m 500 m 400­1000 m 82­300 m 300 m 300­1300 m 2000 m 1160 m 1700­2000 m (?) 2500­2600 m 1600 m 1570­3250 m 2000­2600 m 450­550 m 2500­2600 m

bv se se se se se se wr se bv se? sv/bv bv se bv

Vestimentifera, Frenulata and Sclerolinum together into a single polychaete clade Siboglinidae. McHugh (1997) found molecular support for Rouse & Fauchald's (1997) morphology-based analysis and also advocated the name Siboglinidae. For the taxonomy within the Siboglinidae, Rouse (2001) suggested maintaining the names Frenulata, Monilifera and Vestimentifera. According to his cladistic analyses, the Vestimentifera are nested within the Monilifera. Throughout this paper, the name Sclerolinum will be used as

322

an equivalent to Monilifera sensu Ivanov (1991), but not sensu Rouse (2001). The names Frenulata, Vestimentifera and Siboglinidae will be used as suggested by Rouse (2001), i.e. the Siboglinidae encompassing all three subgroups. The vestimentiferan body can be divided into four regions (Fig. 1A). The anterior region, called the obturaculum, is either circular or Y-shaped in cross-section. Unlike the frenulate cephalic lobe, it develops dorsally to the first branchial filaments and has no obvious equivalent in frenulates

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Fig. 1 A­C. Morphology of Vestimentifera, Frenulata and Sclerolinum [modified from Southward et al. (1995) and Southward (1972, 2000)]. --A. Anterior body regions and opisthosome of the vestimentiferan Ridgeia piscesae; left: dorsal view; middle: ventral view; right: opisthosoma. --B. Generalized frenulate. --C. Sclerolinum major.

(Southward 1988). The obturaculum is enveloped in the branchial filaments that originate from the anterior vestimentum, the second body region. The branchial filaments are organized in paired half-circular lamellae. The vestimentum is characterized by two dorsolateral folds continuous with an anterior collar. The ventral side bears a pear-shaped or elongate ciliated field. The brain, excretory organs and

heart are all located in the anterior vestimentum. Adjacent to the vestimentum is the elongated trunk region in which the gonads and the trophosome are enclosed. The posterior body region, or opisthosome, consists of up to 95 segments, with rows of uncini on the anterior ones. The anterior body region in Frenulata is the cephalic lobe. The branchial filaments, or tentacles, are generally fewer in

323

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

number than in vestimentiferans and originate near the base of the cephalic lobe. Adjacent to the cephalic lobe is the forepart, characterized by a cuticular structure that runs obliquely around it. This structure, termed frenulum, was responsible for the name of the group (Webb 1969). In contrast to the vestimentiferan trunk, the trunk of frenulates shows regional differentiation (Fig. 1B). A segmented opisthosoma is also present, but differs from the vestimentiferan opisthosoma by the presence of only four rod-shaped chaetae per segment instead of uncinal rows. Sclerolinum is characterized by a very small cephalic lobe. Although Ivanov (1991) reports a muscular diaphragm between the forepart (= mesosoma) and the trunk (= metasoma) in two Sclerolinum species, an external distinction between these body regions is absent (Fig. 1C). The trunk is not divided into specialized regions. The posterior end of the trunk and the anterior opisthosomal segments bear rows of uncini. This paper examines the phylogeny among the vestimentiferan species on three levels: (1) relationships among the vestimentiferan species, (2) relationships among the Siboglinidae and (3) phylogenetic affinities of Siboglinidae to within the Sabellida clade. (2) and (3) show some overlap with Rouse & Fauchald (1997) and Rouse (2001). However, the character sets, coding methods and taxon choice are not equivalent with either of the studies, so the analyses presented here give a new outlook on the problem in question. The phylogeny resulting from the analysis of vestimentiferan species will be compared with the classification suggested by Jones (1985) and with phylogenetic studies based on molecular data. The phylogeny will serve as a basis for inferences on biogeographical distribution patterns and ancestral habitats.

Phylogeny of Siboglinidae The analysis was based on 20 characters (Appendices III, IV). The ingroup comprised the ancestral vestimentiferan as reconstructed in the previous step, a sclerolinid (Sclerolinum minor) and five frenulate species (Oligobrachia gracilis, Siboglinum ekmani, Siphonobrachia lauensis, Plybrachia canadensis, Spirobrachia beklemishevi), representing all of the frenulate families (Appendix IV). Oweniidae and Sabellidae have been suggested previously as close relatives of the Siboglinidae (Liwanow & Porfirjewa 1967; Bartolomaeus 1995, 1997; Rouse & Fauchald 1997; Colgan et al. 2001), therefore one representative species of each (Sabella pavonina and Owenia fusiformis) were chosen as the outgroup. Family level analysis The ingroup consisted of all families within the Sabellida as defined by Rouse & Fauchald (1997): Oweniidae, Sabellidae, Serpulidae, Sabellariidae, Siboglinidae. The outgroups were Terebellidae and Spionidae, representatives of the Terebellida and Spionida, respectively, the two clades that, together with the Sabellida, form the Canalipalpata (Rouse & Fauchald 1997) (Appendices V, VI). It was assumed that all included families are monophyletic [for evidence of monophyly see Fauchald & Rouse (1997)]. If polymorphism within taxa occurred, taxa were scored for the ancestral state. Most characters used in the present analysis were unambiguous within the taxa. In two cases, ancestral states (developmental mode in serpulids and spionids) were unknown for the species in question. The taxa were scored as polymorphic for this character. The characters were mainly extracted from Rouse & Fauchald's (1997) data set, but were sometimes recoded (Appendices V, VI). Additional characters, not used by Rouse & Fauchald (1997) are the excretory pores (single/paired), intraepidermal nerve cord, extracellular haemoglobin, spirally grooved nucleus of sperm and a lecithotrophic larva. Some character states used by Rouse & Fauchald (1997) were modified, because homologies were unclear. Coding strategies The method applied here was C-coding sensu Pleijel (1995), equivalent to `conventional' coding sensu Hawkins et al. (1997). To date, no algorithm is available to account for hierarchical character linkage in cladistic analyses of morphological data sets. Rouse & Fauchald (1997) suggested assigning full weight to primary characters, a weight of 0.5 to subsidiary characters and a weight of 0.25 to any characters subsidiary to those. This approach was adopted in addition to performing the analyses with equal weights. Quantitative characters Different solutions for coding quantitative data (Archie 1985; Thiele 1993; Zelditch et al. 1995; Sosa & De Luna 1998) have

Methods

Phylogeny of vestimentiferan species The data set included 15 vestimentiferan, three outgroup species and 24 characters (Table 1 and Appendices I, II). Outgroups were chosen among the Frenulata and Sclerolinum. One representative each of the two frenulate orders Thecanephria and Athecanephria (according to Ivanov 1963) and of Sclerolinum were selected. Siphonobrachia lauensis (Thecanephria) and Oligobrachia gracilis (Athecanephria) were specifically chosen because of their thorough descriptions in the literature, including ultrastructural data (Southward 1978, 1991, 1993). The choice of Sclerolinum minor was arbitrary. Only one Sclerolinum specimen (Sclerolinum brattstromi) has been sectioned for light microscopy to date, but the sections were not informative with regard to the excretory organs (Southward, pers. comm.) on which six characters from the present data set are based. The character states assigned to Sclerolinum minor would apply for other species of Sclerolinum as well (Southward 1972).

324

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Table 2 Vestimentiferan species ranked according to mean

Table 4 Vestimentiferan species ranked according to their number of

obturaculum/vestimentum length; range and character states are indicated.

Mean obturaculum/ vestimentum length 0.22 0.28 0.29 0.30 0.32 0.34 0.34 0.35 0.40 0.41 0.54 0.67 1.04 1.09 1.57

branchial lamellae; approximate range or maximum number and character states are indicated.

Number of branchial lamellae 4 18 up to 19 ca. 20 up to 20 around 20 20­25 25 up to 25 ca. 25 30­35 up to 33 up to 35 up to 63 up to 335

Species

n

1 64 8 3 1 1 4 6 2 7 8 6 1 16 20

Range

State 0 0 0 0 0 0 0 0 0 0 1 1 2 2 2

Species

n

2 1 76 ca. 13 6 4 12 ? 1 1 10 11 26 9 27

State 0 1 1 1 1 1 1 1 1 1 2 2 2 2 2

Lamellibrachia victori Lamellibrachia satsuma Lamellibrachia barhami Seepiophila jonesi Lamellibrachia luymesi Lamellibrachia columna Paraescarpia echinospica Escarpia spicata Alaysia spiralis Escarpia laminata Tevnia jerichonana Oasisia alvinae Arcovestia ivanovi Ridgeia piscesae Riftia pachyptila

? 0.13­0.40 0.23­0.37

0.29­0.41 0.25­0.50 0.37­0.42 0.34­0.56 0.34­0.70 0.40­0.81 0.55­1.81 0.98­3.09

Alaysia spiralis Seepiophila jonesi Lamellibrachia satsuma Lamellibrachia columna Oasisia alvinae Paraescarpia echinospica Arcovestia ivanovi Lamellibrachia barhami Lamellibrachia luymesi Lamellibrachia victori Escarpia laminata Tevnia jerichonana Ridgeia piscesae Escarpia spicata Riftia pachyptila

Table 3 Vestimentiferan species ranked according to mean

Table 5 Vestimentiferan species ranked according to mean number

vestimentum length/diameter; range and character states are indicated.

Mean vestimentum length/diameter 8.00 7.40 6.5 + 6.25 5.79 5.46 5.17 4.65 4.40 3.63 2.88 2.63 2.58 2.47

of opisthosomal segments; range and character states are indicated.

Mean number of opisthosomal segments 8 37

Species Range State 0 0 0 0 0 0 0 1 1 2 2 2 2 2

n

1 1

Range

State 0 1 ? ? ? 1 ? 1 ? 1 1 1 1 ? 1

Species

n

1 2 1 1 6 4 8 1 12 7 18 12 6 8

Lamellibrachia victori Alaysia spiralis Lamellibrachia columna Lamellibrachia luymesi Escarpia laminata Paraescarpia echinospica Lamellibrachia barhami Seepiophila jonesi Arcovestia ivanovi Escarpia spicata Riftia pachyptila Ridgeia piscesae Oasisia alvinae Tevnia jerichonana

4.80­10.00 6.50­13.00 4.29­7.78 4.00­7.00 2.82­6.67

2.44­4.92 0.96­3.88 1.33­4.00 1.95­3.31 2.00­3.46

Alaysia spiralis Arcovestia ivanovi Escarpia spicata Escarpia laminata Lamellibrachia barhami Lamellibrachia columna Lamellibrachia luymesi Lamellibrachia satsuma Lamellibrachia victori Oasisia alvinae Paraescarpia echinospica Ridgeia piscesae Riftia pachyptila Seepiophila jonesi Tevnia jerichonana

2 1 4 1 9 18 6

33.5 33 30 21 35.44 60.67 25.83

27­40

19­36 9­62 2­103 17­38

been suggested. However, the methods are all dependent on a relatively large sample size and were impractical for the current data set. The strategy chosen to code the characters `obturaculum /vestimentum length' and `vestimentum length /diameter' was to rank the means of the species, find the two largest gaps between adjacent taxa (unless they would result in autapomorphies) and use those to delimit one character state from the next (Tables 2 and 3). The differences among the groups with a particular character state were tested with a single factor ANOVA. In both cases, the differences among the groups were significant (P < 0.0005). Data from Jones (1981) indicate that the ratio between obturacu-

lum and vestimentum length is not correlated with specimen size in Riftia pachyptila. Two other quantitative characters were included; the number of branchial lamellae and the number of opisthosomal segments correlate strongly with specimen size in Riftia pachyptila (P « 0.01 in both cases). However, at least for the number of branchial lamellae, there seems to be a species-specific maximum number. Insufficient data are available to perform analyses similar to obturaculum and vestimentum length. In both cases, character states were assigned intuitively (Tables 4 and 5). The number of opisthosomal segments shows great overlap among the species and the information is missing for many

325

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

species. Among the ingroup taxa, only Alaysia spiralis seems to have a significantly lower number of opisthosomal segments and was assigned a separate character state. To test the impact of the quantitative characters on the results, the analyses were also performed excluding the three characters `number of branchial lamellae (nbl)', `size of obturaculum (sob)' and `size of vestimentum (svt)'.

Analyses For the analysis of vestimentiferan relationships, the following options were applied in all combinations: equal weights/ differential weights and including/excluding quantitative characters. The data sets for relationships among siboglinids and the family level analysis had no quantitative characters. The data set for the family level analysis had no nested characters, therefore differential weights were not implied. The program PAUP* version 4.0b4a (Swofford 2000) for Windows 95 or for MacPPC was used for all tree reconstructions. Searches were performed with the branch and bound method, guaranteed to find all most parsimonious reconstructions (MPRs). The addition sequence was `furthest'. Zero length branches were collapsed and the multrees option was in effect. The consistency index (CI), retention index (RI) and rescaled consistency index (RC) were calculated for all MPRs. Within one analysis, these indices are the same for all MPRs. MPRs were rooted with the outgroups. The decay index (Bremer 1988, 1994), calculated with AUTODECAY, version 4.0 (Eriksson 1998) in association with PAUP*, was used as a measure for branch support. For each of the constraint trees generated by AUTODECAY, 10 random addition searches were performed. PAUP* and MACCLADE version 4.0 (Maddison & Maddison 2000) were used to reconstruct character states. Both options for character state reconstructions available in PAUP*, ACCTRAN and DELTRAN, were invoked and only unambiguous states were considered. MACCLADE 4.0 (Maddison & Maddison 2000) was used to map depth range, habitat type and geographical area on to the tree. Abbreviations used in the figures ar, axial rod; bf, branchial filaments; ch, chaetae; cl, cephalic lobe; cr, crust; dg, dorsal groove; dLf, dorsolateral folds; epa, enlarged papillae; fp, forepart; fr, frenulum; g, girdles; mpa, metameric papillae; obt, obturaculum; op, opisthosoma; ov, obturacular vessel; pco, peristomial collar; sau, saucers; seg, segment; t, tentacle; thm, thoracic membrane; tr, trunk; vcf, ventral ciliated field; ves, vestimentum.

lum, a vestimentum and branchial lamellae) and (2) the monophyly of a clade including Arcovestia ivanovi, Oasisia alvinae, Ridgeia piscesae, Tevnia jerichonana and Riftia pachyptila. The latter, from now on referred to as the vent clade, has the following synapomorphies: paired excretory pores, an enlarged obturaculum and excretory papillae. Three of the four analyses support a basal position of Arcovestia ivanovi in the vent clade. The following points are supported in both analyses with differential weights and in the analysis with equal weights excluding quantitative characters: (1) Lamellibrachia has a basal position in the Vestimentifera clade and appears paraphyletic. (2) One lineage gives rise to both the vent clade and another group formed by the two Escarpia species, Paraescarpia echinospica and Seepiophila jonesi (from now on referred to as the Escarpia complex). The two Escarpia species are sister species, as are Paraescarpia echinospica and Seepiophila jonesi. (3) Arcovestia ivanovi is the sister taxon to a clade formed by Ridgeia piscesae, Tevnia jerichonana, Riftia pachyptila and Oasisia alvinae. Within this clade, the topology varies depending on the inclusion or exclusion of quantitative characters and the weighting scheme (Figs 2 and 3). The analysis with equal weights including quantitative characters resulted in 748 MPRs and therefore a consensus tree with low resolution.

Phylogeny among siboglinid species The analysis of phylogenetic relationships among Vestimentifera, Frenulata and Sclerolinum resulted in one most parsimonious tree of length 29 using equal weights (CI = 0.689, RI = 0.727, RC = 0.502), of length 27 using differential weights (CI = 0.667, RI = 0.705, RC = 0.469) (Fig. 4). The cladogram supports the monophyly of the Frenulata species, and within those, the Thecanephria and Athecanephria as sister groups. Sclerolinum is a sister group to the Frenulata. The Frenulata and the Sclerolinum together form the sister group to the Vestimentifera. The monophyly of the Frenulata species is well supported by three character state changes and the highest decay index in the cladogram. With one exception (association of the excretory system with the ventral blood vessel), character states at the ancestral node of siboglinids could be reconstructed unambiguously. According to this analysis, the `ancestral' siboglinid had a chitinous tube without rings or collars. A frenulum was absent. The excretory system was characterized by a single excretory pore and the absence of a nephrostome. Multiple palps (as tentacles or branchial filaments) were present. A long, unsegmented trunk region was present without any specializations. Both trunk and opisthosome bore bands of uncini in single rows. The digestive tract was degenerated in the adult. Spermatophores were not produced. Sperm had a spirally grooved nucleus and eggs were relatively small.

Results

Phylogeny among vestimentiferan species All analyses (Figs 2 and 3) agree (1) on the monophyly of the Vestimentifera (supported by the presence of an obturacu326

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Fig. 2 A, B. Phylogeny among vestimentiferans. --A. Equal weights, including quantitative characters; strict consensus of 748 most parsimonious trees [length 48, consistency index (CI) = 0.604, retention index (RI) = 0.768, rescaled consistency index (RC) = 0.4642]. --B. Equal weights, excluding quantitative characters; strict consensus of 36 most parsimonious trees (length 37, CI = 0.621, RI = 0.800, RC = 0.497). Numbers at branches: Bremer support index. Character state changes are indicated; for abbreviations see Appendix I.

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

327

Cladistic analysis of Vestimentifera · A. Schulze

Fig. 3 A, B. Topology of vent clade as reconstructed with differential

weights. The remainder of the tree is identical to Fig. 2B. Character state change symbols as in Fig. 2. --A. Including quantitative characters [length 34, consistency index (CI) = 0.618, retention index (RI) = 0.783, rescaled consistency index (RC) = 0.484]. --B. Excluding quantitative characters (length 28.75, CI = 0.626, RI = 0.801, RC = 0.501).

Fig. 4 Single most parsimonious tree from the analysis of siboglinid

Family level analysis The family level analysis resulted in one most parsimonious tree of length 23 (CI = 0.652, RC = 0.345, RI = 0.529) (Fig. 5). The topology is identical to the one suggested by Rouse & Fauchald (1997). The Siboglinidae appear as the sister group to the clade formed by the Sabellariidae, Sabellidae and Serpulidae. The Oweniidae form a basal branch to this clade. Three character states for the most recent common ancestor of the Siboglinidae with the other families of the clade could not be reconstructed unambiguously. These were the presence or absence of nuchal organs, intraepidermal nerve cord and extracellular haemoglobin. Unambiguous character states suggest that the last common ancestor had peristomial palps and a prostomium was fused to the peristomium. The animal lacked chaetal inversion, parapodial branchiae, buccal organ and intravasal body. Its only pair of anterior nephridia was separate from the gonoducts and led to the exterior by a single, dorsal excretory pore. Development was lecithotrophic.

phylogeny; character state changes (right of branches) and Bremer support values (left of branches) are indicated. Character state change symbols as in Fig. 2B. For character abbreviations and descriptions see Appendix III.

and Lamellibrachia columna from the rest of the vestimentiferan species is based on tube morphology (no prominent growth collars in Lamellibrachia luymesi and Lamellibrachia columna). However, this character changes frequently over the phylogeny and may vary intraspecifically depending on habitat characteristics (Southward et al. 1995). The paraphyly/monophyly of Lamellibrachia needs to be re-evaluated using other sources of evidence. Jones (1985), regarding Vestimentifera as a phylum separate from the Frenulata, erected two classes within the group: the Axonobranchia with Riftia pachyptila as its only representative and the Basibranchia which included all other species. This classification is not supported by the present analyses. Even though Riftia pachyptila shows many unique characteristics (Appendix I), there is no indication that the species should be separated from the other species as a distinct class. Comparison with molecular phylogenies. Molecular studies on vestimentiferan phylogeny have used the nuclear 28S rRNA (Williams et al. 1993) and 18S rRNA genes (Halanych et al. 1998, 2001) and the mitochondrial genes for cytochrome c

Discussion

Phylogeny among vestimentiferan species There is no morphological synapomorphy for the genus Lamellibrachia. In fact, it appears as paraphyletic in the present analysis. The differentiation of Lamellibrachia luymesi

328

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Fig. 5 Most parsimonious tree (length 23) from the family level analysis; character state changes (right of branches) and Bremer support values (left of branches) are indicated. For character abbreviations and descriptions see Appendix III.

1 There is strong support for a monophyletic Lamellibrachia in the molecular, but not in the morphological data. Morphological differences among the Lamellibrachia species are too few and in some species too poorly examined to achieve better resolution in the phylogeny at the present stage. 2 The Escarpia complex appears as monophyletic in both analyses (although not strongly supported by morphological data). Escarpia spicata and Escarpia laminata are sister species. 3 Both analyses strongly support the monophyletic status of the species inhabiting the east Pacific vents. The topology within the clade is not fully resolved with morphological data and does not have strong bootstrap support in the molecular analysis, but sister group relationships of Tevnia and Riftia as well as Oasisia and Ridgeia find some support in both analyses. 4 The position of Alaysia differs between the analyses: whereas Alaysia spiralis is the sister group to the clade consisting of the vent clade and the Escarpia complex according to the morphological data, Alaysia sp. appears as the sister species to Arcovestia ivanovi in one molecular phylogeny (Kojima et al. 2000). One possible explanation is that Alaysia sp. and Alaysia spiralis are not actually congeners. It is also possible that the morphological data on Alaysia spiralis are insufficient to determine its position with certainty. Future sampling of Alaysia species and morphological and molecular investigations should enable researchers to determine its phylogenetic affinities with more certainty. Biogeography and evolutionary age. It is impossible to assign a biogeographical origin to the basal node of the vestimentiferans. This is due to low resolution, potential paraphyly and great heterogeneity with regard to depth, habitat and geographical origin in Lamellibrachia. The clade formed by Ridgeia, Riftia, Tevnia and Oasisia shares its most recent common ancestor with a species from the western Pacific (Arcovestia ivanovi from the Manus Basin). Ridgeia, Riftia, Tevnia and Riftia seem to have localized distributions along the northeast Pacific ridges (Ridgeia) or on the east Pacific rise (Tevnia, Oasisia and Riftia). Today, the absence of a ridge pathway between the west and east Pacific prevents the dispersal of species from one location to the other across the Pacific plate. However, the Pacific Kula Ridge, which spanned the north Pacific, might have been a pathway in the Early Tertiary (Tunnicliffe & Fowler 1996). Ridgeia is now the only bare rock-inhabiting vestimentiferan genus in the northeast Pacific, whereas the east Pacific rise vents are inhabited by Riftia, Tevnia and Oasisia. The east Pacific ridge system was continuous until approximately 30 million years ago when its central part was subducted under the North American plate (Tunnicliffe et al. 1996). The separation of the Pacific­Farallon Ridge into the northeast Pacific ridges and the east Pacific rise was probably a major

329

oxidase subunit I (COI) (Black et al. 1997; Kojima et al. 1997, 2001, 2002) and 16S rRNA (Kojima et al. 2000; Halanych et al. 2001). In most molecular analyses, Lamellibrachia appears as a monophyletic clade and is the sister taxon to the remaining species (Williams et al. 1993; Black et al. 1997; Halanych et al. 1998, 2001; Kojima et al. 2000). However, the topology within the second clade differs among the studies. 18S rRNA and 16S rRNA do not resolve the relationships within the Vestimentifera very well; bootstrap values are low (Halanych et al. 1998, 2001; Kojima et al. 2000). Black et al. (1997) found support for the vent clade. However, their analysis did not include Arcovestia ivanovi. The vent clade was not supported by Williams et al. (1993). Furthermore, Williams et al. found strong support for a sister group relationship of Tevnia and Ridgeia, whereas the phylogeny by Black et al. supported Tevnia and Riftia as well as Ridgeia and Oasisia as sister groups. In accordance with the morphological data, none of the molecular analyses supports Jones's (1985) classification into Basibranchia and Axonobranchia. Overall, there is good agreement between the morphological and COI sequence data. The following differences and similarities are most noteworthy:

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

MACCLADE 4.0. Only unambiguous reconstructions are shown; polytomies were regarded as `soft', i.e. unresolved relationships, rather than multiple speciation events. --A. Geographical origin. --B. Habitat type. --C. Depth.

Fig. 6 A­C. Biogeographical data mapped on to the phylogeny with

330

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

vicariance event, separating a once continuous fauna into distinct northern and southern faunas (Tunnicliffe 1988). According to present knowledge, vestimentiferans in the Atlantic are restricted to seep-inhabiting species of the genera Lamellibrachia and Escarpia. Despite increasing sampling efforts, no vestimentiferans have been found at the basaltic vent sites on the northern mid-Atlantic ridge. The low amount of amino acid sequence divergence among extant vestimentiferan species suggests that the Vestimentifera as a group diverged within the last 100 million years (Suzuki et al. 1993; Black et al. 1997; Halanych et al. 1998, 2001). The estimate is consistent with today's distribution patterns and historical pathways by which they may have evolved. If Lamellibrachia is the oldest genus and its distribution today a leftover from an old Tethyan distribution, it indicates an origin of the group in the Late Cretaceous to Early Tertiary. On the other hand, there is a consistent record of Vestimentifera-like tubes since the Silurian in fossil hydrothermal vent assemblages (Little et al. 1998). It is possible that even though the extant species radiated relatively recently, they may be derived from an old lineage that has undergone at least one previous radiation and extinction event.

Phylogeny among Siboglinidae Rouse & Fauchald (1997) predicted that the sister group to the Vestimentifera should be among the Sclerolinum species. In the present analysis, however, Sclerolinum represents the sister group to the Frenulata. The only synapomorphy for the clade formed by the Frenulata and Sclerolinum is the presence of metameric papillae, not present in Vestimentifera. Rouse's (2001) analysis, on the other hand, using a different set of characters, confirms Rouse & Fauchald's (1997) prediction that Sclerolinum is the sister group to the Vestimentifera. He lists five synapomorphies for the Sclerolinum/Vestimentifera clade (absent/present analysis). Three of them are associated with chaetae, one with the excretory system and one with sperm packaging. The character states of the excretory system and sperm packaging, however, are unknown for Sclerolinum. Two of the chaetal characters (`posterior segments with uncini only' and `posterior chaetae form rings') were treated as a single character (`rows of uncini in most segments /absent in opisthosome') in the analysis presented here. The last of Rouse's synapomorphies for Sclerolinum and Vestimentifera (`trunk with uncinal girdles posterior') needs further investigation in Vestimentifera. Rouse interprets the first uncinal row at the posterior end as belonging to the trunk region, because a clear septum between trunk and opisthosome may be absent. However, scanning electron micrographs show a clear demarcation between trunk and opisthosome (e.g. Schulze 2001a; fig. 3B). The trunk of newly settled vestimentiferan juveniles shows claw-like larval chaetae (Southward

1988; Jones & Gardiner 1989) which may be uncini, but they are not arranged in girdles and are later lost. Instead of the first uncinal row, these claw-like chaetae may be homologous to the girdle chaetae in frenulates. The differences in the results between this and Rouse's analysis can thus be traced back to different interpretations of characters. Halanych et al.'s (2001) analysis, based on 16S and 18S rRNA, supports the Sclerolinum/ Vestimentifera association. However, the morphological evidence for this hypothesis is ambiguous. The frenulate groups Thecanephria and Athecanephria, as proposed by Ivanov (1963), were supported in the present analysis. Rouse & Fauchald (1997) scored frenulates for the presence of paired peristomial palps. Irrespective of whether the `palps' are indeed peristomial or not, the present analysis suggests that they were multiple in the ancestral frenulate and that a reduced number of `palps' such as found in Siboglinum is secondary. This is contrary to Rouse's (2001) finding that paired palps are ancestral and the multiple palp condition arose several times independently, once in the Vestimentifera and once in a frenulate clade. Jones (1985) separated Vestimentifera and Frenulata (as Perviata) as phyla. However, they share a number of derived character states (Figs 5 and 6) that strongly suggest phylogenetic links to each other and to other polychaetes of the Sabellida clade. Both Frenulata and Vestimentifera produce a chitinous tube, an elongated trunk region with a trophosome, a reduced digestive tract and sperm with spirally grooved nuclei. Jones's bases for the separation of the two groups were the differences in the anterior body regions, the production of spermatophores in frenulates but not in vestimentiferans and, most importantly, differences in the segmentation of the opisthosome. However, Southward (1975) showed that the ontogeny of segmentation in vestimentiferans and frenulates is very similar. Further developmental similarities became obvious when newly settled juveniles of vestimentiferans were described (Jones & Gardiner 1988, 1989; Southward 1988). Today, neither morphological nor molecular data justify maintaining two separate phyla.

Family level analysis Despite the use of a different data set and the modification of character states, the outcome of the present analysis is identical to the phylogeny presented by Rouse & Fauchald (1997) (Fig. 7). Siboglinidae appear as part of a clade characterized by a prostomium fused to the peristomium, giving support to Rouse & Fauchald's (1997) hypothesis that in Vestimentifera, too, the prostomium is fused to the peristomium. In frenulates, the prostomium may have secondarily become distinct again as the cephalic lobe. The vestimentiferan obturaculum is a paired structure which may be homologous to other paired structures in related families. Ivanov (1989) argued

331

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

Fig. 7 Anterior portions of members of the ingroup families, mapped on to the phylogeny. All drawings modified from Beesley et al. (2000;

figs 1.97A, 1.98A, 1.99A, 1.101A, 3.4 and 3.6A).

that the obturaculum corresponds to the first pair of palps. Other potential homologues may be the opercular lobes of Sabellariids (Fig. 8). These are derived from the parapodia of the first segment whose notochaetae form the operculum proper (Fauchald & Rouse 1997). The elongated trunk region of siboglinids may be derived from a condition similar to the one in oweniids with elongated segments. It is unclear, however, if the trunk is derived from a single, elongated segment or the fusion of a number of segments. Lecithotrophic development is a character used by Bartolomaeus (1997) as a synapomorphy for Siboglinidae and other Sabellidae. All Siboglinidae for which developmental data are available have lecithotrophic larvae, as do all Sabellidae (Rouse & Fitzhugh 1994). However, Oweniidae and Sabellariidae, among the closest relatives of the Siboglinidae, have planktotrophic larvae. Apparently, either planktotrophic or lecithotrophic development arose several times convergently within the Sabellida. Liwanow & Porfirjewa (1967) discuss a number of similarities between frenulates and Owenia fusiformis that they use to indicate a close phylogenetic relationship. The most remarkable similarity is the presence of an intraepidermal nerve cord

332

in both groups, an unusual feature in adult polychaetes and probably a derived condition.

Conclusions

The phylogenetic analyses of vestimentiferan relationships presented here agree well with analyses of molecular sequence data of COI. Once sequence data for Arcovestia ivanovi and Alaysia spiralis are available, a combined analysis of morphological and molecular data would be the next step. Jones's (1985) classification into Basibranchia and Anoxobranchia does not reflect evolutionary relationships. The present analysis supports the monophyly of both Vestimentifera and Frenulata. Morphological evidence for the position of Sclerolinum is ambiguous. The phylogenetic affinities of Siboglinidae with polychaetes of the clade Sabellida contradict Jones's (1985) suggestion of classifying both Frenulata and Vestimentifera as phyla separate from each other and from the polychaetes.

Acknowledgements

This study was part of a dissertation for the University of Victoria (Canada). I thank Verena Tunnicliffe for her advice and constructive criticism during the process of gathering

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

and analysing the data for this article. I am also grateful to Eve Southward (Marine Biological Association of the UK) for her continuous support and encouragement. Kristian Fauchald (Smithsonian Institution) provided helpful comments on an earlier version of this manuscript. The thorough reviews by two anonymous referees greatly improved its quality.

References

Andersen, A. C., Hamraoui, L. & Zaoui, D. (2001). The obturaculum of Riftia pachyptila (Annelida, Vestimentifera): ultrastructure and function of the obturacular muscles and extracellular matrix. Cahiers de Biologie Marine, 42, 219 ­237. Archie, J. W. (1985). Methods for coding variable morphological features for numerical taxonomic analysis. Systematic Zoology, 34, 326­345. Bartolomaeus, T. (1995). Structure and formation of the uncini in Pectinaria koreni, Pectinaria auricoma (Terebellida) and Spirorbis spirorbis (Sabellida): implications for annelid phylogeny and the position of the Pogonophora. Zoomorphology, 115, 161­177. Bartolomaeus, T. (1997). Chaetogenesis in polychaetous Annelida -- significance for annelid systematics and the position of the Pogonophora. Zoology, 100, 348 ­364. Beesley, P. L., Ross, G. J. B. & Glasby, C. J. (Eds) (2000). Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis. Melbourne: CSIRO Publishing. Black, M. B., Halanych, K. M., Maas, P. A. Y., Hoeh, W. R., Hashimoto, J., Desbruyères, D., Lutz, R. A. & Vrijenhoek, R. C. (1997). Molecular systematics of vestimentiferan tubeworms from hydrothermal vents and cold-water seeps. Marine Biology, 130, 141­149. Braunbeck, T. & Dales, R. P. (1984). The role of the heart-body and of the extravasal tissue in disposal of foreign cells in two polychaete annelids. Tissue and Cell, 16, 557­563. Bremer, K. (1988). The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution, 42, 795 ­ 803. Bremer, K. (1994). Branch support and tree stability. Cladistics, 10, 295­304. Cary, S. C., Felbeck, H. & Holland, N. D. (1989). Observations on the reproductive biology of the hydrothermal vent tube worm, Riftia pachyptila. Marine Ecology Progress Series, 52, 89 ­94. Christofferson, M. L. & Araújo-de-Almeida, E. (1994). A phylogenetic framework of the Entoerocoela (Metameria: Coelomata). Revista Nordestina de Biologia, 9, 173­208. Colgan, D. J., Hutchings, P. A. & Brown, S. (2001). Phylogenetic relationships within the Terebellomorpha. Journal of the Marine Biological Association of the UK, 81, 765­773. Dando, P. R., Southward, A. J., Southward, E. C., Dixon, D. R., Crawford, A. & Crawford, M. (1992). Shipwrecked tube worms. Nature, 356, 667. Defretin, R. (1971). The tubes of polychaete annelids. In M. Florkin & E. H. Stolz (Eds) Comprehensive Biochemistry, Extracellular and Supporting Structures, Vol. 26 (pp. 713­747). New York: Elsevier. Eriksson, T. (1998). Autodecay. Stockholm: Department of Botany, Stockholm University. Fauchald, K. & Rouse, G. W. (1997). Polychaete systematics: past and present. Zoologica Scripta, 26, 71­138. Flügel, H. J. & Langhof, I. (1983). A new hermaphroditic pogonophore from the Skagerrak. Sarsia, 68, 131­138.

Franzén, A. (1973). The spermatozoon of Siboglinum (Pogonophora). Acta Zoologica, 54, 179­192. Friedman, M. M. & Weiss, L. (1980). An electron microscopic study of hemoglobin synthesis in the marine annelid, Amphritite ornata (Polychaeta: Terebellidae). Journal of Morphology, 164, 121­138. Gaill, F., Voss-Foucart, M. F., Gerday, C., Compère, P. & Goffinet, G. (1992). Chitin and protein contents in the tubes of vestimentiferans from hydrothermal vents. In C. J. Brine, P. A. Sandford & J. P. Zizakis (Eds) Advances in Chitin and Chitosan (pp. 232­236). New York: Elsevier. Gardiner, S. L. & Jones, M. L. (1993). Vestimentifera. In F. W. Harrison & M. E. Rice (Eds) Microscopic Anatomy of the Invertebrates, Vol. 12. Onychophora, Chilopoda, and Lesser Protostomata (pp. 371­ 460). New York: Wiley-Liss. Gardiner, S. L., McMullin, E. & Fisher, C. R. (2001). Seepiophila jonesi, a new genus and species of vestimentiferan tube worm (Annelida: Pogonophora) from hydrocarbon seep communities in the Gulf of Mexico. Proceedings of the Biological Society of Washington, 114, 694 ­707. Glasby, C. J., Hutchings, P. A., Fauchald, F., Paxton, H., Rouse, G. W., Watson Russell, C. W. & Wilson, R. S. (2000). Class Polychaeta. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 1­296). Melbourne: CSIRO Publishing. Goodrich, E. S. (1945). The study of nephridia and genital ducts since 1895. Quarterly Journal of Microscopical Science, 86, 113­ 392. Halanych, K. M., Feldman, R. A. & Vrijenhoek, R. C. (2001). Molecular evidence that Sclerolinum brattstromi is closely related to vestimentiferans, not frenultae pogonophorans (Siboglinidae, Annelida). Biological Bulletin, 201, 65 ­75. Halanych, K. M., Lutz, R. A. & Vrijenhoek, R. C. (1998). Evolutionary origins and age of vestimentiferan tube worms. Cahiers de Biologie Marine, 39, 355­358. Hawkins, J. A., Hughes, C. E. & Scotland, R. W. (1997). Primary homology assessment, characters and character states. Cladistics, 13, 275 ­283. Hutchings, P. A. (2000a). Familiy Oweniidae. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 173­176). Melbourne: CSIRO Publishing. Hutchings, P. A. (2000b). Family Sabellariidae. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 176 ­180). Melbourne: CSIRO Publishing. Ivanov, A. V. (1963). Pogonophora. London: Academic Press. Ivanov, A. V. (1989). Morphological nature of obturacles in Pogonophora. Dokady Adademii Nauk USSR, 308, 785 [in Russian, English summary]. Ivanov, A. V. (1991). Monilifera -- a new class of Pogonophora. Doklady Akademii Nauk USSR, 319, 505 ­507 [in Russian]. Jones, M. L. (1981). Riftia pachyptila, new genus, new species, the vestimentiferan worm from the Galápagos Rift geothermal vents (Pogonophora). Proceedings of the Biological Society of Washington, 93, 1295­1313. Jones, M. L. (1985). On the Vestimentifera, new phylum: six new species, and other taxa, from hydrothermal vents and elsewhere. Bulletin of the Biological Society of Washington, 6, 117­185.

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

333

Cladistic analysis of Vestimentifera · A. Schulze

Jones, M. L. & Gardiner, S. L. (1988). Evidence for a transient digestive tract in Vestimentifera. Proceedings of the Biological Society of Washington, 101, 423­ 433. Jones, M. L. & Gardiner, S. L. (1989). On the early development of the vestimentiferan tube worm Ridgeia sp. and observations on the nervous system and trophosome of Ridgiea sp. and Riftia pachyptila. Biological Bulletin, 177, 254 ­276. Kojima, S., Ohta, S., Miura, T., Fujiwara, Y. & Hashimoto, J. (2000). Molecular phylogenetic study of chemoautosynthesis-based communities in the Manus Basin. JAMSTEC Journal of Deep Sea Research, 16, 7­13 (in Japanese). Kojima, S., Ohta, S., Yamamoto, T., Miura, T., Fujiwara, Y. & Hashimoto, J. (2001). Molecular taxonomy of vestimentiferans of the western Pacific and their phylogenetic relationship to species of the eastern Pacific. I. Family Lamellibrachiidae. Marine Biology, 139, 211­219. Kojima, S., Ohta, S., Yamamoto, T., Miura, T., Fujiwara, K. & Hashimoto, J. (2002). Molecular taxonomy of vestimentiferans of the western Pacific and their phylogenetic relationship to species of the eastern Pacific. II. Families Escarpiidae and Arcovestiidae. Marine Biology, 141, 57­ 64. Kojima, S., Segawa, R. & Ohta, S. (1997). Molecular phylogeny of vestimentiferans collected around Japan, revealed by the nucleotide sequences of mitochondrial DNA. Marine Biology, 127, 507­513. Little, C. T. S., Herrington, R. J., Maslennikov, V. V. & Zaykov, V. V. (1998). The fossil record of hydrothermal vent communities. In R. A. Mellis & K. Harrison (Eds) Geological Society: Special Publications, Vol. 148. Modern Ocean Floor Processes and the Geological Record (pp. 259­270). London: Geological Society. Liwanow, N. A. & Porfirjewa, N. A. (1967). Die Organisation der Pogonophoren und deren Beziehungen zu den Polychaeten. Biologisches Zentralblatt, 86, 177­204. Maddison, W. P. & Maddison, D. R. (2000). Macclade. Sunderland, MA: Sinauer Associates. Mañé-Garzón, F. & Montero, R. (1986). Sobre una nueve forma de verme tubicola Lamellibrachia victori n. sp. (Vestimentifera), proposition de un nueve phylum: Mesoneurophora. Revista de Biologia del Uruguay, 8, 1­28. McHugh, D. (1993). A comparative study of reproduction and development in the polychaete family Terebellidae. Biological Bulletin, 185, 153­167. McHugh, D. (1997). Molecular evidence that echiurans and pogonophorans are derived annelids. Proceedings of the National Academy of Science, USA, 94, 8006 ­ 8009. Miura, T., Tsukahara, J. & Hashimoto, J. (1997). Lamellibrachia satsuma, a new species of vestimentiferan worms (Annelida: Pogonophora) from a shallow hydrothermal vent in Kagoshima Bay, Japan. Proceedings of the Biological Society of Washington, 110, 447­456. Ouaghi, E. M. & Grasset, M. (1990). Synthèse et exocytose de chlorocruorine et mise en réserve du fer par le tissu périvasculaire d'un annélide polychète. Canadian Journal of Zoology, 69, 2338 ­2348. Pleijel, F. (1995). On character coding for phylogeny reconstruction. Cladistics, 11, 309­315. Potswald, H. (1969). Cytological observations on the so-called neoblasts in the serpulid Spirorbis. Journal of Morphology, 128, 241­260. Rouse, G. W. (1999). Trochophore concepts: ciliary bands and the evolution of larvae in spiralian Metazoa. Biological Journal of the Linnean Society, 66, 411­ 464.

Rouse, G. W. (2000a). Family Serpulidae. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 180 ­184). Melbourne: CSIRO Publishing. Rouse, G. W. (2000b). Morphology and physiology. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 9­32). Melbourne: CSIRO Publishing. Rouse, G. W. (2001). A cladistic analysis of Siboglinidae Caullery, 1914 (Polychaeta, Annelida): formerly the phyla Pogonophora and Vestimentifera. Zoological Journal of the Linnean Society, 132, 55­ 80. Rouse, G. W. & Fauchald, K. (1997). Cladistics and the polychaetes. Zoologica Scripta, 26, 139 ­204. Rouse, G. W. & Fitzhugh, K. (1994). Broadcasting fables: is external fertilization really primitive? Sex, size, and larvae in sabellid polychaetes. Zoologica Scripta, 23, 271­312. Salvini-Plawen, L. V. (2000). What is convergent / homoplastic in Pogonophora. Journal for Zoological Systematics and Evolutionary Research, 38, 133­147. Schulze, A. (2001a). Ultrastructure of opisthosomal chaetae in Vestimentifera (Pogonophora, Obturata) and implications for phylogeny. Acta Zoologica, 82, 127­135. Schulze, A. (2001b). Comparative anatomy of the excretory organs of vestimentiferan tube worms (Pogonophora, Obturata). Journal of Morphology, 205, 1­11. Sibuet, M. & Olu, K. (1998). Biogeography, biodiversity, and fluid dependence of deep-sea cold-seep communities at active and passive margins. Deep-Sea Research, II (45), 517­567. Smirnov, R. (2000). Two new species of Pogonophora from the arctic mud volcano off northwestern Norway. Sarsia, 85, 141­ 150. Sosa, V. & De Luna, E. (1998). Morphometrics and character state recognition for cladistic analyses in the Bletia reflexa complex (Orchidaceae). Plant Systematics and Evolution, 212, 185 ­213. Southward, E. C. (1969). New Pogonophora from the northeast Pacific Ocean. Canadian Journal of Zoology, 47, 395 ­ 403. Southward, E. C. (1972). On some Pogonophora from the Caribbean and the Gulf of Mexico. Bulletin of Marine Science, 22, 739­ 776. Southward, E. C. (1975). A study of the structure of the opisthosoma of Siboglinum fiordicum. Zeitschrift für Zoologische Systematik und Evolutionsforschung, Sonderheft, 64 ­76. Southward, E. C. (1978). Description of a new species of Oligobrachia (Pogonophora) from the North Atlantic, with a survey of the Oligobrachiidae. Journal of the Marine Biological Association of the UK, 58, 357­365. Southward, E. C. (1988). Development of the gut and segmentation of newly settled stages of Ridgeia ( Vestimentifera): implications for the relationship between Vestimentifera and Pogonophora. Journal of the Marine Biological Association of the UK, 68, 465­ 487. Southward, E. C. (1991). Three new species of Pogonophora, including two vestimentiferans, from hydrothermal sites in the Lau Back-arc Basin (Southwest Pacific Ocean). Journal of Natural History, 25, 859­881. Southward, E. C. (1993). Pogonophora. In F. W. Harrison & M. E. Rice (Eds) Microscopic Anatomy of the Invertebrates, Vol. 12. Onychophora, Chilopoda, and Lesser Protostomata (pp. 327­369). New York: Wiley-Liss.

334

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Southward, E. C. (1999). Development of Perviata and Vestimentifera (Pogonophora). Hydrobiologia, 402, 185­ 202. Southward, E. C. (2000). Class Pogonophora. In: P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: the Southern Synthesis (pp. 331­351). Melbourne: CSIRO Publishing. Southward, E. C. & Coates, K. A. (1989). Sperm masses and sperm transfer in a vestimentiferan, Ridgeia piscesae, Jones 1985 (Pogonophora, Obturata). Canadian Journal of Zoology, 67, 2776 ­2781. Southward, E. C. & Galkin, S. V. (1997). A new vestimentiferan (Pogonophora: Obturata) from hydrothermal vent fields in the Manus Back-arc basin (Bismarck Sea, Papua New Guinea, Southwest Pacific Ocean). Journal of Natural History, 31, 43­55. Southward, E. C., Schulze, A. & Tunnicliffe, V. (2002). Vestimentiferans (Pogonophora) in the Pacific and Indian Oceans: a new genus from Lihir Island (Papua New Guinea) and the Java Trench, with the first report of Arcovestia ivanovi from the North Fiji Basin. Journal of Natural History, 36, 1179­1197. Southward, E. C., Tunnicliffe, V. & Black, M. (1995). Revision of the species of Ridgeia from Northeast Pacific hydrothermal vents, with a redescription of Ridgeia piscesae Jones (Pogonophora: Obturata = Vestimentifera). Canadian Journal of Zoology, 73, 282­295. Southward, E. C., Tunnicliffe, V., Black, M., Bixon, D. R. & Dixon, L. R. J. (1996). Ocean-ridge segmentation and vent tubeworms (Vestimentifera) in the NE Pacific. In: C. J. MacLeod, P. A. Tyler & C. L. Walker (Eds) Geological Society: Special Publications, Vol. 118. Tectonic, Magmatic, Hydrothermal and Biological Segmentation of Mid-ocean Ridges (pp. 211­224). London: Geological Society. Storch, V. & Schlötzer-Schrehardt, U. (1988). Sensory structures. In: W. Westheide & C. O. Hermans (Eds) Microfauna Marina, Vol. 4. The Ultrastructure of the Polychaeta (pp. 169­179). New York: Gustav Fisher Verlag. Suzuki, T., Takagi, T. & Ohta, S. (1993). N-terminal amino acid sequences of 440 kDa hemoglobins of the deep-sea tube worms, Lamellibrachia sp. 1, Lamellibrachia sp. 2 and slender vestimentifera gen. sp. 1 evolutionary relationship with annelid hemoglobins. Zoological Science, 10, 141­146. Swofford, D. L. (2000). PAUP*: Phylogenetic Analysis Using Parsimony (and Other Methods). Sunderland: Sinauer Associates. Terwilliger, R. C. & Koppenheffer, T. L. (1973). Coelomic cell hemoglobins of the polychaete annelid, Pista pacifica Berkeley. Comparative Biochemistry and Physiology, 45B, 557­566. Terwilliger, R. C., Terwilliger, N. B., Hughes, G. M., Southward, A. J. & Southward, E. C. (1987). Studies on the haemoglobins of the small Pogonophora. Journal of the Marine Biological Association of the UK, 67, 219­234. Terwilliger, R. C., Terwilliger, N. B. & Schabtach, E. (1980). The structure of hemoglobin from an unusual deep sea worm (Vestimentifera). Comparative Biochemistry and Physiology, 65B, 531­535. Thiele, K. (1993). The holy grail of the perfect character: the cladistic treatment of morphometric data. Cladistics, 9, 275 ­304.

Tunnicliffe, V. (1988). Biogeography and evolution of hydrothermalvent fauna in the eastern Pacific Ocean. Proceedings of the Royal Society of London B, 233, 347­366. Tunnicliffe, V. & Fowler, M. R. (1996). Influence of sea-floor spreading on the global hydrothermal vent fauna. Nature, 379, 531­533. Tunnicliffe, V., Fowler, C. R. & McArthur, A. G. (1996). Plate tectonic history and hot vent biogeography. In: C. J. MacLeod, P. A. Tyler & C. L. Walker (Eds) Geological Society: Special Publications, Vol. 118. Tectonic, Magmatic, Hydrothermal and Biological Segmentation of Mid-ocean Ridges (pp. 225­238). London: Geological Society. Tunnicliffe, V., McArthur, A. G. & McHugh, D. (1998). A biogeographical perspective of the deep-sea hydrothermal vent fauna. Advances in Marine Biology, 34, 353­ 442. Van der Land, J. & Nørrevang, A. (1975). The systematic position of Lamellibrachia (Annelida, Vestimentifera). Zeitschrift für Zoologische Systematik und Evolutionsforschung, Sonderheft, 86 ­101. Van der Land, J. & Nørrevang, A. (1977). Structure and relationships of Lamellibrachia (Annelida, Vestimentifera). Det Kongelige Danske Videnskabernes Selskaps Biologiske Skrifter, 21, 1­102. Van Dover, C. L. (2000). The Ecology of Deep-Sea Hydrothermal Vents. Princeton: Princeton University Press. Webb, M. (1964). A new bitentaculate pogonophoran from Hardangerfjorden, Norway. Sarsia, 15, 49­55. Webb, M. (1969). Lamellibrachia barhami, gen. nov., spec. nov. (Pogonophora) from the Northeast Pacific. Bulletin of Marine Science, 19, 18­47. Weber, R. E., Mangum, C., Steinman, H., Bonaventura, C., Sullivan, B. & Bonaventura, J. (1977). Hemoglobins of two terebellid polychaetes: Enoplobranchus sanguineus and Amphitrite ornata. Comparative Biochemistry and Physiology, 56A, 179­187. Wells, R. G. M., Dales, R. P. & Warren, L. M. (1981). Oxygen equilibrium characteristics of the erythrocruorin (extracellular hemoglobin) from Owenia fusiformis Delle Chiaje (Polychaeta: Oweniidae). Comparative Biochemistry and Physiology, 70A, 11­113. Williams, N. C., Dixon, D. R., Southward, E. C. & Holland, P. W. H. (1993). Molecular evolution and diversification of the vestimentiferan tube worms. Journal of the Marine Biological Association of the UK, 73, 437­452. Wilson, R. S. (2000). Family Spionidae. In P. L. Beesley, G. J. B. Ross & C. J. Glasby (Eds) Fauna of Australia, Vol. 4A. Polychaetes and Allies: The Southern Synthesis (pp. 196­200). Melbourne: CSIRO Publishing. Woodside, J. M. (1997). Chemosynthetic Communities. IOC Technical Series, 48. Neotectonics and Fluid Flow through Seafloor Sediments in the Eastern Mediterranean (pp. 127­128). Paris: UNESCO. Young, C. M., Vázquez, E., Metaxas, A. & Tyler, P. A. (1996). Embryology of vestimentiferan tube worms from deep-sea methane/sulphide seeps. Nature, 381, 514­516. Zelditch, M. L., Fink, W. L. & Swiderski, D. L. (1995). Morphometrics, homology and phylogenetics: quantified characters as synapomorphies. Systematic Biology, 44, 179­189.

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

335

Cladistic analysis of Vestimentifera · A. Schulze

Appendix I

Characters and character states for the analysis of vestimentiferan relationships Collars on tube (col). Tube morphology in Vestimentifera is variable, sometimes within species (Southward et al. 1995). Some species have conspicuous anterior flared openings and more basal collar-shaped rings that represent former openings. Other species, like Riftia pachyptila and Escarpia laminata, have smooth tubes without collars ( Jones 1985). In some frenulate species collars are present, but are usually less prominent than in vestimentiferans. In Sclerolinum, collars are absent (Southward 1972). Character states col: 0 = collars absent, 1 = collars present.

Obturaculum (obt), size of obturaculum (sob). The obturaculum of Vestimentifera is the anterior-most body region which supports the branchial plume. It is formed by two symmetrical halves, each completely covered with cuticle, but fused in the center. The obturaculum is unique to vestimentiferans. Obturaculum tissue mainly consists of connective tissue with collagen fibrils and proteoglycans (Andersen et al. 2001). The mean length of the obturaculum, relative to the vestimentum, varies between 0.22 (Lamellibrachia victori) and 1.57 (Riftia pachyptila) (Table 2). Character states obt: 0 = obturaculum absent, 1 = obturaculum present; sob: 0 = obturaculum short, 1 = obturaculum medium, 2 = obturaculum long. Crust (cru). The anterior face of the obturaculum in many vestimentiferan species is covered by a crust of cuticle, up to 2 mm thick in Escarpia spicata, with a rough apical surface. Character states cru: 0 = crust absent, 1 = crust present. Axial rod (axr). The distal parts of the obturacular halves often enclose a cuticular structure, lenticular in crosssection, that seems to be produced by the medial surfaces of the obturacular halves. Cross-sections show a laminate organization, probably indicating periodic addition of new layers. Character states axr: 0 = axial rod absent, 1 = axial rod present. Saucers (sau). In Oasisia and Ridgeia, the axial rod widens at its anterior end and forms distinct saucer-shaped structures. The outermost layers of the axial rod form the more proximal collars, the inner layers form the most anterior collars. Character states sau: 0 = saucers absent, 1 = saucers present. Spike (spi). In Escarpia spicata, Paraescarpia echinospica and Seepiophila jonesi, the axial rod forms a spike that extends beyond the obturacular halves. Its surface is covered with numerous small spines. Character states spi: 0 = spike absent, 1 = spike present.

336

Dorsal groove in obturaculum/arrangement of obturacular musculature (dgr). Some species are characterized by an ovalshaped cross-section through their obturaculum, others show a distinct dorsal groove in the obturaculum, that gives it a Y or heart shape in cross-section ( Jones 1985) (Fig. 8). Longitudinal muscles are located adjacent to the epithelium in the obturacular halves, in large specimens of Riftia pachyptila up to 500 ( Jones 1981). The number seems to vary greatly with the size of the specimen, but they are most closely spaced in Lamellibrachia and Escarpia. The muscle strands are actually bundles of longitudinal muscle fibres that enclose a central coelomic cavity (Andersen et al. 2001). Jones (1981) describes these muscle strands as rings in parasagittal planes in both obturacular halves of Riftia pachyptila. Jones (1985) uses the orientation of the obturacular ring muscles as a taxonomic character to distinguish between the vestimentiferan families: the Riftiidae, Ridgeiidae and Tevniidae are characterized by parasagittal planes of obturacular ring muscles, whereas the Lamellibrachiidae and Escarpiidae have frontal planes. A re-examination of the base and the apical area of the obturaculum of eight vestimentiferan species using serial cross-sections revealed that there are, indeed, differences in the orientation of the muscles, as seen at the top of the obturaculum (Fig. 8). However, the `rings' never seem to be arranged in a clearly parasagittal plane. Instead they show an oblique orientation. At the base of the obturaculum, the situation is less clear. Towards the base, the muscle strands become less obvious until they can hardly be differentiated from the adjacent tissue. They may be inserted at the bottom of the obturaculum. The lateral muscles are more prominent than the medial ones (Andersen et al. 2001). In the species with a dorsal groove, as shown in Ridgeia piscesae (Fig. 8B), the dorsal extensions of the obturacular halves as seen in cross-section may merely be outgrowths of the dorsal side, derived from a condition such as in Escarpia laminata. The oblique orientation of the `ring muscles' could be interpreted as a slightly contorted version of the situation in Escarpia or Lamellibrachia. Because the arrangement of the musculature is closely linked to the presence of a dorsal groove, it was not included in the analysis as a separate character. Character states dgr: 0 = dorsal groove absent, 1 = dorsal groove present. Branchial lamellae (bla), number of branchial lamellae (nbl). The branchial filaments of vestimentiferans are arranged in a number of semicircles in which the basal parts of the filaments are fused and form branchial lamellae. Even though there is some correlation with size, the available data indicate that in some genera (Alaysia, Arcovestia, Lamellibrachia, Oasisia, Paraescarpia and Seepiophilu) the number of branchial lamellae is generally rather low, i.e. smaller

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Fig. 8 A­F. Cross-sections through the top of the obturaculum of seven vestimentiferan species; the left side shows the arrangement of longitudinal muscles, the right side shows the orientation of `muscle rings', if detectable. --A. Escarpia laminata. --B. Ridgeia piscesae.

--C. Arcovestia ivanovi. --D. Riftia pachyptila. --E. Tevnia jerichonana. --F. Oasisia alvinae.

than 30. Character states bla: 0 = branchial lamellae absent, 1 = branchial lamellae present; nbl: 0 = number of branchial lamellae < 10, 1 = 10­30 branchial lamellae, 2 = > 30 branchial lamellae. Outer sheath lamellae ( lsh). In Alaysia and Lamellibrachia, the outer branchial lamellae consist of filaments that are fused for their complete length and do not bear pinnules ( Jones 1985; Southward 1991). They form a sheath around the inner lamellae consisting of only partly fused filaments with pinnules. Character states lsh: 0 = sheath lamellae absent, 1 = sheath lamellae present. Sensory filaments (sfl), distribution of sensory filaments (dsf). Most of the branchial filaments seem to serve a primarily respiratory function. In most vestimentiferan species, except Alaysia spiralis and the Lamellibrachia species, some filaments seem to be specialized for a sensory function. They lack con-

tinuous rows of multiciliated cells and pinnules. The location of these sensory filaments varies among the species (Gardiner & Jones 1993). Whereas they are randomly distributed in Riftia pachyptila (the character state is an autapomorphy), they are concentrated in the dorsal region of each lamella or in the distal part of the obturaculum in other species. Character states sfl: 0 = sensory filaments absent, 1 = sensory filaments present; dsf: 0 = concentrated dorsally, 1 = concentrated distally, 2 = randomly distributed. Number of excretory pores (exp). It seems that vestimentiferan species with a dorsal groove have two excretory pores, whereas those without have only one (Schulze 2001b). Here, the character has been maintained as separate from the character `dorsal groove' because the number of excretory pores is unknown in Alaysia spiralis and a perfect correlation could thus not be shown (Southward 1991). Frenulates, both Thecanephria and Athecanephria, are characterized by paired

337

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

excretory pores (Southward 1993). Character states exp: 0 = one excretory pore, 1 = two excretory pores. Shape of excretory ducts (exd). Most vestimentiferan species with paired excretory pores have blind-ending excretory ducts (i.e. presence of excretory sacs), whereas all species with single excretory pores have continuous excretory ducts that undergo a sharp bend. An exception to this is Arcovestia ivanovi with paired excretory pores and continuous excretory ducts. The excretory ducts of the frenulates are continuous (Schulze 2001b). Character states exd: 0 = excretory ducts continuous, 1 = excretory ducts sac-like. Excretory grooves (exg ). In Tevnia jerichonana and Ridgeia piscesae conspicuous grooves are found anterior to the excretory pores, formed by the anterior vestimentum. Such excretory grooves are absent in all other species included in the analysis (Schulze 2001b). Character states exg: 0 = excretory grooves absent, 1 = excretory grooves present. Excretory papillae (pap). In all vestimentiferan species with paired excretory pores, the pores lie on more or less conspicuous papillae which are absent in the species with single excretory pores (Schulze 2001b). Distinct excretory papillae seem to be absent in the frenulates and Sclerolinum. Character states pap: 0 = excretory papillae absent, 1 = excretory papillae present. Cuticular lining of distal excretory ducts (cex). In vestimentiferans, the cuticle that overlies the outer epidermis usually extends into the excretory pores and the distal-most part of the excretory pores, except in Ridgeia piscesae (Schulze 2001b) and Lamellibrachia luymesi ( Van der Land & Nørrevang 1977; Jones 1985) where no cuticular lining of the distal excretory ducts could be observed. In Oligobrachia gracilis, a cuticular lining is present, in Siphonobrachia it is absent. The situation in Sclerolinum is unknown. Character states cex: 0 = cuticular lining of distal excretory ducts absent, 1 = cuticular lining of distal excretory ducts present. Number of connections between excretory duct and excretory organ (con). The voluminous excretory sacs are usually connected

to the excretory organ by a single small duct that then branches off to give rise to the excretory tubules. In Riftia pachyptila and Escarpia laminata, several connections have been found (Schulze 2001b). In the frenulates there is only one connection (Ivanov 1963). Character states con: 0 = one connection, 1 = two or more connections. Vestimentum (ves), size of vestiment (svt). One distinguishing characteristic of vestimentiferans is the presence of a muscular region with dorsal folds that apparently serves to hold the position of the animal within the tube. The length of the vestimentum in relation to its diameter varies among the vestimentiferan species. Character states were assigned according to Table 3. Character states ves: 0 = vestimentum absent, 1 = vestimentum present; svt: 0 = vestimentum absent, 1 = vestimentum short, 2 = vestimentum medium, 3 = vestimentum long. Posterior vestimental fold (pvf ). In some species, the posterior vestimental fold is continuous on the ventral side, in other species it shows a gap. Character states pvf: 0 = posterior vestimental fold entire, 1 = posterior vestimental fold divided. Rows of chaetae in opisthosome (cho). The anterior-most opisthosomal segments in Vestimentifera and Sclerolium bear bands of uncini, whereas only a limited number (usually four per segment) of rod-shaped chaetae are present in the frenulate opisthosome. The bands may consist of a single row or multiple rows of uncini. Character states cho: 0 = chaetal rows in opisthosome absent, 1 = chaetal rows in opisthosome present. Total number of opisthosomal segments (ops). Frenulata and Sclerolinum generally have a lower number of opisthosomal segments than Vestimentifera. Among the Vestimentifera, the number of segments varies greatly with the size of the specimen. Riftia pachyptila clearly has the largest number of segments, but the within-group variation is too high and not enough data are available to demarcate character states among the species. Character states ops: 0 = < 10 opisthosomal segments, 1 = > 10 opisthosomal segments.

338

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Appendix II

Species level data set, C-coding. 0­2, character states (several states for one taxon indicates polymorphism); ?, unknown state; x, inapplicable. The weights used in the differential weights analysis are indicated.

col 1 1 1 0 0 1 0 0 1 1 1 1 0,1 0 1 1 0 0 1 obt 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 sob 0.5 0 2 0 0 0 0 0 0 0 1 0 2 2 0 1 x x x cru 0.5 0 0 1 1 0 0 0 0 0 0 1 0 0 1 1 x x x axr 0.5 0 0 1 1 0 0 0 0 0 1 1 1 0 1 1 x x x sau spi dgr 0.25 0.25 0.5 x x 0 0 x x x x x 1 0 1 x 0 0 x x x x x 0 1 x x x x x 0 1 0 x 0 0 x x x 1 1 0 0 0 0 0 0 0 1 0 1 1 0 1 x x x bla 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 nbl 0.5 0 1 2 2 1 1 1 1 1 1 1 2 2 1 2 x x x sfl 1 0 1 1 1 0 0 0 0 0 1 1 1 1 1 1 0 0 0 dsf 0.5 x 0 1 1 x x x x x 0 1 0 2 1 0 x x x lsh 0.5 1 0 0 0 1 1 1 1 1 0 0 0 0 0 0 x x x exp 1 ? 1 0 0 0 0 0 0 0 1 0 1 1 0 1 ? 1 1 exd 1 ? 0 0 ? 0 ? ? ? ? ? 0 1 1 0 1 ? 0 0 exg 1 ? 0 ? 0 0 0 0 0 0 0 0 0 1 0 1 ? 0 0 pap 1 ? 1 0 0 ? ? ? ? ? 1 0 1 1 0 1 ? 0 0 cex 1 ? 1 1 1 1 ? 0 ? ? ? 1 0 1 1 1 ? 0 1 con 1 ? 0 1 ? 0 ? 0 ? ? ? 0 0 1 0 0 ? 0 0 ves 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 svt 0.5 0 1 0 2 0 0 0 0 0 2 0 2 2 1 2 x x x pvf 0.5 0 0 0 0 1 1 1 1 1 0 1 0 0 1 0 x x x cho 1 1 1 1 ? ? 1 ? 1 ? 1 1 1 1 ? 1 1 0 0 ops 1 0 1 ? ? ? 1 ? 1 ? 1 1 1 1 ? 1 0 ? 1

Weights

Alaysia spiralis Arcovestia ivanovi Escarpia laminata Escarpia spicata Lamellibrachia barhami Lamellibrachia columna Lamellibrachia luymesi Lamellibrachia satsuma Lamellibrachia victori Oasisia alvinae Paraescarpia echinospica Ridgeia piscesae Riftia pachyptila Seepiophila jonesi Tevnia jerichonana Sclerolinum minor Oligobrachia gracilis Siphonobrachia lauensis

Appendix III

Characters for the analysis of siboglinid species, data set and character states Chitinous tube (chi). The chitin content in obturate tubes varies between 24% in Riftia pachyptila and 45% in Tevnia jerichonana (Gaill et al. 1992). Chitin has been detected in the tubes of various frenulates (Southward 1993). In Siboglinum sp. it represents approximately 33% of the dry tube mass. There are no published analyses of the tube of Sclerolinum. The tubes of Sabella pavonina and Owenia fusiformis are built of mucus and embedded sand grains. Analysis of the carbohydrate portion in the mucus of Sabella pavonina did not reveal the presence of chitin (Defretin 1971). Character states chi: 0 = chitinous tube absent, 1 = chitinous tube present.

(Rouse & Fauchald 1997). The number of anterior appendages varies among siboglinids and the outgroups. Character states app: 0 = paired appendages, 1 = multiple appendages. Forepart with frenulum (fre). The forepart is a cylindrical body region at the anterior end of frenulates that bears the cephalic lobe and the branchial filaments. Approximately midway on the forepart is a pair of prominent cuticular ridges that run diagonally around it. No corresponding structure is found in Vestimentifera. Sclerolinidae show little differentiation between the forepart and the trunk and instead of a bridle there are rows or patches of papillary plaques (Webb 1964; Southward 1972). Character states fre: 0 = frenulum absent, 1 = frenulum present. Transverse grooves in anterior forepart (gro). In some frenulate species, the anterior portion of the forepart bears one or more transverse grooves behind the tentacle (Southward 1969, 1991). Character states gro: 0 = transverse groove in `peristomium' absent, 1 = transverse groove present. Dorsal furrow (dfu). The anterior portion of the frenulate forepart often bears a longitudinal groove that usually ends at the level of the bridle (Ivanov 1963; Southward 1969, 1991). Character states dfu: 0 = dorsal furrow in peristomium present, 1 = dorsal furrow absent. Trunk (tru), regionation in trunk (rtr). Whereas the trunk region of frenulates can clearly be divided into different

339

Rings (rin). The tubes of frenulates are usually of a brownish colour and often characterized by a pattern of light and dark rings (Ivanov 1963; Southward 1993, 2000). Character states rin: 0 = rings on tubes absent, 1 = rings present. Collars (col). see Appendix I for a description and character states. Number of appendages (app). All polychaete palps (including Vestimentifera and Frenulata), irrespective of whether they are of prostomial or peristomial origin, are considered homologous because of their similar patterns of innervation

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

regions, the trunks of Monilifera and Vestimentifera remain without differentiations. Monilifera show an increased number of papillae in the anterior trunk, but no division into pre- and postannular regions or regions with enlarged papillae. Character states tru: 0 = elongated trunk absent, 1 = trunk present; rtr: 0 = no differentiation of different trunk regions, 1 = differentiated trunk regions. Metameric papillae (mpa), cuticular plaques on metameric papillae (cut). Two rows of pyriform glands on slightly elevated ridges are present on the dorsal side of the anterior trunk in Sclerolinum. The presence of a ciliated field on the ventral side suggests that these ridges are homologous to the metameric papillae in frenulates (Southward 1972, 2000). In some species, the metameric papillae are topped with cuticular plaques. Character states mpa: 0 = metameric papillae absent, 1 = metameric papillae present; cut: 0 = cuticular plaques absent, 1 = cuticular plaques present. Enlarged papillae (epa). Some frenulate species show, just anterior to the girdle region, a number of papillae that are larger than any other papillae on the animal's body (Southward 1969, 1978). They are accompanied by a ventral ciliated field, distant from the ventral ciliated field associated with the metameric papillae. Character states epa: 0 = enlarged papillae absent, 1 = enlarged papillae present. Rows of uncini in different body regions (unc), number of uncinal rows per segment (nun). All species included in the initial analysis are characterized by the presence of rows of uncini. These are found in different body regions. In Oweniidae and Sabellidae, they are present in all segments, except in the first one which does not have neuropodia and in the very last ones that are just being formed. In frenulates, these are developed in the trunk region and reduced in the opisthosome. In Sclerolinidae, rows of uncini are found in both the opisthosome and the posterior trunk. In Vestimentifera, they are absent in the trunk, but present in the opisthosome. Single rows of uncini are present in each opisthosomal segment in the ground pattern of the Vestimentifera (see within-group analysis). Among the frenulates, in Oligobrachia gracilis, Polybrachia canadensis and Spirobrachia beklemishevi, each girdle is formed by multiple rows of uncini. Multiple rows of uncini are also present in each segment of Owenia fusiformis. Character states unc: 0 = rows of uncini in most segments, 1 = absent in opisthosome; nun: 0 = one to two rows of uncini in one band, 1 = multiple rows of uncini in one band. Excretory pores (exp). Frenulates have paired excretory pores, as does Owenia fusiformis. Sabella pavonina has a single dorsal

excretory pore. The number in vestimentiferans varies (Schulze 2001b). Character states: see Appendix I. Nephrostome (nep). Among the frenulates, Siphonobrachia lauensis is the only species in which a nephrostome has unambiguously been shown (Southward 1993). No nephrostome could be detected in Vestimentifera, but it is present in both outgroup species. No information is available on the excretory system of Sclerolinum. Character states nep: 0 = nephrostome absent, 1 = nephrostome present. Complete digestive tract in adult (dig). A functional digestive tract (including mouth and anus) is absent in all adult Siboglinidae. Character states dig: 0 = reduced digestive tract in adult, 1 = functional digestive tract. Spermatophores (spe). Spermatophores, defined as `bundles of sperm enclosed by a sheath or capsule that isolates them from the surrounding environment' (Rouse 1999: 216), are produced by all frenulate species known to date, except Siboglinum poseidoni (Flügel & Langhof 1983; Southward 2000). The situation in Sclerolinum is ambiguous: sperm usually seem to be free in the male gonoducts, but in one species flat spermatophores may have been present (Southward 2000). Vestimentifera do not produce spermatophores, but release sperm in clusters or sticky masses (Cary et al. 1989; Southward & Coates 1989). No spermatophores are known in Sabellidae or Oweniidae. Character states spe: 0 = spermatophores absent, 1 = spermatophores present. Spirally grooved nucleus in sperm (sgn). Sperm of both frenulates and vestimentiferans are characterized by a spirally grooved nucleus (Franzén 1973; Southward & Coates 1989). Sclerolinid sperm have not been examined ultrastructurally. Sabellid and Oweniid sperm do not have spirally grooved nuclei. Character states sgn: 0 = spirally grooved nucleus in sperm present, 1 = spirally grooved nucleus absent. Large, yolky eggs (egg). Vestimentiferan eggs are spherical and, with a diameter of 80­100 µm, relatively small (Southward 2000). In frenulates, egg size and shape vary greatly among the species: Siboglinum eggs are usually elongate and yolk-rich (size up to 650 × 130 µm in Siboglinum caulleryi) (Southward 1993). Oligobrachia gracilis eggs are nearly 500 µm in diameter (Southward 1978). There are no developmental data on Polybrachia, Spirobrachia or Siphonobrachia, but eggs are small and probably develop pelagically (Southward 2000). Sclerolinum produces large eggs (Southward 2000). Character states egg: 0 = large yolky eggs absent, 1 = large yolky eggs present.

340

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

A. Schulze · Cladistic analysis of Vestimentifera

Appendix IV

Data set for phylogeny among siboglinids; 0­1, character states; x, inapplicable; ?, unknown character state. The weights used in the differential weights analysis are indicated.

chi 1 1 1 1 1 1 1 1 0 0 col 1 0 0 0 1 1 0 0 0 0 rin 1 0 1 1 0 1 0 0 0 0 app 1 1 1 0 1 1 1 0 1 0 fre 1 0 1 1 1 1 1 0 0 0 gro 0.5 x 0 0 1 1 0 0 x x dfu 0.5 x 0 0 1 1 1 1 x x tru 1 1 1 1 1 1 1 1 0 0 rtr 0.5 0 1 1 1 1 1 0 x x mpa 1 0 1 1 1 1 1 1 0 0 cut 0.5 x 0 0 1 1 1 0 x x epa 1 0 1 0 0 1 0 0 0 0 unc 1 0 1 1 1 1 1 0 0 0 nun 1 0 1 0 1 1 1 0 0 1 exp 1 0 1 1 1 1 1 ? 0 1 nep 1 0 0 ? 1 ? ? ? 1 0 dig 1 0 0 0 0 0 0 0 1 1 spe 1 0 1 1 1 1 1 ? 0 0 sgn 1 1 1 1 1 1 1 ? 0 0 egg 1 0 1 1 0 0 0 1 ? 0

Weights Vestimentifera Oligobrachia gracilis Siboglinum ekmani Siphonobrachia lauensis Polybrachia canadensis Spirobrachia beklemishevi Sclerolinum minor Sabella pavonina Owenia fusiformis

Appendix V

Characters, character states and data set for the family level analysis Prostomium ( pro). In all ingroup taxa (except possibly the Siboglinidae), the prostomium is completely fused to the peristomium, whereas it is distinct in Terebellidae and Spionidae (Rouse & Fauchald 1997). Character states pro: 0 = prostomium distinct, 1 = prostomium completely fused to peristomium.

chaetes. Of the taxa considered here, no nuchal organs were found in Siboglinidae and Oweniidae. In the other families, nuchal organs are ciliated pits or grooves at the posterior margin of the prostomium (Storch & Schlötzer-Schrehardt 1988; Rouse & Fauchald 1997). Character states nuc: 0 = nuchal organs absent, 1 = nuchal organs present. Chaetal inversion (inv). In Sabellariidae, Sabellidae and Serpulidae, abdominal uncini are located on the notopodia rather than on the neuropodia. Character states inv: 0 = chaetal inversion absent, 1 = chaetal inversion present. Nephridia (nph). In Sabellidae, Serpulidae, Sabellariidae and Siboglinidae, there is only one pair of anterior excretory organs that is completely separate from the posterior gonoducts. Oweniids have one or two pairs of nephridia that also probably serve to release the gametes (Goodrich 1945). Both spionids and terebellids have multiple nephridia. Character states nph: 0 = more than one pair of nephridia, 1 = one anterior pair of nephridia. Excretory pores (exp). Single dorsal excretory pores are present in the ground pattern of the Siboglinidae, in the Serpulidae, Sabellidae and Sabellariidae. Excretory pores are paired and situated more ventrolaterally in Terebellidae, Spionidae and Oweniidae (Goodrich 1945; Bartolomaeus 1997). Character states: see Appendix I. Intraepidermal nerve cord (ien). In both Frenulata and Vestimentifera the nervous system is completely intraepidermal (Gardiner & Jones 1993; Southward 1993), whereas in most other polychaetes it is separate from the epidermis in adults. Oweniidae exceptionally have an intraepidermal nerve cord like Siboglinidae (Hutchings 2000a). Character states ien: 0 = intraepidermal nerve cord absent in adult, 1 = intraepidermal nerve cord present in adult.

341

Peristomium ( per). In Sabellidae, Serpulidae, Oweniidae and Spionidae the peristomium is distinct, while it is reduced to lips in Terebellidae and Sabellariidae (Rouse & Fauchald 1997). Character states per: 0 = peristomium distinct, 1 = peristomium limited to lips. Peristomial grooved palps ( pal). Fauchald & Rouse (1997) and Rouse & Fauchald (1997) assume that vestimentiferan branchial filaments and frenulate tentacles are peristomial grooved palps. In frenulates, the tentacles originate from the forepart (Southward 1993), in vestimentiferans from the anterior vestimentum (Gardiner & Jones 1993). In this study, Siboglinidae were scored with a `?' for the character in order to take a conservative approach. Even though homology with palps of other polychaetes is assumed, their anatomical origin cannot be homologized with certainty with the body divisions in the related polychaete families. Character states pal: 0 = paired palps, 1 = multiple palps. Parapodial branchiae ( pbr). The notopodia of Sabellariidae and Spionidae bear dorsal, well-vascularized appendages that serve for gas exchange. Parapodial branchiae are absent in all other taxa included. Character states pbr: 0 = parapodial branchiae absent, 1 = parapodial branchiae present. Nuchal organs (nuc). Nuchal organs are chemosensory structures that display a variety of morphological forms in poly-

© The Norwegian Academy of Science and Letters · Zoologica Scripta, 32, 4, July 2003, pp321­ 342

Cladistic analysis of Vestimentifera · A. Schulze

Appendix VI

Data set for the family level analysis; last four characters are larval characters.

pro 1 ? 1 1 1 1 0 0 per 1 ? 1 0 0 0 1 0 pal 1 ? 0 1 1 1 1 0 pbr 1 0 1 0 0 0 0 1 nuc 1 0 1 1 1 0 1 1 inv 1 0 1 1 1 0 0 0 nph 1 1 1 1 1 0 0 0 exp 1 1 1 1 1 0 0 0 ien 1 1 0 0 0 1 0 0 hem 1 1 1 0 0 1 1 ? buc 1 0 0 0 0 1 1 ? ivb 1 1 ? 0 1 0 1 0 lec 1 0 0 1 0,1 0 1 0,1 met 1 0 0 1 1 1 0 0 obf 0.5 0 0 0 1 1 0 0 cfg 1 0 0 ? 1 1 0 0 tel 1 1 1 0 0 0 1 1

Weights Siboglinidae Sabellariidae Sabellidae Serpulidae Oweniidae Terebellidae Spionidae

Intravasal body (ivb). Among the taxa included, an intravasal body is present in Terebellidae, Serpulidae and Siboglinidae. It seems to be absent in Sabellidae, Oweniidae and Spionidae. The situation in Sabellariidae is unknown. Character states ivb: 0 = intravasal body absent, 1 = intravasal body present. Extracellular haemoglobin ( hem). Two types of blood pigment are found in polychaetes: red haemoglobin and green chlorocruorin (Rouse 2000b). Both usually occur dissolved in the blood, but have in some cases also been found intracellularly in addition to the dissolved form (e.g. Terebellidae; Weber et al. 1977). Of the taxa studied, extracellular haemoglobin (= erythrocruorin) occurs in Siboglinidae (Terwilliger et al. 1980, 1987), in Sabellariidae (Hutchings 2000b), Terebellidae (Terwilliger & Koppenheffer 1973; Weber et al. 1977; Friedman & Weiss 1980; Braunbeck & Dales 1984) and Oweniidae (Wells et al. 1981), whereas chlorocruorin is found in Sabellidae (Ouaghi & Grasset 1990) and Serpulidae (Potswald 1969). Character states hem: 0 = extracellular haemoglobin absent, 1 = extracellular haemoglobin present.

Buccal organ (buc). The buccal organ is the anterior part of the alimentary canal that is derived from the larval stomodaeum. It is characterized by complex folds, musculature and glands (Glasby et al. 2000). Among the taxa considered here, it is present in Oweniidae and Terebellidae. The situation in Spionidae is unknown. It is absent in all other taxa included. Character states buc: 0 = buccal organ absent, 1 = buccal organ present. Lecithotrophic development (lec). All Siboglinidae of which development has been studied to date have lecithotrophic larvae. Juveniles develop a transient digestive tract only at the time of settlement (Young et al. 1996; Southward 1999). Larvae of Sabellidae (Rouse & Fitzhugh 1994) and Terebellidae (McHugh 1993) are also exclusively lecithotrophic. In serpulids, both lecithotrophic and planktotrophic larvae occur (Rouse 2000a) and it is not clear what the ancestral mode is. The same is true for Spionidae (Wilson 2000). Oweniidae (Hutchings 2000a) and Sabellariidae (Hutchings 2000b) produce free-swimming planktotrophic larvae. Character states lec: 0 = lecitotrophic larvae absent, 1 = lecitotrophic larvae present.

342

Zoologica Scripta, 32, 4, July 2003, pp321­ 342 · © The Norwegian Academy of Science and Letters

Information

22 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

1056985


You might also be interested in

BETA