CLADISTIC ANALYSIS OF METAZOAN RELATIONSHIPS: A REAPPRAISAL

Cladistics (1993) 9:167-181

CLADISTIC ANALYSIS OF METAZOAN RELATIONSHIPS: A REAPPRAISAL

Thierry Backeljau1, Birgitta Winnepenninckx2 and Luc De Bruyn3

' Royal Belgian Institute of Natural Sciences (KBIN), Malacology Section, Vautierstraat 29, B-1040 Brussels, Belgium,

2 University of Antwerp (UIA), Department of Biochemistry, Universiteitsplein 1, B-2610 Wilrijk, Belgium and

3 University of Antwerp (RUCA), Department of Biology, Groenenborgerlaan 171, B-2020 Antwerp, Belgium

Received for publication 3 September 1992; accepted 8 March 1993

Abstract—A recently published cladistic analysis of metazoan relationships based on 77 embryo-logical and morphological characters is shown to be defective with respect to both methodological issues and the interpretation of certain characters. Consequently, many conclusions of this analysis are not supported by the data. Alternative phylogenetic hypotheses are therefore proposed, based on a re-evaluation of an adapted data set.

Introduction

Recently, Schram (1991) proposed a cladogram of metazoan relationships based on a cladistic analysis of anatomical and larval features. From this cladogram (Fig. 1) he concluded that: (1) the Metazoa are monophyletic; (2) Cnidaria and Ctenophora are not sister-groups, but the latter is a sister-group to the Bilateria; (3) the Pseudocoelomata is a sister-group to Acoelomata + Eucoelomata; (4) the Acoelomata is a sister-group to the Eucoelomata; (5) the Eucoelomata consists of two lineages, which are traditionally referred to as Protostomia and Deuterostomia (including the Lophophorata); (6) the Lophophorata (=Phoronida + Ectoprocta + Brachiopoda) and Hemichordata (=Pterobranchia + Enteropneusta) are para-phyletic groups. Schram's (1991) cladogram also suggested that the Entoprocta are deuterostomes and/or lophophorates, the Nemertea is a sister-group to the protostomes and the Chaetognatha are pseudocoelomates.

Schram (1991) remarked that part of his results are conflicting with traditional viewpoints and stressed that "the above postulated relationships are dependent on the particular data matrix used". We will demonstrate, however, that many of these conclusions are not supported by Schram's (1991) data, but rather are artifacts of his analyses.

Schram's Cladistic Analysis

The 77 characters listed in Appendices 1 and 2 were analyzed by Schram (1991) using Farris's Physis Wagner.S program. The resulting tree was rooted to an ancestor scored plesiomorphic for all characters. The same data were reanalyzed with Swofford's PAUP (version not mentioned) and summarized in a consensus tree based on 100 runs. Evaluating this tree is difficult for Schram (1991) did not mention which algorithms and options were used, how many trees were produced, what kind of consensus tree was generated and how the tree in Fig. 1 was selected.

0748-3007/93/020167+15$08.00/0 © 1993 The Willi Hennig Society

168 T. BACKELJAUETAL.

ONYCHOPHORA UNIRAMIA CHEUCERIFORMES TARDIGRADA CRUSTACEA PENTASTOMIDA POGONOPHORA ANNELIDA SIPUNCULA ECHIURA MOLLUSCA NEMERTEA ENTOPROCTA ECTOPROCTA PHORONIDA BRACHIOPODA PTEROBRANCHIA ECHINODERMATA ENTEROPNEUSTA UROCHORDATA CEPHALOCHORDATA GNATHOSTOMULIDA PLATYHELMINTHES CHAETOGNATHA NEMATODA NEMATOMORPHA ACANTHOCEPHALA ROTIFERA GASTROTRICHA PRIAPULIDA LORICIFERA KINORHYNCHA CTENOPHORA CNIDARIA PORIFERA PLACOZOA MESOZOA

Fig. 1. Original cladogram of invertebrate relationships presented by Schram (1991) (L= 140; CI = 0.55).

Anyhow, both PAUP and Physis Wagner.S gave comparable results and yielded a most parsimonious (MP) tree of length (L) 140 and consistency index (CI) 0.55 (Fig. 1).

Empirical comparisons of phylogeny inference packages suggested that Physis and older versions of PAUP may be less reliable in finding MP trees than Farris's (1989) Hennig86 1.5 (Platnick, 1987, 1989; Sanderson, 1990). Therefore, we reanalyzed Schram's (1991) data (treated as unordered binary data) with the heuristic algorithms mhennig* + bb of Hennig86. These generated 100 different MP trees (L = 139; CI = 0.55) and yielded the strict consensus tree of Fig. 2(A) [the nelsen command in Hennig86 derives a strict consensus tree instead of a Nelson consensus tree (Anderberg and Tehler, 1990); the difference between both is discussed by Page (1989), while the utility of strict consensus trees is advocated by Anderberg and Tehler (1990)].

The latter analysis is supposed to be "equivalent" to that of Schram (1991), even though our trees are one step shorter than his. It demonstrates that his data do not provide evidence to conclude that: (1) the Pseudocoelomata is a sister-group to the Acoelomata + Eucoelomata; (2) the Acoelomata is a sister-group to the Eucoelomata; (3) the Eucoelomata consists of only two clades; (4) the Entoprocta are deuterostomes; and (5) the Nemertea is a sister-group to the protostomes. On the other hand, it confirms Schram's (1991) conclusion as to: (1) the monophyly

CLADISTIC ANALYSIS OF METAZOA 169

ij= Ancestor t=p= Mesozoa D'— Placozoa j= Porifera ■—J ij= Cnidaria IM ij= Ctenophora N=4 rj= Nemertea

U= Ancestor \=U= Mesozoa

\= Placozoa

nemertea Entoprocta j= Gnathostomul != Platyhelmint]

N= Placozo ij= Porifi

M rj= Cnidaria IM rj= Ctenophora

^=\= Nemertei nathostomulid Platyhelminthe; Phoronida

ida is

i j = Phoronida = U= Ectoprocta t= Brachiopod; U= Echinodermata U= Enteropneusta t= Pterobranchia t= Urochordata i= Cephalochordata j= Mollusca = [= Sipuncula t= Echiura t= Annelida U= Pogonophora t= Pentastomida != Pentastomida j= Tardigrada U= Onvr -hnn^o ra

iformes

tf j= Tardigradi --U= Onychopho: *r= Uniramia k= Cheliceri: l!= Crustacea

ij= Nematomorphi S=k= Nematoda

Ltenopnor Nemertea Entoprocta if= Gnathostomulida

=5= Platyhelminthes rr= Phoronida

= p= Ectoprocta t= Brachiopoda t= Echinodermata t= Enteropneusta p= Pterobranchia k= Urochordata i= Cephalochordata j= Mollusca

= *F= Sipuncula }= Echiura k= Annelida t= Pogonophora t= Pentastomida if= Tardigrada

*=$= Onychophora f= Uniramia r= Cheliceriformes <= Crustacea

ij= Nematomorpha i=t= Nematoda t= Chaetognatha n= Gastrotricha

M i?= Rotifera

B Fig. 2. Strict consensus trees obtained with Schram's (1991) original data. (A) Based on 100 MP trees

produced by mhennig* + bb (L = 139; CI = 0-55). (B) Based on 741 MP trees produced by mhennig* + bb* before (L = 139; CI = 0.55) and after eliminating 14 autapomorphies (L = 125; CI = 0.50).

of the Metazoa; (2) the affinities of Cnidaria, Ctenophora and Bilateria; (3) the paraphyly of both the Lophophorata and Hemichordata; and (4) the pseudocoelo-mate relationships of the Chaetognatha.

By not specifying the * option for the bb algorithm, only 100 MP trees are saved. Therefore, we performed a second run of the data using the combination mhennig* + bb*, which saves all the topologically different MP trees encountered. In this way, we obtained 741 MP trees (L = 139; CI = 0.55) and the strict consensus tree of Fig. 2(B). It shows that one of Schram's (1991) conclusions confirmed by the previous analysis (Ctenophora is a sister-group to the Bilateria) is no longer supported.

We also implemented the exact ie and the heuristic tread + bb* algorithms of Hennig86. The ie algorithm took too much computer time, since even with the — specification, the analysis could not be completed after a 5-day run on a 386DX 33 MHz microcomputer. The tread + bb* combination (Platnick, 1989), on the contrary, yielded exactly the same results as mhennig* + bb*. These analyses took about 10-20 min. Hereafter, all Hennig86 runs will refer to the mhennig* + bb* combination followed by the nelsen command.

Finally, we briefly assessed possible effects of the data input order. Our "default" order was that of Schram (1991) (Appendix 2). It reflects traditional groupings and patterns in metazoan phylogeny. We disrupted this order as illustrated in

170 T. BACKELJAU ET AL.

Appendix 3 and reanalyzed the data. Each input order yielded different numbers of MP trees (Appendix 3). Nevertheless, the strict consensus trees were all the same [Fig. 2(B)] (seven other input orders yielded no additional information). These results show that at least the numbers of distinct MP trees found by Hennig86 depend on the input order of the data. We expect that for certain data sets this may affect (consensus) tree topologies. Similar and more fundamental observations of this kind were recendy made with PAUP (Barinaga, 1991; Blair Hedges et al., 1991; Templeton, 1991). Whether Schram (1991), who used PAUP, took possible input order effects into account, cannot be deduced from his paper.

Character Interpretation

Although Schram (1991) analyzed 77 characters, only 63 are phylogenetically informative for the following 14 are autapomorphic: 3 = nematogen/infusorigen stages (Mesozoa); 9 = water channels and choanocytes (Porifera); 10 = embryonic inversion (Porifera); 15 = nematocysts (Cnidaria); 17 = colloblasts (Ctenophora); 29 = renette cells, aphids and phasmids (Nematoda); 35 = by cuticula ensheathed cilia (Gastrotricha); 36 = uterus bell (Acanthocephala); 38 = cuticula with several unit membranes (Rotifera); 59 = calcareous endoskeleton (Echinodermata); 60 = water-vascular system (Echinodermata); 65 = rhynchocoel (Nemertea); 68 = apical growth of larvae (Pogonophora); and 70 = anal vesicles (Echiura).

Autapomorphies will not alter the topology of a tree, yet they artificially increase tree lengths and CI values [here L = 139 and CI = 0.55; Fig. 2(B)]. So, after excluding the autapomorphies we obtained the same trees, but now with L = 125 and CI = 0.50.

We also checked the character state distribution proposed by Schram (1991) and found at least seven characters with debatable or erroneous character states (see Appendix 2). We used Barnes (1974), Brusca and Brusca (1990) and Willmer (1990) to adapt them as follows (character numbers in parentheses). Retrocerebral organ (27): is an autapomorphy of Rotifera (Lorenzen, 1985). Eutely (28): Priapulida are not eutelic (Shapeero, 1961). Ecdysis (30): Pentastomida do moult (Osche, 1963). Lorica (40): is present in Rotifera (Clement, 1985). Lophophore (45): Pterobranchia and Entoprocta do not have a true lophophore since their "lophophore" either includes the anus (Entoprocta) or does not completely surround the mouth (Pterobranchia) [but see Gilmour (1979) and Lester (1985)]. Enterocoelic coelom (52): Pogonophora and Chaetognatha are supposed to be enterocoelic (Godeaux, 1974; Ivanov, 1975) [but see also Bakke (1980)]. Hemocoel (67): Pentastomida and Mollusca (except Cephalopoda) have a hemocoel (e.g. Lemche and Wingstrand, 1959; Wingstrand, 1985).

The adapted data were subjected to Hennig86. The ambiguous interpretations of enterocoely and hemocoely were dealt with by evaluating three data sets: one with all characters adapted except for 52 in Pogonophora and 67 in Mollusca, one in which character 52 was adapted for the Pogonophora too and, finally, one in which all characters were adapted. Each analysis yielded more than 2444 different MP trees (respectively L = 125, L = 126 and L = 127 and CI = 0.49 or 0.48), but produced the same strict consensus tree (Fig. 3). This shows that the adapted data: (1) confirm the unresolved position of Ctenophora; (2) do not provide evidence for the monophyly of both deuterostomes and pseudocoelomates; (3) unite protostomes with Acoelomates + Nemertea + Entoprocta; (4) place Pentastomida in a mono-


rf = Ancestor = Mesozoa = Placozoa rj= Porifera

=j| n= Cnidaria l!=M rj= Ctenophora

1!=4= Gastrotricha Nematomorpha Nematoda Chaetognatha Phoronida Ectoprocta Brachiopoda Echinodermata Enteropneusta Pterobranchia Urochordata Cephalochordata r?= Rotifera

'—"— Acanthocephala Kinorhyncha i(= Loricifera J= Priapulida

i= Nemertea — = Entoprocta

Gnathostomulida Platyhelminthes Mollusca Sipuncula Echiura Annelida

= Pogonophora Pentastomida Tardigrada Onychophora Uniramia

= Cheliceriformes Crustacea

Fig. 3. Strict consensus tree constructed from more than 2444 MP trees (L = 125-127; CI = 0.49-0.48) obtained with mhennig* + bb* and based on Schrams's (1991) adapted data.

uc Lc

phyletic arthropod assemblage; and (5) confirm the pseudocoelomate clades Acanthocephala + Rotifera and Kinorhyncha + Loricifera + Priapulida. Part of these results is at variance with Schram's (1991) conclusions.

The topological impact of four controversial characters was investigated by subtracting them from the adapted data set (leaving 61 characters for analysis). In one instance we excluded two characters simultaneously. The strict consensus trees obtained are presented in Fig. 4. These trees differ from that of Fig. 3 as follows (numbers of characters excluded in parentheses). Enterocoel (52) [Fig. 4(A)]: pseudocoelomates, lophophorates and Urochordata + Cephalochordata are mono-phyletic groups. Metanephridia (53) (see Brusca and Brusca, 1990, for a discussion of the controversy about this character) [Fig. 4(B)]: Ctenophora is a sister-group to the Bilateria; Urochordata + Cephalochordata is a monophyletic group within the deuterostomes; the monophyletic clade Nemertea + Entoprocta + Acoelomata + Protostomia is split in an unresolved polychotomy of nine lineages (retaining the monophyletic group Pentastomida + Arthropoda). Hemocoel (67) and ecdysis (30) [Fig. 4(C)]: Arthropoda + Pentastomida is split in six unresolved lineages.

172 T. BACKELJAUETAL.

C=

Echinodermata Enteropneusta Pterobranchia Urochordata Cephalochordata Phoronida Ectoprocta Brachiopoda Nemertea Entoprocta

I |j= Gnathostoinulida LJ== Platyhelminthes

Mollusca Sipuncula

?= Echiura L= Annelida U= Pogonophora

= Pentastomida = Tardigrada = Onychophora = Uniramia = Cheliceriformes = Crustacea

Nematomorpha Neraatoda

D Chaetognatha i= Gastrotricha

— = Rotifera r=^= Acanthocephala rj= Kinorhyncha

l!=j [F= Loricifera I*—^ Priapulida

Nemertea Sipuncula Echiura Annelida Pogonophora Entoprocta Gnathostomulida Platyhelminthes Pentastomida

=!!= Tardigrada = Onychophora

Uniramia Cheliceriformes Crustacea Phoronida Ectoprocta Brachiopoda Echinodermata Enteropneusta Pterobranchia Urochordata Cephalochordata

rj= Nematomorpha

U E

LJC

B

Nematoda Chaetognatha [f= Gastrotricha

Rotifera Acanthocephala Kinorhyncha Loricifera Priapulida

j= Ancestor = t= Hesozoa

\= Placozoa

U= Chaetognatha U= Phoronida — Ectoprocta i= Brachiopoda = Echinodermata = Enteropneusta = Pterobranchia = Urochordata = Cephalochordata rj= Rotifera

=2= Acanthocephala fp Rotifera

=£= Acanthoceph; [f= Kinorhyncha

j c

.nornyncha fj= Loricifera == Priapulida Nemertea Entoprocta Gnathostomulida Platyhelminthes

r?= Mollusca = p= Sipuncula U= Echiura f= Annelida = Pogonophora = Pentastomida = Tardigrada = Onychophora = Uniramia = Cheliceriformes == Crustacea

j= Mesozoa != Placozoa

rifera jj= Cnidaria

l==| if= Ctenophora L! j= Crustacea

= Cheliceriformes Uniramia Onychophora

=jj= Tardigrada I l== Pentastomida = Pogonophora

Annelida Echiura Sipuncula Mollusca

ij= Platyhelminthes == Gnathostomulida Entoprocta Nemertea

rf= Phoronida

Urochordata Cephalochordata

Nematoda Chaetognatha = Gastrotricha j ij= Rotifera li=J== Acanthocephala n= Kinorhyncha

R= Loricifera m*— Priapulida

Fig. 4. Strict consensus trees obtained for four different data sets. (A) Minus character 52 (enterocoel) (1902 MP trees; L = 121; CI - 0.50); (B) minus character 53 (metanephridia) (834 MP trees; L = 123; CI = 0.49); (C) minus characters 30 and 67 (ecdysis and hemocoel) (>2447 MP trees; L = 123; CI = 0.48); (D) with chitin as additional character (ancestor *= 0: >2446 MP trees; L = 134; CI = 0.47; ancestor = 1: 1176 MP trees; L = 133; CI = 0.47). The trees were constructed with mhennig* + bb*.


We finally evaluated the effect of adding a new character to the data by considering the presence/absence of chitin (Ghiselin, 1989) [see Willmer (1990) for a critique on this]. The presence of this molecule among Metazoa was inferred from Shapeero (1962), Abele et al. (1989), Brusca and Brusca (1990) and Willmer (1990) (Appendix 2). We then applied Hennig86 on two (adapted) data sets: one where the ancestor had chitin, and one where it did not. This yielded, respectively, 1176 and over 2446 different MP trees (L = 133 and L = 134; CI = 0.47). The strict consensus trees, however, were identical [Fig. 4(D)]. They show the following differences with respect to the tree in Fig. 3: (1) Ctenophora is a sister-group to the Bilateria; (2) Pseudocoelomata and Deuterostomia (+Lophophorata) are mono-phyletic assemblages; and (3) Enteropneusta + Urochordata + Cephalochordata is a monophyletic group, with the latter two as sister taxa.

Discussion

Recently, Eernisse et al. (1992) reanalyzed Schram's (1991) data set with PAUP 3.0. They confirmed Schram's cladogram of L = 140 and found in addition 1422 alternative topologies of L = 139. The strict consensus tree derived from these latter is identical to the one we obtained from the 741 MP trees generated by the mhennig* + bb* algorithms of Hennig86 applied on Schram's original data [Fig. 2(B)].

Our analyses illustrate that single characters, data input orders and individual character states may profoundly affect the lengths, topologies and numbers of the MP trees found. Schram's (1991) cladogram suffered from such effects, for the number of MP trees surveyed was too small and the inclusion of autapomorphies artificially increased tree lengths and CI values. Moreover, there is no reason to prefer the tree in Fig. 1 above other MP trees, and at least seven characters used by Schram (1991) reveal debatable or erroneous character states. In addition, the homology of some characters is highly questionable. This is amongst others the case for the "proboscis" (34) of Nemertea, Acanthocephala, Echiura and Enteropneusta, and the "Coiling of the gut, anus anterior" (44) of Sipuncula, Lophophorata and Urochordata. Schram's (1991) phylogenetic presuppositions about these "homologies" are evidenced by the fact that he did not consider the "coiled gut" of Mollusca (Gastropoda, Scaphopoda, Cephalopoda) and Echinodermata (Crinoidea), nor included the "proboscis" [or "buccal introvert" (25)] of Platyhelminthes (Kalyptorhynchia), Annelida (Polychaeta), Sipuncula, Mollusca (Scaphopoda, some Gastropoda) and Pterobranchia. These doubtful and by function or analogy defined "homologies" emphasize the inadequacy of homologizing, for example, the nemertean and echiuran "proboscis". Note that if Pterobranchia are endowed with a "proboscis" (Gilmour, 1979; Ghiselin, in litt.) and Schram's (1991) adapted data matrix is corrected accordingly, then the Hemichordata become monophyletic with the Echinodermata as sister-group, while Lophophorata becomes the sister taxon of the Echinodermata + Hemichordata clade (Fig. 5).

In summary, many of Schram's (1991) conclusions about metazoan phylogeny are not warranted. Actually, only two of them can be maintained: (1) Metazoa are monophyletic; and (2) Cnidaria and Ctenophora are not sister-groups. This does not mean that Schram's (1991) cladogram is a priori wrong. Some of the branching patterns implied by his tree (Fig. 1) also occur in our trees (Figs 3-5). After correction and reanalysis, Schram's (1991) data suggest in particular that (Fig. 3):


|j= Ancestor =4= Mesozoa = Placozoa ij= Porifera

=| |j= Cnidaria ij= Cnxaaria il i^ Ctenophora

Gastrotricha

hala

iia

Gastrotnchi Nematomorphi Nematoda Chaetognatha n= Rot if era

—"— Acanthocephe ij= Urochordata

!—"— Cephalochordata ij= Kinorhyncha =j i?= Loricifera "—"— Priapulida

ip= Brachiopoda —'[= Ectoprocta

— Phoronida F= Echinodermata

—I ij= Enteropneust. '—N— Pterobranchi;

= Nemertea = = Entoprocta

|r= Gnathostomulida =J!= Platyhelminthes (r= Mollusca

= p= Sipuncula [== Echiura i= Annelida r= Pogonophora if= Pentastomida

±=±= Tardigrada N= Onychophora t= Uniramia t= Cheliceriformes — Crustacea

Fig. 5. Strict consensus tree based on 1428 MP trees (L = 127; CI = 0.48) obtained with mhennig* + bb* applied on Schram's (1991) adapted data (see Fig. 3) corrected for the presence of a proboscis in Pterobranchia.

(1) the position of Ctenophora as sister-group to the Bilateria (e.g. Ax, 1989; Schram, 1991) is neither supported, nor rejected; (2) Platyhelminthes and Gnathostomulida are sister taxa, forming the Acoelomata or Platyhelminthomorpha; (3) both Arthropoda and the "traditional" protostomes are monophyletic; (4) the Pentastomida belong to the Arthropoda; (5) the Acoelomata + Nemertea + Entoprocta + Protostomia constitute a monophyletic assemblage; (6) the deuterostomes and pseudocoelomates are unresolved polychotomies, containing Rotifera + Acanthocephala, as well as Kinorhyncha + Loricifera + Priapulida as monophyletic groups; and (7) the position of chaetognaths is still unresolved. Figure 5 further suggests the existence of two monophyletic deuterostome clades: Lophophorata + Echinodermata + Hemichordata and Cephalochordata + Urochordata.

Several of these observations are corroborated by previous morphological, embryological and paleontological studies (e.g. Ax, 1985; Clement, 1985; Lorenzen, 1985; Van der Land and Norrevang, 1985; Brusca and Brusca, 1990; Conway Morris, 1993). Yet, contradictory views also exist (e.g. Nursall, 1962; Anderson, 1983; Inglis, 1985; Bergstrom, 1986; Willmer, 1990; Conway Morris, 1993). It is beyond the scope of this paper to expand on these issues, the more since the character analysis (Fig. 4) revealed that some of our inferences are far from stable. Moreover, our adapted data matrix still contains several doubtful homologies and debatable character states [e.g. the presence of a pseudocoel (21) and the absence of ecdysis (30) in Priapulida [contradicted by Shapeero (1961, 1962)] or the absence of ecdysis


(30) in Kinorhyncha [contradicted by Nyholm (1947)] . Hence , further testing of metazoan relationships using other, i ndependen t data sets is needed. Small-subunit ribosomal RNA sequences, for example, are such data. They support: (1) the monophyly of Metazoa (Lake, 1989, 1990, 1991; Ghiselin, 1989; Patterson, 1989; Erwin, 1991; Winnepenninckx et al., 1992) [but see Christen et al. (1991a,b) for tentative evidence suggesting polyphyly]; (2) the monophyly of Ar thropoda (Turbeville et al., 1991; Winnepenninckx et al., 1992) [but see Lake (1989) for an alternative viewpoint]; (3) the inclusion of the Pentastomida in the Ar thropoda (Abele et al., 1989; Winnepennincks et al., 1992); and (4) the unresolved affinities of protos tome phyla (Field et al., 1988; Raff et al., 1989; Lake, 1989, 1990; Patterson, 1989). In addit ion, 18S rRNA data suggest that Nemer tea are protostomes (Turbeville et al., 1992) and that Echinodermata diverged before the Hemichordata (Ente ropneus ta ) -Chorda ta splitting (Holland et al., 1991). Recently, 12S rRNA sequence data provided evidence for the inclusion of the Onychophora in the Ar thropoda (Ballard et al., 1992). We will no t expand on these molecular data.

Acknowledgments

We are indebted to M. T. Ghiselin (California Academy of Sciences), J. Van Goethem, K. Wouters (both KBIN, Brussels) and an anonymous referee for their comments on the manuscript . Financial support was received from FMSR grant No. 3.0080.89 and FJBR grant No. 2.0004.91. LDB is senior Research Assistant at the National Fund for Scientific Research (Belgium).

REFERENCES

ANDERSON, D. T. 1983. Origins and relationships among the animal phyla. Proc. Linn. Soc. NSW 106: 151-166.

Ax, P. 1985. The position of the Gnathostomulida and Platyhelminthes in the phylogenetic system of the Bilateria. In: S. Conway Morris, J. D. George, R. Gibson and H. M. Platt (eds). The Origins and Relationships of Lower Invertebrates. The Systematic Association Special Volume No. 28, Oxford Univ. Press, Oxford, pp. 168-180.

Ax, P. 1989. Basic phylogenetic systematization of the Metazoa. In: B. Fernholm, K. Bremer and H. Jornvall (eds). The Hierarchy of Life. Excerpta Medica (Elsevier), Amsterdam, pp. 229-245.

ABELE, L. G., W. KIM AND B. E. FELGENHAUER. 1989. Molecular evidence for inclusion of the phylum Pentastomida in the Crustacea. Mol. Biol. Evol. 6: 685-691.

ANDERBERG, A. AND A. TEHLER. 1990. Consensus trees, a necessity in taxonomic practice. Cladistics 6: 399-402.

BAKKE, T. 1980. Embryonic and post-embryonic development in the Pogonophora. Zool. Jb. Anat. 103: 276-284.

BALLARD, J. W. O., G. J. OLSEN, D. P. FAITH, W. A. ODGERS, D. M. ROWELL AND P. W. ATKINSON. 1992. Evidence from 12S ribosomal RNA sequences that onychophorans are modified arthropods. Science 258: 1345-1348.

BARINAGA, M. 1991. "African Eve" backers beat a retreat. Science 255: 686-687. BARNES, R. D. 1974. Invertebrate Zoology, third edition. W. B. Saunders, Philadelphia. BERGSTR6M,J. 1986. Metazoan evolution—a new model. Zool. Scr. 15: 189-200. BLAIR HEDGES, S., S. KUMAR, K. TAMURA AND M. STONEKING. 1991. Human origins and analysis

of mitochondrial DNA sequences. Science 255: 737-739. BRUSCA, R. C. AND G.J. BRUSCA. 1990. Invertebrates. Sinauer Associates, Sunderland. CHRISTEN, R. A., A. RATTO, A. BAROIN, R. PERASSO, K. G. GRELL AND A. ADOUTTE. 1991a. Origin

of metazoans—A phylogeny deduced from sequences of the 28S ribosomal RNA.


In: A. M. Simonetta and S. Conway Morris (eds). The Early Evolution of Metazoa and the Significance of Problematic Taxa. Cambridge Univ. Press, Cambridge, pp. 1-9.

CHRISTEN, R. A., A. RATTO, A. BAROIN, R. PERASSO, K. G. GRELL AND A. ADOUTTE. 1991b. An analysis of the origin of metazoans, using comparisons of partial sequences of the 28S RNA, reveals an early emergence of triploblasts. EMBOJ. 10: 499-503.

CLEMENT, P. 1985. The relationships of rotifers. In: S. Conway Morris, J. D. George, R. Gibson and H. M. Platt (eds). The origins and Relationships of Lower Invertebrates. The Systematic Association Special Volume No. 28, Oxford Univ. Press, Oxford, pp. 224-247.

CONWAY MORRIS, S. 1993. The fossil record and the early evolution of the Metazoa. Nature 361:219-225.

EERNISSE, D. J., J. S. ALBERT AND F. E. ANDERSON. 1992. Annelida and Arthropoda are not sister taxa: a phylogenetic analysis of spiralian metazoan morphology. Syst. Biol. 41: 305-330.

ERWIN, D. H. 1991. Metazoan phylogeny and the Cambrian radiation. Trends Ecol. Evol. 6: 131-134.

FARRIS, J. S. 1989. Hennig86: A PC-DOS program for phylogenetic analysis. Cladistics 5: 163. FIELD, K. G., G. J. OLSEN, D. J. LANE, S. J. GIOVANNONI, M. T. GHISELIN, E. C. RAFF, N. R. PACE

AND R. A. RAFF. 1988. Molecular phylogeny of the animal kingdom. Science 239: 748-753.

GHISELIN, M. T. 1989. Summary of our present knowledge of metazoan phylogeny. In: B. Fernholm, K. Bremer and H. Jornvall (eds). The Hierarchy of Life. Excerpta Medica (Elsevier), Amsterdam, pp. 261-272.

GILMOUR, T. H. J. 1979. Feeding in pterobranch hemichordates and the evolution of gill slits. Can.J.Zool. 57:1136-1142.

GODEAUX, J. E. A. 1974. Introduction to the morphology, phylogenesis, and systematics of lower Deuterostomia. In: M. Florkin and B. T. Scheer (eds). Chemical Zoology, Volume VIII. Deuterostomians, Cyclostomes, and Fishes. Academic Press, New York/London, pp. 3-60.

HOLIAND, P. W. H., A. M. HACKER AND N. A. WILLIAMS. 1991. A molecular analysis of the phylogenetic affinities of Saccoglossus cambrensis Brambell & Cole (Hemichordata). Phil. Trans. R. Soc. Lond. B 332: 185-189.

INGLIS, W. G. 1985. Evolutionary waves: patterns in the origins of animal phyla. Aust. J. Zool. 33: 153-178.

IVANOV, A. V. 1975. Embryonalentwicklung der Pogonophora und ihre systematische Stellung. Z. zool. Syst. Evolut.-forsch. Sonderheft 1: 10-44.

LAKE, J. A. 1989. Origin of the multicellular animals. In: B. Fernholm, K. Bremer and H. Jornvall (eds). The Hierarchy of Life. Excerpta Medica (Elsevier), Amsterdam, pp. 261-272.

IAKE, J. A. 1990. Origin of the Metazoa. Proc. Natl. Acad. Sci. USA 87: 763-766. IAKE, J. A. 1991. Tracing origins with molecular sequences: metazoan and eukaryotic

beginnings. Trends Biochem. Sci. 16: 46-50. LEMCHE, H. AND K. G. WINGSTRAND. 1959. The anatomy of Neopilina galatheae Lemche, 1957.

GalatheaRep. 1-71. LESTER, S. M. 1985. Cephalodiscus sp. (Hemichordata: Pterobranchia): observations of func

tional morphology, behavior and occurrence in shallow water around Bermuda. Mar. Biol. 85: 263-268.

LORENZEN, S. 1985. Phylogenetic aspects of pseudocoelomate evolution. In: S. Conway Morris, J. D. George, R. Gibson and H. M. Platt (eds). The Origins and Relationships of Lower Invertebrates. The Systematic Association Special Volume No. 28, Oxford Univ. Press, Oxford, pp. 210-223.

NURSALL.J. R. 1962. On the origin of the major groups of animals. Evolution 16: 118-123. NYHOLM, K.-G. 1947. Contributions to the knowledge of the postembryonic development in

Echinoderida Cyclorhagae. Zool. Bidr. Uppsala 25: 423-428. OSCHE, G. 1963. Die systematische Stellung und Phylogenie der Pentastomida. Z. Morph.

Okol. Tiere 52: 487-596.


PAGE, R. D. M. 1989. Comments on component-compatibility in historical biogeography. Cladistics 5: 167-182.

PATTERSON, C. 1989. Phylogenetic relations of major groups: conclusions and prospects. In: B. Fernholm, K. Bremer and H. Jornvall (eds). The Hierarchy of Life. Excerpta Medica (Elsevier), Amsterdam, pp. 471-488.

PLATNICK, N. I. 1987. An empirical comparison of microcomputer parsimony programs. Cladistics 3:121-144.

PLATNICK, N. I. 1989. An empirical comparison of microcomputer parsimony programs, II. Cladistics 5: 145-161.

RAFF, R. A., K. G. FIELD, G. J. OLSEN, S.J. GIOVANNONI, D.J. LANE, M. T. GHISELIN, N. R. PACE AND E. C. RAFF. 1989. Metazoan phylogeny based on analysis of 18S ribosomal RNA. In: B. Fernholm, K. Bremer and H. Jornvall (eds). The Hierarchy of Life. Excerpta Medica (Elsevier), Amsterdam, pp. 247-260.

SANDERSON, M. J. 1990. Flexible phylogeny reconstruction: a review of phylogenetic inference packages using parsimony. Syst. Zool. 39: 414—420.

SHAPEERO, W. L. 1961. Phylogeny of Priapulida. Science 133: 879-880. SHAPEERO, W. L. 1962. The epidermis and cuticle of Priapulus caudatus Lamarck. Trans. Am.

Microsc. Soc. 81: 352-355. SCHRAM, F. R. 1991. Cladistic analysis of metazoan phyla and the placement of fossil

problematica. In: A. M. Simonetta and S. Conway Morris (eds). The Early Evolution of Metazoa and the Significance of Problematic Taxa. Cambridge Univ. Press, Cambridge, pp. 35-46.

TEMPLETON, A. R. 1991. Human origins and analysis of mitochondrial DNA sequences. Science 255: 737.

TURBEVILLE, J. M., D. M. PFEIFER, K. G. FIELD AND R. A. RAFF. 1991. The phylogenetic status of arthropods, as inferred from 18S rRNA sequences. Mol. Biol. Evol. 8: 669-686.

TURBEVILLE, J. M., K. G. FIELD AND R. A. RAFF. 1992. Phylogenetic position of phylum Nemertini, inferred from 18S rRNA sequences: molecular data as a test of morphological character homology. Mol. Biol. Evol. 9: 235-249.

VAN DER LAND, J. AND A. NORREVANG. 1985. Affinities and intraphyletic relationships of the Priapulida. In: S. Conway Morris, J. D. George, R. Gibson and H. M. Platt (eds). The Origins and Relationships of Lower Invertebrates. The Systematic Association Special Volume No. 28, Oxford Univ. Press, Oxford, pp. 261-273.

WILLMER, P. 1990. Invertebrate relationships—Patterns in animal evolution. Cambridge Univ. Press, Cambridge.

WINGSTRAND, K. G. 1985. On the anatomy and relationships of recent Monoplacophora. GalatheaRep. 16: 1-94.

WINNEPF.NNINCKX, B., T. BACKELJAU, Y. VAN DE PEER AND R. DE WACHTER. 1992. Structure of the small ribosomal subunit RNA of the pulmonate snail Limicolaria kambeul and phylogenetic analysis of the Metazoa. FEBS Lett. 309: 123-126.

Appendix 1

Characters and character states used by Schram (1991) to infer metazoan phylogeny. For each character we give first the primitive state (0) and then the derived one (1). Additions supposed to be implied, but no t ment ioned , by Schram (1991) are indicated in [ ] .

1. No collagen, n o acetylcholine/[cholinesterase] transmission, reproductive cells (if separate) on surface of exterior (0); collagen, acetylcholine/ cholinesterase system, flagellated or ciliated sperm with condensed chromatin and mitochondria , reproductive cells or tissues internal (1).

2. N o gut (0); gut or special cells or areas for digestion (1). 3. Generalized free-swimming larva (0); nematogen/ infusor igens (1). 4. No symmetry (0); bilateral symmetry (1).


5. No alteration of generations (0); metagenesis of some type (1). 6. Locomotion by action of cilia or flagella (0); locomotion by muscle action (1). 7. No basal membranes, nerve cells or gap junctions (0); epithelia with basal

nerve synapses, gap junctions between cells (1). 8. No endoderm (0); endoderm in embryo (1). 9. No special water channels or choanocytes (0); pores and canals for water

circulation facilitated by choanocytes (1). 10. Embryonic layers do not invert (0); embryonic inversion (1). 11. No mesoderm (0); mesoderm in embryo (1). 12. Cleavage indeterminate (0); cleavage determinate (1). 13. Muscles ectodermal in origin (0); muscles subepidermal (1). 14. No symmetry (0); radial or biradial (1). 15. No nematocysts (0); nematocysts (1). 16. No anus (0); anus (1). 17. No colloblasts (0); colloblasts (1). 18. Radial cleavage (0); spiral quartet cleavage (1). 19. Ectomesoderm (0); 4d mesoderm (1). 20. Nervous system as poorly polarized nerve net (0); orthogonal nervous system,

with anterior nerve ring and several longitudinal cords (1). 21. No body cavity (0); pseudocoel (1). 22. Generalized free-swimming larvae (0); no primary larva (1). 23. Body wall muscles both longitudinal and circular, when present (0); longitudinal

body wall [muscles] only (1). 24. Radial cleavage (0); aberrant spiral cleavage (monets and duets) (1). 25. Buccal chamber not eversible (0); buccal introvert (1). 26. Cuticle solid, if present (0); cuticularized epidermis with tubules (1). 27. No retrocerebral organ (0); retrocerebral organ (1). 28. Noeutely (0);eutely (1). 29. No renettes, amphids, or phasmids (0); renette cells, amphids, and phasmids (1). 30. No moulting (0); ecdysis with ecdysone (1). 31. No special excretory organ (0); protonephridia (1). 32. Flame cells (ciliate) (0); flame bulbs (flagellate) (1). 33. No lemnisci (0); lemnisci (1). 34. No proboscis (0); proboscis (1). 35. Free cilia on epidermis (0); cilia ensheathed with cuticle (1). 36. No uterus bell (0); uterus bell (1). 37. Epidermis solid (0); epidermis lacunae (1). 38. Cuticle uniform (0); cuticle with several unit membranes (1). 39. No scales (0); scalids on introvert (1). 40. No lorica (0); lorica at some stage (1). 41. No caudal appendages (0); caudal appendages at some stage (1). 42. No body cavity (0); coelom (metacoel) (1). 43. Mesoderm not segmented (0); segmented or serial structures derived from

mesoderm (1). 44. Gut straight (0); gut coiled or looped, anus anterior (1). 45. No lophophore (0); lophophore (1). 46. Downstream particle [capture] (0); upstream particle capture in adults (1). 47. Downstream particle capture (0); upstream particle capture in larvae (1).


48. Larval apical organ does not degenerate (0); larval apical organ degenerates (1). 49. Generalized free-swimming larva (0); larvae as trochophore or trochophore-

like (1). 50. No special photoreceptor (0); photoreceptor cell with tuft of tightly packed

cilia (1). 51. Coelom undivided (0); archimeric coelom (1). 52. Coelom as a schizocoel, if present (0); enterocoel if present (1). 53. No special excretory organ (0); metanephridia (1). 54. Generalized free-swimming larva (0); tornaria/bipinaria larva (1). 55. No pharyngeal slits (0); pharyngeal slits (1). 56. No buccal diverticulum (0); buccal diverticulum (1). 57. No single dorsal nerve cord (0); dorsal nerve cord from tube-like infolding of

ectoderm (1). 58. Pharyngeal slits simple (0); pharyngeal slits divided (1). 59. No calcareous endoskeleton (0); calcareous endoskeleton (1). 60. No water-vascular system (0); water-vascular system (1). 61. No notochord (0); notochord (1). 62. Brain not derived from any part of larval apical organ (0); brain in part derived

from larval apical organ; main nerve cord ventral (1). 63. No particular cell source for mesoderm and coelom (0); teloblasts give rise to

mesoderm, coelom mesoderm and coelom (1). 64. Coelom large and pervasive (0); coelom restricted to around circulatory system

(1)-65. No rhynchocoel (0); rhynchocoel (1). 66. No larval prototrochal lobes (0); larva with prototroch developed as ciliated

lobes (pilum or velum) (1). 67. No hemocoel (0); hemocoel (unlined cavity in mesoderm) (1). 68. Body grows uniformly (0); apical growth of larva (1). 69. Only circular and longitudinal body muscles (0); oblique body wall muscles (1). 70. No anal vesicles (0); anal vesicles (1). 71. No appendages (0); lobopods or uniramous limbs (1). 72. No antennae (0); deutocerebral antennae (1). 73. No special excretory organ (0); segmental, excretory glands (1). 74. No appendages (0); biramous limbs (1). 75. No tracheae (0); tendency to develop tracheae (1). 76. No special excretory organ (0); tendency to develop Malpighian tubules (1). 77. No special jaws (0); whole limb jaw (1).


Apendix2

Original character states of 77 characters used by Schram (1991) (1-77: 0 = plesiomorphic state, 1 = apomorphic state, - = state unknown) and distribution of chitin among Metazoa (78: 0 = absent, 1 = present; - = unknown, X = present on absent).

Ancestor Mesozoa Placozoa Porifera Cnidaria Ctenophora Gnathostomulida Plathyhelminthes Gastrotricha Rotifera Acanthocephala Loricifera Kinorhyncha Priapulida Nematomorpha Nematoda Chaetognatha Mollusca Nemertea Sipuncula Echiura Annelida Pogonophora Pentastomida Tardigrada Onychophora Uniramia Cheliceriformes Crustacea Phoronida Ectoprocta Entoprocta Brachiopoda Echinodermata Enteropneusta Pterobranchia Urochordata Cephalochordata

1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 78 I I I I I I I I I I I I I I I I I 00000000000000000000000000000000000000000000000000000000000000000000000000000 X 101110(>00000<X)WOOOOO(H)000000()00000<)000(K)00()000^^ -10-000000000000000000-0000000000000000000000000000000000000000000000000000000 -1100000011000(>00000000()00(>0<)0<WO<)00000(H)000000000()0^ 1 1100111100000110(KX)0000()0000000000000000000000000(H)000000()000000000000000()000 1 11000111001111(»100000()0(H)00()0000000000{)000000()00000()0000000(H)0000()OOOOOOOW 0 11010011001-100001-0010000000010000000000000000000000000000000000000000000000 1 1101001100111000011000(M)00000010000000()00000()000000000000000000()0000000000000 0 11010011001110010001010100010011001000000000000000000000000000000000000000000 1 11010111001110010001110101110011110001000000000000000000000000000000000000000 1 10010111001110010001110101010011110110000000000000000000000000000000000000000 -11010111001110010-0- -10-100-00—000000111000000000000000000000000000000000000 -11010111001110010-01110-10010011000000100000000000000000000000000000000000000 -11010111001110010001110010010011000000111000000000000000000000000000000000000 1 110101110011100100011111100-0000000000000000000000000000000000000000000000000 -11010111001110010001111110011100000000000000000000000000000000000000000000000 1 11010111001010010001111001100000000000000000000000000000000000000000000000000 1 11010111001110010110000000000000000000000110000010001000000001010100000000000 1 11010111001110010110000000000010010000000110000000000000000000011100000000000 0 11010111001110010110000000000000000010000101000010001000000001000000000000000 0 11010111001110010110000000000000010000000100000010001000000001000000110000000 1 11010111001110010110000000000000000000000110000010001000000001100000000000000 1 100101110011100001-0000000000010000000000110000010000000000001100001000000000 1 110101110011100101-0000000000000000000000110000000000000000001100000000000000 1 110101110011100101-0000000010100000000000110000000010000000001100010001000000 1 110101110011100101-0000000000100000000000110000000000000000001100010101000101 1 110101110011100101-0000000000100000000000110000000000000000001100010001100111 1 110101110011100101-000000000010000000000011000000000000000000110001000-01-110 1 11010111001110010100000000000100000000000110000000000000000001100010000111000 1 11010111001110010000000000000000000000000101111100101000000000000000000000000 1 11010111001110010000000000000000000000000101111001-00000000000000000000000000 1 11010111001110010110100000000010000000000101100101000000000000000000000000000 1 11010111001010010000000000000000000000000101111100111000000000000000000000000 1 11000111001011010000000000000000000000000100001100110100001100000000000000000 0 11010111001010010000001000000000010000000100001100110111110000000000000000000 0 1101011100101001000000000000000000000000010111-1001-0111000000000000000000000 0 11011111001010010000000000000000000000000001000100000010110010000000000000000 0 11010111001010010000000000000000000000000110000100010010110010000000000000000 1 I I I I I I I I I I I I I I I I I 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 78

CLADISTIC ANAIA5IS OF METAZOA 181

Appendix 3

Alternative input orders of Schram's (1991) data (Appendix 2) and their effect on the mhennig* (mh*) + bb* analysis.

Ancestor Pterobranchia Placozoa Sipuncula Porifera Cnidaria Ectoprocta Platyhelminthes Onychophora Gastrotricha Rotifera Acanthocephala Pentastomida Priapulida Nematomorpha Nematoda Chaetognatha Echiura Annelida Gnathostomulida Pogonophora Tardigrada Uniramia Ctenophora Cheliceriformes Crustacea Phoronida Loricifera Entoprocta Mesozoa Nemertea Brachiopoda Echinodermata Enteropneusta Kinorhyncha Urochordata Mollusca Cephalochordata

mh* 4 trees bb* 735 trees

Ancestor Pterobranchia Sipuncula Porifera Cheliceriformes Cnidaria Ectoprocta Platyhelminthes Onychophora Gastrotricha Annelida Acanthocephala Pentastomida Priapulida Nematomorpha Nematoda Placozoa Chaetognatha Rotifera Echiura Gnathostomulida Pogonophora Uniramia Ctenophora Crustacea Phoronida Loricifera Entoprocta Mesozoa Nemertea Tardigrada Brachiopoda Echinodermata Enteropneusta Kinorhyncha Urochordata Mollusca Cephalocordata

mh* 2 trees bb* 735 trees

Ancestor Pterobranchia Sipuncula Porifera Phoronida Cheliceriformes Ectoprocta Onychophora Gastrotricha Annelida Acanthocephala Pentastomida Priapulida Nematoda Placozoa Rotifera Echiura Gnathostomulida Pogonophora Cephalochordata Uniramia Ctenophora Crustacea Cnidaria Loricifera Entoprocta Mesozoa Nemertea Tardigrada Brachiopoda Nematomorpha Echinodermata Chaetognatha Enteropneusta Kinorhyncha Urochordata Mollusca Platyhelminthes

mh* 3 trees bb» 768 trees

CLADISTIC ANALYSIS OF METAZOAN RELATIONSHIPS: A REAPPRAISAL

Documents