Metazoan Relationships on the Basis of 18S rRNA Sequences: A Few Years Later...'
BIRGITTA M. H. WINNEPENNINCKX,2 YVES VAN DE PEER,* AND THIERRY BACKELJAU
SYNOPSIS. The 18S rRNA database is continuously growing and new tree construction methods are being developed. The present study aims at assessing what effect the addition of recently determined animal 18S rRNA sequences and the use of a recently developed distance matrix calculation method have on the results of some previously published case studies on metazoan phylogeny. When re-assessing three case studies, part of their conclusions was confirmed on the basis of the present 18S rRNA data set: 1) the monophyly of Arthropoda; 2) the monophyly of the Vestimentifera-Pogonophora and their protostome character; 3) the doubt about the monophyletic origin of Echiura-Sipuncula and 4) the coelomate character of Nemertea. Yet, it is also demonstrated that some of the previous results are no longer warranted when updating the analyses: 1) the monophyly of both the Annelida and the Eutrochozoa; 2) the sister-group relationship of Echiura to Vestimentifera-Pogonophora and 3) the polyphyly of the Mesozoa and their close relationship to Myxozoa and Nematodes. In addition, some new very preliminary evidence is provided for: 1) a common ancestry of Platyhelminthes and Mesozoa and the monophyly of the latter group and 2) the monophyly of Clitellata, Hirudinida and Oligochaeta. Finally, doubt is casted on the monophyly of the Polychaeta and the polychaete orders Spionida, Phyllodocida, and Sabellidae. Of course, these hypotheses also need further testing.
Despite several decades of morphological and anatomical research, numerous aspects of metazoan relationships have remained uncertain (for an overview see Brusca and Brusca, 1990; Willmer, 1990; Meglitsch and Schram, 1991). In the 80s, molecular data were introduced as independent markers to trace phylogenetic relationships. Several properties make small ribosomal subunit RNA (SSU rRNA) an excellent tool for this purpose, such as its ubiquity, homology, length and functional constancy (e.g., Raff, 1988; Raff et al., 1989; Hillis and Dixon, 1991; Solignac et al., 1991; Woese, 1991). Moreover, as there is a difference in variability between sites, the molecule can be used to trace phylogenetic relationships at a broad range of taxonomical levels. Finally, several very conservative regions in the molecule allow the development of "universal" PCR and sequencing primers. Therefore, SSU rRNA sequences have had an overwhelming impact on phylogenetic studies in general and have yielded several new insights in evolutionary biology (e.g., Woese and Fox, 1977; Cedergen et al., 1988; Gouy and Li, 1989; Sogin et al., 1989; Hasegawa et al., 1993; Doolittle and Brown, 1994; Halanych et al., 1995).
A decade ago, mainly due to the work of Field et al. (1988), SSU rRNA (18S rRNA) was also introduced to trace metazoan relationships. Since then there has been an "18S rRNA sequence-boom," 413 complete or nearly complete metazoan 18S rRNA sequences were available this year in the database of Van de Peer et al. (1998). Nearly all metazoan phyla are now represented and for some phyla tens of species have been studied (e.g., Arthropoda, Mollusca, Annelida, Nematoda; for an overview of sequences available see Van de Peer et al., 1998).
Most of the 18S rRNA studies focus on only one or a few aspects of metazoan phylogeny and very often include only a small part of the available 18S rRNA sequence data. Relevant taxa are sometimes even not considered without giving a valid explanation. Arbitrarily, we regard a particular taxon to be relevant for a certain problem if there is any previous indication that this taxon is closely related to the organisms under consideration or that this taxon belongs to the clade which is likely to contain one of the species under consideration.
Not only the number of available 18S rRNA sequences has increased dramatically but also alignments, tree construction methods and stability tests have improved over the course of time. Yet, earlier results are seldomly re-assessed using more recent alignments and more advanced tree construction methods (e.g., Adoutte and Philippe, 1993). In a previous paper (Winnepenninckx and Backeljau, 1996), we reported on the effect that alignment may have on phylogenetic results. The present paper assesses the impact of broader taxon sampling and/or the use of a more advanced distance matrix calculation method on previously published analyses. In particular we intend to: 1) examine the hypotheses of Kim et aL (1996) on annelid, molluscan, and arthropod monophyly by adding relevant sequences that were originally omitted from their analyses; 2) give an update of our previously published paper on the relationships of protostome worms (Winnepenninckx et al., 1995a) by adding recently determined sequences to the data set; 3) assess the effect of using a more advanced, recently developed distance matrix calculation method and of including additional sequences on the results on mesozoan relationships (Katayama et al., 1995, Hanelt et al., 1996; Pawlowski et al., 1996).
Kim et al. (1996) focused on molluscan, annelid and arthropod relationships and tested their monophyly but did not include all available relevant 18S rRNA sequences (e.g., Van de Peer et al., 1994; Winnepenninckx et al., 1995a). Using four mollusc (Gastropoda and Bivalvia), six arthropod and eight annelid sequences, Kim et al. (1996) concluded that Annelida, Arthropoda and Eutrochozoa are all monophyletic groups. Molluscan non-monophyly is suggested in his molecular analyses but could not be proven with significant support. We extended their data set with the sequences of Winnepenninckx et al. (1995a), to assess what effect the addition of these taxa has on Kim et al.'s (1996) conclusions.
Winnepenninckx et al. (1995a) studied the relationships of six worm phyla (Vestimentifera, Pogonophora, Echiura, Sipuncula, Nemertea and Annelida), using complete 18S rRNA sequences. At that time however only a single complete 18S rRNA sequence was available for each of these phyla, except for the Annelida of which two sequences were known. Vestimentifera and Pogonophora have long been a subject of debate in several morphological/anatomical studies which considered them as proto- or deuterostomes and arguments pro and contra their sister group relationship were raised (e.g., Ivanov, 1955; Southward, 1963, 1988; Webb, 1964; Norrevang, 1970; Godeaux, 1974; Gupta and Little, 1975; Van der Land andNorrevang, 1975; Jones 1985; Brusca and Brusca, 1990; Conway Morris, 1993). On the basis of neighborjoining (NJ) and maximum parsimony (MP) analyses of 22 complete metazoan 18S rRNA sequences, Winnepenninckx et al. (1995a) suggested that Vestimentifera and Pogonophora constitute a monophyletic protostome clade with the unsegmented worm phylum Echiura as sister group. Yet, the Echiura have a strong morphological, larval and embryological resemblance to Sipuncula. Therefore, it has been suggested that these two phyla are closely related (Clark, 1969; Rice, 1985; Brusca and Brusca, 1990). Some authors consider both groups to be related to annelids (Clark, 1969). These viewpoints were not confirmed by the 18S rRNA study of Winnepenninckx et al. (1995a). Furthermore, the 18S rRNA study suggested that Nemertea, a phylum of dorso-ventrally flattened, cephalized unsegmented worms, are protostomes. In anatomical and embryological studies they have been considered either as acoelomates related to Turbellaria (e.g., Norenburg, 1985; Bartolomaeus, 1988; Willmer, 1990) or as protostomes (e.g., Berg, 1985; Turbeville and Ruppert, 1985; Turbeville, 1986). Partial 18S rRNA sequences also supported the protostome character of Nemertea (Turbeville et al., 1992). In the present study we tested all these hypotheses on protostome worm relationships by adding 26 annelids and a second nemertean. Moreover, as lophophorates (e.g., Halanych et al., 1995; Conway Morris et al., 1996; Mackey et al., 1996) and some "pseudocoelomate" groups (Winnepenninckx et al., 1995b; Garey et al., 1996) belong to the same clade as protostome worms, they were also included.
The third case study concerns the relationships of the mesozoan taxa Orthonectida and Rhombozoa, whose evolutionary origin has been controversial for many years. Some consider Mesozoa as very primitive organisms that link protists and animals (e.g., Lapan and Morowitz, 1972; Margulis and Schwartz, 1988), while others believe that the simplicity of Mesozoa is secondary and regard them as degenerate flatworms because of their morphological resemblance with flatworm larvae and their complex life cycle (Stunkard, 1954, 1972). Still others consider Mesozoa as multicellular protists unrelated to the animals (Cavalier-Smith, 1993). Also Mesozoa monophyly is debated. Indeed, some consider the morphological similarity between both Orthonectida and Rhombozoa to be purely superficial and the result of convergence due to similar lifestyles (e.g., Brusca and Brusca, 1990). Two recent 18S rRNA studies support this view. Both Pawlowski et al. (1996) and Hanelt et al. (1996), determined the 18S rRNA sequence of the orthonectid Rhopalura ophiocomae, while the former authors also determined the sequence of the rhombozoid Dicyema sp. Previously, the 18S rRNA sequences of two other dicyemids, viz. D. orientale and D. acuticephalum had been published (Katayama et al., 1995). On the basis of pairwise distance, maximum parsimony, and maximum likelihood analyses, a separate origin for orthonectids and rhombozoids was suggested. Furthermore, in most analyses, the Mesozoa diverged early in the animal evolution, close to myxozoans and nematodes. We reanalyzed the 18S rRNA data of Hanelt et al. (1996) using the substitution rate calibration (SRC) method of Van de Peer et al. (1996a), which takes into account substitution rate variation among sites. The data set was also extended with additional sequences, including the Mesozoa sequences of Pawlowski et al. (1996) and Katayama et al. (1995).
MATERIAL AND METHODS
The alignment of Hanelt et al. (1996) was taken from their internet site (http://biology.unm.edu/-esloker/hanelt96.html). For all other analyses, we used the alignment of Van de Peer et al. (1998) (website http://rrna.uia.ac.be/ssu/), which is based on the secondary structure of the molecules. For the re-analyses of the data of Winnepenninckx et al. (1995a) and of Kim et al. (1996), we considered only the same nucleotides as in the original analyses. For the re-analyses of the Hanelt et al. (1996) data, domain 23 was excluded from the alignment of Van de Peer et aL (1998). Two tree construction methods were applied: NJ (Saitou and Nei, 1987) and MP analysis. Distance matrix analyses were performed using the computer program TREECON (Van de Peer and De Wachter, 1997a). Gaps were not taken into account. Apart from the Jukes and Cantor (JC) (1969) and Kimura (1980) (transition:transversion ratio 2:1) formulas for calculating distances, the SRC method of Van de Peer et aL (1996a) was also used. This method estimates evolutionary distances by considering substitution rates of individual nucleotides. One of the main problems in considering among-site rate variation is the quantitative estimation of the substitution rates or variabilities of nucleotide sites. MP estimates can be heavily biased (Wakeley, 1993; Yang, 1996; Tourasse and Gouy, 1997), while maximum likelihood (ML) estimates may experience computational difficulties (Yang, 1996). The SRC method measures the relative substitution rate of individual sites in a nucleotide sequence alignment on the basis of a distance approach (see Van de Peer et al., 1993 and Van de Peer et al., 1996a for details). The main advantage of this approach is that nucleotide variability estimates can be based on very large numbers of sequences and that they do not depend upon a given tree topology, contrary to estimates inferred from MP or ML methods (Sullivan et al., 1996; Yang, 1996). The final equations obtained to convert dissimilarity into distance are similar to the equations used to compute gamma distances (Jin and Nei, 1990; Rzhetsky and Nei, 1994), but the novelty of the SRC lies in the computation of the substitution rate of individual nucleotides. It has been proven that trees constructed on the basis of SRC distances may suffer less from anomalies caused by the presence of extremely long branches (Van de Peer et al., 1996a).
MP trees were constructed using the heuristich search option of PAUP version 3.1 (Swofford et al., 1993) with the tree bisection-reconnection branch swapping option invoked. Although this approach does not guarantee optimality, it is known to be very effective (Swofford et al., 1993). Random addition of sequences (50 times) was used to enhance the effectiveness of the heuristic search (Swofford et al., 1993). Only parsimony informative sites were included and gaps were treated as missing data. Stability was tested using bootstrapping (Felsenstein, 1985) (1000 replicates). Nodes were considered to be significantly supported if bootstrap values exceeded 70% (Hillis and Bull, 1993). RESULTS
Re-analysis of the data of Kim et. al. (1996)
The NJ tree obtained on the basis of the same sequence sampling as Kim et al. (1996) but aligned according to Van de Peer et al. (1998), supports arthropod monophyly (Fig. IA). Yet, neither annelids nor molluscs are monophyletic groups. Eutrochozoa is a monophyletic group, but without significant bootstrap support. These results are confirmed using Kimura or SRC distances (results not shown). The MP tree (Fig. lB) supports arthropod and eutrochozoan monophyly. In agreement with the MP analysis of Kim et al. (1996), Mollusca are not monophyletic, although this result is not significantly supported. However, one has to keep in mind that not all tree topologies could be evaluated. For 20 taxa, there are 8.2 102 possible rooted trees (Cavalli-Sforza and Edwards, 1967). However, only 3.5 105 rearrangements were considered. Figure 2A shows the NJ tree obtained when the data set of Kim et al. (1996) is extended with the echiuran (1), sipunculid (1), vestimentiferan (1), pogonophoran (1), nemertean (1) and annelid (2) sequences of Winnepenninckx et al. (1995a). Arthropoda are once again monophyletic. Yet, there is no support for the monophyly of Annelida, Mollusca or Eutrochozoa. These results are confirmed when the Kimura or SRC method are used instead of the JC method. In the MP tree (Fig. 2B), both Arthropoda and Eutrochozoa are monophyletic groups, yet the latter is not significantly supported by bootstrap analysis. Mollusca and Annelida are polyphyletic groups, albeit with bootstrap values of less than 50%. The Sipuncula are included in an Annelida clade, though with a very low bootstrap value. Although 2.98 1033 rooted topologies are potentially present for 27 taxa, only 1.5 106 rearrangements were tried. A summary of the results obtained on the basis of the different data sets is given in Table 1. Re-analysis of the data of Winnepenninckx et al. (1995a)
Figure 3 shows the NJ tree derived from the JC distances of 57 complete or nearly complete metazoan 18S rRNA sequences. Opisthorchis viverrini was used as outgroup. The "pseudocoelomate" groups Rotifera and Acanthocephala form a monophyletic clade that branches off first. Arthropoda, Vertebrata and Chordata are monophyletic groups, although the latter is not supported by a significant bootstrap value. As previously reported by Winnepenninckx et al. (1995b), the pseudocoelomate group Priapulida is a sister group to the Arthropoda. The cluster of Arthropoda, Priapulida and Chordata is separated from the "eutrochozoan" clade, though with low bootstrap support. Within the Eutrochozoa, it is suggested that: 1) Vestimentifera and Pogonophora form a monophyletic clade but not with Echiura as a sister group; 2) the nemerteans, represented by the orders Enopla and Anopla, are monophyletic, but this result is not significantly supported by bootstrapping; 3) there is no indication for a close relationship between Echiura and Sipuncula, nor for a relationship of one of these phyla to the Annelida; 4) neither Annelida nor Mollusca is a monophyletic group; 5) Hirudinida and Clitellata are both monophyletic groups; 6) within the Hirudinida, the subclasses Branchiobdellida (Xironogiton victoriensis and Sathodrilus attenuatus) and Hirudinea (Haemopis sanguisuga and Glossiphonia sp.) are both monophyletic groups; 7) the seven different polychaete annelid orders (Phyllodocida, Spionida, Capitellida, Terebellida, Sabellida, Orbinda, Chaetopterida) included in the analysis do not have a common origin; 8) none of the polychaete orders for which more than one sequence is available (Spionida, Phyllodocida, and Sabellida), are monophyletic; 9) Brachiopoda, Phoronida, Brachiopoda + Phoronida, Bivalvia, Ectoprocta, and Gastropoda are all monophyletic groups. Except for the position of the Clitellata (Oligochaeta + Hirudinida), the same topology was obtained with the twoparameter method of Kimura (1980) (resuits not shown). The SRC method (Van de Peer et al., 1996a) confirms all conclusions derived on the basis of JC distances, except for nemertean monophyly. Instead, Prostoma eilhardi branches off first within the eutrochozoan cluster.
The MP analysis yielded two equally parsimonious trees. Yet, with this large number of taxa (N=57), only a very small fraction of all possible topologies could be examined (45 106 rearrangements; >1099 possible rooted trees). Hence, it is not guaranteed that these are indeed the shortest possible trees. The strict consensus topology (not shown) confirms all significantly supported nodes of Figure 3, except for the bivalve monophyly and the existence of a Protula-Capitella clade. Moreover, there is no longer significant bootstrap support for the monophyly of Clitellata, Hirudinida, the Phyllodocida group including Harmothoe, Aphrodita, and Nereis, Brachiopoda, and for the sister group relationship of Priapulida to Arthropoda. Yet, as only a heuristic search was possible, one has to regard these results with caution. Table 2 compares the hypotheses on protostome worm phylogeny, derived from the limited and more extended 18S rRNA data set.
Re-analysis of Hanelt et al. (1996)
Figure 4 shows the NJ tree obtained on the basis of the alignment of Hanelt et al. (1996) using the SRC method (Van de Peer et al., 1993, 1996a). In contrast to the findings of Hanelt et al. (1996), it clusters Dicyema with Rhopalura, yet without significant bootstrap support. Surprisingly, Convoluta naikensis is not clustered with the other flatworms but rather with the Myxozoa. Yet, doubt about the classification was already previously expressed (Hanelt et al., 1996) as well as the possibility of an artificial clustering (Katayama et al., 1995). Although Hanelt and coworkers also considered among-site rate variation, by applying a discrete approximation to the gamma distribution (Yang, 1994), our approach is quite different (see Van de Peer et al., 1993; 1996a), which may account for the different tree topologies. In an analysis on the same taxa but aligned according to Van de Peer et al. (1998), all nodes with significant bootstrap support were confirmed (not shown) while Dicyema and Rhopalura were once again clustered together.
In a second approach, we used a more extended data set including a broad range of relevant organisms including additional platyhelminthes, nematodes, Myxozoa and Mesozoa. Figure 5 shows a NJ tree of 78 complete or nearly complete animal 18S rRNA sequences also based on the SRC method. The first diverging lineages are formed by the successive emergence of poriferans-which appear to be paraphyletic by including ctenophores-cnidarians, Placozoa and the Myxozoa, which are characterized by long branch lenghts. Myxozoa are clearly separated from the triploblastic animals (Bilateria). Nematodes are monophyletic and form the first diverging cluster within the Bilateria. Mesozoa branches off after the nematodes, but without significant bootstrap support. The platyhelminthes are monophyletic, but this is not supported by bootstrap analysis. The Mesozoa are monophyletic, but only with a bootstrap value of 63%. Yet, the same result was consistently found in other analyses using different sets of sequences (results not shown). When either Dicyema or Rhopalura were omitted from the analysis, the remaining mesozoans usually diverged from within the flatworms, albeit with a low bootstrap value. Although the SRC method has shown to be less sensitive to unequal evolutionary rates than classical methods (e.g., Van de Peer et al., 1996a, b; Van de Peer and De Wachter 1997b), we are not sure whether it can completely avoid the attraction effect of very long branches. Therefore, as the branches leading to Myxozoa are extremely long (Fig. 5) and as the Myxozoa always remain clearly separated from the Mesozoa, we assessed the effect of removing them from our analysis (Fig. 6). When doing so, Nematoda branch off before the other Bilateria. Mesozoa form a monophyletic grouping and cluster with the flatworms, although only weakly supported by bootstrap analysis (bootstrap value of 63%). The platyhelminthes form a monophyletic lineage, but this is not supported by bootstrap analysis either. However, when the flatworm Planocera is omitted from the analysis (not shown), bootstrap values rise to over 75%. In large trees containing hundreds of 18S rRNA sequences of different eukaryotic "crown" taxa (Knoll, 1992), in addition to representatives of most metazoan lineages, Dicyema and Rhopalura cluster consistently with the flatworms at a bootstrap level of >80%, when rate variation among sites is considered (Van de Peer and De Wachter, 1997b). When the flatworms are not included in the distance analyses, the Mesozoa tend to form independent metazoan evolutionary lineages (results not shown). An overview of the hypotheses obtained on mesozoan evolution by Hanelt et al. (1996) and in the present study, is given in Table 3.
Re-analysis of the data of Kim et al. (1996)
As expected, the inclusion of recently available 18S rRNA sequences and the use of more advanced distance estimation methods may influence phylogenetic results. The present analyses suggest that part of the conclusions of Kim et al. (1996) may have been biased by limited taxon sampling. Our more extended data set confirmed the finding of Kim et al. (1996) on arthropod monophyly, a finding which is also in agreement with several previous 18S rRNA studies (e.g. Turbeville et al., 1991; Abele et al., 1992) and other molecular studies (e.g., Ballard et al., 1992; Boore et al., 1995). Yet, in contrast to the conclusions of Kim et al. (1996), there is no convincing evidence for annelid or eutrochozoan monophyly. Also previous 18S rRNA studies are at variance with the results of Kim et al. (1996) and have reported on the inability of 18S rRNA to give persuasive evidence on these issues (e.g., Adoutte and Philippe, 1993; Winnepenninckx et al., 1995a, b, 1996; Mackey et al., 1996; Maley and Marshall, 1998). We will not expand on these points, but rather wish to point the reader to the possible consequences of omitting relevant sequences. The need to include as many sequences as possible was already previously pointed out by Lecointre et al. (1993). Of course, the number of taxa included in an analysis is sometimes constrained by limits imposed by computer time and software (e.g., MP and ML methods). Moreover, the inclusion of many taxa makes it more difficult to find optimal solutions. There is no way to decide whether enough taxa are considered to allow a definitive conclusion to be formulated. Yet, it is instructive to include all relevant sequences and to re-assess previous problems as soon as more sequences become available. We also observed an alignment effect (Winnepenninckx and Backeljau, 1996), when comparing the trees of Figure 1 with the results of Kim et al. (1996). Of course, it is often difficult to say when an alignment is superior to another and therefore, we do not claim that the topologies of Figure 1 are more reliable than those of Kim et al. (1996).
Re-analysis of the data of Winnepenninckx et al. (1995a)
A taxon sampling effect was also shown by the re-assessment of the Winnepenninckx et al. (1995a) study. Our extended data set (Fig. 3) confirmed most of their previous findings: 1) Pogonophora and Vestimentifera are a monophyletic group belonging to the Eutrochozoa cluster; 2) there is no indication for a common ancestor of Pogonophora and Vestimentifera with the Annelida; 3) there is no indication for a sister-group relationship of Sipuncula and Echiura; 4) Nemertea are coelomates; 5) protostome eutrochozoans form a clade, albeit not with significant bootstrap support and 6) there is a lack of support for a sister group relationship between the Arthropoda and the protostome eutrochozoans. As to the exact position of the Arthropoda, results remain indecisive. All our present findings are also in agreement with the results of several other studies, listed by Winnepenninckx et al. (1995a). Moreover, the monophyly of Vestimentifera and Pogonophora is in agreement with the findings of Southward (1993) and Rouse and Fauchald (1995) and the protostomous nature of Pogonophora and Vestimentifera agrees with previous findings on the basis of elongation factor-lo sequences (McHugh, 1997; Kojima, 1998), although their conclusions on the annelid relationship of Vestimentifera could not be confirmed. The conclusion of Winnepenninckx et al. (1995a) on the sister group relationship of Echiura to Pogonophora + Vestimentifera, is not confirmed by the present analyses. Therefore, we suspect that this result of Winnepenninckx et al. (1995a) must have been due to a taxon sampling effect.
Surprisingly, although nemertean monophyly is supported by the presence of the highly specialized and unique proboscis (rhynchocoel), their monophyly is not significantly supported by bootstrap analysis in our trees.
The present study provides independent support for some previously expressed hypotheses on annelid relationships: 1) Clitellata is monophyletic; 2) Hirudinida and Oligochaeta each have a monophyletic origin; 3) the two hirudinid subclasses Branchiobdellida and Hirudinea are each monophyletic; 4) there is no indication for the monophyly of the Annelida or the Polychaeta. Our study confirms the monophyly of the Clitellata, which is now generally accepted and is supported by several morphological and anatomical characters, such as the presence of a clitellum, absence of free larvae, a fixed number of body segments and some features of the nervous and reproductive system (e.g., Dales, 1967; Clark, 1969; Brusca and Brusca, 1990; Meglitsch and Schram, 1991; Purschke, 1997). Our finding on the existence of a separate Hirudinida and Oligochaeta clade is corroborated by the fact that hirudinids differ from Oligochaeta by, for example, the reduction of the coelom, loss of septa, absence of chaeta, the development of circumoral and posterior suckers, a dorsal subterminal anus, a muscular sucking pharynx and jaws, and the relative position of the testes and ovaries (e.g., Clark, 1969; Brusca and Brusca, 1990; Purschke, 1997). 18S rRNA yields independent support for the idea that polychaetes may be non-monophyletic but have arisen from several ancestral stocks (e.g., Meglitsch and Schram, 1991; Westheide, 1997) and that Clitellata arose independently from polychaetes (e.g., Sawyer, 1986). Also Brinkhurst (1982) indicated that oligochaetes may have arisen independently from polychaetes. Furthermore, our study is not the first that questions annelid monophyly. Annelid monophyly has been questioned previously by e.g., Sawyer (1986), Purschke (1997) and Westheide (1997). Finally, our analyses indicated that none of the three polychaete orders of which more than one taxon is present (Phyllodocida, Sabellida, Spionida), is monophyletic. This may be of interest to the current discussion about polychaete taxonomy and classification (e.g., Fauchald, 1974; Rouse and Fauchald, 1995; Meglitsch and Schram, 1991; Westheide, 1997). Our findings on the polyphyly of Polychaeta and of the order Phyllodocida and on the monophyly of Clitellata, Hirudinida and Oligochaeta are also in agreement with the results of Kojima (1998) on the basis of elongation factor1(alpha). However, one has to keep in mind that our findings on annelid and polychaete polyphyly are only weakly supported by bootstrap analysis and therefore may also have been caused by a lack of information in the 18S rRNA molecule. As annelids and polychaetes most probably arose during the explosive Cambrian radiation (e.g., Fauchald, 1974), this may indeed be a plausible explanation. Moreover, although the current data set is more extensive than previous analyses, the coverage of available relevant taxa remains defective. Therefore, it remains questionable whether the current 18S rRNA based results will stand time. Re-analysis of the data of Hanelt et al. (1996)
Unlike the results of Pawlowski et al. (1996) and Hanelt et al. (1996), our analyses do not support a different origin for orthonectids and rhombozoids. On the contrary, there is even some weak evidence for a monophyletic origin of the Mesozoa. Furthermore, of all metazoan taxa included, the Mesozoa seem to be most closely related to the flatworms. The Mesozoa always remain clearly separated from the Myxozoa and nematodes. Although our results using the SRC method are quite different from those of Pawlowski et al. (1996) and Hanelt et al. (1996), we do not claim that they are superior. A common origin for Mesozoa and flatworms was also confirmed by a recent study where we compared 500 SSU rRNA sequences of eukaryotes belonging to the so-called "crown" taxa (Van de Peer and De Wachter, 1997b). These findings seem to support the old view that Mesozoa and flatworms share a common ancestor and that the former are most probably secondarily simplified organisms, degraded as a result of parasitism (see Stunkard, 1954 and references cited there). To our knowledge, the only other molecular marker that has been applied to mesozoan phylogeny is SS rRNA, a molecule that is rarely used these days to study phylogenetic relationships since it is generally regarded as too short to infer sufficient and reliable phylogenetic information (e.g., Halanych, 1991; Steele et al., 1991). The 5S rRNA studies either suggested that the dicyemids form the first diverging animal lineage (Ohama et al., 1984; Hendriks et al., 1986) or did not find Dicyema clustered with the other animals at all (Hori and Osawa, 1986). Apart from the limited information content of SS rRNA, the reliability of these findings can be doubted due to the tree construction algorithms used. In the mid-eighties the most current methods to infer phylogenetic trees were clustering methods such as UPGMA and WPGMA, which are particularly sensitive to unequal evolutionary rates in different lineages and as seen in Figures 4-6, the animal kingdom is characterized by large differences in evolutionary rate. NJ analyses consistently cluster Dicyema and flatworm SS rRNAs together.
Of course, we do not claim that our trees in Figures 4-6 give a better picture of metazoan evolution than previous trees. Yet, in contrast to previous trees, the SRC method considers rate variation among sites, which is particularly important when evolutionary distances between sequences are large. Indeed, the larger the evolutionary distances, the more these are underestimated when an inappropriate substitution model is used to convert dissimilarity into distance (Van de Peer et al., 1996a). It is known that substitution rates among individual nucleotides in ribosomal RNA differ considerably (Gutell et al., 1985; Hillis and Dixon, 1991; Egebjerg et al., 1990; Van de Peer et al., 1993, 1996c). Ignoring this rate variation in the computation of evolutionary distances can lead to serious artifacts in tree topology (Olsen, 1987; Van de Peer et al., 1996a, b). Katayama et al. (1995) already suspected that long branch instabilities and long branch attraction were obscuring the phylogenetic position of the Mesozoa, since in their trees all long branches consistently clustered together. In trees we constructed on the basis of JC (1969) distances (not shown), we noticed the same phenomenon and Mesozoa clustered with the nematodes or formed an independent lineage close to the nematodes. Furthermore, as in the analyses of Pawlowski et al. (1996) and Hanelt et al. (1996), orthonectids and rhombozoids often diverged separately and formed independent evolutionary lineages. This was also the case in MP trees we inferred on the basis of our sequence alignment (not shown).
On the basis of the present re-analyses several hypotheses on metazoan relationships emerged. However, as our analyses were mainly meant to demonstrate the existence of a taxon sampling effect and to test the influence of using the SRC method, no additional sequences were determined and we included only the available relevant 18S rRNA sequences. Still others may be missing and therefore our hypotheses have to be considered to be preliminary ones.
The relationships among the eutrochozoan phyla and the monophyly of several metazoan phyla and classes still remain uncertain. It is indeed generally acknowledged that the phylogenetic relationships and divergence order between animal taxa are hard to resolve (e.g., Adoutte and Philippe, 1993; Winnepenninckx et al., 1995a, 1996; Mackey et al., 1996; Pawlowski et al., 1996; Maley and Marshall, 1998). The main reasons are (1) the short internodes between most of the animal phyla, which may be due to a massive radiation of new evolutionary lineages within a relatively small time interval during the Cambrian (Erwin, 1991; Adoutte and Philippe, 1993; Philippe et al., 1994; Ayala et al., 1998; but see Wray et al., 1996 for a different opinion) and (2) the high variation in evolutionary rates between lineages. Although not superior to previous results, the present study shows clearly that taxon sampling may have an effect on phylogeny inference. Therefore, as previously suggested (Lecointre et al., 1993; Milinkovitch et al., 1996), the addition of more sequences may be expected to further stabilize the animal tree. Finally, it is shown that the use of more accurate substitution models in the inference of evolutionary trees may also have an influence on the results and can perhaps help in resolving complex phylogenies.
We are indebted to Dr. Kim who kindly provided their alignment, to An Caers for sequence alignment and to two anonymous referees for giving their valuable comments on this work. This study was supported by FWO grant 2.0023.94. Y. Van de Peer is a Postdoctoral Fellow of the Fund for Scientific Research - Flanders.
Abele, L. G., T Spears, W. Kim, and M. Applegate. 1992. Phylogeny of selected maxillopodan and other crustacean taxa based on 18S ribosomal nucleotide sequences: A preliminary analysis. Acta Zool. 73:373-382.
Adoutte, A. and H. Philippe. 1993. The major lines of metazoan evolution: Summary of traditional evidence and lessons from ribosomal RNA sequence analysis. In Y. Pichon (ed.), Comparative molecular neurobiology, pp. 1-30. Birkhauser Verlag, Basel.
Ayala, E J., A. Rzhetsky, and E J. Ayala. 1998. Origin of the metazoan phyla: Molecular clocks confirm paleontological estimates. Proc. Natl. Acad. Sci. U.S.A. 95:606-611.
Ballard, J. W., G. J. Olsen, D. P Faith, W. A. Odger, D. M. Powell, and P W. Atkinson. 1992. Evidence from 12S ribosomal RNA sequences that ony
chophorans are modified arthropods. Science 258: 1345-1348.
Bartolomaeus, T 1988. No direct contact between the excretory system and the circulatory system in Prostomatella arenicola Friedrich (Hoplonemertini). Hydrobiologia 156:175-181.
Berg, G. 1985. Annulonemertes gen. Nov., a new segmented hoplonemertean. In Conway Morris, S., J. D. George, R. Gibson, and H. M. Platt (eds.), The orgins and relationships of lower invertebrates, pp. 200-209. Clarendon Press, Oxford.
Boore, J. L., T M. Collins, D. Stanton, L. L. Daehler, and W. M. Brown. 1995. Deducing the pattern of arthropod phylogeny from mitochondrial DNA rearrangements. Nature 376:163-165.
Brinkhurst, R. O. 1982. Evolution in the Annelida. Can. J. Zool. 60:1043-1059.
Brusca, R. C. and G. J. Brusca. 1990. Invertebrates. Sinauer Associates, Sunderland.
Cavalier-Smith, T 1993. Kingdom Protozoa and its 18 phyla. Microbiol. Rev. 57:953-994.
Cavalli-Sforza, L. L. and A. W. E Edwards. 1967. Phylogenetic analysis: Models and estimation procedures. Amer. J. Hum. Genet. 19:233-257.
Cedergen R., M. W. Gray, Y. Abel, and D. Sankoff. 1988. The evolutionary relationships among known life forms. J. Mol. Evol. 28:98-112.
Clark, R. B. 1969. Systematics and phylogeny: Annelida, Echiura and Sipuncula. In Florkin, M. and B. T Scheer (eds.), Chemical zoology Vol. IV, pp. 1-68. Academic Press, New York.
Conway Morris, S. 1993. The fossil record and the early evolution of the Metazoa. Nature 361:219225.
Conway Morris, S., B. L. Cohen, A. Gawthrop, T Cavalier-Smith, and B. Winnepenninckx. 1996. Lophophorate phylogeny. Science 272:282.
Dales, R. P 1967. Annelids. Hutchinson, London. Doolittle, W. F and J. R. Brown. 1994. Tempo, mode, the progenote, and the universal root. Proc. Natl. Acad. Sci. USA 91:6721-6728.
Egebjerg, J., N. Larsen, and R. A. Garrett. 1990. Structural map of 23S rRNA. In W. E. Hill, A. Dahlberg, R. A. Garrett, P B. Moore, D. Schlessinger and J. R. Warner (eds.), The ribosome: structure, function, and evolution, pp. 168-179. American Society for Microbiology, Washington, D.C. Erwin, D. H. 1991. Metazoan phylogeny and the cam
brian radiation. Trends Ecol. Evol. 6:131-134. Fauchald, K. 1974. Polychaete phylogeny: A problem in protostome evolution. Syst. Zool. 23:493-506. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39: 783-791.
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.
Garey, J. R., T. J. Near, M. R. Nonnemacher, and Nadler, S. A. 1996. Molecular evidence for Acanthocephala as a subtaxon of Rotifera. J. Mol. Evol. 43:287-292.
Godeaux, J. E. A. 1974. Introduction to the morphology, phylogenesis and systematics of lower Deu
terostomia. In Florkin, M. and B. T. Scheer (eds.), Chemical zoology: volume VIII, pp. 3-60. Academic Press, New York.
Gouy M. and W.-H. Li. 1989. Molecular phylogeny of the kingdoms Animalia, Plantae and Fungi. Mol. Biol. Evol. 6:109-122.
Gupta, B. L. and C. Little. 1975. Ultrastructure, phylogeny and Pogonophora. Z. zool. Syst. Evolut.forsch. Sonderheft 1:45-63.
Gutell, R. R., B. Weiser, C. R. Woese, and H. E Noller. 1985. Comparative anatomy of 16-S-like ribosomal RNA. Prog. Nucl. Acids Res. Mol. Biol. 32: 155-216.
Halanych, K. M. 1991. 5S ribosomal RNA sequences inappropriate for phylogenetic reconstruction. Mol. Biol. Evol. 8:249-253.
Halanych, K. M., J. D. Bacheller, A. M. A. Aguinaldo, S. M. Liva, D. M. Hillis, and J. A. Lake. 1995. Evidence from 18S ribosomal DNA that the lophophorates are protostome animals. Science 267: 1641-1643.
Hanelt, B., D. Van Schyndel, C. M. Adema, L. A. Lewis, and E. S. Loker. 1996. The phylogenetic position of Rhopalura ophiocomae (Orthonectida) based on 18S ribosomal DNA sequence analysis. Mol. Biol. Evol. 13:1187-1191.
Hasegawa M., T. Hashimoto, J. Adachi, N. Iwabe, and T. Miyata. 1993. Early branchings in the evolution of eukaryotes: Ancient divergence of Entamoeba that lacks mitochondria revealed by protein sequence data. J. Mol. Evol. 36:380-388. Hendriks, L., E. Huysmans, A. Vandenberghe, and R. De Wachter. 1986. Primary structures of the SS ribosomal RNAs of 11 arthropods and applicability of SS RNA to the study of metazoan evolution. J. Mol. Evol. 24:103-109.
Hillis, D. M. and M. T. Dixon. 1991. Ribosomal DNA: Molecular evolution and phylogenetic inference. Quart. Rev. Biol. 66:411-453.
Hillis, D. M. and J. J. Bull. 1993. An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Syst. Biol. 42: 182-192.
Hori, H. and S. Osawa. 1986. Evolutionary change in SS rRNA secondary structure and a phylogenetic tree of 325 SS rRNA species. BioSystems 19:163172.
Ivanov, A. 1955. On the assignment of the class Pogonophora to a separate phylum of Deuterostomia-Brachiata A. Ivanov. Phyl. Nov. Syst. Zool. 4:177-178.
Jin, L. and M. Nei. 1990. Limitations of the evolutionary parsimony method of phylogenetic analysis. Mol. Biol. Evol. 7:82-102. Jones, M. L. 1985. On the Vestimentifera, new phylum: Six new species, and other taxa, from hydrothermal vents and elsewhere. Biol. Soc. Wash. Bull. 6:117-158.
Jukes, T. H. and C. R. Cantor. 1969. Evolution of protein molecules. In H. H. Munro (ed.), Mammalian protein metabolism, pp. 21-132. Academic Press, New York.
Katayama, T, H. Wada, H. Furuya, N. Satoh, and M. Yamamoto. 1995. Phylogenetic position of the di
cyemid Mesozoa inferred from 18S rDNA sequences. Biol. Bull. 189:81-90. Kim, C. B., S. Y Moon, S. R. Gelder, and W. Kim. 1996. Phylogenetic relationships of annelids, molluscs, and arthropods evidenced from molecules and morphology. J. Mol. Evol. 43:207-215. Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16:111-120.
Knoll, A. H. 1992. The early evolution of eukaryotes:
A geological perspective. Science 256:622-627. Kojima, S. 1998. Paraphyletic status of Polychaeta suggested by phylogenetic analysis based on the amino acid sequences of elongation factor-la. Mol. Phyl. Evol. 9:255-261.
Lapan, E. A. and H. Morowitz. 1972. The Mesozoa. Sci. Am. 227:94-101.
Lecointre, G., H. Philippe, H. L. V. Le., and H. Le Guyader. 1993. Species sampling has a major impact on phylogenetic inference. Mol. Phyl. Evol. 2:205-224.
Mackey, L. Y., B. Winnepenninckx, R. De Wachter, T Backeljau, P. Emschermann, and J. R. Garey. 1996. 18S rRNA suggests that Entoprocta are protostomes, unrelated to Ectoprocta. J. Mol. Evol. 42:552-559.
Maley, L. E. and C. R. Marshall. 1998. The coming age of molecular systematics. Science 279:505506.
Margulis, L. and K. V. Schwartz (eds.). 1988. Five kingdoms. An illustrated guide to the phyla of life on Earth. Freeman, New York.
McHugh, D. 1997. Molecular evidence that echiurans and pogonophorans are derived annelids. Proc. Natl. Acad. Sci. U.S.A. 94:80068009. Meglitsch, P. A. and E R. Schram. 1991. Invertebrate zoology. Oxford University Press, New York. Milinkovitch, M. C., R. G. LeDuc, J. Adachi, E Farnir, M. Georges, and M. Hasegawa. 1996. Effects of character weighting and species sampling on phylogeny reconstruction: A case study based on DNA sequence data in Cetaceans. Genetics 144: 1817-1833.
Norenburg, J. L. 1985. Structure of the nemertine integument with consideration of its ecological and phylogenetic significance. Amer. Zool. 25:37-51. N6rrevang, A. 1970. The position of Pogonophora in the phylogenetic system. Z. zool. Syst. Evolut. -forsch. 8:161-172.
Ohama, T, T Kumazaki, H. Hori, and S. Osawa. 1984. Evolution of the multicellular animals as deduced from SS rRNA sequences: A possible early emergence of the Mesozoa. Nucleic Acids Res. 12: 5105-5108.
Olsen, G. J. 1987. Earliest phylogenetic branchings: Comparing rRNA-based evolutionary trees inferred with various techniques. Cold Spring Harbor Symposia on Quantitative Biology LII:825837.
Pawlowski, J., J.-I. Montoya-Burgos, J. E Fahrni, J. West, and L. Zaninetti. 1996. Origin of the Mesozoa inferred from 18S rRNA gene sequences. Mol. Biol. Evol. 13:1128-1132.
Philippe, H., A. Chenuil, and A. Adoutte. 1994. Can the Cambrian explosion be inferred through molecular phylogeny? Development Suppl.:15-25. Purschke, G. 1997. Ultrastructure of the nuchal organs in polychaetes (Annelida)-new results and review. Acta Zool. 78:123-143.
Raff, R. A. 1988. Ribosomal RNA sequences and the early history of the Metazoa. In B. Runnegar, and J. W. Schopf (eds.), Molecular evolution and the fossil record. Short courses in paleontology: Molecular evolution and the fossil record, Volume 1. pp. 63-74. The Paleontological Society, Knoxville, Tennessee.
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, pp. 247-261. Elsevier Science Publishers, Amsterdam.
Rice, M. E. 1985. Sipuncula: Developmental evidence for phylogenetic inference. In S. C. Morris, George, J. D., R. Gibson, and H. M. Platt (eds.), The origins, and relationships of lower invertebrates. Systematics Association, Clarendon Press, Oxford.
Rouse, G. W. and Fauchald, K. 1995. The articulation of annelids. Zool. Scripta 24:269-301. Rzhetsky, A. and M. Nei. 1994. Unbiased estimates of the number of nucleotide substitutions when substitution rate varies among different sites. J. Mol. Evol. 38:295-299.
Saitou, N. and M. Nei. 1987. The Neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406-425. Sawyer R. T. 1986. Leech biology and behavior. Clarendon Press, Oxford. Sneath, P H. A. and R. R. Sokal (eds.). 1973. Numerical taxonomy. W. H. Freeman. San Francisco.
Sogin, M. L., J. H. Gunderson, H. J. Elwood, R. A. Alonso, and D. A. Peattie. 1989. Phylogenetic meaning of the kingdom concept: An unusual ribosomal RNA from Giardia lamblia. Science 243: 75-77.
Solignac, M., M. Pelandakis, M. Rousset, and A. Cheneuil. 1991. Ribosomal RNA phylogenies. In Hewitt, G. M., A. W. B. Johnston, and J. P. W. Young (eds.), NATO ASI series, vol. H57, Molecular techniques in taxonomy, pp. 73-85. SpringerVerlag, Berlin, Heidelberg.
Southward, E. C. 1963. Pogonophora. Oceanogr. Mar. Biol. Annu. Rev. 1:405-428.
Southward, E. C. 1988. Development of the gut and segmentation of newly settled stages of Ridgeia (Vestimentifera): Implications for relationships between Vestimentifera and Pogonophora. J. Mar. Biol. Ass. U. K. 68:465-487.
Southward, E. C. 1993. Pogonophora. In E W. Harrison and M. E. Rice, (eds.), Microscopic anatomy of invertebrates, vol. 12. Onychophora, Chilopoda and lesser Protostomata, pp. 327-369. WileyLiss, New York.
Steele, K. P, K. E. Holsinger, R. K. Jansen, and D. W. Taylor. 1991. Assessing the reliability of SS rRNA
sequence data for phylogenetic analysis in green plants. Mol. Biol. Evol. 8:240-249. Stunkard, H. W. 1954. The life-history and systematic relations of the Mesozoa. Quart. Rev. Biol. 29: 230-244.
Stunkard, H. W. 1972. Clarification of taxonomy in the
Mesozoa. Syst. Zool. 21:210-214. Sullivan, J., K. E. Holsinger, C. Simon. 1996. The effect of topology on estimation of among-site rate variation. J. Mol. Evol. 42:308-312. Swofford, D. L. 1993. PAUP: Phylogenetic Analysis Using Parsimony, version 3.1. Illinois Natural History Survey, Champaign.
Tourasse, N. J. and M. Gouy. 1997. Evolutionary distances between nucleotide sequences based on the distribution of substitution rates among sites as estimated by parsimony. Mol. Biol. Evol. 14:287298.
Turbeville, J. M. 1986. An ultrastructural analysis of coelomogenesis in the hoplonemertine Prosorhochmus americanus and the polychaete Magelona sp. J. Morphol. 187:51-60. Turbeville, J. M. and E. E. Ruppert. 1985. Comparative ultrastructure and the evolution of nemertines. Amer. Zool. 25:53-71.
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 the phylum Nemertini, inferred from 18S rRNA sequences: Molecular character homology. Mol. Biol. Evol. 9:235-249. Van de Peer Y., A. Caers, P. De Rijk, and R. De Wachter. 1998. Database on the structure of small ribosomal subunit RNA. Nucleic Acids Res. 26: 179-182.
Van de Peer, Y., S. Chapelle, and R. De Wachter 1996c. A quantitative map of nucleotide substitution rates in bacterial ribosomal subunit RNA. Nucl. Acids Res. 24:3381-3391.
Van de Peer, Y., and R. De Wachter. 1997a. Construction of evolutionary distance trees with TREECON for Windows: Accounting for variation in nucleotide substitution rate among sites. Comput. Applic. Biosci. 13:227-230.
Van de Peer, Y., and R. De Wachter. 1997b. Evolutionary relationships among the eukaryotic crown taxa taking into account site to site rate variation in 18S rRNA. J. Mol. Evol. 45:619-631. Van de Peer, Y., J.-M. Neefs, P. De Rijk, and R. De Wachter. 1993. Reconstructing evolution from eukaryotic small-ribosomal-subunit RNA sequences: Calibration of the molecular clock. J. Mol. Evol. 37:221-232.
Van de Peer, Y., S. Rensing, U.-G. Maier, and R. De Wachter 1996b. Substitution rate calibration of small ribosomal subunit RNA identifies chlorarachniophyte endosymbionts as remnants of green algae. Proc. Natl. Acad. Sci. USA 93:7732-7736. Van de Peer, Y., I. Van den Broeck, P. De Rijk, and R. De Wachter. 1994. Database on the structure of small ribosomal subunit RNA. Nucl. Acids Res. 22:3488-3494.
Van de Peer, Y., G. Van der Auwera, and R. De Wachter. 1996a. The evolution of stramenopiles and alveolates as derived by "substitution rate calibration" of small ribosomal subunit RNA. J. Mol. Evol. 42:201-210.
Van der Land, J. and A. Norrevang. 1975. The systematic position of Lamellibrachia (Annelida, Vestimentifera). Z. zool. Syst. Evolut. -forsch. Sonderheft 1:86101.
Wakeley, J. 1993. Substitution rate variation among sites in hypervariable region 1 of human mitochondrial DNA. J. Mol. Evol. 37:613-623. Webb, M. 1964. Evolutionary pathways within the phylum Pogonophora. Sarsia 16:59-64. Westheide, W. 1997. The direction of evolution within Polychaeta. J. Nat. Hist. 31:1-15. Willmer, P 1990. Invertebrate relationships. Cam
bridge University Press, Cambridge. Winnepenninckx, B. and T Backeljau. 1996. 18S rRNA alignments derived from different secondary structure models can produce alternative phylogenies. J. Zool. Syst. Evol. Res. 34:135-143. Winnepenninckx, B., T Backeljau, and R. De Wachter. 1995a. Phylogeny of protostome worms derived from 18S rRNA sequences. Mol. Biol. Evol. 12: 641-649.
Winnepenninckx, B., T. Backeljau, and R. De Wachter. 1996. Investigation of molluscan phylogeny on the basis of 18S rRNA sequences. Mol. Biol. Evol. 13:13061317.
Winnepenninckx, B., T. Backeljau, L. Y. Mackey, J. M. Brooks, R. De Wachter, S. Kumar, and J. R. Garey. 1995b. 18S rRNA data indicate that aschelminthes are polyphyletic in origin and consist of at least three distinct clades. Mol. Biol. Evol. 12:1132-1137.
Woese, C. R. 1987. Bacterial evolution. Microbiol. Rev. 51:221-271.
Woese, C. R. 1991. The use of ribosomal RNA in reconstructing evolutionary relationships among bacteria. In Selander, R. K., A. G. Clark and T. S. Whittam (eds.), Evolution at the molecular level, pp. 1-25. Sinauer Associates Inc., Sunderland, Massachusetts.
Woese C. R. and G. E. Fox. 1977. Phylogenetic structure of the prokaryotic domain: the primary kingdoms. Proc. Natl. Acad. Sci. U.S.A. 4:5088-5090. Wray, G. A., J. S. Levinton, and L. H. Shapiro. 1996. Molecular evidence for deep precambrian divergences among metazoan phyla. Science 274:568573.
Yang Z. 1994 Maximum likelihood pylogenetic estimation from DAN sequences with variables rates over sites approximate methods J mol Evol. 39. 306-314
Yang, Z. 1996. Among-site rate variation and its impact on phylogenetic analyses. Trends Ecol. Evol. 11:367-372.
Corresponding Editor: Douglas H. Erwin
Royal Belgian Institute of Natural Sciences, Vautierstraat 29, B-1000 Brussels, Belgium *Department of Biochemistry, University of Antwerp (UIA), Universiteitsplein 1, B-2610 Wilrijk, Belgium
I From the symposium Evolutionary Relationships of Metazoan Phyla: Advances, Problems, and Approaches presented at the Annual Meeting of the Society for Integrative and Comparative Biology 3-7 January 1998, at Boston, Massachusetts
2 E-mail: firstname.lastname@example.org.
Copyright Society for Integrative and Comparative Biology Dec 1998
Provided by ProQuest Information and Learning Company. All rights Reserved