Polychaeta, Annelida, and Articulata are not monophyletic ......(1998) suggested that the data used by MCHUGH (1997) were insufficient to decide the position of Echiura and Pogonophora,

Revista Brasileira de Zoologia 20 (1): 23–57, março 2003

The phylogenetic history of the polychaetes is long and diverse.The first efforts may be characterized as strictly taxonomical.Various authors have differed in the emphases given to theimportance of general body shape and feeding habits, someattention being provided to the anatomy of particular structures(LINNAEUS 1758; LAMARCK 1801; CUVIER 1816; BLAINVILLE 1816;AUDOUIN & MILNE EDWARDS 1832; GRUBE 1850; QUATREFAGES 1866).The first evolutionary proposal for the group is that of HATSCHEK(1878, 1893). He expounds a progressive view, which starts withthe simpler archiannelids.

More recently, DALES (1962, 1963) contributed withevolutionary considerations based on the structure of the

pharynx. The system is rather conservative and maintains thegeneral lines of HATSCHEK (1878, 1893), with archiannelids atthe base. CLARK (1963, 1964) innovated considerably byinterpreting the origin of segmentation as an adaptation forexcavation. He proposed an oligochaetoid basic plan andcontested either the unity or the basal position of thearchiannelids. HERMANS (1969), ORRHAGE (1974), TRUEMAN (1975),ALÓS (1982), PURSCHKE (1985a, b, 1987a, b, 1988a), and PURSCHKE& JOUIN (1988) have corroborated this latter point.

MILEIKOVSKY (1968, 1977) proposed a new classificationof the Polychaeta Grube, 1850 based largely on larval traits. Init he incorporated some of the ideas of DALES (1962, 1963).

PolPolPolPolPolychaeta,ychaeta,ychaeta,ychaeta,ychaeta, Annelida,Annelida,Annelida,Annelida,Annelida, and and and and and Articulata arArticulata arArticulata arArticulata arArticulata are not monophe not monophe not monophe not monophe not monophyletic:yletic:yletic:yletic:yletic:articulating the Metamerarticulating the Metamerarticulating the Metamerarticulating the Metamerarticulating the Metameria (Metazia (Metazia (Metazia (Metazia (Metazoa,oa,oa,oa,oa, Coelomata) Coelomata) Coelomata) Coelomata) Coelomata)

Waltécio de Oliveira Almeida 1, Martin Lindsey Christoffersen 1, Dalton de Souza Amorim 2,André Rinaldo Senna Garraffoni 3 & Gustavo Sene Silva 3

1 Departamento de Sistemática e Ecologia, Universidade Federal da Paraíba. 58051-900 João Pessoa, Paraíba, Brasil.Respective e-mails: [email protected] and [email protected] Departamento de Biologia, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo.14040-901 Ribeirão Preto, São Paulo, Brasil. E-mail: [email protected] Centro de Estudos do Mar, Universidade Federal do Paraná. 83255-000 Pontal do Paraná, Paraná, Brasil.Respective e-mails: [email protected] and [email protected]

ABSTRACT. Polychaetes are metameric worms recognized for having parapodia, chaetae, and nuchal organs.Some authors have extended the Annelida to include Pogonophora, Echiura, and Clitellata. These suggestions areinsufficient to generate a monophyletic group. They do not take into account two very large and importantclades that in a cladistic analysis at a higher level are shown to be nested within the Annelida: the Ecdysozoa(arthropods and related taxa) and Enterocoela (deuterostomes and related taxa). Evolutionary histories of mostcharacters across metazoan phyla are still very poorly known. Metameres and coeloms have been consideredhomoplastic in the literature, and yet the homeobox genes responsible for the expression of metamerism and ofpaired appendages, at least, are very largely distributed among the Metazoa. A phylogenetic analysis was performedfor the ingroups of Polychaeta, including Clitellata, Enterocoela, and Ecdysozoa as terminal taxa. The remainingnon-metameric phyla Platyhelminthes, Nemertea, Mollusca, and Sipuncula were included to root the tree withinthe Bilateria. Empirical data was obtained from the literature and run with the software Hennig86 with twocomparative interpretations of a priori hypotheses of primary homology: one with negative characters (codinglosses) and another considering only positive characters (without assumptions about losses). The most relevantconclusions are: (1) Annelida and Polychaeta are non-monophyletic, even when including Echiura, Clitellata, andPogonophora; (2) Articulata, as traditionally circumscribed for Annelida and Arthropoda, is also not monophyletic;(3) Metameria becomes monophyletic only when Ecdysozoa and Enterocoela are included in addition to thetraditional annelid taxa; (4) Ecdysozoa are the sister group of Aphrodita; (5) Clitellata are related to deposit-feeding sedentary polychaetes (scolecids), and Questidae represent their sister group; (6) Owenia plus Enterocoelaform a monophyletic group related to the tubicolous polychaetes.

KEY WORDS. Bilateria, Clitellata, Ecdysozoa, Enterocoela, Metazoa, phylogeny, Polychaeta.

24 Almeida et al.


STORCH (1968) introduced further changes. After analyzing themuscular pattern of several polychaetes, he pointed to anaphroditid-like groundplan for the group. This was contraryto CLARK (1963, 1964), who maintained a simple groundplanwithout locomotory or sensory appendages. Polychaete familiescontinued to be treated as natural units without any detectableinterrelationships by HARTMAN (1968, 1969). METTAM (1971)reexamined the muscular ultrastructure of Aphrodita aculeataLinnaeus, 1761 and concluded that this taxon was specialized,contrary to STORCH’s (1968) position.

FAUCHALD (1974, 1977) proposed non-arbitrary familyrelationships for the first time. He positioned some of thescolecids, such as Capitellidae Grube, 1862, OrbiniidaeHartman, 1942 and Questidae Hartman, 1966, at the base ofhis system. PILATO (1981) resurrected STORCH’s (1968) ideas onthe primitiveness of the errant polychaetes and suggested thatthey could be derived directly from a flatworm-like ancestor.

CHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA (1994) first proposedthe paraphyletic nature of the polychaetes. They indicated thatEnterocoela Huxley, 1875 (Pogonophora Ivanov, 1949, thelophophororates, and Deuterostomia Huxley, 1875) would bethe sister group of part of the Polychaeta, most probably ofthe Oweniidae Rioja, 1917. NIELSEN (1995) also interpretedPolychaeta as paraphyletic, but suggesting that Pogonophora,Echiura Newby, 1940, Lobatocerebridae Rieger, 1980,Gnathostomulida Ax, 1956, and Sipuncula Sedgwick, 1898 alsobelong to this taxon.

ROUSE & FAUCHALD (1995) argued for the monophyly ofthe Polychaeta. The presence of nuchal organs was chosen astheir main autapomorphy. They questioned the validity of theAnnelida Lamarck, 1801, but EIBYE-JACOBSEN & NIELSEN (1996)did not agree. The latter authors argument that the inclusionof groups such as Clitellata Michaelsen, 1928 and Pogonophorashould be made a priori. Otherwise Polychaeta would becomeparaphyletic. However, ROUSE (1997) criticized the supposedlyunnecessary and non-parsimonious assumptions that such aprocedure would entail. He suggested that it might be preferableto analyze paraphyletic groups than to arbitrarily includequestionable taxa such as Pogonophora into the analysis,regardless of whether the final results ended with more or fewersteps.

WESTHEIDE (1997) questioned the validity of thePolychaeta, discussing different possibilities for the origin ofsegmentation and parapodia. He argued that Clitellata is aningroup of the Polychaeta and indicated a probable homologyof parapodia and arthropod legs. ROUSE & FAUCHALD (1997)published a detailed phylogeny of the families of polychaetes,with an extensive discussion of characters. The evidence forthe monophyly of some terminal taxa was inconclusive.

PURSCHKE (1997) demonstrated that nuchal organs areabsent in several polychaetes specialized for a terrestrial wayof life (e.g., Hrabeiella periglandulata Pizl & Chalupský, 1984,Parergodrilus heideri Reisinger, 1925, and Stygocapitella subterraneaKnöllner, 1934). He speculated that a similar adaptation mayhave occurred in the Clitellata.

MCHUGH (1997) did not accept the monophyletic statusof the Polychaeta. Her molecular data indicated that Echiurabelongs to this group. Furthermore, she dissented from ROUSE& FAUCHALD (1997) by placing Harmothoe Kinberg, 1855 andNereis Linnaeus, 1758 near the base of the cladogram, and byproducing phylogenetic trees that are distinctly incongruent

from those of the latter authors in several details. SIDDAL et al.(1998) suggested that the data used by MCHUGH (1997) wereinsufficient to decide the position of Echiura and Pogonophora,and thus inadequate as a basis to reclassify the Annelida. Thiscriticism has been vigorously rebutted by MCHUGH (1999).

KOJIMA (1998) used molecular data to position theClitellata among the polychaetes. Once again, the molecularresults were totally incongruent with those obtained frommorphology by ROUSE & FAUCHALD (1997).

GIANGRANDE & GAMBI (1998) developed a hypothesis forthe origin of the Polychaeta that is similar to that of HATSCHEK(1878, 1893). They considered both the presences of a standardtrochophore larva and the post-larval development in thearchiannelid Polygordius Schneider, 1868 as possible indicationsof primitiveness. Consequently, they interpreted the polychaeteancestor as being interstitial and polygordiid-like, from whichall the remaining polychaetes would have been derived.

ROUSE & FAUCHALD (1998) revised the heuristic value oftheir previous cladistic results (ROUSE & FAUCHALD 1995, 1997).They concluded that the alternative proposals of WESTHEIDE(1997), MCHUGH (1997), PURSCHKE (1997), and GIANGRANDE &GAMBI (1998) required an additional set of aprioristic assertionsthat exceeded the present knowledge available for thePolychaeta and Annelida. These conclusions have beenreaffirmed in the synthesis of WESTHEIDE et al. (1999).

Although ROUSE & FAUCHALD (1998) and ROUSE (inWESTHEIDE et al. 1999) persist in defending the monophyly ofthe Polychaeta, new contrary evidence was presented byHESSLING & WESTHEIDE (1999). These authors argued that thedevelopment of the supraoesophageal ganglion of Enchytraeuscrypticus Westheide & Graefe, 1992 clearly demonstrated thatthe Clitellata were derived from a polychaetoid pattern anddid not represent a primitive condition as predicted by thecladistic hypothesis of ROUSE & FAUCHALD (1995, 1997).

PURSCHKE (1999) investigated several further adaptiveapomorphies occurring in the specialized terrestrial polychaetes(Hrabeiella periglandulata, Parergodrilus heideri and Stygocapitellasubterranea). These adaptive traits included the simplificationor loss of prostomial appendages, parapodia, and nuchalorgans. Further similarities between the terrestrial polychaetesand the clitellates were encountered in the supraoesophagealganglion, development patterns, and reproductive strategies.

BROWN et al. (1999) provide new molecular data fromhistone H3, U2 snRNA, and 28S rDNA in support of theinclusion of Echiura, Pogonophora, and Clitellata within thePolychaeta. As with MCHUGH (1997) and KOJIMA (1998), severalof their proposed clades are incongruent with those suggestedby ROUSE & FAUCHALD (1997).

A phylogenetic revision of the errant polychaetes ledALMEIDA & CHRISTOFFERSEN (2000) to argue once more for theparaphyly of the Polychaeta. MANTON (1967) produced the maintentative arguments against considering the parapodia ofpolychaetes homologous to the lobopodia and arthropodia ofthe Onychophora Guilding, 1826 and Arthropoda Siebold,1848, respectively. ALMEIDA & CHRISTOFFERSEN (2000) argumentedthat the musculature of Aphrodita Linnaeus, 1758 is alreadyquite advanced for walking on the substrate and thuspinpointed Aphrodita as the most likely sister group ofArthropoda and related groups. This enhances the problem ofthe paraphyly of the Polychaeta by requiring the interpretationthat the Ecdysozoa Aguinaldo, Turbeville, Linford, Rivera,

25Polychaeta, Annelida, and Articulata are not monophyletic...


Garey, Raff & Lake, 1997 represent an “annelid specialization”(ALMEIDA & CHRISTOFFERSEN 2000).

More recently MCHUGH (2000) reviewed the moleculardata pertaining to annelid phylogeny. Annelida and Polychaetawere reaffirmed to be non-monophyletic because of theexclusion of Echiura, Pogonophora, and Clitellata. Thesegroups are probably modified from polychaetes.

The contribution provided in this paper is twofold. First,the Ecdysozoa, Enterocoela, and Clitellata are placed togetherwith their closest relatives among the polychaetes. Second, anumber of characters that have been insufficiently consideredup to this date are discussed under a phylogenetic perspective.As a result, some novel transformation series are suggested.Moreover, it is hoped that some of the present hypotheses ofhomology of different structures proposed herein will have thepotential to change the consensual, old views on metazoanphylogeny in a significant way.

MATERIAL AND METHODSThe literature was used as the main source of empirical

data in this work. Therefore the results should be consideredpreliminary and dependent on future anatomical andmolecular studies for corroboration of the present hypothesesof homology. However, this work is not alone in usingdescriptive data in the literature as a logical first step forproducing broad phylogenies. For example, SCHRAM (1991) andROUSE & FAUCHALD (1995, 1997), among others, present resultsthat are broadly accepted or actively debated by the scientificcommunity.

The analyses included most families of Polychaeta andthe basal groups of Ecdysozoa (including AcanthocephalaKoelreuter, 1771, Arthropoda Siebold, 1848, GastrotrichaMetschnikoff, 1864, Gnasthostomulida Ax, 1956, KinorhynchaReinhard, 1887, Loricifera Kristensen, 1983, NematodaRudolphi, 1793, and Nematomorpha Vejdovsky, 1886,Onychophora Guilding, 1826, Pentastomida Rudolphi, 1819,Priapulida Delage & Herouard, 1897, Rotifera Cuvier, 1798, andTardigrada Spallanzani, 1776) (SCHMIDT-RHAESA et al. 1998;ALMEIDA & CHRISTOFFERSEN 2000: 19-21). Many lobopodian fossilswere considered as belonging to a more basal position inrelation to the arthropod lineage (MONGE-NÁJERA 1995; MONGE-NÁJERA & HOU 1999).

Characters of the Enterocoela Huxley, 1875 (includingBrachiopoda Cuvier, 1802, Cephalochordata Owen, 1846,Cephalodiscida Fowler, 1892, Chaetognatha Huxley, 1875,Echinodermata Klein, 1734, Ectoprocta Nitsche, 1869,Enteropneusta Huxley, 1875, Phoronida Wright, 1856,Pogonophora Ivanov, 1949, Rhabdopleurida Schmiketsch,1890, Tunicata Lamarck, 1815, and Vertebrata Cuvier, 1817)are based on CHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA (1994). Theprocedures described in CHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA(1994) were followed for organizing and analyzing the availabledescriptive information.

The main emphasis in this paper has been to provide ageneral framework at the higher levels of generality of theMetameria Christoffersen & Araújo-de-Almeida, 1994, in orderto reduce excessive convergence perceived in the traditionalschemes of metazoan phylogeny. For this reason, like ROUSE &FAUCHALD (1997), several interstitial, commensal, parasitic, andpelagic groups were not included in analyses: Aberrantidae

Wolf, 1987, Aelosomatidae Beddard, 1895, Alciopidae Ehlers,1864, Ctenodrilidae Kennel, 1882, Diurodrilidae Kristensen &Niilonen, 1982, Fauveliopsidae Hartman, 1971, HartmaniellidaeImajima, 1977, Histriobdellidae Vaillant, 1890, IchthyotomidaeEisig, 1906, Iospillidae Bergström, 1914, LopadorhynchidaeClaparède, 1868, Myzostomidae Benham, 1896, Nautillienel-lidae Miura & Laubier, 1990, Nerillidae Levinsen, 1883,Oenonidae Kinberg, 1865, Parergodrilidae Reisinger, 1925,Poeobiidae Heath, 1930, Polygordiidae Czerniavsky, 1881,Pontodoridae Bergström, 1914, Potamodrilidae Bunke, 1967,Protodrilidae Czerniavsky, 1881, Protodriloidea Purschke &Jouin, 1988, Psammodrilidae Swedmark, 1952, SaccocirridaeCzerniavsky, 1881, Sternaspidae Carus, 1863, SpintheridaeJohnston, 1865, Tomopteridae Johnston, 1865, Typhlosco-lecidae Uljanin, 1878, and Uncispionidae Green, 1982. Thesegroups are assumed to be highly derived, yet closely related tothe taxa included in the analyses, but their exact positions arebeyond the purpose of this paper.

Another taxon not included in these analyses was theEchiura. Similarly to the above taxa, they may be highlymodified polychaetes. The resolution of their phylogeneticposition demands analyses at much lower levels of generalitythan has been possible to deal with herein.

Chrysopetalidae Ehlers, 1864 and Hesionidae Grube,1850 were also excluded from our analyses because themonophyly of the first taxon is uncertain (ALMEIDA &CHRISTOFFERSEN 2000), considering that the analyses of PLEIJEL &DAHLGREN (1998) is still insufficient to resolve the position ofthe Chrysopetalidae.

Finally, Sphaerodoridae Malmgren, 1867 andLumbrineridae Schmarda, 1861 were not mentioned in thispaper, because there is a possibility that these taxa representinternal groups of Syllidae Grube, 1850 and Eunicidae Berthold,1827, respectively (ALMEIDA & CHRISTOFFERSEN 2000).

The cladistic analyses were based on Hennigian principles(HENNIG 1966; WILEY 1981; WÄGELE 1994, 1995, 1996a, b, 1999;AMORIM 2002). For phylogenetic reconstructions we used theprogram Hennig86 version 1.5 (FARRIS 1989). The commands“ie*” was used in order to obtain all parsimonious trees. Whenthe number of trees obtained was larger than 100, “x sw” wasused for successive weighting. The most parsimonious treesobtained from the best characters, as indicated by their “ri”,and “ne”, where then used to construct strict consensus trees.As a screen interface to construct the data matrix and to bettervisualize the resultant trees in Windows, the programTreeGardener v. 2.2 was used (RAMOS 1996).

To polarize characters the method of comparison withmultiple outgroups was used (NIXON & CARPENTER 1993; CHRISTOF-FERSEN & ARAÚJO-DE-ALMEIDA 1994; MOURA & CHRISTOFFERSEN 1996;AMORIM 2002; VON STERNBERG 1997). Flatworms, nemerteans,molluscs and sipunculids were selected as outgroups.

One of the most difficult tasks in an analysis is to obtaina matrix with good characters (AMORIM 2002). This meansmaking correct hypotheses of primary homology, establishingadequate subdivisions between different character states,deciding upon an appropriate order for these states, verifyingtheir occurrences in different terminal taxa, and inferring theirpresence in stem-lineages. In short, it does not seem verysensible to perform numerical analysis of provisional matrices(WÄGELE 1994, 1996a, 1999; MIKKELSEN 1998). To avoid theseproblems, characters were grouped into general systems,

26 Almeida et al.


ordered so as to form biologically interdependenttransformation series of character states whenever possible. Forexample, it has been recognized that, in a single transformationseries of primary homology, larvae are originally trochophore-like, then become gradually tornaria-like, and finally becomelost in several specialized lineages when development hasbecome direct. This procedure is very different from simplycoding different larval types as typologically distinct characters(which results in different tree topologies). To minimize casesof “pseudoplesiomorphy” (WILEY 1981), two alternative codingstrategies were adopted. In the “negative series”, following thecriteria of MIKKELSEN (1998) and PURSCHKE et al. (2000), absencesfor which there are clear evidences of secondary reductionswere considered as further valid character states. Cases in whichit can be shown that there is structure reduction (because ofmarks, sutures, or whatever positive evidences) cannot be codedas absences (coded as zeros). In the “positive series”, charactershave not been coded as losses. Both coding strategies have beenanalyzed as non-ordered and as ordered transformations. Theresults of the four independent analyses were then compared.

Despite this particular treatment of special characters,the use of explicit subjectivity as methodologically adequateis not advocated. Being objective does not mean closing oneseyes for subtle evidences of secondary reductions (see PURSCHKEet al. 2000). Such oversights frequently result in themisplacement of apical taxa to basal positions. Problems ofprimary homology, ordering, and coding may nevertheless stillaffect the interpretation of some characters. Such errors areintensified by the fact that comprehension of the evolution ofsome structures, such as nephridia and larvae (see BARTOLOMAEUS& AX 1992; ROUSE & FAUCHALD 1997; ROUSE 1999), among manyothers, must still be considered fragmentary. Evolutionarychanges in the biology of many groups may dramatically affectthem in parallel ways. For example, colonization of terrestrialenvironments has occurred in some specialized polychaetes,such as Hrabeiella periglandulata, and the clitellates (PURSCHKE1999; HESSLING & WESTHEIDE 1999; PURSCHKE et al. 2000).

Several noisy characters, still cursorily described in theliterature or of uncertain phylogenetic content (e.g., numberof eyes, number of segments, length of animal, ultrastructuralfeatures of sperm, etc.) were not utilized in the analyses inorder to preserve the stronger phylogenetic signal provided bythe more congruent data. This certainly affects the overallresult, since it impoverishes a matrix that desperately needsmore reliable data.

RESULTS

During the analyses, closely related family-level taxa (ofuncertain monophyletic status) were grouped into largerterminal groups: Sigalionoidea (including Pholoidae Kinberg,1858, and Sigalionidae Malmgren, 1867), Polynoidea (includingAcoetidae Kinberg, 1856, Eulepethidae Chamberlin, 1919, Poly-noidae Malmgren, 1867, and part of the Aphroditidae Malm-gren, 1867), Spionida (including Apistobranchidae Mesnil &Caullery, 1898, Chaetopteridae Audouin & Milne Edwards,1833, Longosomatidae Hartman, 1944, Magelonidae Cunnin-gham & Ramage, 1888, Poecilochaetidae Hannerz, 1956,Spionidae Grube, 1850, and Trochochaetidae Pettibone, 1963),Terebellida (including Acrocirridae Banse, 1969, AlvinellidaeDesbruyères & Laubier, 1986, Ampharetidae Malmgren, 1866,

Cirratulidae Carus, 1863, Flabelligeridae Saint-Joseph, 1894,Pectinaridae Quatrefages, 1866, Terebellidae Malmgren, 1867,and Trichobranchidae Malmgren, 1866), Sabellida (includingSabellariidae Johnston, 1865, Sabellidae Malmgren, 1867,Serpulidae Johnston, 1865, and part of the Owenidae Rioja,1917) and Paraonida (including Cossuridae Day, 1963,Orbiniidae Hartman, 1942, Paraonidae Cerruti, 1909, andScalibregmatidae Malmgren, 1867). Although these groupingsdo not always represent the most widely accepted taxa inpolychaete taxonomy, these groups were chosen in order toreduce to a minimum the potential influence of paraphyletictaxa in the analyses. Nevertheless, Sigalionoidea, Polynoidea,Spionida, and Paraonida remain paraphyletic. The ingroupresolution of these terminals is not the scope of this work,which aims to provide a general framework for the higher levelsof generality of the metameric metazoans.

The 60 characters used in our analyses are presented intables I-III. A total of 34 of these characters were marked asuncertain (“?”) for Ecdysozoa. This coding is a result of the useof the fossil Cambrian lobopodians as the basal group ofEcdysozoa. This choice strongly influences the positioning ofthe Ecdysozoa in our analyses. Nevertheless, this position wasmaintained, because any other form of coding the characters(for example, as primary absences or secondary losses) wouldalso significantly affect the results of the analyses, either byexcluding the Ecdysozoa from the polychaetes, or by attractingthem to more specialized branches such as the Clitellata. Theevolutionary transitions necessary for either of these two latterhypotheses could only have occurred by way of the fossillobopodians. The Recent onychophorans are morphologicallytoo distant from the Cambrian fossils, and their adaptationsfor the terrestrial environment may significantly obscure theirancestral morphology and anatomy. By opting for a middle-groud-solution it is hoped that the large number of questionmarks will serve to instigate future research, which maycorroborate or refute the results.

The positive series were analyzed as ordered, two treesbeing obtained (length 129, ci 0.65, and ri 0.88), the consensusof which is presented in figure 1a. The unordered analysis ofthe proposed transformation series resulted in 1575 trees(length 114, ci 0.73, and ri 0.90). The successive weightingoption was used and the number of trees obtained was reducedto 310 (length 784, ci 0.88, and ri 0.96), the consensus of whichis presented in figure 1b.

The ordered negative series resulted in three parsimo-nious trees (length 128, ci 0.78, and ri 0.94) (Fig. 2a). The unor-dered negative series resulted in 330 parsimonious trees (length121, ci 0.83, and ri 0.93), which were reduced to 120 trees(length 946, ci 0.91, ri 0.97) by successive weighting (Fig. 2a).

The main differences among these analyses are that thepositive series resulted in more parsimonious cladograms, whilethe negative and ordered analyses produced the most resolvedtopologies.

Much more significantly than the differences, however,is that all four analyses are congruent in indicating that thePolychaeta become monophyletic only when Ecdysozoa,Clitellata, and Enterocoela are included among the remainingtaxa of metameric worms. The clades Aphrodita + Ecdysozoa(Holopodia, tax. nov.), Questidae + Clitellata (Apoclitellata,tax. nov.), and Owenia + Enterocoela (Neotrochozoa, tax. nov.)are maintained throughout our analyses.



Figure 1. (A) Strict consensus tree (length 129, ci 0.65, ri 0.88) of the 2 trees obtained from analysis of the positive ordered series. (B)Consensus tree (lenght 910, ci 0.76, ri 0.90) of the 310 trees obtained with successive weighting from our analysis of our positiveunordered series.

Figure 2. (A) Strict consensus (Length 128, ci 0.83, ri 0.93) of the 3 trees obtained from the analysis of ordered negative series. (B) Strictconsensus tree (length 963, ci 0.90, ri 0.96) of the 120 trees obtained with successive weighting from analysis of negative unorderedseries.

28 Almeida et al.


DISCUSSION

About hypotheses of homologies and phylogeneticrelationships

The following discussion presents hypotheses of primaryhomology based on the totality of the results. As in any cladisticanalysis, it is evident that hypotheses of primary homologyinfluence the results. However, all hypotheses of relationshipspresented below were a product of the cladograms. The twomost polemical results refer to the positions of the Ecdysozoaand of the Enterocoela among the polychaetes. These ideashave received a high degree of a priori rejection and havegenerated strong opposition among many researchers.

Much empirical support for the inclusion of Ecdysozoaand Enterocola among the polychaetes is presented below(synapomorphies #23(2), #24(2), #35, #44(2), #52, #59, #60;Tab. I). Independent empirical data derived from Hox genesrepresent the most surprising discovery of recent molecularstudies. Much of the genetic machinary that patterns theappendages of polychaetes, arthropods, and vertebrates ishomologous (SHUBIN et al. 1997; PANGANIBAN et al. 1997).Furthermore, it has been amply demonstrated that thehomologous Hox genes are responsible for the definition ofthe dorso-ventral and anterior-posterior body axes, and forsegmentation in polychaetes, arthropods and vertebrates(MCGINNIS et al. 1984; LAWRENCE 1990; FRANÇOIS & BIER 1995;HOLLEY et al. 1995; JONES & SMITH 1995; HOLLAND et al. 1997).New developmental data on Hox genes further indicates notonly that deuterostomes derive from a protostome ancestor,but also that a polarity inversion of both body axes has occurredin this process (NÜBLER-JUNG & ARENDT 1994; 1999; ARENDT &NÜBLER-JUNG 1994, 1997, 1999).

Consequently, the hypotheses presented below reinforcethe idea that some scale worms have become armored and that,in the case of Aphrodita, their parapodia have developedfunctionally and anatomically to a condition approaching thelobopodia. On the other hand, there are other polychaetes suchas Owenia that “mysteriously” present deuterostome conditionsin their larvae and during embryonic development.

Must all these coincidences be explained away as theresult of developmental constraints (see SOMMER 1999)? Evenadmitting the possibility of a positive response to this question,Aphrodita, Questidae and Owenia still serve as models forunderstanding how Ecdysozoa, Clitellata and Enterocoela mayrealistically have evolved from segmented vermiform ancestors.Under this perspective it is no longer necessary to create“imaginary ancestors”, such as the hemocoelomate worms ofVALENTINE (1989), ERWIN et al. (1997), and FRYER (1996, 1998), oreven non-existent structures such as the antenniformappendage of PANGANIBAN et al. (1997).

Characters

Mollusk-like cleavageMollusca and Sipuncula share a peculiar pattern of

cleavage, in which the cross-shaped center of the apical end ofthe embryo is formed by blastomeres 1a12-1d12 (see SCHELTEMA1993) (Tab. 1, #1). ROUSE & FAUCHALD (1995) also used this cha-racter, but SALVINI-PLAWEN (1988) strongly objected to the validityof this supposed synapomorphy of Mollusca+ Sipuncula. There

is uncertainty about the validity of this character and of thevalidity of the clade Mollusca+Sipuncula, but the character wasincluded in the analyses to provide some data for the resolutionof the multiple outgroups. An equally parsimonious inter-pretation for this character would be to consider the sharedcondition of Mollusca + Sipuncula symplesiomorphic inrelation to a modified cross in the Annelida.

Homologies related to coeloms and metameresWÄGELE et al. (1999) note that the reduction of the

annelid-like hydrostatic coelom is possible in Arthropodabecause when segmental appendages develop into effective legsthe hydraulic pressure needed for peristaltic movementsbecome superfluous (this reasoning is also valid for non-euarthropods such as Onychophora). Actually, as ALMEIDA &CHRISTOFFERSEN (2000) have shown, the coelom condition ofAphroditidae (with reduced peritomium septae) is intermediarybetween that of typical errant polychaetes and that ofonychophorans, as is also the case of the locomotory system.

Coelom development in Metameria (Tab. I, #2) is closelylinked to a gradual production of metameres (compartmentedsegments are associated with a ganglioned nervous system (Tab.I, #3). The aschelminths also present evidences of metamerism,particularly in some clades (e.g., ganglioned nervous systemof the Rotifera and Kinoryncha) (see ALMEIDA & CHRISTOFFERSEN2000: fig. 2, tab. 2). The recognition of these metameres hasbeen underscored in the literature due to an emphasis on theso-called “typical” (usually parasitic) representatives of eachgroup. In the highly modified forms for a parasitic or interstitialway of life, the metameric organization becomes lessconspicuous.

Traditionally, theories about the origin of metamerismare associated with those on the origin of the coelom. Themost frequent ones found in textbooks are the enterocoelictheory (REMANE 1956) and the hydrostatic theory (CLARK 1963,1964). The enterocoelic theory envisions a simultaneous originof the coelom and of segmentation, as a consequence of theontogeny of the mesoderm. However, besides difficultiesencountered in establishing a direct relation of homologybetween different types of mesoderm and body cavities, oneof the most inelegant consequences of this proposal is toconsider flatworms as belonging to the clade of metameric andcoelomate animals.

The hydrostatic theory avoids generalizations such as theabove, being by far the most generally accepted theory. ForCLARK (1963, 1964), the origin of the coelom is linked to theevolution of a burrowing way of life. Animals with a hydrostaticskeleton supposedly would have an advantage over acoelomateforms for burrowing. From this point of view, body cavitiescould have been under selective pressure to evolve several timesindependently for a fossorial habitat within the MetazoaHaeckel, 1874. The body division into segments wouldrepresent an additional improvement for living within thesubstratum. Again, such a favorable selection pressure wouldhave produced metamery several times independently withinthe Metazoa (CLARK 1963, 1964).

The results of the present cladistic analyses reject bothof these scenarios. The available evidence does not permit therejection of an initial hypothesis of homology of the bodycavities. Compartimented body segments with fluid cavitiesdelimited by peritomium are present in all metamerians (except



.setatscihpromopaylevisseccus,4-1;etatscihpromoiselp,0.desusretcarahcfotsiL.IelbaT

rebmunretcarahC setatsdnasretcarahC

1 .tneserp,1;tnesba,0:egavaelcekil-csulloM

2 .tneserp,1;tnesba,0:moleoC

3 .tneserp,1;tnesba,0:yremateM

4 .tneserp,1,tnesba,0:sisomgaT

5desufyllatot,3;muimotsorpotdesufyllaitrap,2;muimotsorpotdesuftontubtneserp,1;tnesba,0:tnemgestsriF

.muimotsorpot

6 .tneserp,1;tnesba,0:stnemgessuodopA

7 .tneserp,1;tnesba,0:muimotsorP

8 .tneserp,1;tnesba,0:muimotsirepotdesufmuimotsorP

9 .suocipsnocdnatneserp,2;tneipicnitub,tneserp,1;tnesba,0:)snoitalunna(muimotsorP

01 .tnesbayliradnoces,2;tneserp,1;tnesba,0:kaeplaimotsorP

11 .deihportrepyh,2;deihportrepyhtondnatneserp1;tnesba,0:)spillaimotsirep(muimotsireP

21 .tnesbayliradnoces,2;tneserp,1;stnesba0:splaP

31 .tnesbayliradnoces,4;lasrod-oretal,3;lartnev-oretal,2;lartnev,1;tnesba,0:)noitisop(splaP

41 .tneserp,1;tnesba,0:splapyrosnesgnoL

51 .tneserp,1;tnesba,0:splaptuotS

61 .tneserp,1;tnesba,0:)noitalucitra(splaP

71 .)tnesbayliradnoces,3(;sevoorgelpitlumhtiw,2;tneserpsevoorgforiapeno1;tnesba,0:splapdevoorG

81 .tnesbayliradnoces,4;yrotaripserdnagnideef,3;gnideef,2;yrosnes,1;tnesba,0:)noitcnuf(splaP

91 .tnesbayliradnoces,2;tneserp,1;tnesba,0:eannetnaderiaP

02 .tnesbayliradnoces,2;tneserp,1;tnesba,0:eannetnanaideM

12,4;decuder,3;ralimisaidoporuendna-otonhtiw,2;)aidoporuengnitcejorp(tneserp,1;tnesba,0:aidoparaP

.eateahcfosdnabowttsuj,5;irotgnimrof

22 .tneserp,1,tnesba,0:aidopotoN

32 .ekil-aidopobol,2;serutcurtsyrotomocolsatneserp,1;tnesba,0:aidoporueN

42 .tnesba,1;tneserp,0:snoitaludnularetaL

52otnideifidom,3;)eartyle(selucrebutlasrodotnideifidom,2;mrofirricdnatneserp1;tnesba,0:irriclasroD

.tnesbayliradnoces,4;eaihcnarb

62 .tnesbayliradnoces,2;tneserp,1;tnesba,0:irriclartneV

72 .tneserpsriaperomroowt,2;tneserpriapeno,1;tnesba,0:irricralucatneT

82 .tneserp,1;tnesba,0:eateahC

92 .tneserp,1;tnesba,0:3OCaChtiweateahC

03 .tneserp,1;tnesba,0:eateahcykliS

13 .tnesbayliradnoces,3;inicnuotnideifidom,2;tneserp,1;tnesba,0:skoohdedooH

23 .tnesbayliradnoces,2;tneserp,1;tnesba,0:ealucicA

33 .tnesbayliradnoces,2,tneserp,1,tnesba,0:irriclaidigyP

43 .tneserp,1;tnesba,0:elcituC

53 .tneserp,1;tnesba,0:eallipapcimredipemrofitidorhpA

63 .tneserp,1;tnesba,0:gnirevreN

73 drocranoilgnaglartnevhtiw,1;drocranoilgnaglartnevtuohtiw,0:metsyssuovreN

.tnoC

30 Almeida et al.


in many ecdysozoans in which they may be totally reduced toa “pseudocoel” or absent). Furthermore, the enterocoelic typeof coelom formation may be considered an apomorphic,secondary modification of the schizocoelic pattern withteloblastic development (CHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA1994). The present hypothesis differs from the enterocoelictheory because we invert the polarity of body cavities(enterocoely becomes derived instead of primitive) and by ourunlinking the characters coelom and metamerism. Body cavitiesappear before metamerism in the history of the Metazoa.

Optimization of ecological features over the phylogenyfurnishes no evidence whatsoever for the hypotheses that thecoelom and metamerism are associated with a burrowing habit(CLARK 1963, 1964). The first body cavities are seen in Nemertea,where they are used as hydrostatic devices (rhynchocoels) foreverting the proboscis (BRUSCA & BRUSCA 1990), even thoughthis may not be the original function in the groundplan of theCoelomata Christoffersen & Araújo-de-Almeida, 1994. Thebody cavity in Nemertea may be considered an exaptation (pre-adaptation), subsequently modified for other functions, suchas providing an hydrostatic body skeleton, promoting the

circulation of body fluids, producing space for the maturationof reproductive cells, etc. (HYMAN 1951; BRUSCA & BRUSCA 1990;TURBEVILLE 1991). Most metameric marine animals that dig intothe substrate have the septae reduced or totally absent (e.g.,Sipuncula Sedgwick, 1898, Glyceridae Grube, 1850, GoniadidaeKinberg, 1866, and Arenicolidae Johnston, 1835 among others).On the other hand, the mesenteric septae are well developedin some tubicolous forms and in most errant polychaetes (CLARK1962; WILLMER 1990). These observations give little support fora correlation between septae and particular life styles. Theanalysis of ALMEIDA & CHRISTOFFERSEN (2000) favors the hypothesisof BONICK et al. (1976) and WESTHEIDE (1997) for some sort ofassociation between the metameric septae and the developmentof parapodia (Fig. 3). The septae contain blood vessels for theirrigation of the parapodia and muscles that may facilitatelateral body undulations when the animals move (WESTHEIDE1997). However, we further sustain that this scenario of theorigin of metamerism in Polychaeta and Arthropoda is alsoshared with the deuterostomes (Enterocoela). Owenia delleChiaje, 1841 and Pogonophora provide clear successiveconnecting links between the traditional protostome and

.deunitnoC.IelbaT

rebmunretcarahC setatsdnasretcarahC

83 .tnesbayliradnoces,2;tneserp,1;tnesba,0:snagrolahcuN

93 .tneserp,1;tnesba,0:)elcnurac(snagrolahcuN

04 .tnesbayliradnoces,2;tneserp,1;tnesba,0:snagrolaretaL

14tondnalartnev,elbisreve3;deziralucsumdnalartnev,elbisreve,2;laixadnaelbisreve,1;elbisreveton,0:xnyrahP

.ekil-casdnaelbisreve,4;deziralucsum

24 .tneserp,1,tnesba,0:xnyrahpehtfoevoorglasrevsnarT

34 .tneserp,1;tnesba,0:)noitisopsidtelcric(swaJ

44 .tnesbayliradnocesrodecuder,2;tneserp,1;tnesba,0:)noitisopsidlartnev-osrod(swaJ

54 .htangonetcdnadecuder,2;htangodbaldnatneserp,1;tnesba,0:stroppusdnaealixaM

64 .tneserp,1,tnesba,0:xnyrahpdegdiR

74 .tneserp,1,tnesba,0:nagrolaccuB

84 .tneserp,1;tnesba,0:etelpmocmetsysevitsegiD

94 .laidirhpenotorpyliradnoces,2;laidirhpenatem,1;laidirhpenotorp,0:metsysyrotercxE

05 .tneserp,1;tnesba,0:aidirhpenroiretnaforiapA

15 .tneserp,1;tnesba,0:eropoidirhpenlasroD

25 .tneserp,1;tnesba,0:ekil-motsoretuedaidirhpenlavraL

35 .tneserp,1;tnesba,0:mulletilC

45 .tneserp,1;tnesba,0:tnemgesdeificepsnisnagroevitcudorpeR

55 .tneserp,1;tnesba,0:suoiravoroiretnaehtforiapowT

65 .tneserp,1;tnesba,0:41tnemgesnitcudonogehtfogninepO

75 .tneserp,1;tnesba,0:eacetamrepS

85 .)tceridtnempoleved(tnesbaeavral,4;airanrot,3;airartim,2;erohpocort,1;alunalp,0:tnempoleveddnaavraL

95 .tneserp,1;tnesba,0:ekil-motsoretuedeavralnitnemegnarrarailiC

06 .tneserp,1;tnesba0:sisohpromatemcihportsataC



deuterostome conditions. The present hypotheses ofrelationships have antecedents in the proposals of GEOFFROYSAINT-HILLAIRE (1822) and DOHRN (1875), both of whom proposedthe origin of the Vertebrata from annelid-like ancestors.

The morphological hypothesis detailed above is stronglycorroborated by two other types of empirical data: (1) Theburgeoning evidence of homology of the Hox genes thatcontrol segment formation in protostomes and deuterostomes(MCGINNIS et al. 1984; LAWRENCE 1990; FRANÇOIS & BIER 1995;HOLLEY et al. 1995; JONES & SMITH 1995; HOLLAND et al. 1997; SHUBINet al. 1997; PANGANIBAN et al. 1997; PETERSON & DAVIDSON 2000;PETERSON et al. 2000a, b); and (2) the growing evidence forhomologous developmental patterns shared by protostomesand deuterostomes (WEISBLAT et al. 1993; NÜBLER-JUNG & ARENDT1994, 1999; ARENDT & NÜBLER-JUNG 1994, 1997, 1999).

Sequence data (e.g., 18S rRNA, 12S rRNA, 18S rDNA, etc.)have produced molecular phylogenies which maintain thetraditional divergence between Protostomia Grobben, 1908 andDeuterostomia Grobben, 1908, and suggest a distantrelationship between Annelida and Arthropoda (FIELD et al.

1988; BALLARD et al. 1992; WINNEPENNINCKX et al. 1992, 1998;EERNISSE et al. 1992; KIM et al. 1996; GAREY et al. 1996, 1998;AGUINALDO et al. 1997; EERNISSE 1997; GIRIBET & RIBERA 1998; ZRZAVÝet al. 1998; SCHIMIDT-RHAESA et al. 1998). The heuristic value ofthese sequence data is clearly inferior to the genotypic-phenotypic data produced from shared orthologous genes. AsWÄGELE (1995, 1996b, 1999), WÄGELE & WETZEL (1994) andWÄGELE et al. (1999) have stressed, many analyses of sequencedata are not based on explicit phylogenetic criteria and resultessentially from numerical analyses of similarities. There isample evidence that these results are strongly dependent onsample size, selected organisms, types of sequences used, andchosen methods of analyses (WÄGELE & WETZEL 1994; WÄGELE1995; SIDDALL et al. 1998; ALESHIN et al. 1998; WÄGELE et al. 1999;ALESHIN & PETROV 1999). PHILIPPE et al. (1994) furtherdemonstrated that, for cladogenetic events occurring before400 Myr ago, data from 18S rRNA are definitely not reliable.ALESHIN & PETROV (1999: 196) and GIRIBET & RIBERA (2000: 225)also presented results indicating that phylogenies based onrDNA sequence may not contain enough information to givea satisfactory explanation for the large and complicatedevolutionary history of arthropods. For all these reasons askeptical posture was assumed regarding the phylogeneticsignificance of sequence data that differ from other types ofcomparative information analyzed in a phylogenetic context.Like the phenetic analyses of morphological data, many of thetaxonomic arrangements proposed by molecular biologists maybe based on plesiomorphic similarities and/or convergences.

Homeobox genes have recently been discovered inPlatyhelminthes (BALAVOINE & TELFORD 1995; BALAVOINE 1996).This led BALAVOINE (1997, 1998) to develop the idea thatPlatyhelminthes are highly modified coelomates. Thishypothesis was not adopted here, because it is possible thatthe homeoboxes of Platyhelminthes are only responsible forthe expression of the antero-posterior body axis, and possiblyalso of the serial organs (e.g., gonads in triclad tuberlarians;HYMAN 1951: 26). There is no direct evidence thatPlatyhelminthes have coeloms or derive from coelomate ormetameric ancestors. The generality of Hox genes simplyextends to the base of the Metazoa (PETERSON & DAVIDSON 2000).The discovery of homeobox genes in Platyhelminthes is veryinteresting, but may be misleading if interpreted under pre-established evolutionary scenarios (such as the gonocoelic orenterocoelic theories). The present hypotheses on metazoanphylogeny have very old roots (e.g., GEOFFROY SAINT-HILLAIRE 1822;DOHRN 1875), but must also include the latest evidence, andthe older views must be modified accordingly. The case of thePlatyhelminthes provides a good example of this problem. Thecoelom appears at the base of Coelomata and in Nemertea it isused for everting the proboscis, while mesenteric septae and aventral ganglionated nerve cord appear in Metameria associatedwith the development of parapodia (BONICK et al. 1976;WESTHEIDE 1997). Thus the origin of the coelom and the originof metamerism are decoupled in evolution. The presence ofhomeobox genes in Platyhelminthes is not incongruent withthis scenario. A repetition of structures may be found also inthe scoleces of cestode flatworms, in Mollusca Linnaeus, 1758(HASZPRUNAR 1992; SCHELTEMA et al. 1994), and apparently alsoin the Precambrian metazoan Dickinsonia Sprigg, 1947(RUNNEGAR 1982). A repetition of structures thus seems toprecede both the origin of the mesodermic body cavities andthe origin of the mesodermic metameres. The true novelty of

Figure 3. Groundplan of Metameria obtained by addingsynapomorphies from ALMEIDA & CHRISTOFFERSEN (2000). An, lateralantennae; Pci, pigidial cirri; Pl, palps; Prj, projecting neuropodia.

32 Almeida et al.


metameric animals would consist in their reorganizing theirbody for a new system of locomotion that relies on metamericappendages. With this functional reorganization, preexistingplesiomorphies, such as homeobox genes, a hydrostaticskeleton, and a repetition of internal organs, have all been co-opted for the walking-invention. Data suggests, hence, thathomeobox genes are plesiomorphic for the Coelomata. Saidin another way, metamerism is not to be interpreted as thedirect phenotypic expression of homeobox genes.

Both tagmosis (functional specialization of groups ofmetameres) (Tab. I, #4) and reduction of the total number ofmetameres are accepted as apomorphic. The latter trend occursin several tubicolous lineages such as Terebellida and Sabellida,and in burrowing forms such as Arenicolidae Johnston, 1835and Maldanidae Malmgren, 1867. These polarities reestablishthe traditional view that heteronomous polychaetes are derivedin relation to those with a largely homonomous pattern ofsegmentation (FAUVEL 1923). With Owenia positioned as thesister group of Enterocoela, several genetic patterns of tagmosisfound in basal groups of deuterostomes (PETERSON et al. 2000a)are also seen to be found in polychaetes, providing furtherempirical support for the present hypothesis.

The first segment (Tab. I, #5) may be partially or totallyfused to the cephalic tagma in Phyllodocida and in the basalgroups of eunicids and related taxa (Nephtyidae Grube, 1850and Amphinomidae Savigny, 1818) (Fig. 4). The followingsegments may also be cephalized, as in Phyllodocidae Örsted,1843 (DAY 1967; SCHRÖDER & HERMANS 1975; FAUCHALD 1977;GLASBY 1993; FAUCHALD & ROUSE 1997). These characters areconsidered apomorphies for these clades, which is congruentwith the interpretations of GLASBY (1993), ROUSE & FAUCHALD(1997), PLEIJEL & DAHLGREN (1998), and ALMEIDA & CHRISTOFFERSEN(2000). Evidence for or against the homology of the head tagmaof Polychaeta with that of Ecdysozoa is still lacking.

With the total cephalization of the first metamere,parapodial structures such as neuropodia, notopodia andchaetae are not expressed. The only structures that indicatetheir metameric origin are the parapodial cirri (now calledtentacular cirri) (Fig. 4). In Nereididae, the first pair of tentacularcirri appears on the peristomial region (e.g., BLAKE 1975) andtheir inervation patterns are indicative of a segmental origin(ORRHAGE 1993). If this first pair of cirri were not present, noevidence would be available to indicate that the anterior collarregion in Nereididae Johnston, 1865 is metameric. Suchsegmental inervation could have been lost in DorvilleidaeChamberlin, 1919 (see EIBYE-JACOBSEN 1994), if the peristomialcirri of Onuphidae Kinberg, 1865 and Eunicidae Berthold, 1827are equivalent to the tentacular cirri of Phyllodocida (Fig. 4).In any case, many representatives of Sedentaria have one ortwo anterior apodous rings (Tab. 1, #6): Spionida (Magelo-nidae), Terebellida (Cirratulidae, Alvinellidae, Ampharetidae,Terebellidae, and Trichobranchidae), and scolecids(Arenicolidae, Capitellidae, Maldanidae, and Questidae) (seeFAUCHALD & ROUSE 1997).

Prostomium and peristomium have dynamicboundaries

A traditional definition of the prostomium (Tab. I, #7) isthe following: situated anteriorly, delimited from trunk by agroove, of pre-trochal origin, and containing in the adult atleast part of the supra-oesophageal ganglion. Sensory

appendages (antennae and palps) and photo-sensory organs(ocelli) are regularly present on the prostomium (FAUVEL 1959;FAUCHALD 1977; PETTIBONE 1982; GEORGE & HARTMANN-SCHRÖDER1985; FAUCHALD & ROUSE 1997) (Figs 4 and 5).

Typological descriptions such as the above, wheninterpreted as immutable patterns, may generate confusion inthe interpretation and delimitation of this structure. Forexample, if the prostomium corresponds to the pre-trochalanterior region of the trochophore larvae, then Mollusca andSipuncula have a cephalic region of identical origin (see HYMAN1959; SCHELTEMA et al. 1994; CUTLER 1994) and thus all threetaxa should have homologous prostomia.

The groove that delimits the prostomium of annelidsrepresents a redundant structure, because it results from bodysegmentation. In other words, non-metameric groups obviouslydo not have a head separated from the trunk by a groove. Itmay also be difficult to characterize the prostomium as theregion that contains the supra-oesophageal ganglion, becausein Clitellata this ganglion may be significantly displacedposteriorly (HESSLING & WESTHEIDE 1999).

The presence of sensory structures is also not a goodsolution to characterize the prostomium, because manyspecialized Sedentaria, such as the scolecids, usually do nothave palps, antennae or ocelli, and this absence does notcharacterize them as polychaetes without a prostomium.

The more basal forms of Ecdysozoa (except KerigmachelaBudd, 1993) also do not present sensory cephalic structures(BUDD 1993, 1996). On the other hand, the anterior region, oracron, of Crustacea Pennant, 1777 bears cephalic appendagesand contains part of the supra-oesophageal ganglion (SCHRAM1986; SCHRAM & EMERSON 1991). In both cases it is very difficultto establish either the presence or the absence of a typicalprostomium in Ecdysozoa. Because of the above difficulties,presence of a pre-prototrochal cephalic region may be asynapomorphy of the clade Mollusca + Sipuncula + Metameria(Tabs. II and III).

The different degrees of fusion between the pre-prototrochal (prostomial) and prototrochal (peristomial)regions (Tab. I, #8), sometimes also involving the firstmetamere, are derived conditions in Metameria. This fusionwas coded as a single character. This may form a “distinct head”in Maldanidae and Paraonidae (here provisionally included inParaonida), a “tentacular crown” in the Sabellida or a “jointfoliose, saddle-shaped, structure” in the Terebellida (PILGRIM1966; HOLTHE 1986; ROUSE & FITZHUGH 1994) (Fig. 5). The lattertwo conditions are clearly autapomorphies of the Sabellida andTerebellida, respectively. But more information is needed toresolve if the simpler head of Maldanidae and Paraonidae couldrepresent a homologous stage towards the latter conditions –avoiding an essentialistic coding of the character.

Despite much variation in the shape of the prostomium– e.g., “T-shaped” in Spionida or forming an “inverted T” inNereididae (DAY 1967; FAUCHALD 1977) –,only the annulationof the prostomium in Glyceridae and Goniadidae wasconsidered as forming a synapomorphy (see HARTMAN 1950)(Tab. I, #9). This hypotheses agrees with that of HANSTRÖM (1928,1930), in which the prostomium of the Glyceridae and of theGoniadidae are very complex and derived structures. In thefigures of PETTIBONE (1963: fig. 46) and DAY (1967: fig. 15e), anapparent annulation is also observed in ParalacydoniidaePettibone, 1963. These annuli may represent a synapomorphy



Figure 4. General scenario for the main modifications of many anterior morphological characters of the errant metameric lineages (indorsal view). (A) Paralacydoniidae represented by Paralacydonia paradoxa Fauvel, 1913 (modified from PETTIBONE 1963); (B) Scale wormsrepresented by Lepidonotus caelorus Kinberg, 1866 (Polynoidea) (modified from IMAJIMA 1997); (C) Phyllodocyformia represented byNereis diversicolor Müller, 1776 (Nereididae) (modified from BÖGGEMANN 1997); (D) Basal group of Eunicida represented by Aglaophamusgippslandicus Rainer & Hutchings, 1977 (Nephtyidae) (modified from IMAJIMA & TAKEDA 1985); (E) Eunicidae represented by Hyalinoeciatubicola Müller, 1766 (Onuphidae) (modified from GEORGE & HARTMANN-SCHRÖDER 1985); (F) Dorvilleidae represented by Protodorvilleakefersteini McIntosh, 1869 (modified from GEORGE & HARTMANN-SCHRÖDER 1985). An, lateral antennae; Anp, annulated prostomium; Ap,amphinomid-like parapodia; Dc, dorsal cirri; Ely, elytrophore; Fs, first segment; Hi, hypertrophied peristomial lips; Lpl, long sensorypalps; Lvp, latero-ventral palps; Ma, median antennae; Pl, palps; Pr, peristomial ring; Prj, projecting neuropodia; Spi, spionimorphpalps; Spl, stout articulated palps; Tc, tentacular cirri.

Figure 5. Further scenario for main modifications of many anterior morphological characters of the sedentary metameric lineages (inlateral view). (G) Spionida represented by Aonides oxycephala (Sars 1862) (Spionidae) (modified from IMAJIMA 1989); (H) Terebellidarepresented by Nicolea amnis Hutchings & Murray, 1984 (Terebellidae) (modified from HUTCHINGS & MURRAY 1984); (I) Sabellida representedby Dasynema chrysogyrus (Grube 1876) (Sabellidae) (modified from IMAJIMA & HOVE 1984); (J) Owenia represened by Owenia fusiformisdelle Chiaje, 1842 (Oweniidae) (modified from IMAJIMA & MORITA 1987); (K) Capitellidae represented by Notomastus estuarius Hutchings& Murray, 1984 (HUTCHINGS & MURRAY 1984); (L) Questidae represented by Questa caudicirra Hartman, 1966 (modified from FAUCHALD1977); (M) Clitellata represented by Phallodrilus riparius Giani & Martinez-Ansemil, 1981 (Tubificidae) (modified from GIANI & MARTINEZ-ANSEMIL 1981). Fs, first segment; Gpl, grooved palps; Mp, multiple palps; Pb, parapodial branchiae; Pbn, parapodial chaetal bundles; Pr,peristomial ring; Pro, prostomium; Sp, spionimorph parapodia; Tcr, tentacular crown, Tno, truncate notopodia; To, tori.

34 Almeida et al.


rotnesbanoitamrofniretcarahc)?(,seiresnoitamrofsnartderedrO)4-0(.seiresevitisopehtfosesylanacitsidalcnidesuxirtamataD.IIelbaT.suoibud

axaTsretcarahC

1 2 3 4 5 6 7 8 910

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

sehtnimlehytalP 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

aetremeN 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

alucnupiS 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

acsulloM 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiinodycalaraP 0 1 1 0 1 0 1 0 1 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 ? 0 0 0 0 0 0 0 0 1 0 0

eadirecylG 0 1 1 0 1 0 1 0 2 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0

eadidainoG 0 1 1 0 1 0 1 0 2 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0

eadinoisiP 0 1 1 0 2 0 1 0 0 0 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0

aedionoilagiS 0 1 1 0 2 0 1 0 0 0 1 1 1 1 0 0 0 1 1 1 1 1 1 1 2 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

aedionyloP 0 1 1 0 2 0 1 0 0 1 1 1 1 1 0 0 0 1 1 1 1 1 1 1 2 1 1 1 0 1 0 1 1 1 0 1 1 1 0 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

atidorhpA 0 1 1 0 2 0 1 0 0 0 1 1 1 1 0 0 0 1 0 1 1 1 2 0 2 1 1 1 0 1 0 1 1 1 1 1 1 1 0 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0

aozosydcE 0 1 1 ? ? ? 1 ? ? 0 ? ? ? ? ? ? ? ? 0 ? ? ? 2 0 2 ? ? ? ? ? ? ? ? 1 1 1 1 0 0 ? ? ? 0 0 0 0 ? 1 1 0 0 0 0 ? ? ? ? 0 ? ?

eadiinodycaL 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 ? 0 0 0 0 0 0 0 0 1 0 0

eadicodollyhP 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0

eadillyS 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadidiereN 0 1 1 0 3 0 1 0 0 0 1 1 1 0 1 1 0 1 1 0 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadigraliP 0 1 1 0 3 0 1 0 0 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiythpeN 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 0 2 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0

eadimonihpmA 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 1 2 1 1 1 3 1 0 1 1 0 0 1 1 1 0 1 1 1 1 0 2 0 0 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadinisorhpuE 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 1 2 1 1 1 3 1 0 1 1 0 0 1 1 1 0 1 1 1 1 0 2 0 0 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadicinuE 0 1 1 0 ? ? 1 0 0 0 2 1 2 0 0 0 0 1 1 1 1 1 1 1 3 1 ? 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eadihpunO 0 1 1 0 ? ? 1 0 0 0 2 1 2 0 0 0 0 1 1 1 1 1 1 1 3 1 ? 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiellivroD 0 1 1 0 ? ? 1 0 0 0 1 1 2 0 0 0 1 1 1 1 1 1 1 1 3 1 0 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

adinoipS 0 1 1 0 ? 1 1 0 0 0 1 1 3 0 0 0 1 2 0 1 3 1 1 1 3 0 0 1 0 0 1 0 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

adillebereT 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 3 0 0 4 1 1 1 3 0 0 1 0 0 2 0 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 ? 0 0 0 0 0 0 1 0 0

adillebaS 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 3 0 0 4 1 1 1 3 0 0 1 0 0 2 0 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 1 0 0

ainewO 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 2 0 0 4 1 1 1 0 0 0 1 0 0 2 0 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 2 1 1

aleocoretnE 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 2 0 0 4 1 1 1 0 0 0 1 0 0 2 0 ? ? 0 1 1 0 0 1 ? 0 0 0 0 0 ? 1 1 1 1 1 0 0 0 0 0 3 1 1

adinoaraP 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 4 1 1 1 3 0 0 1 0 0 1 0 1 1 0 1 1 0 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eaditseuQ 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 1 1 1 3 0 0 1 0 0 1 0 1 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 1 1 1 1 1 4 0 0

atalletilC 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 5 1 1 1 0 0 0 1 0 0 0 0 1 1 0 1 1 0 0 1 ? 0 0 0 0 0 ? 1 1 0 0 0 1 1 1 1 1 4 0 0

eadillehpO 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 1 1 1 3 0 0 1 0 0 0 0 0 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadilletipaC 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 1 1 1 0 0 0 1 0 0 1 0 0 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadilocinerA 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 1 1 1 3 0 0 1 0 0 1 0 0 1 0 1 1 1 0 0 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadinadlaM 0 1 1 1 ? 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 1 1 1 3 0 0 1 0 0 1 0 0 1 0 1 1 1 0 0 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0



rotnesbanoitamrofniretcarahc)?(,seiresnoitamrofsnartderedrO)4-0(.seiresevitagenehtfosesylanacitsidalcnidesuxirtamataD.IIIelbaT.suoibud

axaT sretcarahC

1 2 3 4 5 6 7 8 910

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0

sehtnimlehytalP 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

aetremeN 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

alucnupiS 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

acsulloM 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiinodycalaraP 0 1 1 0 1 0 1 0 1 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 ? 0 0 0 0 0 0 0 0 1 0 0

eadirecylG 0 1 1 0 1 0 1 0 2 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 1 0 0 0 0 1 2 0 0 0 0 0 0 0 0 1 0 0

eadidainoG 0 1 1 0 1 0 1 0 2 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 1 0 0 0 0 1 2 0 0 0 0 0 0 0 0 1 0 0

eadinoisiP 0 1 1 0 2 0 1 0 0 0 1 1 1 1 0 0 0 1 1 2 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 2 0 0 1 0 0 1 0 0 0 1 2 0 0 0 0 0 0 0 0 1 0 0

aedionoilagiS 0 1 1 0 2 0 1 0 0 0 1 1 1 1 0 0 0 1 1 1 1 1 1 1 2 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

aedionyloP 0 1 1 0 2 0 1 0 0 1 1 1 1 1 0 0 0 1 1 1 1 1 1 1 2 1 1 1 0 1 0 1 1 1 0 1 1 1 0 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

atidorhpA 0 1 1 0 2 0 1 0 0 2 1 1 1 1 0 0 0 1 2 1 1 1 2 2 2 1 1 1 0 1 0 1 1 1 1 1 1 1 0 0 1 0 0 2 0 0 0 1 1 0 0 0 0 0 0 0 0 4 0 0

aozosydcE 0 1 1 ? ? ? 1 ? ? 2 ? ? ? ? ? ? ? ? 2 ? ? ? 2 2 2 ? ? ? ? ? ? ? ? 1 1 1 1 2 0 ? ? ? 0 2 0 0 ? 1 1 0 0 0 0 ? ? ? ? 4 ? ?

eadiinodycaL 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 ? 0 0 0 0 0 0 0 0 1 0 0

eadicodollyhP 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 1 0 0

eadillyS 0 1 1 0 3 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadidiereN 0 1 1 0 3 0 1 0 0 0 1 1 1 0 1 1 0 1 1 2 1 1 1 1 1 1 2 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadigraliP 0 1 1 0 3 0 1 0 0 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiythpeN 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 2 2 1 1 1 1 1 0 1 0 0 0 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 1 0 0

eadimonihpmA 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 1 2 1 1 1 3 1 0 1 1 0 0 1 1 1 0 1 1 1 1 0 2 0 0 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadinisorhpuE 0 1 1 0 1 0 1 0 0 0 1 1 2 0 0 0 0 1 1 1 2 1 1 1 3 1 0 1 1 0 0 1 1 1 0 1 1 1 1 0 2 0 0 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 0 0

eadicinuE 0 1 1 0 ? ? 1 0 0 0 2 1 2 0 0 0 0 1 1 1 1 1 1 1 3 1 ? 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eadihpunO 0 1 1 0 ? ? 1 0 0 0 2 1 2 0 0 0 0 1 1 1 1 1 1 1 3 1 ? 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eadiellivroD 0 1 1 0 ? ? 1 0 0 0 1 1 2 0 0 0 1 1 1 1 1 1 1 1 3 1 0 1 0 0 1 1 1 1 0 1 1 1 0 0 2 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

adinoipS 0 1 1 0 ? 1 1 0 0 0 1 1 3 0 0 0 1 2 2 1 3 1 1 1 3 2 0 1 0 0 1 2 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

adillebereT 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 3 2 2 4 1 1 1 3 2 0 1 0 0 2 2 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 ? 0 0 0 0 0 0 1 0 0

adillebaS 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 3 2 2 4 1 1 1 3 2 0 1 0 0 2 2 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 1 0 0

ainewO 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 2 2 2 4 1 1 1 4 2 0 1 0 0 2 2 1 1 0 1 1 1 0 1 3 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 2 1 1

aleocoretnE 0 1 1 1 ? 1 1 1 0 0 1 1 3 0 0 0 2 2 2 2 4 1 1 1 4 2 0 1 0 0 2 2 ? ? 0 1 1 2 0 1 ? 0 0 0 0 0 ? 1 1 1 1 1 0 0 0 0 0 3 1 1

adinoaraP 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 1 4 1 1 1 3 2 0 1 0 0 1 2 1 1 0 1 1 2 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 0 0

eaditseuQ 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 4 1 1 1 3 2 0 1 0 0 1 2 1 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 1 1 1 1 1 4 0 0

atalletilC 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 5 1 1 1 4 2 0 1 0 0 3 2 1 1 0 1 1 2 0 1 ? 0 0 0 0 0 ? 1 1 0 0 0 1 1 1 1 1 4 0 0

eadillehpO 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 4 1 1 1 3 2 0 1 0 0 3 2 2 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadilletipaC 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 4 1 1 1 4 2 0 1 0 0 1 2 2 1 0 1 1 1 0 1 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadilocinerA 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 4 1 1 1 3 2 0 1 0 0 1 2 2 1 0 1 1 1 0 2 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

eadinadlaM 0 1 1 1 ? 1 1 0 0 0 1 2 4 0 0 0 3 4 2 2 4 1 1 1 3 2 0 1 0 0 1 2 2 1 0 1 1 1 0 2 4 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0

36 Almeida et al.


for these three taxa, although further confirmation inParalacydoniidae is needed.

In some Polynoidae (here included in Polynoidea) theprostomium has small projections, or “prostomial peaks”(PETTIBONE 1963, 1989; DAY 1967) (Tab. I, #10). There is noinformation on the function of these prostomial peaks, butfind it reasonable to suppose that they have a tactile-sensorialfunction. In Polynoidea, prostomial projections are also foundin acoetids and aphroditids (PETTIBONE 1989; WATSON RUSSELL1989; HUTCHINGS & MCRAE 1993). In the latter groups thesesensitive structures occur in the same position, but maysometimes be photosensorial, because of the presence of ocelliin some acoetids (see PETTIBONE 1989), and in other cases onlytactile-sensitive, as in Laetmonice producta Grube, 1878 (seeHUTCHINGS & MCRAE 1993). Although the resolution ofinterrelationships between all the scaly worms lies beyond theobjectives of this paper, some decisions regarding the basalplan of Aphrodita have made (ALMEIDA & CHRISTOFFERSEN 2000).Eupanthalis McIntosh, 1876 may be one of the most basalgroups among the Acoetidae, because of the high number ofbody segments, lateral antennae, and absence of ocularpeduncles. These characters contrast in particular with manypolynoids. On the other hand, some genera of Polynoidea (e.g.Lepidonotopodium Pettibone, 1983; PETTIBONE 1983, 1984;DESBRUYÈRES & HOURDEZ 2000a, b) may have less than 60 bodysegments, an oval body shape, while lateral antennae areabsent. These characters are also shared with Aphrodita, andthus genera such as Lepidonotopodium may represent a transitionbetween polynoids and aphroditids, while acoetids could haveother still undetermined relationships among the scaly worms.

ROUSE & FAUCHALD (1997) considered the spinning glandsto be synapomorphic for Acoetidae and Aphroditidae. However,they may represent independent acquisitions in these two taxa.The same applies to the ocular peduncles, which do not appearto be present in the basic plan of the Acoetidae.

Comparing the ocular peduncles of Aphroditidae withthe prostomial peaks present in genera of polynoids such asLepidonotopodium (see PETTIBONE 1983, 1984; DESBRUYÈRES &HOURDEZ 2000a, b), the following hypotheses may be made: (1)they represent independent structures; or (2) they arehomologous. The first hypothesis begs the question of whygroups so close as polynoids with prostomial peaks andaphroditids with ocular peduncles would have independentlydeveloped structures which are so similar in position andfunction. It becomes perfectly plausible to consider that theprostomial peaks represent an evolutionary step that precedesthe formation of ocular peduncles. For example, the developedocular peduncles, such as those of Pontogenia Claparède, 1868(HUTCHINGS & MCRAE, 1993: 310, fig. 52a), could not have arisenin a single step without a precursor. The minute ocularpeduncles such as those of Aphrodita terraereginae Haswell, 1883(HUTCHINGS & MCRAE 1993: 310, fig. 25a) could represent anintermediary evolutionary step between prostomial peaks andwell developped ocular peduncles. Instead of creatinghypothetical structures for the origin of ocular peduncles inAphroditidae, it is preferable simply to search their closestpolynoid relatives (e.g., Lepidonotopodium; PETTIBONE 1983, 1984)for plausible structures of similar function and relative positionto reach a hypothesis of homology. By character congruence,it is more parsimonious to hypothesize a transformation seriesgoing from prostomium without sensitive structures, passing

through a stage of prostomial peaks present, and culminatingwith ocular peduncles, than to accept the alternative scenario:sensory structures absent, present, lost (in the ancestralpolynoid of the Aphroditidae), and the reinvention of sensitivestructures in the same position as the prostomial peaks.

The absence of ocular peduncles in Aphrodita (HUTCHINGS& MCRAE 1993) may represent a loss shared with Ecdysozoa,which would account for their absence in the latter taxon.Nevertheless, as discussed above, homologies between theanterior regions of Polychaeta and Ecdysozoa need furtherstudy.

The definition of the peristomium is also confusing inFAUCHALD & ROUSE (1997: 77) because it varies a great deal inshape in adults (or may be restricted to an area surroundingthe mouth, or to the roof of the mouth, or lips, or may formone elongated segment). The peristomium is also described asof prototrochal origin (FAUCHALD 1977). In the adults it may berestricted to the epidermal folds (lips) surrounding the mouth(FAUCHALD & ROUSE 1997). They may also be identified as ring-shaped (BRINKHURST & JAMIESON 1971; EIBYE-JACOBSEN 1994). Finally,different degrees of fusion with the prostomium and/or withthe first metamere have also been characterized (FAUCHALD &ROUSE 1997).

The problem may lie in the notion that everything thatis situated in a prototrochal region should be non-metameric(GILPIN-BROWN 1958). The example of the first pair of tentacularcirri may be illuminating. These structures are identical to theremaining tentacular cirri (ORRHAGE 1993). However, becauseof their prototrochal embryonic origin they have for a longtime been interpreted as pre-segmental structures (SCHRÖDER &HERMANS 1975: 161). But if segmental structures, and consequen-tly representing the first metamere, can change positions duringembryonic development, as appears to be the case of the secondtentacular cirri pair of Nereididae (GILPIN-BROWN 1958), thenthe apodous rings in this group and many Eunicidae, SedentariaAudouin & Milne Edwards, 1833 and Clitellata (Figs 4 and 5),considered to be peristomial, may actually represent cephalizedmetameres. This would explain why peristomial lips “strangely”simulate apodous metameres (in the form of the rings).

The idea that embryonic regions would be static andimmutable during development is not new. For example, untilrecently the differences among embryological fate maps forseveral phyla have been taken as evidences for the polyphyleticorigins of the Metazoa (ANDERSON 1982). COLLAZO & FRASER (1996)and COLLAZO (2000) have stressed, however, that embryologicalvariations are expected during the development of anyindividual. In their words, “evolution depends on genotypicvariation to proceed”. Thus, valid characters form dynamicspectra of particular semaphoronts. The first tentacular cirri inNereididae, although having neither chaetae nor parapodia,should not be interpreted as non-metameric. Comparativelythe development of other tentacular cirri show ventral ganglia,parapodia, and chaetae that are lost and the appendages moveforwards, fusing with the mouth region (GILPIN-BROWN 1958;SCHRÖDER & HERMANS 1975). The first pair of tentacular cirri ofthe Nereididae should be a step towards total cephalizationfrom a state similar to other tentacular cirri. Hence, theperistomium is homologized exclusively with the peristomial lips(Tab. I, #11), which is not the same as considering rings orsegments to be cephalized. The peristomium structure mustrepresent a synapomorphy for all metamerians. The peristomial



labiae become subsequently hypertrophied in Eunicidae andOnuphidae. Tentacular cirri, as mentioned above, must bereferred to as cephalized appendages of metameric origin.

In Ecdysozoa there is apparently no structure equivalentto the peristomium. Nevertheless, Xenusion auerswaldaePompeckj, 1927 has the buccal region surrounded by papillae(DZIK & KRUMBIEGEL 1989). These may correspond to theperistomial labiae or even to the buccal papillae of thePolychaeta. Because there is no strong evidence for thishomology, this characters was coded with a “?” for theEcdysozoa.

Prostomial and peristomial appendages: palps andlophophores

There has been much uncertainty regarding thehomology of the cephalic appendages of Polychaeta. The palpsof Nephtyidae Grube, 1850 were considered to be an extra pairof antennae (see FAUCHALD 1977), or the presence of sensorypalps were believed to be restricted to the Nereidoidea George& Hartman-Schröder, 1985 (GLASBY 1993). Much of thisconfusion has been clarified by anatomical investigations ofthe cerebral nerves (ORRHAGE 1966, 1974, 1978, 1980, 1990,1991, 1993, 1995, 1996; PURSCHKE 1993).

From these studies it has been recognized that all palpsare homologous structures (Tab. I, #12). Contrary to ROUSE &FAUCHALD (1997), optimization analyses show that the sensitive,ventral position of these organs shall represent theplesiomorphic condition (Tab. I, #13). The presence of ventro-lateral palps in Eunicida must represent an intermediary statein relation to the grooved feeding palps of the Spionida. Longand robust sensitive palps are a synapomorphy of the scaleworms (Tab. I, #14). On the other hand, stout palps and bi-articulated palps (Tab. I, #15, #16) are interpreted here asautapomorphies respectively of Phyllodocidae and Nereididae.The subsequent state of this series is represented by the ventro-lateral palps beginning in the Nephtyidae (Fig. 4).

Grooved palps with the functions of feeding andrespiration (Tab. I, #17, #18) represent a strong synapomorphyof the sedentarians, associated with a radical change in theirmode of life. The palps of Dorvilleidae, despite not having aciliated groove as in Spionida, Terebellida, and Serpulida, mustrepresent a step immediately preceding the grooved palps,considering their shape and relative size. There is no greatobstacles to this proposal of homology, because Magelonidae(here included in Spionida), also do not have a ciliated groovein their palps.

Losses of palps occur several times in Metameria inconnection with habitat shifts. This should not appearsuspicious. In several groups in which palps are already absent(Scalibregmatidae and Paraonidae), there are still clearindications of vestigial inervation of these palps (ORRHAGE 1966,1993). The inverse hypothesis, that the presence of nerves is aprevious condition to the occurrence of palps, is not feasible,because the innervation found in Scalibregmatidae andParaonidae do not innervate particular prostomial structures.This inervation becomes completely lost in the Clitellata.PURSCHKE et al. (1993) recognize that these losses of theprostomial appendages represented an autapomorphy for theClitellata (PURSCHKE et al.,2000). Stout and articulated palpsbecome a nested synapomorphy of the Nereididae Johnston,1865 and Pilargidae Saint-Joseph, 1899 (Figs 4 and 5).

Due to hypotheses of primary homology and to charactercongruence, short and non-articulate prostomial antennaemust be present in the groundplan of the Metameria (Tab. I,#19). These antennae change their position from frontal tooccipital in the Eunicida. They may be reduced to occipitalpapillae in many specialized groups such as the Eunicidae, ormay be totally absent in others, such as the Myzostomidae,Ichthyotomidae, Hartmaniellidae, and Aeolosomatidae.Prostomial antennae are also absent in Sabellida, Terebellida,Owenia, and the Enterocoela (the basal group of which are thepogonophorans having a prostomium without antenna, seeCHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA 1994) (Figs 4 and 5).

The median antenna does not have the same generalityas the lateral antennae (Tab. I, #20). This structure is absent inthe outgroup and in the most basal taxa of Paralacydoniidae,Glyceridae, Goniadidae, Pisionidae, and Nereididae, forexample.

A possible homology between the cephalic appendagesof Polychaeta and the mesocoelic appendages of Enterocoelais still controversial. They have a prostomial origin in mostpolychaetes (FAUCHALD 1977), while they originate from the“peristomium” (or mesocoelic region) in the phoronids (SALVINI-PLAWEN 1982). The first pair of tentacular cirri in Nereididae isalso peristomial, having a similar inervation pattern to theprostomial appendages of other polychaetes (BLAKE 1975; GLASBY1993; ORRHAGE 1993). In other words, the first somite ofNereididae was added to the cephalic region, appearingconnected to the peristomium already in the embryo. Maybea similar embryonic shift in position has also occurred fromthe prostomial palps of Pogonophora to the peristomial(mesocoelic) lophophores of the Radialia Ax, 1989. Actually,the entire transformation series involved in the shift from anerrant to a burrowing mode, and then to a tubicolous habit isrelated to a gradual reduction of the posterior parapodia and aparallel development of the anterior appendages in the cephalicregion. Hence, the lophophore appears to be homologous tothe well-developed tentacles seen in the Sabellida and inOwenia.

Comparing the cephalic region of polychaetes with thebasal groups of Ecdysozoa is still more difficult. The lobopodCambrian fossils Xenusion Pompeckj, 1927, HallucigeniaConway Morris, 1977, and Microdyction Bengtson, Matthews& Missarzhevsky, 1981, Onychodictyon Hou, Ramsköld &Bergström, 1991, for example, have no cephalic appendages.Aphrodita does not have lateral antennae, but still has arudimentary median antenna (HUTCHINGS & MCRAE 1993). Thisindicates that these antennae are gradually lost in this lineage.We thus interpret the prostomial antennae as being reducedin Aphrodita and lost in Ecdysozoa. In fact, there is little supportfor the presence of prostomial antennae in ecdysozoan lineages.

The cephalic palps may also have been lost in theEcdysozoa, although these structures still appear to be welldeveloped in scale worms. In another Cambrian lobopod,Kerigmachela, however, there are two frontal tentacles (BUDD1993) that could represent modified palps (ALMEIDA & CHRISTOF-FERSEN 2000: 43).

Although these suggestions of secondary reductions ofcephalic appendages sound speculative, they may be less far-fetched than trying to establish homologies between suchdistant structures as the palps of Aphrodita and the first pair ofantennae in Crustacea.

38 Almeida et al.


Parapodia and lobopodia: homologous locomotoryappendages

Parapodia are extensions of the body wall functioningin locomotion and for protection (FAUCHALD 1974; PETTIBONE1982; BRUSCA & BRUSCA 1990). The appearance of parapodia inanimal evolution may be related to the development of themesenteric septa and of the ventral ganglia. As a result of theseinnovations we obtain a functional complex that includesblood irrigation and the nervous control of parapodial muscles(BONICK et al. 1976; WESTHEIDE 1997). These correlations arecorroborated by the congruence of all these individualcharacters in the present analyses.

The most generalized parapodium in our analysis has aneuropodium more developed than the notopodium(“projecting neuropodia”) (Tab. I, #21). This condition occursin most of our basal groups of Metameria. It had been expectedthat a basal parapodium similar to that hypothesized byFAUCHALD (1974) would be found, in which both notopodia (Tab.1, #22) and neuropodia (Tab. 1, #23) were equally developed.This opinion turned out to be equivocated, however, since suchsymmetrical parapodia are only found in Nephtyidae andAmphinomidae. Equal rami of the parapodia thus become aconvergence in Nephtyidae and Amphinomidae.

Among polychaetes, the scale worms (e.g., HarmothoeKinberg, 1855, Lepidonotus Leach, 1816, and Aphrodita) havethe most efficient parapodia for locomotion (METTAM 1971:510). While most ‘errant polychaetes’ depend on lateralundulations of the body for the functioning of their parapodia,Aphrodita use only the parapodia for movement (METTAM 1971:512) (Tab. I, #24). Furthermore, Aphrodita also develops a seriesof diagonal parapodial muscles, which dramatically increasethe muscular complexity of these parapodia (STORCH 1968;METTAM 1971; PILATO 1981). ALMEIDA & CHRISTOFFERSEN (2000)argumented these characters are possibly synapomorphiesshared by Aphrodita and Ecdysozoa (Fig. 6). They demonstratedthat MANTON (1952, 1967, 1969, 1973, and 1977) based herarguments for the homology of lobopodia and arthropodia onthe presence of these new diagonal muscles (STORCH 1968;METTAM 1971; PILATO 1981). Arguments of Manton against thehomology of parapodia and lobopodia were based on thesupposed absence of these muscles in Polychaeta (MANTON 1967:11). The muscle structure of the parapodia of the Aphroditashows that the main argument used to criticize the derivationof Arthropoda from Polychaeta is based on incorrect premises(ALMEIDA & CHRISTOFFERSEN 2000).

In proposals suggesting the derivation of arthropodiafrom parapodia (WALTON 1927; SHAROV 1966; LAUTERBACH 1978)the notopodia of polychaetes are homologous to the exopoditesof arthropods, and the neuropodia are homologous to theendopodites. The matter regarding which lobopodiansrepresent the most basal ecdysozoans is still unresolved. If formswith biramous appendages such as Kerygmachela (BUDD 1993)turn out to be most basal, no changes in our scheme should benecessary. On the other hand, if uniramous forms such asXenusion auerswaldae turn out to be most basal (BUDD 1996), itwould become necessary to consider that only theneuropodium is homologous to the lobopodium andarthropodium. In this case the biramous condition found indifferent lobopodian and arthropod clades would not behomologous. BUDD (1996) sustains this last scenario for hisphylogeny of the lobopodians. However, in order to make sense

of these two conflicting hypotheses, if the homology betweenelytrae and dorsal plates is accepted, it would be necessary tosustain that the ancestral ecdysozoan was uniramous andarmored, the opposite of what BUDD (1996) suggested. Theresolution of these problems requires a joint analysis of thephylogeny of the Ecdysozoa with the Lobopodia.

ZRZAVÝ & STYS (1995) provided other arguments againstthe homology of parapodia and arthropodia. They analyzedthe expression of gene en (related to the expression of parapodiaand arthropodia) and wg (related only to arthropodia). Asalready considered in ALMEIDA & CHRISTOFFERSEN (2000), thereare disagreements with the conclusions of ZRZAVÝ & STYS (1995).Despite detecting differences among taxa, these authors didnot prove that parapodia and arthropodia have a non-homologous genetic basis. More recent molecular data

Figure 6. Hypothesis of ALMEIDA & CHRISTOFFERSEN (2000) with mainsynapomorphies of the Holopodia (Aphrodita + Ecdysozoa). (A)Errant polychaetes represented by parapodia of Aglaophamusgippslandicus (Nephtyidae) (modified from IMAJIMA & TAKEDA, 1985);(B) Aphrodita represented by neuropodia of Aphrodita aculeataLinnaeus, 1761 (modified from HARTMAN 1965); (C) Ecdysozoarepresented by lobopod of Peripatus sp. (modified from HARMER etal. 1997). Ch, chaetae; Cw, claws; Ep, epidermal papillae; Lb,lobopod; Ne, neuropodia; No, notopodia.



presented by PANGANIBAN et al. (1997) show that arthropodiaand parapodia share the same genetic expression for the geneDistal-less (Dll). Even so, PANGANIBAN et al. (1997) do not proposethe homology of parapodia with arthropodia, supposedlybecause of the contrary opinions of FIELD et al. (1988) andWINNEPENNINCKX et al. (1995). As discussed above, sequence datahave an inferior heuristic value in relation to data that correlategenes with phenotypes.

The homology between parapodia of polychaetes andchaetal bundles of oligochaetes has also been a matter of somequestioning. According to BRINKHURST (1984a), the arrangementof chaetae into dorso-lateral and ventro-lateral bundles is themost efficient disposition for burrowing and should beconsidered the most derived state. In this scenario the mostbasal annelid should have the body entirely surrounded bychaetae. This pattern would be congruent with the pattern ofspicules and hooks present in aplacophoran molluscs(HASZPRUNAR 1992; SCHELTEMA et al. 1994) and sipunculans (STEPHEN& EDMUNDS 1972; RICE 1993; CUTLER 1994). Unfortunately forthis hypothesis, chaetae surrounding the body are only foundin derived forms of terrestrial Oligochaeta. TIMM (1981) indicatesthat groups such as Haplotaxidae Michaelsen, 1900 andTubificinae Eisen, 1879, in which the chaetal bundles shouldbe vestiges of parapodia, represent the basal forms of Clitellata.The basal groups of oligochaetes indicated by the analysis ofBRINKHURST (1971, 1982, 1984a, b, 1988, 1989, 1991a, b, 1992 a,b, 1994), BRINKHURST & NEMEC (1987), and OMODEO (1998) allhave dorso-lateral and ventro-lateral chaetal bundles. Suchchaetal bundles do not occur in the burrowing outgroups ofthe Metameria (Sipuncula and Mollusca), and thus do notappear to be “advantageous” for excavation after all, assuggested by BRINKHURST (1984a). Consequently, there are noreal obstacles for considering parapodia of polychaetes andchaetal bundles of clitellates as homologous structures. A likelycandidate for the closest outgroup of the Clitellata, theQuestidae also has parapodia considerably reduced as aconsequence of their burrowing habit. Chaetal bundles areinferred to represent one more successive character state inthe evolution towards fossorial habitats, which began inlineages of marine polychaetes and culminated in the conquestof land in some lineages of clitellates. In other burrowing andtubicolous polychaetes (scolecids and Spionida), the parapodiaare also reduced when compared to those found in the errantpolychaetes. The transformation series affecting parapodia maybe ordered from (1) projecting parapodia, to (2) spioniformparapodia, to (3) parapodia reduced to tori (which includesPogonophora, the most basal group of Enterocoela)(CHRISTOFFERSEN & ARAÚJO-DE-ALMEIDA 1994) to (4) the exclusivepresence of chaetal bundles in the scolecids and clitellates(originally in two pairs per segment) (Fig. 5). All these statesrepresent successive stages in the conquest of a sedentary wayof life and then of continental environments.

Parapodial cirri, parapodial branchiae, elytrae, andlobopodian dorsal plates

Parapodial cirri are structures fundamentally involvedin sensory functions (PETTIBONE 1982). Both dorsal (Tab. I, #25)and ventral cirri (Tab. I, #26) are quite generally distributed inthe Metameria. It is hypothesized that they were already presentin the basal plan of this taxon, and their presence in glyceridsgives additional support to this point of view.

Sometimes the evolution of the parapodial cirri mayoccur by parallel transformation series. For example, i

Documents

Polychaeta, Annelida, and Articulata are not monophyletic ......(1998) suggested that the data used by MCHUGH (1997) were insufficient to decide the position of Echiura and Pogonophora,