Zootaxa 1668:245–264 (2007) ISSN 1175-5326 (print edition) www.mapress.com/zootaxa/ ZOOTAXA Copyright © 2007 · Magnolia Press ISSN 1175-5334 (online edition) Annelida* GREG W. ROUSE1 & FREDRIK PLEIJEL2 1Scripps Institution of Oceanography, UCSD, 9500 Gilman Drive, La Jolla CA, 92093-0202, USA. E-mail: [email protected] 2Department of Marine Ecology, Tjärnö Marine Biological Laboratory, Göteborg University, SE-452 96 Strömstad, Sweden. E-mail: [email protected] *In: Zhang, Z.-Q. & Shear, W.A. (Eds) (2007) Linnaeus Tercentenary: Progress in Invertebrate Taxonomy. Zootaxa, 1668, 1–766. Table of contents Abstract . .245 Introduction . .245 Major polychaete taxa . .250 Monophyly of Annelida . .255 Molecular sequence data . 258 Rooting the annelid tree . .259 References . 261 Abstract The first annelids were formally described by Linnaeus (1758) and we here briefly review the history and composition of the group. The traditionally recognized classes were Polychaeta, Oligochaeta and Hirudinea. The latter two are now viewed as the taxon Clitellata, since recognizing Hirudinea with class rank renders Oligochaeta paraphyletic. Polychaeta appears to contain Clitellata, and so may be synonymous with Annelida. Current consensus would place previously rec- ognized phyla such as Echiura, Pogonophora, Sipuncula and Vestimentifera as annelids, though relationships among these and the various other annelid lineages are still unresolved. Key words: Polychaeta, Oligochaeta, Clitellata, Echiura, Pogonophora, Vestimentifera, Sipuncula, phylogeny, review Introduction Annelida is a group commonly referred to as segmented worms, found worldwide in terrestrial, freshwater and marine habitats. The first annelids were formally named by Linnaeus, including well-known forms such as the earthworm Lumbricus terrestris Linnaeus, 1758, the medicinal leech Hirudo medicinalis Linnaeus, 1758, and the sea-mouse Aphrodite aculeata Linnaeus, 1758. Today we estimate that the current number of accepted species level taxa is around 14 000 (Rouse & Pleijel 2006), though several thousand more have been named and are considered invalid. Lamarck (1802) first used the term ‘Annélides’ when naming a group of organisms taken from the broad taxon Vermes erected by Linnaeus. The name Annélides was based on the Latin word anellus, meaning a little ring, in reference to the presence of ring-like segments that characterize the group. Earthworms are the most familiar annelids to people, but the bulk of the annelid diversity lies among the marine representatives, which are found in nearly every marine habitat, from beach sands and Accepted by Z.-Q. Zhang: 27 Nov. 2007; published: 21 Dec. 2007 245 intertidal zones, to the plankton and down in the deep-sea sediments. Through much of the 19th and 20th cen- tury, Annelida was usually divided into four classes; Archiannelida erected by Hatschek (1878), Polychaeta by Grube (1850), Oligochaeta by Grube (1850), and Hirudinea by Lamarck (1818). A number of authors used the name Clitellata, introduced by Michaelsen (1919), recognizing that Oligochaeta (earthworms and close relatives) and Hirudinea (leeches) were actually a clade. Other taxa that have been considered to be annelids at various times and with varying ranks include Echiura Newby, 1940, Myzostomida Graff, 1877, Pogonophora Johansson, 1939 and most recently Sipuncula Rafinesque, 1814. Archiannelida was erected for a group of generally minute annelids, such as Dinophilus Schmidt, 1848 (Fig 1A), which were presumed to be primitive because of their simple body-structure. This perceived sim- plicity is now regarded as secondary, related to the interstitial habitat and associated small size of the animals (Hermans, 1969; Westheide, 1985). This limits the number of currently accepted annelid classes to two, Poly- chaeta and Clitellata, but recently serious doubts about the monophyly of the former taxon have been raised and will be reviewed here. Polychaeta and the subdivision of the group has been unstable for many years. Over eighty families of polychaetes are currently recognized (Rouse & Pleijel 2001). Until relatively recently the system most commonly used for classifying polychaetes into higher groups was derived from a pre-evolu- tionary system used in the early 19th century, where Polychaeta was split into Errantia (wandering forms) and Sedentaria (sedentary forms), mainly based on whether they were mobile, or lived in tubes or burrows (e.g., Day 1967a, b; Hartman 1968, 1969). This utilitarian classification was gradually supplanted by ones that split the Polychaeta into as many as 22 orders but without any real linkage between them (Dales, 1962; Fauchald, 1977). This situation was also unsatisfactory and gave no useful insight into polychaete evolution. A morpho- logical cladistic analysis of Annelida and other groups resulted in a new classification of polychaetes (Fauchald & Rouse 1997; Rouse & Fauchald 1997), which is outlined here (see below) with modifications (Rouse & Pleijel 2001), with the caveat that much of this is now being challenged by results from the analysis of molecular sequence data. There is increasing evidence that Clitellata (Fig. 2A, B), along with Echiura (Fig. 1C), Pogonophora (Fig. 5A), Sipuncula (Fig. 1I), and other groups, may well belong inside Polychaeta (see below), thus making the name Polychaeta synonymous with Annelida (McHugh 1997; Westheide 1997; Westheide et al. 1999). Recent molecular studies have all shown that Clitellata are nested among polychaetes (Bleidorn et al. 2003a, b; Brown et al. 1999; Jördens et al. 2004; McHugh 1997; Struck et al. 2002), but none to date have provided robust support for any sister group relationship with a particular polychaete group. No doubt such a relation- ships will be recovered soon and either the name Polychaeta or Annelida will be redundant. The monophyly of Clitellata [oligochaetes and leeches (Fig. 2A, B)] is strongly supported on both morphological and molecu- lar grounds (Martin 2001; Siddall et al. 2001). Comprehensive phylogenetic studies using molecular sequence data and morphology provide strong support that Lumbriculida is the sister group to the ectoparasitic clade comprised of Hirudinida, Acanthobdellida and Branchiobdellida (Erséus & Källersjö 2004; Martin 2001; Sid- dall et al. 2001) and should be referred to either as Oligochaeta (Siddall et al. 2001), or Clitellata (Erséus & Källersjö 2004; Martin 2001). There are arguments for using either name with respect to a monophyletic taxon but we have used the name Clitellata here. Morphological features supporting monophyly include the presence of a clitellum, the organization of the reproductive system, and features of sperm ultrastructure (Jamieson 2006). FIGURE 1. —Various annelid groups A. Dinophilus sp. from California (Dinophilidae), dorsal view. —B. Aeolosoma sp. (Aeolosomatidae), dorsal view. —C. Metabonellia haswelli (Johnston and Tiegs, 1920) (Echiura) from the Australia, female specimen. Inset shows Capitella sp. (Capitellidae) anterior, now thought to be the sister group to Echiura. —D. Hypomyzostoma dodecephalcis Grygier, 1992 from Australia, on crinoid host, lateral view. —E. Hypomyzostoma dode- cephalcis, ventral view of entire specimen showing parapodia and everted pharynx. F. Polygordius lacteus, lateral view of head and anterior body. —G. Protodrilus rubropharyngeus, whole animal. —H. Nerilla antennata, whole animal. —I. A phascolionid sipunculid. All photos G. Rouse. 246 · Zootaxa 1668 © 2007 Magnolia Press LINNAEUS TERCENTENARY: PROGRESS IN INVERTEBRATE TAXONOMY ROUSE & PLEIJEL: ANNELIDA Zootaxa 1668 © 2007 Magnolia Press · 247 FIGURE 2.—Clitellata and Scolecida. A. Lumbricus terrestris (Clitellata, Lumbricidae) from USA, dorsal view of entire specimen. —B. Macrobdella decora Say, 1824 (Clitellata, Hirudinidae) from USA, dorsal view of entire specimen. —C. Scoloplos arimiger (O.F. Müller, 1776) (Scolecida, Orbiniidae) from Sweden, whole specimen. —D. Euzonus mucrona- tus (Treadwell, 1914) (Scolecida, Opheliidae) from California, whole specimen. —E. Clymenella sp. (Scolecida, Orbini- idae) from Belize. All photos G. Rouse, except 2B. courtesy of Mark Siddall. 248 · Zootaxa 1668 © 2007 Magnolia Press LINNAEUS TERCENTENARY: PROGRESS IN INVERTEBRATE TAXONOMY Echiura (Fig. 1C) were originally considered members of the same group as the annelids (Cuvier 1817; Lamarck 1818). They were grouped with sipunculids and priapulids in the Gephyrea by Quatrefages (1847), a group subsequently not considered valid by Sedgewick (1898). Sedgewick considered echiurids as a group of annelids, but Newby (1940) proposed a separate phylum, Echiura, based on a detailed embryological study of Urechis caupo. His proposal was generally accepted (e.g., Stephen & Edmonds 1972), until McHugh (1997) provided molecular evidence that they should again be considered as annelids. Subsequent morphological studies support this conclusion (Hessling 2002; Hessling & Westheide 2002) and Echiura appear to now be closely related to capitellid polychaetes (Fig. 1C) (Colgan et al. 2006; Rousset et al. 2007; Struck et al. 2007). The taxon Myzostomida was erected by Graff (1877). During the 20th century these animals (Fig. 1D, E) were treated as polychaete annelids with a rank of family or order (Hartman 1969; Pettibone 1982), or treated, not as polychaetes, but as a class of annelids (Dales 1963). Rouse & Fauchald (1995; 1997) argued that Myzostomida were a polychaete group and placed them, after their cladistic analyses, as part of Phyllodocida. Zrzav! et al.