A review of the classification of the Class Polychaeta (Annelida) with comments on the characters used to identify the different included taxa has led to the recognition of seventeen orders. All taxa down to the generic level are defined and a phylogenetic sequence suggested. Keys are presented to the families and genera of the Polychaetes. and the setae, if present, only rarely occur in bundles. These two groups were considered more advanced than the marine, dioecious polychaetes. The polychaetes have been defined for the last seventy years as dioecious, marine annelids with parapodia bearing numerous setae. They also should have anterior appendages of various sorts (antennae, palps, tentacular cirri) and the gonadal ducts should be simple. These definitions work if some of the smaller groups are disregarded. If these groups are taken into account, as they must, the only separation that consistently can be made between the oligochaetes/leeches and the polychaetes, is the presence in the former grouping of hermaphroditic gonads limited to a few segments. Some hermaphroditic polychaetes are known, but these usually have gonads in a large fraction of the total number of segments. It is then difficult to give a good, consistent and practically useful definition of what is meant by a polychaete, but a definition along the lines suggested below, should separate them from the other annelids with reasonable accuracy. The polychaetes are multi-segmented annelids with parapodia; setae are present in distinct fascicles. They are dioecious and have simple exit ducts from the 1 REVIEW COMMITTEE FOR THIS SCIENCE SERIES ROBERT J. LAVENBERG DONALD REISCH ‘Allan Hancock Foundation, University of Southern California, Los Angeles, California, 90007. Contribution Number 358 of the Allan Hancock Foundation. 2 NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY Science Series 28 gonads. They are usually marine, more rarely freshwater and only rarely terrestrial or parasitic in habitat. Any of these features need not be present and none of them is essential for the recognition of an animal as a polychaete. This topic has been treated in considerably greater detail by Clark (1969) and to a lesser extent by Fauchald (1974a). A key morphological feature and at the same time one of the most important taxonomic characters of the polychaetes is the setal (chaetal) construction. The setae are ectodermal derivatives, formed by ectodermal cells that during the development have migrated to a position well below the rest of the ectodermally derived epidermis. Each seta consists of a bundle of filaments laid down by a basal chaetoblast and up to several lateral cells. The material in the setae is a glycoprotein, consisting of chitin (a polysaccharide) and a protein cross-linked at the time of formation. The formation of structural details in the setae is very well controlled, but exactly how this takes place has only partially been clarified. The best current review of this topic was made by O’Clair and Cloney (1974) from which most of the above information has been gleaned. Polychaetes traditionally are separated into two large orders, ERRANTIA and SEDENTARIA (Audouin and Milne Edwards 1834, pp. 24-26). The separation is based on the development of the anterior end and the life habits of the included species. The errants are supposed to have a large number of equal body-segments. The anterior appendages are few in number and differentiated into palps, antennae, tentacular cirri, etc. These worms are considered freeliving and, generally, should be rapacious in habits. All polychaetes with jaws are included in this order; thus the onuphids, despite their tubicolous habits, are considered errants since their large jaw-apparatus resembles the jaw-apparatus in other, non-tubicolous eunicidlike animals. The sedentaries are supposed to have a limited number of body segments. The body may be separated into different regions. Anterior appendages may be absent or a few to many similar appendages may be present. The sedentaries have short parapodia associated with their tubicolous or burrowing habits and are usually depositor filter-feeders. These definitions have not changed much over time (cf. Grube 1850, p. 281 and tables; Fauvel 1923a, pp. 27-29; Hartmann-Schroder 1971, p. 29). The advantage of the system is that the bulk of the 8,000+ described species of polychaetes separates into two roughly similar groups in terms of numbers of species and genera as well as families. The separation is otherwise unsatisfactory since neither order can clearly be defined. Several attempts have been made to subdivide the polychaetes in a more acceptable manner (Dales 1 962, pp. 424-425; Clark 1969, p. 47). Polychaete taxonomists have tended to disregard these attempts and have continued to treat the polychaetes as if the class consisted of two orders (Fauvel 1958, pp. 1661 90; Hartmann-Schroder 1971, p. 29) or subclasses (Uschakov 1955a), or have treated the group as if it consisted of about 75 distinct and unrelated families (Hartman 1968, 1969). The problem with all proposed schemes is that they are internally inconsistent. Furthermore, they give no better solutions to classificatory problems than the old, admittedly artificial, separation into two orders. The three most ambitious recent proposals were by Dales (1962), Storch (1968) and Clark (1969). Dales used the variable structures of the eversible stomodeal region (pharynx) to separate different groups. The arrangement of the body-wall musculature was used by Storch. Clark used a variety of different structures to characterize his eight orders. These authors gave no formal definition of any taxon above the family level (except by inference from contained taxa) and it has been difficult to evaluate their schemes. CHARACTERS USED TO DEFINE HIGHER TAXA Major anatomical and morphological features were reviewed during a study of the phylogenesis of the polychaetes (Fauchald 1974a). Below is given a survey of the findings with an expanded discussion of their taxonomic aspects. A. Prostomium. The prostomium usually is distinct and may have or lack appendages. In several families it is more or less fused with the peristomium and the first segments. The degree of fusion is difficult to determine even in an examination of the nervous system so the degree of distinctness of the prostomium is a character that can have no great taxonomic value (see Benham 1894, 1896). Prostomial appendages include antennae and palps. Antennae are innervated through single roots directly from the brain; palps always have double roots, either from the brain or from the circumesophageal ring (Akesson 1963; Orrhage 1966). Antennae are always sensory; palps may be sensory or may be used as feeding appendages. The presence of either one or both categories of appendages is considered here of great importance. The position of the palps varies from ventral to dorsal, from frontal to occipital. The position and function of the palps furnish important taxonomic characters. It is impossible to distinguish any other classes of prostomial appendages either on morphological or anatomical grounds. B. Peristomium.
