ZoologicalJournal ofthe Linnean Society, 65: 367-384. With 6 figures April 1979 A reconsideration of the sipunculan taxa Fisherana Stephen, Mitosiphon Fisher and Apionsoma Sluiter EDWARD B. CUTLER Utica College of Syracuse University, Utica, New York 13502, U S .A. Acceptedfor publication February 1978 The sipunculan taxon Fisheram Stephen, erected as a genus, is critically examined. Type material of three of the four originally assigned species is reviewed and Colfing~~ warini is shown to be a junior synonym of G. lobostoma. Colfingia saualoui is reduced to a junior synonym of G . capitata. The subgenus Mitosiphon Fisher is considered in light of the recent discovery of the tentacular arrangement of G . murinae and G . misakiam. The tentacles do not surround the mouth bur lie dorsal to it as in Phascolosoma. Following an analysis of hook and papilla morphology, including scanning electron microscopy, F i s b a n a and Milosiphon are combined into one subgenus of Golfingia and the senior available name is applied: Apionsoma Sluiter. Keys to the species in this subgenus and to the subgenera of Golfingio are included. KEY WORDS: - Sipuncula - Fisheram - Mitosiphon - Apionsoma - taxonomy - Golfingiidae. CONTENTS . . . . . . . . . . . . . . . . . . . . . . Introduction Systematicsection . . . . . . . . . . . . . . . . . . . . . Mitosiphon species . . . . . . . . . . . . . . . . . . . . Golfingiamurim . . . . . . . . . . . . . . . . . . . Golfingia muakiana . . . . . . . . . . . . . . . . . . . Golfingia recondita . . . . . . . . . . . . . . . . . . . Golfingia trichocephola . . . . . . . . . . . . . . . . . . Fisherana species . . . . . . . . . . . . . . . . . . . . Colfingia saualoui . . . . . . . . . . . . . . . . . . . Golfingia capitata . . . . . . . . . . . . . . . . . . . Golfingiapapillqera . . . . . . . . . . . . . . . . . . . Golfingaa lobostoma . . . . . . . . . . . . . . . . . . . Golfingiawarini . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . Key to Apionsuma species . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . 367 369 369 369 370 3 72 3 13 313 373 373 374 316 376 377 38 I 383 385 383 INTRODUCTION The genus Fisherana was first erected in 1964 in a short paper by A. C. Stephen. It was subsequently included in Stephen 8c Edmonds’s 1972 monograph with this 367 0024-4082/79/040367 -1 8/$02.00/0 @ 1979 The Lirinean Society of London E. B. CUTLER 368 definition: “Small species with trunk spindle to subcylindrical in shape. Hooks on introvert with structure and shape like those of the genus Phascolosoma. Papillae dome-like to conical in shape, often capped with a small protuberance, and usually most densely packed at the anterior and posterior extremities of the trunk as in Phascolosoma. Longitudinal musculature of the body wall continuous as in Golfingza. One or two pairs of retractor muscles. Tentacles in some species arranged in a crescent or near circle which lies dorsal to the mouth as in Phascolosoma (the arrangement of the tentacles of the type species is not known)” (Stephen & Edmonds, 1972). Fisherana, type-species Phascolosoma papillijierum Keferstein, was erected to include a few species which seemed intermediate between the genera GoEfingia Lankester and Phascolosoma Leuckart. In Stephen’s (1964) key he distinguished it from Golfingia on the sole feature of hook complexity. Edmonds, who took on the task of completing the monograph after Stephen’s death, stressed the importance of the tentacular arrangement and on this ground placed the group in the newly established family Phascolosomatidae. Fisherana is further discussed by Murina ( 1974-a new species and a key to the species-and 1975, in a discussion of the evolution and phylogeny within the Sipuncula). In both papers Fisheranu is treated as a subgenus of Golfingia with some phasocolosomatid features. I concur in the placement of this taxon in the Golfingiidae. The subgenus Mitosiphon, erected in 1950 by W. K. Fisher to include G. hespera (Chamberlain) and G. misakiana (Ikeda), was defined as follows: “Retractors 4; longitudinal muscle layer not divided into bundles; nephridia bilobed; spindle muscle attached to posterior end of body; introvert hooks with an accessory comb of spinelets at base.” In Cutler (19731, where G. murinae Cutler and G. Table 1. Material upon which this analysis is based A. Museum specimens 1. Golfingiu cupitutu-Gerould’s type from National Museum of Natural History, Washington, D.C. 2. Golfingiu hesperu-Chamberlain’s type from Museum of Comparative Zoology, Harvard University 3. Golfingia hespera-from California, identified by W. K. Fisher, from National Museum of Natural History, Washington, D.C. 4. Go~ngialo~ostoma-Fischer’stypefromZoologischesMuseumUniversitAtHamburg 5. Golfrngia longirostris-identified by E. Wesenberg-Lund, from Universitetets Zoologiske Museum, Copenhagen 6. Golfingia misakiam-from Australia, identified by Fischer, from Museum fur Naturkunde HumboldtUniversitAt zu Berlin 7. Golfingia papibjera (sic)-identified by Sluiter, from Institut Ockanographique, Monaco 8. ColJingia reconditum-Sluiter’s type (external only)from Institut Ockanographique, Monaco 9. Goljngia lenuissima-identified by E. Wesenberg-Lund 10. Colfingia trichocephala-Sluiter’s type from Institut OcCanographique, Monaco I 1. Golfingia wnrini-Lanchester’s type from British Museum (Natural History), London B. More recent collections 1. Golfingiucapitata from the northwestern Atlantic 2. Golfingiu misuhiana-384 from Madagascar and Mombassa 3. Golfingiamzsahiuna-from Tanzania (courtesy of P. K. B. Menon) 4. Golfingiumisakianu-from East Africa (courtesy of G. Murina) 5. GolJingiu trichocephala-from southeastern United States (Cutler, 1973) 6. Go@ngiatrichocephala-off Durban (Stephen& Cutler, 1969) 7 . Golfingia trichocephala-267 from Madagascar and Arabian Sea 8. GolfingiaIrichocephala-from Australia (courtesy of S . J. Edmonds) 9. Golfingia trichocephula-from Texas and Florida Gulf coasts SIPUNCULAN TAXONOMY 369 trichocephala (Sluiter)were added to this subgenus, the diagnosis was modified/expanded as follows: “ Four retractors; nephridia either bilobed o r single lobed; spindle muscle attached to posterior end of body; introvert hooks, if present, possibly with accessory comb of spinelets at base; tentacles possibly reduced; introvert longer than trunk. ” In Cutler 8c Murina (1977) it was proposed that G. murinae should be moved to Fisherana on the basis of its now known tentacular arrangement. G. recondita (Sluiter)was also shifted into Mitosiphon but with some uncertainty as the type is unavailable for inspection. Apionsoma was erected as a genus in 1902 by Sluiter for his new species trichocephala. This taxon was transferred into Golfingia by Wesenberg-Lund (1959b)as a separate subgenus. Stephen & Edmonds (1972) subsequently placed it in the subgenus Golfingia sensu stricto. Sluiter’s description had several shortcomings. The type material was re-examined and redescribed by Cutler (1973). The analysis of a large number of specimens from several areas has provided me with important new data relevant to these m a . Table 1 lists the material analysed for this study. Reference collections of G. capitata (Gerould), G. misakiana, and G. trichocephala have been deposited in the American Museum of Natural History, New York. SYSTEMATIC SECTION Mitosiphon species Golfingza murinae Cutler, 1969 When this species (and its two subspecies) was first described it was based on animals with incompletely extended introverts. A reference was made to one individual which seemed to show the phascolosomatid tentacle arrangement. Subsequent material has produced several more specimens with completely extended introverts and confirms the presence of tentacles dorsal to the mouth (Fig. lA, B). Complex hooks and papillae are present. The suggestion made in Figure 1. GoIfingia murinue. A. Whole animal (scale 1 mm). B. Tip of introvert with oesophagus protruding through mouth ventral to tentacles (scale 1 mm). 22 370 E. B. CUTLER Cutler 8c Murina (1977) to move this into Fisherana was an error for reasons discussed later. Gogngia mzsakzana (Ikeda, 1904) Phascolosoma hespera Chamberlain, 1920 : 3 1 . Golfingia hespera Fisher, 1952: 392-395, pl. 24, figs 1-3. Stephen & Edmonds, 1972: 113. Amor, 1975: 115-1 16. The identity of Go&ngta misakiana has been one of the more difficult problems confronting sipunculan biologists. There have been five species names in the literature. Cutler (1967, 1973) discussed this problem and synonymized G. longzrostris and G. tenuissima with G. trichocephala. He also proposed informally that G. hespera might be a junior synonym of G. misakiana and concluded that G. trichocephala and G. misakiana are very similar. Since then, the acquisition of additional specimens has required a re-examination of this complex. From a collection taken at Dar es Salaam, P. K. B. Menon was able to determine the true tentacular arrangement of G. misakiana (personal communication). Menon’s specimens show the phascolosomatid array of tentacles dorsal to the mouth (Fig. 2). A large Indian Ocean collection (over 650 specimens) recently analysed by Cutler 8c Cutler (1979) also contributed to our knowledge. Unfortunately Ikeda’s types of G . misakiana cannot be located and no recent collections have been made around the type locality in Japan. The most recent published work on this species is by Amor (1975) in which she concurs with Cutler (1973) on most points but still retains the two species G. misakiana and G. hespera. The sole distinguishing feature she uses is the number of denticles/spinelets on the base of the hook. She maintains that G. hespera has 5-9 spinelets but some have no spinelets, while G. misakiana has 1-5 spinelets but some lack them. Amor also describes how the rows of hooks show differences-most spinelets being on the most distal rows of hooks while proximally the number of spinelets decreases to zero. This division of a population, otherwise identical, so that those with five or more spinelets are one species and those with five or less are another seems unnatural and arbitrary. The hooks from Fischer’s (1927)Australian specimens of G. misakiana exhibit 4 , 5 , and 6 spinelets. Several interesting points grew out of a study of hook morphology using the Figure 2. Go&ngia misakiam. Tip of introvert with oesophagus protruding through mouth ventral to tentacles (scale 0.5 mm). SIPUNCULAN TAXONOMY 371 scanning electron microscope and stimulated the following hypothesis. Considering the ontogeny of these structures and the introvert itself, it is proposed that when very young, the introvert is devoid of hooks, but at some point hooks begin to be produced near the growing tip. These early hooks are scattered, not in rows or rings, and these early hooks do not have spinelets (Fig. 3C). Only later are spinelets produced on the hooks, at first only a few, the full complement not appearing until maturity. Also, the growth of any one hook is a gradual process, growing out of the dermal tissue much the same way as a fingernail. This would mean that the most distal ring of hooks is in the early stages of growth and not yet fully erupted so that even if it may eventually have seven or eight spinelets, there is a time when only two or three are visible (Fig. 3A,B). Some appropriate supportive evidence comes from a recent study of radular teeth in opisthobranch gastropods by Bertsch ( 1976) wherein he recommends that knowledge about variations should replace typological thinking with modern systematics. He suggests, “The presence o r absence of a certain feature, such as denticles, may indicate a second species or it may show simply an ontogenetic stage.” Figure 3. CdJrngia m’suhiam hooks (see text for interpretation) (scale lOpm). A. First three rows behind tentacles showing hooks in different stages of development. B. Fully formed hooks from about the tenth row. C. Hook from zone of scattered hooks proximal to rows,without spinelets. E. B. CUTLER 572 Table 2. Comparisons of two similar species of Mitosiphon G . misakiam G . murinae Habitat Shallow water (<50 m), tropical or subtropical, sand Deep water (>50 m), cold, mud Epidermis Thin, transparent Not usually thin or transparent Retractor muscles Two pairs, short, thin, and equal in size and distance from ventral nerve cord Two pairs, longer, stout, of unequal size and distance from ventral nerve cord Nephridia Bilobed and both lobes of equal length One subspecies unilobed, other with secondary lobe less than 1/3 the main lobe size Introvert length From 10 to 12 times the trunk length From 4 to 6 times the trunk length Amor (1975) also commented on G. murinae and Murina’s G. hespera. She suggested (without corresponding with this author or comparing material) that G. murinae is really G. hespera or G. misakiana. The distinctions, discussed at length in Cutler (1969: 214; 1973: 142-143), are not just in the size and density of the papillae as Amor states, but include trunk size and shape, introvert length, nephridia, retractors, pigmentation, and habitat (see Table 2). The Indian Ocean collections provided me with a large series ofworms from a single population and after some initial confusion I was able to accept the fact that G. misakiana and G. trichocephda do co-exist in the same community. How their niches differ is still unclear but may have to do with their feeding mode. These populations are discussed more fully in Cutler 8c Cutler (1978). I t is concluded that G. hespera contains only larger, older individuals of G. misakiana, with larger papillae and more spinelets on the hooks, and is therefore a junior synonym. Goljngia recondita (Sluiter, 1900) The only two identified specimens of this form are on public display in the Oceanographic Museum at Monaco. In 1970 I was able to view the specimens but not to remove them from the bottle. Recent attempts to obtain them for closer inspection were fruitless because they were permanently fixed to a glass plate inside the jar for display purposes. It is probable that the length of the introvert exceeds the trunk length. As the introverts of Sluiter’s specimens were retracted, his statement about the tentacles surrounding the mouth is open to question, as discussed elsewhere in this paper. He says that a fully developed, posteriorly anchored spindle muscle is absent but that the posterior end of the gut coil is attached to the body wall by a muscle divided into rays that come together near the most posterior convolution. With no specimen it is difficult to interpret this but I am treating this as a spindle SIPUNCULAN TAXONOMY 313 muscle. The peculiar nephridial state (apparently only described from one worm) of a unilobed right nephridium and a bilobed left nephridium is also impossible to verify. Sluiter compares it to the condition in G. abnorme (Sluiter). This latter species has been considered a junior synonym of Phascolosmapectinatum by Cutler 8c Murina (1977).When inspected by me, G. abnorme clearly showed longitudinal muscle bands and in all ways (except one unilobed nephridium) was identical to P. pectinatum. I t is tempting to change the status of this species because of the unavailability of the type (and only) material. For the time being that temptation shall be resisted. Golfingia trichocephala (Sluiter, 1902) As noted earlier this species has been discussed at length in Cutler ( 1973) and little more need be said. In size, shape, form, and habitat it closely resembles G. misakiana but lacks the posterior papillae, hooks, and tentacles. Murina leaves this species in its own monotypic subgenus Apionsomu (personal communication) because it lacks the hooks and tentacles. It is my opinion that these differences are outweighed by the remainder of the traits which unite it with the other species in this complex. Fisherana species Golfingia savalovi Murina, 1974 This species was described in detail by Murina as a new form distinguished from G. capitata on an erroneous interpretation of Gerould’s description. She thought the distal region of the introvert, described as an orange head, was unique in G. capitata. After comparison of the two forms it became evident that they were the same and after subsequent exchange of specimens, Murina now concurs in this. Therefore, G. savalovi is ajunior synonym of G. capitata. A few comments on the four species originally assigned to Fisherana by Stephen are in order. All but the first were originally placed by their describers in the genus currently known as Golfingia. Golfingia capitata (Gerould, 19 13) Forty specimens identified by Gerould have been examined as well as several dozen individuals from recent collections. This material was provided by four different sources as shown in Table 3. Gerould’s records ranged in depth from about 1220-3200 m along the eastern United States from latitude 3 8 O 29’ to 41° 53’ N. Taken together this clearly shows a slope inhabitating species, absent from the shelf and abyssal depths. The extreme variability in the density and size of the trunk papillae make statements like, “Trunk . . . with conspicuous scattered papillae which are thickly scattered over the posterior end” (Stephen 8c Edmonds, 1972: 331) rather misleading. Many worms have small and few papillae. Stephen 8c Edmonds’s (1972: 331) comments about the point of origin of the retractor muscles in the posterior end omit Gerould’s elaboration ( 19 13 : 423-424) on the extreme E. B.CUTLER 314 Table 3. Recent collections of Go&ngiu cupitutu Source Howard Sanders, Woods Hole Oceanographic Institution Latitude Longitude Depth (m) Date spec. 58 202 38" 34' N 56' S 39" 47' N 72" 55'W 12" 15' E 70" 50' W 1925 1535 1600 7/IX/63 23/V/68 22/11/69 40 6 5 28" 21' N 39"31'N 28" 50'W 31"05'W 1009 650 10/IX/71 25/X/ll 2 39" 42' N 70° 45'W 1860 28/VIII/59 I 3 9 O 12" 71" 48'W 2116 15IXII73 2 209 R/V Charcot, Paris Museum Natural History 136 R. Wigley, NOAA, NMFS, Woods Hole Delaware 59-10 Tow 18 G . Rowe, Woods Hole No. of Station 25 Knorr 1 312 Oceanographic Institution variability of this feature. The author concurs with Gerould's statement, viz: "This species varies much in regard to the position of the points of attachment of the retractor muscles to the body wall. Thus in one specimen both pairs were attached near together near the middle of the trunk. In a small, young specimen the ventral retractors have the usual attachment near the posterior end of the body, whereas the slender dorsal muscles are joined to the body wall much further forward than in the individual which I have figured (fig. 141, and underneath the nephridia. In general, in a few specimens which I have dissected, the position of the attachment of the dorsal retractors varies from about the posterior end of the first third of the trunk backward to the posterior end of the third quarter." The epidermis is often very loose and fragile, rubbing off easily to give a very different appearance, with the pale whitish underlying muscle layer and large papillae the prevalent features (Fig. 4). Unless the introvert is extended and the phascolosomatid arrangement of the tentacles is visible, one can easily place this form in the subgenus Golfingzella Stephen, and if one believes Wesenberg-Lund ( 1959a)about the spindle muscle, it easily keys out to G. pudicu (Selenka). This problem of accurate determination of tentacular arrangement will be returned to. Golfingtapapillifera (Keferstein, 1865) This was first described from a single specimen with a 9mm trunk (not introvert as in Stephen & Edmonds, 1972: 322) and a 15mm introvert. The arrangement of the 12 oval, leaf-like tentacles is neither commented on nor illustrated and Stephen 8c Edmonds note (1972: 329): "How the tentacles are ordered relative to the mouth is not known for the type F. papilltjera nor have we been able to locate the type specimen. From a comparison with F. capitata and F. wasini it seems possible that the tentacles of F. papillifera are arranged in the typical Phuscolosoma-manner. If ever this is shown not to be so, the position of SIPUNCULAN TAXONOMY 375 Figure 4. GoFngia capitata. Trunk papillae from specimen with sloughed off skin. A. General view of posterior region (scale 0.2 mm). B. Two papillae (scale 0.1 mm). C. Tip of one papilla (scale 0.01 mm). Fisheranu in the family Phascolosomatidae will be untenable.” It seems to me that if the tentacles were visible to Keferstein, as they must have been for him to count them, he would certainly have noted their arrangement if they were not typically golfingiid. To assume otherwise is very risky. In addition to Keferstein’s worm, there are only two other original records of this species in the literature. Shipley (1902: 132) found “several specimens” in coral and stone masses in the Laccadive Islands. His animals were longer (30 mm) and all of the specimens had ruptured body walls with the retractors and intestine protruding-not prime material. Attempts to locate his specimens at Cambridge and London were not productive. The only other known record is that of Sluiter (1912) from Cape Verde Islands. This single small worm ( 5 mm trunk, 4 mm introvert) was obtained from the Monaco Museum and proved to have not two but four well developed retractor muscles-clearly not Keferstein’s species. Also, the longitudinal musculature while not obviously forming bands showed upon close inspection a tendency towards splitting up; a normal condition for a 5 mm Phcolosoma, possibly P . nigrescens Keferstein. 376 E. B. CUTLER When Stephen (1964) first described Fisheranu he did not mention the tentacle arrangement, which may account for his choosing this poorly known species for the type. The appearance of the hook seems to have been important to Stephen, but with no possibility of first hand verification, there is room for doubt. Keferstein included it in the complex of species that are currently referred to as Go@ngza and his comments on it did not suggest any thought that it was significantly dissimilar from those such as G. pellucida Keferstein, which also has hooks and large papillae. Therefore, the species selected as the type for Fisheranu has no traceable representatives in any museum. The only published descriptions are either not verifiable (Shipley)or incorrect (Sluiter).Despite extensive collections in the area of the type locality, Rice (1975a)does not report finding any representatives of this species. This is hardly a firm foundation for a genus. One could make an argument for placing this species in the subgenus Siphonoides Murina. Goljingia lobostoma (Fischer, 1895) This species has been recorded twice, the second only a note of the occurrence of a single worm from Madagascar with no description other than a figure of its hook. Recent collections of several hundred sipunculans from Madagascar (Cutler 8c Cutler, 1979) failed to produce any representatives of this taxon. The abbreviated translation of Fischer’s description found in Stephen 8c Edmonds (1972: 333) says the introvert is 1.5 mm long-it should read 1.5 cm. It also omits the fact that the spindle muscle is attached posteriorly. The original description of the tentacles is unclear but does not sound like that proposed as typical for Fisheranu, i.e. dorsal to the mouth. Fischer’s statement (1895: 14) is: “Die Tentakel scheinen bundelweise vereinigt zu sein, 8-10 kleinere und 2 grossere konnte ich bemerken.” Goljingia wasini (Lanchester, 1905) This species, from off East Africa, has not been recorded since the original description over 70 years ago, despite numerous collections in the area. The distinctions between this form and G. lobostoma found on adjacent Madagascar are ill-defined. The location of the nephridiopores “just anterior to’’ or “just posterior to” the anus as found in Stephen 8c Edmonds’s key ( 1972)to this group is not a very significant distinction. G . lobostoma has a posterior spindle muscle, while G . wasini has a fine fastening mesentery attaching the gut convolutions to the posterior end of the trunk according to Stephen 8c Edmonds (1972). In their key this translates into “spindle muscle absent” in G. wasini. However, Lanchester’s own words are ( 1905: 33) “The intestine is held to the hind end of the body by a fine muscle strand. . .”. When I examined Lanchester’s specimens I had no doubt that it is a normal spindle muscle. The brown hooks matched the published figures well and measure 35-50 pm high, the clear streak often a bit narrower than that shown in fig. 40F of Stephen 8c Edmonds (1972: 330). The trunk length of these worms ranges from 3-9mm. The large papillae are restricted to the two ends of the trunk and the midregion is quite smooth. It is conceiuable that these may simply be juvenile Phascolosoma in which the longitudinal musculature is still in its early ontogenetic state, i.e., undivided and continuous. SIPUNCULAN TAXONOMY 311 Edmonds re-examined the type material and says that the tentacles lie dorsal to the mouth. In the four specimens I examined none of the introverts were extended so this could not be verified. A side by side comparison of hooks and papillae of G. lobostoma and G. wasini showed no significant differences in size, shape or appearance. Externally the animals show no differences. Internally the on1 detectable distinction (acknowledging their poor condition) is the minor dif? erence in anudnephridiopore placement. Based on these observations it is concluded that these two “species”, neither having been collected for a very long time, are synonymous under the older name G. lobostoma. DISCUSSION One basic biological question which has not recently been addressed in the sipunculan literature is: what is our concept of a genus? The biological or population species concept has been touched on but no explicit consensus has been reached among different workers. This is less of a problem now than it has been, but it is not yet resolved, nor shall this be an attempt to do so. Stephen’s ( 1964) short preliminary work provided no philosophical basis for his decisions except that the previous arrangement was “inconsistent at several points”. The problem he saw was the longitudinal musculature and the way it had been used as a taxonomic character. He felt it should be weighted more heavily than it had been, thereby requiring the splitting of Aspidosiphon Diesing into two genera and elevating Phuscolopsis Fisher with its single species to generic rank. He also created a new subgenus (Goljngiella), resurrected the all but forgotten name of Themiste Gray for the familiar Dendrostomum Griibe, and created the genus Fisheranu. The striving for internal consistency within natural systems is an admirable trait, but, as an example, the action on Themiste and Phascolopsis did not make any friends for systematists in the larger community of biologists who fail to see the necessity for such changes. The reality and consistency of the AspidosiphonJParaspidosiphon division is very doubtful as the body musculature is not always either continuous or divided into distinct bundles (Cutler, 1973; Cutler 8c Cutler, 1979; Ditadi, 1975; Murina, 1975).While it may be a convenient, often dependable tool for identification, a subgeneric ranking for this character seems more than adequate, which brings us back to Fisheranu and the genus concept. I t is generally acknowledged that the rank given a taxon above the species level is a subjective matter. A look at the history of the Sipuncula over the last 100 years easily illustrates this (Stephen & Edmonds, 1972).A genus has been defined variously but one standard version from Mayr, Linsley & Usinger (1953:48)is: “A genus is a systematic category including one species or a group of species of presumably common phylogenetic origin, which is separated from other similar units by a decided gap.” The focal point here is the “decided gap”; this is also the point where judgments vary. This task was easier when we had less information and fewer specimens. As the number and size of collections increase we gain a better idea of how much natural variation occurs within a group (population, species or genus) due to size, age, environmental conditions or type of anesthesia/preservativesused. This requires a constant re-evaluation of earlier constructs as the gaps become less distinct. 918 E. B. CUTLER The creation of a genus is more synthetic than analytic and usually involves evaluating several characters, not single ones. Which characters we select for this purpose is crucial. Blackwelder ( 1967 : 202-203) comments: “There is no question that all characters to be used are chosen by the taxonomist for that particular use. This is a personal decision, regardless of what basis he uses for his choice. The choice is, therefore, highly subjective. The choosing of the characters is one of the major functions of the taxonomist.’’ He goes on to discuss good and bad characters: “A good character at any level is one which ‘works’, which produces a useful distinction or grouping. This means that it is ( 1 ) readily detectable, (2)clear-cut in its segregation of the group possessing it, (3)complete in its uniting of the members of the group, and (4)evidently correlated with other aspects of the life of this group, so that the grouping itself is useful in a variety of situations.’’ Blackwelder also comments on how good characters should not be influenced by environmental factors and then concludes that section saying (1967: 204): “These can be identified only in practice, by trial and recognition of failure.” I t is my contention that the groupings proposed by Stephen are poor because the characters he uses are not significant. Let us review them for Fisherana . Stephen (1964) used two characters, the hooks and papillae which are supposedly complex and like those found in Phascolosoma. In Stephen & Edmonds (1972) a third character is added, the arrangement of the tentacles. Looking first at hooks, the assumption seems to be that all members of the genus Phascolosoma have hooks of complex internal structure and that on any single worm all of the hooks exhibit this complexity. In fact, there are at least six species of Phascolosoma which have been described as lacking hooks altogether. Furthermore, in those bearing hooks, the internal complexity is correlated to size (age) of the hooks. Often those rings most distal (most recently formed) are smaller, less pigmented, and without the “typical” internal characteristics. This is not peculiar to Sipuncula as Bertsch (1976: 120) points o u t for radular teeth in opisthobranchs : “Newly formed teeth have a different morphology than fully formed teeth. Young teeth are weaker, thinner, and more transparent. . . .” It is possible to select from a typical Phascolosoma a number of hooks which look ve golfingiid o r to find none at all. While it is true that hooks in most GoCfFngia loo different from most Phuscolosoma hooks and those known for Fisheranu are closer to the phascolosomatid morph, it is not in my judgment a clear, distinct, good generic character. As Bertsch (1976: 121) points out, “The opisthobranch radula has an important use in taxonomy but it must be treated as a biological entity subject to variation and not as the product of a typological mind.” The papillae can also be misused. The papillae of a single worm taken from different body regions may vary considerably in structure. Even more variation is shown in a population of different-sized worms (e.g. G. jiagr$?ra (Selenka) in Cutler, 1973). Supposedly Fisheranu have PhmcoloJma-likepapillae, i.e., dome-like or cone-shaped with chitinous platelets, generally more pronounced and complex than in Goljingia. To examine this more closely the scanning electron microscope was again used and Figs 4-6 show some of the results. Other species within each “genus” show quite different shapes and sizes, Phascolosoma pacijicum having among the largest in that group. My conclusion is that the papillae of Fisheranu species are not clearly distinct from many GoCfngia. A distinct “gap” is hard to demonstrate, therefore this is a poor character. u SIPUNCULAN TAXONOMY 379 Figure 5. Papillae from two specimens of Golfingta capitafa. A and B from G . Rowe’scollection with “normal” skin; C and D from H . Sander’s collection with “loose” skin. Difference possibly due to post mortem chemical environment or nature of microhabitat while alive (scale 0. I mm). In the definition of Fishmanu, one finds this phrase: “Papillae . , . usually most densely packed at the anterior and posterior extremities of the trunk as in Phcolosoma.” In the many species of Golfingta which bear trunk papillae, the same distribution pattern holds true. In other words, this character, while used in Stephen 8c Edmonds (1972) to relate Fisheranu to Phascolosomu, can just as legitimately be used to relate it to GoLjngia. In Blackwelder’s terms neither the hooks nor papillae are clear-cut in segregating Fisheranu from Golfingia. An analysis of these characters must include some reflection on an old and debated biological principle of ontogeny recapitulating phylogeny. There is general agreement that Golfingza represents the early, primitive generalized morph (Murina, 1975; Stephen & Edmonds, 1972: 15-18). What one sees during the ontogeny of Pharcofosoma hooks, papillae, and musculature is a much more golfingiid (generalized) condition, prior to maturity. In mature Fisheranu, one sees an intermediate situation-a step in the Phascolosoma direction, a part of a Figure 6. Comparison of Go@zgia and fhcolosoma papillae. The lack of chitinous platelets on the surface of' the Pharcolosoma species is noteworthy and not restricted to these two species. I t may be that this technology obscures these or that they are artifacts of light microscopy. Papillae of Phascolosuma do do not appear very dissimilar to those of Golfingia on the surface. Much variation in size and shape could be found on a single worm. A, B. Golfingia pellucida (scale 0.05 mm). C , D. Pharcolosoma scolops (scale0.1 mm). E. fharcolosomapacijicum (scale0.1 mm). SIPUNCULAN TAXONOMY 38 1 divergent evolutionary process. This is consistent with Murina ( 1975) but does not imply that Pharcolosma evolved from any extant Fisherana. The tentacular arrangement added to the definition of Fisherana by Edmonds is an improvement but it is troublesome for different reasons. Using this character requires that the introvert be completely extended and in the more than twenty thousand sipunculans I have studied, probably one or two per cent have met this criterion. Secondly, even with tentacles showing, the task of precisely locating the mouth is very difficult unless the esophagus happens to be protruding through it. This is rare. The likelihood of misidentifying a worm for this reason is great. It is of course true that the classification of a group and the process of identification are, or ought to be, separate processes. A practical problem remains which relates to one of Blackwelder’s points, i.e. it simply is not readily detectable. It is unfortunate that when they established their families for this phylum, Stephen 8c Edmonds (1972: 19) invested heavily in this often hidden character. Table4 shows what remains from the definition of Fisheranu. Table 4. Characters used in definition of Fisheranu Character ~~ Genus which it resembles ~~~~ ~ ~~ Golfingia and Phcolosoma leaning towards Phcolosoma Golfingia and Phaxolosoma Go&ngia Golfingia and Phcolosoma uncertain or P h c o l o s m a Golfingia Trunk size and shape Hook structureand shape Papillae shape and distribution Longitudinal musculature Retractor muscles Tentacle arrangement Egg shape’ ’ Character not used by Stephen & Edmonds (1972). I have added an additional character to the list-egg shape. Most phylogeneticists assume that similarities in type of egg, cleavage patterns, mesoderm development, etc. are more instructive than adult external form. Leaning heavily on external characters which are more likely to vary in an adaptive fashion (for tasks like food gathering) can be misleading (Cutler, 1975). Gerould (1913: 422) points out that “the eggs [of G. capitatal are spherical, small, and transparent, covered with a yolk membrane pierced with distinct pore canals.” Rice (1975b: 152) tells us that “Although eggs of most sipunculans are spherical . . . those of the genus Phascolosoma are characteristically flattened ellipsoids, frequently with depressed apices.” These quotes have been confirmed by my own observations and I would, therefore, suggest that this is a strong argument against a close relationship between Fijherana and Phascolosoma. CONCLUSIONS In reviewing the evidence from gross morphology, light and scanning electron microscopy on all of the material listed in Table 1, I conclude that the species discussed herein are related and should be grouped together but d o not merit generic rank. Rather, the taxon Fisheranu might be viewed as a subgenus of GolJingia, acknowledging that it probably represents an offshoot of the main branch in the direction of Phcolosoma. E. B. CUTLER 382 Keturning to the Mitosiphon species: if we interpret the latest diagnosis of Fisherana in one way it would behove us to move G. misakiana and G. murinae into it as we now know them to have the phascolosomatid tentacle arrangement. This was my first inclination, as evidenced in Cutler 8c Murina (1977). This would leave only G. trichocephala and G. recondita in Mitosiphon. However, as G. hespera (now G. misakiana) is the type for Mitosiphon, additional problems arise, i.e. Mitosiphon would have seniority over Fisherana. Alternatively, the two groups of species could remain separate if we simply modified the concept of Mitosiphon to include its “proper” tentacle arrangement. After reconsidering the grouped characters, this second alternative seemed logical and rational. However, at this point two problems remain. These.two groupings are not very distinct on paper and the argument for keeping them separate appears weak. There is also the fact that Murina (1975) retains Apionsoma as a monotypic subgenus. After some helpful correspondence with P. E. Gibbs, the most sensible solution seems to be to place all of these species into a single subgenus and apply the senior available name, which is Apionsoma. Those differences we might have used for separating the taxa can now be used for constructing a key to the species (below). There remains as much internal unity as within most other subgenera. A problem with this idea is that the original concept bears little resemblance to this revised version and as trichocephala now becomes the typespecies, it is externally the least “typical” of the group (i.e., lacking hooks, tentacles, and papillae). But we now have some type material available for study. There is reason to move in this direction as a solution to a perplexing and troublesome situation. The revised subgeneric diagnosis proposed is : Goljngia (Apionsoma) Sluiter, 1902 sensu Cutler Two pairs of retractor muscles (except G. papillfera, which cannot be verified); spindle muscle attached to posterior end of trunk; nephridia single or bilobed; introvert hooks usually present; tentacles, where known, arranged in crescent or near circle dorsal to mouth; introvert of varying lengths; papillae, when present, most obvious at posterior end of trunk and often dome-like or conical. Includes G. capitata, G. lobostoma (=G. wasini), G. misakiana (=G. hespera), G. murinae. G. papillifera, G. recondita, and G. trichocephala (type-species). Table 5 , showing some of the diagnostic characters of the six Goljzngia subgenera, can serve as a key to the groups. It is noted there that Phascoloides Table 5. Diagnostic table for Goljngia subgenera Contractile vessel villi Number of retractors Posterior spindle muscle Tentacles, if present Thysanocardia present 2 absent around mouth Siphanoides absent 2 present around mouth Nephcrroma (Phascoloides) absent 2 absent around mouth Golfingia absent 4 absent around mouth Golfingiella absent 4 present around mouth Apionsoma absent 4 present dorsal to mouth (1 with 2) SIPUNCULAN TAXONOMY 383 should now be replaced by Nephasoma Pergament. Informal consensus now suggests that to retain Phuscoloides as suggested in Cutler & Murina (1976) would be a violation of the International Code. KEY TO Al'IdNSOMA SPECIES 1A. Introvert shorter than trunk; hooks without accessory comb of spineletsatbase . . . . . . . . . . . . . . . . 2 1B. Introvert equal to or longer than trunk; hooks, when present, with . . . . . . . . . . 4 2A. Two retractor muscles . . . . . . . . . . . G . papillfera 2B. Fourretractormuscles . . . . . . . . . . . . . . 3 3A. Ventral retractor muscles arise in posterior half of trunk . . G . capitata accessory comb of spinelets at base 3B. Ventral retractor muscles arise in middle or anterior part of trunk . . . . . . . . . . . . . . . . . G.lobostoma 4A. Papillae, hooks, tentacles absent . . . . . . . . G . trichocephala 4B. Papillae, hooks, tentacles present . . . . . . . . . . . 5 5A. Introvert more than 9 times trunk length; body wall thin, transparent, with small papillae . . . . . . . . . . G . misakiana 5B. Introvert less than 7 times trunk length; body wall thick, opaque, with large papillae . . . . . . . . . . . . . G . murinae ACKNOWLEDGEMENTS The completion of this work was dependent on the co-operation of those institutions listed in Table 1 for the loan of pertinent specimens. I am also indebted to G. Murina, Sevastopol; P. K. B. Menon, Dar-es-Salaam; and S. J. Edmonds, Adelaide, for the loan of specimens and for dialogues on this problem. Dr Murina and Dr P. E. Gibbs, Plymouth, have read an earlier version of the manuscript and offered helpful comments. Dr Anthony Checchi, Utica College, provided valuable assistance with the scanning electron microscopy which was performed in the Biology Department, University of Massachusetts, Boston. Mr Henry Iwanicki made the drawings. Ms Norma Cutler provided many forms of assistance throughout this project. REFERENCES AMOR, A., 1975. Notas sobre Sipuncula de la Argentina, Brad y Peru. Physis Buenos Aires (A), 3 4 : 113-120. BLACKWELDER, R. E., 1967. Taxonomy. A Text and Reference Book. New York:John Wiley. BERTSCH, H., 1976. Intraspecific and ontogenetic radular variation in opisthobranch systematics (Mollusca: Gastropoda). Systemdic Zoology, 25: 117-122. CHAMBERLAIN, R . V., 1920. Notes on the sipunculids of Laguna Beach.Jouma1 of Entomology and Zoology, 12: 30-3 1 . CUTLER, E. B., 1967. The Sipuncula j o m the Western North Atlantic; Their Systema!rcs, Ecology and Distribution. University of Rhode Island, unpublished dissertation. (Microfilm, University !dicrofilms, Ann Arbor, Mich. Dissertation number 69-1030.) CUTLER, E. B., 1969. New species of Sipuncula from the western North Atlantic. Proceedings ofthe Biological Society of Warhington, 82: 209-2 18. CUTLER, E. B., 1973. Sipuncula of the Western North Atlantic. Bulletin ofthe American Museum ofNatural History, 152: 105-204. 384 E. B. CUTLER CUTLER, E. B., 1975. Pogonophora and Protostomia-a Procrustean Bed? Zeitschn? f u r Zwlogische Systematih und Evolutionsforschung, 13: 112-122. CUTLER, E. B. & CUTLER, N. J., 1979. Madagascar and Indian Ocean Sipuncula. Bulletin du Mutium d’Hutoirc Naturelle, in press. CUTLER, E. B. & MURINA, V. V., 1977. O n the sipunculan genus Golfingia Lankester, 1885. ZwlogicalJouml of the Linnean Society, 60: 173-187. DITADI, A. S., 1975. Aspidosiphon schneghageni (Sipuncula) inhabitating Tmnatina shells. Veliger, 18: 200-202. FISCHER, W., 1895. Die Gephyreen des Naturhistorischen Museums zu Hamburg. Abhandlungen aur dem CA’rt der Naturm’ssenschaja, 13: 1-24. FISCHER, W., 1927. Sipunculoidea und Echiuroidea. Fauna Sadwest-Auttralim, herausgeben von W . Muhaelsen, 5: 199-216. FISHER, W. K., 1950. The sipunculid genus Phascolosoma. Annals and Magazine ofNatural History (Series 12), 3 : 547-552. FISHER, W. K., 1952. The sipunculid worms of California and Baja California. Proceedings ofthe United States National Museum, 102: 37 1-450. GEROULD, J . H., 1913. The Sipunculids of the Eastern Coast of North America. Proceedings of the United States National Museum, 44: 373-437. IKEDA, I . , 1904. The Gephyrea ofJapan. Journal ofthe College ofscience, Imperial University of Tohyo, 20: 1-87. KEFERSTEIN, W., 1865. Beitrxge zur anatomischen und systematischen Kenntniss der Sipunculiden. Nachrichten von der Kaniglicha Gesellschaj der Wissenschajen, Gottinga, 1865 : 189-209. L4NCHESTER, W. F., 1905. Gephyrea in the marine fauna of Zanzibar and British East Africa from the collections made by C. Crossland in 1901-02. Proceedings of the Zoological Society ofLondon, 1: 28-35. MAYR, E., LINSLEY, E. G. & USINGER, R. L., 1953. Methods and Principles ofSystematic Zwlogy. New York: McGraw-Hill. MURINA, V. V., 1974. Contributions to the knowledge of the fauna of sipunculid worms from the South Atlantic based on the data of the “Akademik Kurchatov” expedition in 197 1. Trudj Instituta Oheanologii, Akademia Nauk SSSR, 98: 228-239. MURINA, V. V., 1975. Ways of evolution and phylogeny of the phylum Sipuncula. Z o o l o p c h e s h i f Z h u d , 54: 1747-1758. RICE, M. E., 1975a. Survey of the Sipuncula of the coral and beachrock communities of the Caribbean Sea. In M. E. Rice & M. Todorovic (Eds), Proceedings of the International Sgmposium on the Biology of S i p u d a and Echiura, 1: 35-50. Institute for Biological Research: Belgrade. RICE, M. E., 1975b. Observations on the development of six species of Caribbean Sipuncula with a review of development in the Phylum. In M. E. Rice & M. Todorovic (Eds), Proceedings of the International Symposium on the Biology of Sipuncula and Echiura, 1: 14 1-160. Institute tor Biological Research: Belgrade. SHIPLEY, A. E., 1902. Sipunculoidea, with an account of a new genus Lithacrosiphon. Fauna and Geography of the Maldiveand Laccadive Archipelago, 1: 131-140. SLUITER, C. P., 1900. Gkphyriens(Sipunculideset Echiurides)provenant des campagnes de I’Hirondelleet de la Princesse Alice. 1886-1887. Risultatsdes CarnpagnesScientijques Accomfiliesparle Prince AlbertI, 15: 1-30. SLUITER, C. P., 1902. Die Sipunculiden und Echiuriden der “Siboga” Expedition. Siboga-Expeditie, 25: 1-53. SLUITER, C. P., 19 12. Gkphyriens (Sipunculideset Echiurides) provenant des campagnes de la Princesse Alice, 1898-1 9 10. Rbultats des Campagnes Scientijques Accompliespar la Prince Albert I,36: 1-36. STEPHEN, A. C., 1964. A revision of the classification of the phylum Sipuncula. Annuls and Magazinc ofNaturd History, (Series 13), 7: 457-462. STEPHEN, A. C. & CUTLER, E. B., 1969. On a collection of Sipuncula, Echiura and Priapulida from South African waters. Transactions of the Royal Society ofSouth Af;rca, 38: 11 1-121. STEPHEN, A. C. & EDMONDS, S. J., 1972. The Phyla Sipunncla and Echiura. London: British Museum (Natural History). WESENBERG-LUND, E., 1959a. Sipunculids and Echiuridae from Mauritius. Videnshabelige Meddelelser j a Danske Naturhistorish Forening i Kjrbenhavn, 121: 53-13. WESENBERG-LUND, E., 1959b. Sipunculoidea and Echiuroidea from Tropical West Africa. Atlantide Report, 5 : 177-210.
© Copyright 2025 Paperzz