genus

: The enigmatic sea pen, Gyrophyllum sibogae Hickson, 1916, is redescribed from Tasmanian material, extending the range of the species from the Malay Archipelago and Madagascar. A phylogenetic analysis of several related pennatulacean taxa attempts to resolve some problematic aspects of the literature regarding the systematics of the genus Gyrophyllum. Six taxa are compared in the analysis: Pennatula inflata, Ptilosarcus undulatus, Gyrophyllum sibogae, Sarcoptilus grandis, Crassophyllum cristatum

Introduction tiated the two species as follows-G. sibogae At least fifteen species in nine genera of sea Wlth , two ^ly* teeth per polyp and larger colony pens have been recorded or are known to occur ■" K l u ° 295 T m ^f ™* G ' hl ™* u _ a * v wu-■ \ tu del lei without calyx teeth and smaller colony in southern Australia (Williams, in press). The . , ." J , , T ... . , \ recent discovery of Gyrophyllum sibogae off Tas-™* ^ to ' 2 J Tu ^ ^T' "m ania represents a new record to the Australian distnbutional and bathymetnc differences-G. fauna. Southern Australia is defined here as the sibogae (Indonesia at 567 m) compared to G. t -♦ i u ir c *u \\r * hirondellei (vicinity of the Azores at 1222continental shelf region of southern Western *: v J Australia, South Australia, Victoria, Tasmania, m ^' and southern New South Wales (i.e., the southern The enigmatic nature of the genus Gyrophylcoastal waters extending from Perth to Sydney), ^m has resulted in some disputation by previous Hickson (1916) originally described Gyro-authors regarding the systematics of the taxon. phyllum sibogae from the Malay Archipelago and It is shown here that Gyrophyllum represents a differentiated it from the type species G. hiron-morphological intermediate between the Pendellei of the northern Atlantic, the only other natulidae and Pteroeididae thereby making the species known for the genus. Hickson differen-distinction between the two families tenuous. [315]
Description (Figs. 1-3, 5C).-The seven whole colonies that were examined range in length from 185-250 mm. The colonies excluding the polyp leaves are somewhat clavate, with the distal portion of the rachis distinctly wider than the rest of the colony. The rachis comprises roughly 30%-40% of the total colony length, while the peduncle makes up approximately 60%-70%. The maximum width of the rachis varies from 12-1 7 mm, while that of the peduncle ranges from 8-12 mm.
The number of polyp leaves per side varies from 6-10, but most colonies have 6 or 7 per side. The polyp leaves are fanlike, thick and fleshy (3-6 mm thick and 1 5-50 mm in breadth). The autozooids are confined to the outer margin of the leaves, where they are congested and disposed roughly in two adjacent rows.
Each polyp leaf contains 20-50 autozooids. The autozooids are 1.5-2.5 mm in diameter and protrude from a fleshy calyx-like base that usually contains two more-or-less distinct teeth or pointed lobes on opposite sides or adjacent to one another. These lobes contain relatively sparsely-set polyp leaf sclerites.
The siphonozooids are numerous but scattered (not particularly densely-set) on both sides of the polyp leaves between the autozooids and onto the faces of the leaves below the autozooids. Each siphonozooid is approximately 0.5 mm in diameter, and appears as an inconspicuous hemispherical protuberance, sparsely spiculated. The siphonozooids are scattered over the surface of the polyp leaves, are not congested or restricted in distribution, and thus do not form zones or pads.
The sclerites of the tentacles are short, blunt rods (0.10-0.24 mm in length), smooth to somewhat longitudinally grooved but not threeflanged. The sclerites of the polyp leaves, rachis and peduncle are elongated, three-flanged rods. Those from the polyp leaves are 0.28-0.48 mm long. Sclerites from the surface and interior of the rachis are 0.27-0.45 mm in length, while those   Color of the colonies varies from tan to light brown. Distribution (Fig. 4).-Indo-West Pacific and southeastern Australia, 520-585 m in depth: Ceram (Malay Archipelago), 567 m depth (Hick-son, 1916); Nosy Be (Madagascar), 580-585 m depth (Tixier-Duri vault and d'Hondt, 1973); Tasmania, 520 m depth (present study).
Remarks. -Hickson (1916:254) reported numerous but inconspicuous siphonozooids on the dorsal surface of the rachis of the type specimen from Indonesia-these were discerned only as a result of thin sectioning and staining. He records the diameter of the siponozooids to be 5 mm. This is apparently mistaken however, since his figure 48 of plate VIII shows each siphonozooid to be approximately 0.4-0.5 mm in diameter. Siphonozooids have not been observed on the dorsal surface of the rachis in the Tasmanian material.
The known specimens of Gyrophyllum sibogae, although from disparate geographical regions of the southern hemisphere (Madagascar, Indonesia, and Tasmania), show a remarkable similarity and restriction of depth distribution (520-585 m).

Discussion-Phylogenetics
Gyrophyllum sibogae is morphologically intermediate between the pennatulids (Pennatula and Ptilosarcus) and the pteroeidids (Sarcoptilus, Crassophyllum, and Pteroeides), based on the presence of three-flanged sclerites (as in the pennatulids) and the presence of siphonozooids on the polyp leaves (as in the pteroeidids). This latter character is considered derived and defines a clade separating Sawcoptilus, Crassophyllum, and Pteroeides from other pennatulaceans. Williams (1993: 732-733) provides a summary of previous work on pennatulacean phylogeny, and summarizes the difficulty regarding analyses of phylogenetic systematics in such groups as sponges and coelenterates.
Historical Context (Fig. 6). -Kolliker ( 1 869) originally unified the subfamilies Pennatulinae and Pteroeidinae into one family. Subsequently, Kolliker (1880) elevated the status of the subfamilies to separate families. Studer (1901: 34) placed the new genus Gyrophyllum in the family Pteroeididae. Kukenthal and Broch (1911: 394) stated that Gyrophyllum should not be placed in the Pteroeididae due to the presence of threeflanged sclerites, which are characteristic of the family Pennatulidae. Kukenthal and Broch (1911: 463) and Kukenthal (1915: 120) considered the taxon to be a systematic enigma and consequently labeled it as "incertae sedis" at the end of their accounts of the Pennatulacea (Williams, 1995). Hickson (1916: 252-253), disagreed with Kukenthal, being of the opinion that Gyrophyllum shows stronger affinities to the Pteroeididae than to the Pennatulidae, based on the presence of siphonozooids on the polyp leaves, and that the presence of three-flanged sclerites should not be considered as important a clue to affinity as the placement of siphonozooids. Tixier-Durivault and D'Hondt (1973: 263) concurred with Studer and Hickson, and considered the genus to be a member of the family Pteroeididae. Hickson (1916) also placed Scytalium and Acanthoptilum in the Pennatulidae along with Pennatula and Leioptilus (a synonym of Ptilosarcus). In addition, Kolliker (1880) placed Halisceptrum (a synonym of Virgularia) in the Pennatulidae. Scytalium, Acanthoptilum, and Virgularia all have minute and sparsely distributed siphonozooids located on the rachis between the polyp leaves. They have therefore been placed in the Virgulariidae (Williams, 1995).
Outgroup. - Williams (1993) has shown that the presence of polyp leaves is considered a synapomorphic feature in the Pennatulacea and serves to distinguish a derived clade composed of the virgulariids (such as Virgularia and Scytalium), the pennatulids (Pennatula and Ptilosarcus), and the pteroeidids (such as Pteroeides). The pteroeidids are here considered the most highly-derived pennatulaceans on the basis of both the presence of well-developed polyp leaves as well as the restriction of the siphonozooids to the polyp leaves. All other sea pens have siphonozooids present on the rachis. The genus Pennatula was here chosen as the outgroup to the pteroeidid genera on the following basis. The well-developed polyp leaves of Pennatula are similar to those of the pteroeidids in being large, often thick, fleshy, and opaque, as well as containing numerous autozooids in several layers within the leaves. Most often, the polyps are imbedded within the fleshy matrix composing the leaves. The virgulariids, on the other hand, have polyp leaves that are not as highly developed as the pennatulids and pteroeidids. The virgulariid polyp leaves are often small, thin, and translucent, being composed of the fused polyp Figure 6. Summary of classification schemes for pennatulid and pteroeidid taxa. A. Kolliker, 1869. B. Kolliker, 1880. C. Kukenthal, 1915. D. Hickson, 1916. E. Present study with cladogram showing overlapping characters that unify the six taxa. Williams (1995)  bodies of a single layer of the autozooids themselves. Consequently, the pennatulids are considered the pennatulacean taxa closest to the pteroeidid clade.
Characters and Polarity (refer to Table 1).
1 . Polyp leaf mesozooids: Pennatula injlata has conspicuous mesozooids, each with two spiculated terminal teeth, densely-distributed in a row along the basal dorsal margin of each polyp leaf, as well as in two longitudinal rows along the rachis (Fig. 5A). This character is unique to Pennatula, and is therefore considered an autapomorphy. 2. Calyx teeth: In the genus Pennatula, eight distinctly spiculated teeth are present on the distal portions of each autozooid calyx (Fig.  5A). In the other pennatulid genus, Ptilosarcus, only one or two spiculated teeth are present, which are sometimes indistinct (Fig.  5B). In Gyrophyllum, two fleshy and very sparsely spiculated teeth occur, or teeth are absent altogether (Fig. 5C). In Sarcoptilus, Crassophyllum, and Pteroeides, calyx teeth are mostly absent. Consequently, the presence of eight distinct calyx teeth is consid-ered a plesiomorphic character, while a reduction of the number of teeth or of the degree of spiculation is considered the derived state. 3. Dorsal tract siphonozooids: In species of Pennatula, siphonozooids are present on the dorsal surface of the rachis as well as the lateral portions of the rachis between the polyp leaves (Fig. 5A). In Ptilosarcus, the siphonozooids are restricted to the dorsal surface of the rachis and are absent from the lateral margins of the rachis between the polyp leaves. In Gyrophyllum, siphonozooids may be scattered on the dorsal surface of the rachis in addition to the polyp leaves (Hickson, 1916: 254). In Sarcoptilus, Crassophyllum, and Pteroeides, the siphonozooids are restricted to the polyp leaves-being absent from the rachis altogether (Figs. 5D-F). The restriction of the sipnonozooids to the dorsal region of the rachis is here considered autapomorphic for Ptilosarcus while the disposition of siphonozooids on the lateral portions as well as the dorsal surface of the rachis, as in Pennatula, is considered plesiomorphic. 4. Polyp leaf siphonozooids: In species of the pennatulid genera Pennatula and Ptilosarcus, the siphonozooids are restricted to the surface of the rachis, and are not present on the polyp leaves. In Pennatula firmbriata, the entire dorsal surface of the rachis is covered with siphonozooids. In the pteroeidid genera, siphonozooids are either restricted to the polyp leaves, or in Gyrophyllum may be present on the leaves as well as part of the rachis (Fig. 5C). Also, in other pennatulacean genera with polyp leaves such as Virgularia, Scytalium, and Stylatula, siphonozooids are present on the rachis. Consequently, the presence of siphonozooids on the rachis is here considered plesiomorphic, while the presence of siphonozooids on the polyp leaves is considered the apomorphic state. 5. Distribution of leaf siphonozooids: Gyrophyllum is the only pteroeidid taxon to have siphonozooids present on both surfaces of each polyp leaf. In Pennatula and Ptilosarcus, siphonozooids are not present on the polyp leaves and in Sarcoptilus, Crassophyllum, and Pteroeides, they are present only on the lower side of each leaf. The presence of siphonozooids on both sides of each leaf is therefore considered autapomorphic for Gyrophyllum. 6. Siphonozooid zones: The restriction of siphonozooids to distinct regions or zones in the proximal region of each polyp leaf is found only in three pennatulacean taxa-Sarcoptilus, Crassophyllum, and Pteroeides (Figs. 5D-F). In the pennatulids, as well as Gyrophyllum, the siphonozooids are not organized into distinct zones on the leaves. The presence of siphonozooid zones is considered a synapomorphy for Sarcoptilus, Crassophyllum, and Pteroeides, while their absence is plesiomorphic. In Crassophyllum and Pteroeides, the zones take the form of siphonozooid plates, while in Sarcoptilus they are modified as siphonozooid pads. 7. Sclerites: Three-flanged sclerites, composed of three longitudinal grooves along the surface of an otherwise smooth and elongated sclerite, are present in many pennatulacean taxa including Pennatula, Ptilosarcus, and Gyrophyllum, as well as many less-derived taxa (Figs. 2, 3). The three-flanged state is not found in Sarcoptilus, Crassophyllum, and Pteroeides. Regarding the pteroeidids, the presence of three-flanged sclerites is consequently considered to be a plesiomorphic character state, while the presence of smooth sclerites is considered the derived state. 8. Distal rachis mesozooids: Members of the genus Sarcoptilus are the only sea pens for which mesozooids have been reported as present on the distal ventral portion of the rachis (Williams, in press). This character is therefore considered an autapomorphy for Sarcoptilus, while its absence is considered plesiomorphic. 9. Siphonozooid pads: The arrangement of the siphonozooids into distinct ovoid swollen pads in the proximal region of each polyp leaf adjacent to the dorsal surface of the rachis, is unique to one pennatulacean taxon -Sarcoptilus (Fig. 5D). This character is considered a further modification of the zonation and restriction of siphonozooids in the proximal portions of the polyp leaves. rays are found in the genus Pteroeides and are not present in any other pennatulacean taxon (Fig 5F). These rays are composed of elongate needle-like sclerites aligned in parallel and in close proximity. They act as supporting rays to give a state of rigidity to the leaves. The rays radiate outward from the proximal siphonozooid zone to (and often beyond) the distal margins of the polyp leaves often resulting in a spiny appearance. In Pteroeides dofleini Balss, 1909, a single conspicuous ray is present along the dorsal mar-gin of each leaf, but multiple rays throughout the surface of the leaf are absent. Because of the uniqueness of this character, its presence is considered autapomorphic to Pteroeides, and its absence is considered the plesiomorphic state.

Conclusion
The cladogram in Figure 7 was produced from the character matrix in Table 1 using PAUP (Phylogenetic Analysis Using Parsimony) version 3.1. A single tree was generated with a CI (Consistency Index) of 0.923.
A detailed comparative examination of the taxa Pennatula, Ptilosarcus, Gyrophyllum, Sarcoptilus, Crassophyllum, and Pteroeides, shows that Gyrophyllum clearly shares characters with both families and therefore bridges the morphological gap between the two (Fig. 6E). In addition, a preliminary cladistic analysis suggests that these six genera formerly assigned to the two families Pennatulidae and Pteroeididae, can best be viewed as representing a single monophyletic group (Fig. 7), thereby making the distinction and retention of the two families, as one of only nominal significance (Williams, 1995).
As a result of the present study, there is no justification for recognizing two distinct families. I therefore consider the Pteroeididae Kolliker, 1880, to be synonymous with the Pennatulidae Ehrenberg, 1 834, which has priority. The six genera listed above are here considered members of the family Pennatulidae.
The family Pennatulidae is here defined as follows: bilaterally symmetrical sea pens with large, mostly fleshy polyp leaves; autozooids in one or more rows along the margins of the polyp leaves; anthocodiae retractile into the base of the polyps, which may be fleshy or in the form of spiculated calyces with terminal teeth; mesozooids present or absent; siphonozooids are usually conspicuous and numerous in bands or longitudinal rows on the rachis or polyp leaves, or in pads or plates at the base of the polyp leaves; sclerites are threeflanged or smooth needles, rods, plates, or ovals.