Revisionary systematics of the endemic soft coral fauna (Octocorallia: Alcyonacea: Alcyoniina) of the Agulhas Bioregion, South Africa Mcfadden, Catherine S. Van Ofwegen, Leen P. Zootaxa 2017 2017-12-13 4363 4 451 488 5TVXM Mcfadden & Van Ofwegen, 2017 Mcfadden & Van Ofwegen 2017 [151,375,152,178] Anthozoa Alcyoniidae Alcyonium GBIF Animalia Alcyonacea 12 463 Cnidaria species dolium sp. nov.  ?  Alcyonium sarcophytoides(non Burchardt, 1903) Thomson 1921: 157–158.     Materialexamined.  Holotype. RMNH Coel.40204 (SAF292), South Africa, KwaZulu-Natal, Park Rynie, Lander’s Reef, 30º19.973'S, 30º47.523'E, depth  22–28 m, coll. C.S. McFadden,  18 March 2008.   Othermaterial.  Alcyonium variabile: RMNH Coel.40202, South Africa,  EasternCape, Algoa Bay, Riy Banks; 33º59.097'S, 25º51.774'E, depth  15–20 m, coll. C.S. McFadden,  11 March 2008.  RMNH Coel.41530,  RMNH Coel.41531, South Africa,  EasternCape, Algoa Bay, Evans Peak, 33º50.578'S, 25º48.998'E, depth  22–28 m, coll. C.S. McFadden,  14 March 2008.  RMNH Coel.40203,  South Africa,   WesternCape, CapePeninsula, Vulcan Rock, 34º03.970'S, 18º18.627'E, depth  15–27 m, coll. C.S. McFadden,  23 March 2008.  RMNH Coel.40800, South Africa,  EasternCape, Algoa Bay, Riy Banks, depth  18–20 m, coll. J. Starmer,  12 October 1998.   Description. The holotypeis a firm, club-shaped colony, 4.0 cm tall ( Fig. 5b). The stalk is 1.5 cmlong and 1.2 cmin diameter at its widest point, and is smooth except for a few small bumps and longitudinal wrinkles. The polyparium is smooth and bulbous, 1.8 cmin diameter at its widest point near the distal end. The stalk and polyparium are demarcated only by the presence of polyps in the latter, and by a slight difference in color. Polyps are distributed regularly over the surface of the polyparium, spaced about 1 mmapart. The polyps are retracted flush with the colony surface, but remain visible as small, pale bumps. The sclerites in the polyps are spindles, 0.14–0.22 mmlong, sparsely tuberculated and often with small terminal spines ( Fig. 7a), arranged in a collaret and points. The tentacles have small, flat rods ( Fig. 7b). The sclerites in the surface of the polyparium are stout spindles, clubs and radiates, 0.06–0.14 mmlong, with thorny tubercles ( Fig. 7c–e). The interior of the polyparium contains needle-like spindles up to 0.47 mmlong ( Fig. 7f–g), very sparsely distributed. The surface and interior of the stalk have tuberculated spheroids with a median waist, 0.09–0.11 mmlong and up to 0.08 mmwide ( Fig. 7h), barrels, 0.15 mmlong by 0.07 mmwide ( Fig. 7i), and a few small radiates, 0.05–0.08 mmlong ( Fig. 7j), all with thorny tubercles.  Color. Golden yellow with a paler yellow stalk and white polyps. The sclerites are colored, mostly yellow with a few red spindles in the colony interior. Azooxanthellate.   Etymology. From the Latin  dolium, a cask or tun, in reference to the barrel-shaped sclerites in the surface of the stalk.   Remarks.  A. dolium  n. sp.is superficially very similar to the yellow color morph of  Alcyonium variabile(Thomson) (Williams 1986b). It can be distinguished from that species by differences in the colony growth form, which in  A. dolium  n. sp.is clavate without a strong distinction between stalk and polyparium, while  A. variabileis typically capitate, with a strongly demarcated stalk and polyparium. The form of the sclerites in the surface and interior of the stalk differs, with  A. variablehaving radiates while  A. dolium  n. sp.has barrels and spheroids with a median waist. The clubs in the polyparium surface also differ in shape from those of  A. variabile. These distinctions can be seen by comparing Fig. 7to Fig. 8, which depicts the sclerites typically found in  A. variabile. Phylogenetic analyses place  A. dolium  n. sp.as a very close sister taxon to  A. variabile( Fig. 1) but slightly differentiated from it genetically at all of the loci we sequenced. Pairwise genetic distances (K2p) between the two species were 0.4–0.6% at mtMutS, 0.1% at COI, and 1.5–1.7% at 28SrDNA.   A. compactofestucumVerseveldt & Ofwegendiffers from both  A. dolium  n. sp.and  A. variabileby the unique shape of the sclerites in the stalk interior, which are spindles with long, oblique spinose processes (Verseveldt & Ofwegen 1992: Fig. 6, r–t) rather than spheroids, barrels or radiates. Likewise,  A. studeri(Thomson)can be distinguished from each of these other similar species by the presence of needle-like spindles in the interior of the stalk. Thomson (1921) identified as  A. sarcophytoides(non Burchardt) a specimen he described as “irregularly cylindrical in shape and yellow in color,” with a distinct stalk and polyparium, and sclerites of a form he termed “Zwirnrollen” (spools). Williams (1992a) reported, however, that this specimen has long, slender spindles in the surface of the polyparium, a character that would distinguish it from  A. dolium  n. sp.Neither author, however, gives any illustrations of the colony or its sclerites, and we were unable to examine the specimen.   A. dolium  n. sp.occurs syntopically with  Parasphaerasclera aurea(Benayahu & Schleyer), which shares a similar golden-yellow color and unbranched growth form. Colonies of  P. aureaare, however, digitiform rather than clavate, and typically have a shorter stalk, the polyps lack sclerites, and the only sclerite forms present in the colony surface and interior are tuberculate spheroids and radiates (McFadden & Ofwegen 2013). In addition to the four very similar species discussed above, the only other South African species that appears to belong to  Alcyonium sensu strictois  A. elegans(Kükenthal).  A. eleganshas an unbranched colony form with polyps restricted to the distal end of a broad and compressed sterile stalk. Like other  Alcyoniumspecies (McFadden & Ofwegen 2013) it has spindles arranged to form a collaret and points in the polyp, radiates in the colony surface, both radiates and slender spindles in the interior coenenchyme, and the sclerites are permanently colored (Williams 1992a). Attempts to obtain DNA sequences from specimen SAM H3802 were unsuccessful, so at present we cannot confirm this species' status as a member of  Alcyonium sensu stricto. Several other South African species that are currently placed in  Alcyoniumdo not, however, share the characters of that genus. We reassign  A. fauriThomson, A. distinctumWilliamsand  A. wilsoni(Thomson)to other genera below. The correct generic placement of  A. foliatumThomson, a membranous species with foliate clubs in the coenenchyme, is not clear at present, but it does not appear to belong to  Alcyonium sensu stricto. Attempts to sequence specimen SAM H3690 failed. Williams (1992a) discussed six additional species of  Alcyoniumthat have been reported from South African waters, but whose occurrence and identity remain unverified.  Other Alcyoniidae.Four additional genera endemic to South Africaare currently assigned to family Alcyoniidae. These include the monotypic  Dimorphophyton mutabiliformeWilliams,  Lanthanocephalus clandestinusWilliams & Starmerand  Malacacanthus capensisHickson, and two species of  VerseveldtiaWilliams.Phylogenetic analysis of a 28SrDNA sequence obtained from an uncatalogued specimen from SAM (SAF492) suggests that  Dimorphophytonbelongs not to Alcyoniidaebut rather to Paralcyoniidae, where it is a sister taxon to  Studeriotesand [  Paralcyonium+  Ceeceenus] ( Fig. 1). Like these other genera,  Dimorphophytonhas a colony form consisting of a distinct, stout pedicel into which the fleshy polyparium or trunk can be partially or completely retracted. The simple, irregular rod-like sclerites of  Dimorphophytonare also similar in form to those found in the polyps and trunk of other paralcyoniid genera (e.g., Ofwegen & Benayahu 2006), but it differs from them in lacking supporting spindles in the pedicel. Based on the molecular phylogenetic result and these morphological similarities, we hereby re-assign  Dimorphophytonto Paralcyoniidae. Among the other three endemic genera assigned to Alcyoniidae, molecular data are available only for  Malacacanthus capensis.This species occupies a unique phylogenetic position in the Holaxonia-Alcyoniina clade of octocorals, not closely related to any other genus of Alcyoniidae(or any other family) for which sequence data are available ( Fig. 1). Better phylogenetic resolution of the basal relationships within the Holaxonia-Alcyoniina clade is needed, however, to determine if  Malacacanthusmerits assignment to a monotypic family. Although  Lanthanocephalusand  Verseveldtiaare unlike any other members of Alcyoniidae, molecular phylogenetic datanot currently available for either genus—will be necessary to confirm their familial placement.  Lanthanocephalusresembles  Malacacanthusmost closely in colony growth form, but unlike that genus it has sclerites (spindles and radiates) in the colony surface (Williams & Starmer 2000). The sclerites of  Verseveldtiaare permanently colored spheroids and radiates similar in form to those found in family Parasphaerascleridae, but it differs from  Parasphaerasclerain having dimorphic polyps and sclerites in the autozooids (Williams 1990). 1705513503 2008-03-18 RMNH C. S. McFadden South Africa Material 25 -30.332884 Park Rynie 1 30.79205 Coel. 12 463 1 KwaZulu-Natal holotype 1705513525 2008-03-11 RMNH C. S. McFadden South Africa Other 18 -33.98495 Algoa Bay 1 25.8629 Coel. 12 463 1 Eastern Cape holotype 1705513516 [944,1181,417,442] RMNH South Africa Coel. 12 463 1 Eastern Cape holotype 1705513514 2008-03-14 RMNH C. S. McFadden South Africa Coel. 25 -33.842968 Evans Peak 1 25.816633 Algoa Bay 12 463 1 Eastern Cape holotype 1705513504 2008-03-23 RMNH C. S. McFadden South Africa Coel. 21 -34.066166 Western Cape 1 18.31045 South Africa 12 463 1 Western Cape holotype 1705513519 1998-10-12 RMNH J. Starmer South Africa Coel. 19 Riy Banks Algoa Bay 12 463 1 Eastern Cape holotype