PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 48 Published 1992-1995 by the Calfiornia Academy of Sciences Golden Gate Park San Francisco, California SCIENTIFIC PUBLICATIONS COMMITTEE Thomas F. Daniel, Editor Ann Senuta, Managing editor Terrence M. Gosliner Tomio Iwamoto Thomas Moritz Wojciech J. Pulawski ISSN 0068-547X © California Academy of Sciences Golden Gate Park San Francisco, California 94118 PRINTED IN THE UNITED STATES OF AMERICA BY ALLEN PRESS, INC., LAWRENCE, KANSAS 66044 TABLE OF CONTENTS VOLUME 48 1. WILLLIAMS, GARY C. Revision of the soft coral genus Minabea (Octocorallia: Alcyoniidae) With new taxa from the Indo-West Pacific. (Illus.) July 28, 1992.................. 1-26 2. SAZONOV, YURI I. AND TOMIO IWAMOTO. Grenadiers (Pisces, Gadiformes) of the Nazca and Sala y Gomez ridges, southeastern Pacific, (Illus.) July 28,1992 ..........4...... 27-95 3. SHCHERBACHEV, YURI N., YURI I. SAZONOV, AND TOMIO IWAMOTO. Synopsis of the grenadier genus Kuronezumia (Pisces: Gadiformes: Macrouridae), with description of a new species. PU APTIe MINN ENIADETa ee LO 92 5 2Ns ie ctr x 2 Ber epee (a8 rRPRo ay zor Does Sed 2 eee ee rer 97-108 4. LAMBERT, GRETCHEN. Three new species of stolidobranch ascidians (Chordata: Ascidiacea) inom the California continental shelf. (illus:) January 21,1993............2...... 109-118 5. DANIEL, THOMAS F. Taxonomic and geographic notes on Central American Acanthaceae. iN MM ANN AT 9 Ss en sheen veh, eeeeh eu e Lak Por ae REPS eal «2 S's eee eee 119-130 6. MAYR, ERNST. Fifty years of progress in research on species and speciation. ESIAIDE EM OO ye oy SE aK, Glace eee ee Gn tae eae oa ale a RNs ernae ee 131-140 7. ALMEDA, FRANK. An evaluation of the Mesoamerican species of Meriania (Melastomataceae: Seeniemeac s(luins.emecember2 1. 1993 2. As eke esas oe. os has one oe Ba ees 141-152 8. MCCOSKER, JOHN E. AND RICHARD H. ROSENBLATT. A revision of the snake eel genus Myrichthys (Anguilliformes: Ophichthidae) with the description of a new eastern Pacific Paeeeneciiloseccempen 2 1993 ca eet te ee ees et ae 153-169 9. GOSLINER, TERRENCE M. New records of Flabellinidae (Opisthobranchia: Aeolidacea) from the tropical Americas, with descriptions of two new species. (IIllus.) PMN) ee ee Sc Sven 5s ase apg Cn ohaie, Syerape tetas ein tie Sy = eee 171-183 10. BROWN, WALTER C. AND ANGEL C. ALCALA. Philippine frogs of the family Rhacophoridae. TSS) AAS 1G. IO eS ORO ey Re gine eee earner 185-220 11. IWAMOTO, TOMIO AND YURI I. SAZONOV. Revision of the genus Kumba (Pisces, Gadiformes, Macrouridae), with the description of three new species. (IIlus.) DLT OS) ty Sr see eS Sie ease 2 eaten eileen Glin Gael gue loisge Gig csue iets, oo 221-237. 12. GOSLINER, TERRENCE M. New species of Chromodoris and Noumea (Nudibranchia: Chromodorididae) from the western Indian Ocean and southern Africa. (IIlus.) Ree Ae Re ey ee LEE cel IN al 3 ciara reese he tna ears eso ana snaione « ad enee 239-252 13. DANIEL, THOMAS F. New and reconsidered Mexican Acanthaceae. VI. Chiapas. (Illus.) ceva 2. WOO 6 ooh Be ii St Ao mie ny Une aan Ee SA PRI a aeirth e BOISE aici: Sc 253-284 14. SHCHERBACHEV, YURI N. AND TOMIO IWAMOTO. Indian Ocean grenadiers of the subgenus Coryphaenoides, Genus Coryphaenoides (Macrouridae, Gadiformes, Pisces). (Illus.) LEIS De GIS eS AI sR Si a ea Re Pree Oa thr 285-314 15. WILLIAMS, GARY C. The enigmatic sea pen genus Gyrophyllum— A phylogenetic reassess- ment and description of G. sibogae from Tasmanian waters (Coelenterata: Octocorallia). (IIlus.) 2 SEVEUSIL | MISS 2 8 esas cite Ns Pan ea OP See PORE Ont ae ee APN 315-328 16. DE NEVERS, GREG. Notes on Panama Palms. (Illus.) August 1, 1995.............. 329-342 Dagestan ome 45. December 9: 1996... ene eT owe ae bs HOP eee we oe oe ee 343-352 *. i plsitittye, t Me + 7¥feié vai) CATS) sie a } eoMinir | ey . af gip Cty oS Nerina? (tytn nth Cai Tite - eS | 5 eke es » eles @oneies Be i f lie ae J ehinlat “ine 7 Gk! Sat mec tani ie Nat Fi 4 Tilia way (PRA +). WI hisk stove" One : ad nary . PROCEEDINGS Bi Wane OFTHE jie EAST acre ITEC ; ; ANOTALOL Y al Laporatt ry LIBRARY CALIFORNIA ACADEMY OF SCIEN GES ——— Vol. 48, No. 1, pp. 1-26, 16 figs., 2 tables. = ys July 28, 1992 f Woods Fo 5 em creme REVISION OF THE SOFT CORAL GENUS MINABEA (OCTOCORALLIA: ALCYONITDAE) WITH NEW TAXA FROM THE INDO-WEST PACIFIC By Gary C. Williams Department of Invertebrate Zoology, California Academy of Sciences Golden Gate Park, San Francisco, California 94118 Asstract: A revised diagnosis for the soft coral genus Minabea Utinomi, 1957 (family Alcyoniidae) is provided as a result of recently collected material representing five new species from a wide geographic region in the Indo-West Pacific. At the present time the genus is thought to consist of nine species distributed from Africa to Fiji, and Japan to New Zealand, from the shallow waters of coral reefs to regions of deep water up to 370 meters. A comparison of the world species, related genera, and geographic distributions is included. Received May 28, 1991. Accepted November 10, 1991. INTRODUCTION The genus Minabea Utinomi, 1957 (family Al- cyoniidae) was formerly applied to five digiti- form soft corals with dimorphic polyps. These included M. ozakii Utinomi, 1957 and M. ro- busta Utinomi and Imahara, 1976, from Japan. Anthomastus phalloides Benham, 1928, from New Zealand and 4A. agilis Tixier-Durivault, 1970, from New Caledonia, were transferred by Utinomi and Imahara (1976) to the genus Mina- bea. Anthomastus agilis differs from other spe- cies of both Anthomastus and Minabea, as well as those of other alcyoniid genera, and therefore is considered a member of a new genus that has not yet been published (P. N. Alderslade, pers. comm.). Bellonella indica Thomson and Hen- derson, 1905, from Sri Lanka, can be transferred to Minabea as well (F. M. Bayer, pers. comm. and the present study). Recent explorations in the Indo-West Pacific have provided additional material that can be [1] allocated to the genus Minabea but not to any of the previously described species. This addi- tional material has been collected from the In- dian Ocean coast of South Africa, northwestern Australia and the Great Barrier Reef, Fiji, the Bismarck Sea of Papua New Guinea, Belau, and Guam in the Mariana Islands. The material rep- resents five previously undescribed species, which are presented here as new species. Thus, the range of this genus of nine species is extended to in- clude the Indo-West Pacific from the southeast- ern coast of Africa to Fiji in the western Pacific, and Japan in the north to New Zealand in the south. A comparison of Minabea with the superfi- cially similar genera Acrophytum Hickson, 1900, and Verseveldtia Williams, 1990 (both from South Africa) is presented, along with biogeographic considerations. A morphological and faunistic comparison of all species of Minabea is also in- cluded. 2 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 MATERIALS AND METHODS All material was collected by means of SCU- BA or steel dredge. Colonies were preserved in 70% ethanol. Sodium hypochlorite was used to dissociate sclerites from tissue. Drawing tubes attached to a dissecting microscope and a com- pound microscope were utilized to depict colo- nies and sclerites. A Hitachi S-510 scanning elec- tron microscope was used to make micrographs of sclerites. Institutions that sent or stored ma- terial for study are abbreviated in the text as follows: BMNH: British Museum (Natural His- tory), London, England; CASIZ: California Academy of Sciences, Department of Inverte- brate Zoology, San Francisco, California; MSM: Marine Science Museum, Tokai University, Shi- zuoka, Japan; NTM: Northern Territory Muse- um of Arts and Sciences, Darwin, Australia; SAM: South African Museum, Cape Town, South Af- rica; USNM: United States National Museum (Smithsonian Institution), Washington, D.C. SYSTEMATIC ACCOUNT Key to the World Species of Minabea 1. Colonies digitiform (fingerlike to cylindri- Callomelavate ie 2 — Colonies dome-shaped (hemispherical to COMIC) perenne 8 2. Sclerites of the colony include spindles or OO. Se ee ee 3 Spindlesion rods absent... 5 3. Sclerites include slender, relatively smooth rodsofthe mterior _ M. ozakii Ev SMIOO¢MOGS:ADSEML.- 4 4. Colonies elongate fingerlike, tapering dis- tally. Sclerites as radiates and thorny spin- CE ue ta Se eee Atte M. phalloides — Colonies short cylindrical. Sclerites as eight radiates and thorny spindles ... M. kosiensis 5. Surface of coenenchyme with two distinct types and sizes of sclerites: small radiates and large robust barrels _. M. cosmarioides — Sclerites throughout the colony not of con- spicuously disparate sizes 6 6. Polyparium forms >50% total colony length. Colonies from shallow water (< 30 INELETSIOE Nth) wee AOU o ms be Least) i - Polyparium forms <50% total colony length. Colonies from deep water A Ba a ens, a Roh Cac M. indica 7. Sclerites as eight-radiates and elongate bar- Tels: 2.2200 eee M. aldersladei — Sclerites as radiates, robust barrels and tu- berculate’spheroids 2 22 es M. robusta 8. Colonies hemispherical, usually longer than wide. Stalk prominent. Polyps restricted to extreme distal end of colony, which may be somewhat flattened or rounded to con- ical. Sclerites as barrels and eight-radiates ett WS 0, a M. acronocephala — Colonies conical, wider than long. Stalk in- conspicuous. Polyps completely cover most of the colony, which tapers distally to form a central apex. Sclerites as six- or eight- radiates and double stars M. goslineri Minabea Utinomi, 1957:139; Utinomi and Ima- hara, 1976:206. Bayer, 1981:913. Tixier-Du- rivault, 1987:154. REVISED GENERIC DIAGNOsIS.—Colonies un- branched and hemispherical to digitiform: low and dome-shaped to elongate and finger-like. Polyps dimorphic. The distal polyparium arises from a proximal stalk that is variable in length. Polyps evenly distributed over distal portion of colony and capable of complete retraction. Si- phonozooids are minute, numerous and sur- round the autozooids, which are larger and fewer in number. Polyps not forming calyces. Polyps without sclerites. Sclerites of surface and interior of colony densely set: mostly barrels and six- or eight-radiates with spindles, rods, tuberculate spheroids; seven-radiates or double stars some- times occurring. Color red, orange, yellow or pinkish-white to cream-white. An alcyoniid ge- nus of at least nine species from the Indo-West Pacific, 1-370 meters in depth. DISTRIBUTION. — New information presented in this study as well as in previously published distributional data shows that the genus Minabea has a wide Indo-West Pacific distribution that includes southeastern Africa, Sri Lanka, north- western Australia, the Great Barrier Reef, New Zealand, Fiji, Papua New Guinea, Belau, Guam, and Japan (Fig. 16). A triangle roughly formed by Guam, northwestern Australia and Fiji has the highest number of species with three. In ad- dition there are four outlying regions: Japan with two species to the north, Sri Lanka with one species and southern Africa with two species to the west, and New Zealand with one species to the south. WILLIAMS: SOFT CORAL GENUS MINABEA Minabea acronocephala sp. nov. (Figs. 1A, 2, 3) Type MATERIAL. — Holotype: CASIZ 078414, in small caves along vertical surface of outer barrier reef, The Pinnacle (5°10’S, 145°50’E), PAPUA NEW GUINEA. Madang Province: Bis- marck Sea, 9-18 meters depth, 15 November 1990, Scuba, G. C. Williams. Paratypes: CASIZ 078415, two colonies (each cut longitudinally into two halves), same data as holotype. OTHER MATERIAL.—CASIZ 078416, 15 colonies, same data as holotype. CASIZ 078417, 9 colonies, in small caves and overhangs on outer wall of barrier reef, north of Wongat Island (5°10’S, 145°50’'E), PAPUA NEW GUINEA. Madang Proy- ince: Bismarck Sea, 6-21 meters depth, 25 November 1990, G. C. Williams. CASIZ 078418, 2 colonies, in small caves and overhangs along walls, The Quarry (4°52’S, 145°48’E), PAPUA NEW GUINEA. Madang Province: Bismarck Sea, 8—15 meters depth, 20 November 1990, Scuba, G. C. Williams. NTM-C. 2976, 1 colony, FIJI. Viti Levu: Suva Harbor (18°15’S, 178°30'E), 5-10 meters depth, November 1984, C. Ireland. DiAGnosis.— Colonies dome-shaped. Stalk conspicuous and longer than polyparium. Polyps restricted to the distal extremity of the colony. Sclerites eight-radiates and barrels (0.05—0.13 mm long), with short medial waists. Color variable: yellow, orange, red. DESCRIPTION.— The thirty colonies that were examined range in height from 4 to 14 mm and from 4 to 9 mm in width. The colonies are dome- shaped (hemispherical to sub-digitiform), the to- tal height usually being somewhat greater than the diameter, although some colonies may be slightly wider than tall. The stalk is usually con- spicuous and longer than the polyparium, being distinctly demarcated from it. The stalk arises from the basal holdfast, which may be somewhat spreading, and gives rise to the slightly enlarged distal polyparium, as the distal half of the col- onies represent the greatest width. The polyps are restricted to the extreme distal end of the colony, which is often somewhat flattened to low- rounded or sometimes conical. In preserved ma- terial, the siphonozooids often are indicated by nothing more than minute pores approximately 0.07 mm in diameter surrounding the markedly larger retracted autozooids, which measure ap- proximately 1 mm in diameter. The sclerites are densely distributed in the surface coenenchyme as well as throughout the interior of the colonies. These sclerites are primarily eight-radiates and barrels 0.05-0.13 mm long, relatively finely tu- berculated, with short medial waists (0.007-0.015 mm in length). A few sclerites may possess more elongate medial regions. Sclerites of the polypary surface range in length from 0.05 to 0.13 mm. The interior of the polyparium contains sclerites that are 0.08—-0.12 mm long. Sclerites from the surface and interior of the stalk measure 0.05— 0.11 mm in length. The colonies are variable in color, ranging from entirely yellow or yellow with reddish polyparies and deeper red autozooids, to red-orange with yellow polyparies and auto- zooids. Some colonies are entirely reddish with deeper red autozooids. The holdfast region varies from yellow to orange or salmon. DISTRIBUTION.— This species is known from the vicinity of Madang, Papua New Guinea, in the Bismarck Sea, and Suva Harbor, Fiji, be- tween 5 and 21 meters in depth. At Madang, colonies are encountered mostly along the outer walls of barrier reefs: on the ceiling or sides of caves, alcoves, or depressions found along areas of vertical relief. EtyMoLocy.—The specific epithet is derived from the Greek akron = extremity, summit, highest point, and kephale = a head; in reference to the restriction of the polyp-bearing portion to the distal-most extremity of the colony. REMARKS.— Minabea acronocephala can be distinguished from other species of the genus by having a low, dome-shaped growth form with a somewhat flattened or low-rounded to conical distal region and prominent stalk, polyps re- stricted to the extreme distal end of the colony, and sclerites that are predominantly eight-radi- ates or barrels with short medial waists. Minabea aldersladei sp. nov. (Figs. 1B, 4, 5) Bellonella indica (non Thomson and Henderson, 1905): Bayer, 1974:261; Faulkner and Chesher, 1979:267, pl. 22. Type MATERIAL. — Holotype: CASIZ 078419, in small caves on vertical relief, Magic Passage (5°10’S, 145°50’E), PAPUA NEW GUINEA. Madang Province: Bismarck Sea, 10-15 me- ters depth, 14 November 1990, G. C. Williams. Paratypes: CASIZ 078420, 2 colonies (one of which is cut longitudinally into two halves), same data as holotype. Paratypes: NTM C-10796 and C-10797, 2 colonies, same data as holotype. OTHER MATERIAL. —Same data as holotype, 3 colonies CAS- IZ 078421. PAPUA NEW GUINEA. Madang Province: Bis- marck Sea, 5°10’S, 145°50’E): Outer Pig Island, 12 m, Scuba, T. M. Gosliner, 17 November 1990, | colony (distal part only) from a vertical wall, CASIZ 078422. Wongat Island, 14 m, Scuba, G. C. Williams, 13 November 1990, 1 colony from vertical surfaces shoreward of the island, CASIZ 078423. Daphne’s Reef, 10 m, Scuba, G. C. Williams, 13 November 1990, 1 colony from depression in wall surface, CASIZ 078424. AUSTRALIA. Queensland: Great Barrier Reef, Low Isles, 16°18’S, 145°35’E, P. Murphy, 11 June 1976, 2 colonies, one of them cut in half longitudinally, SAM-H4263; M. Strepher + PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 WILLIAMS: SOFT CORAL GENUS MINABEA 5 1560 km Figure 2. Minabeaacronocephala A, A single colony, 5.4 mm in height. B, Longitudinal section through colony in A showing gastric cavities of autozooids and siphonozooids. C, Detail of a single autozooid surrounded by nine siphonozooids, length of figure represents 1.6 mm. D, Fifteen colonies showing variability of colony shape, scale bar represents 10 mm. E, Map of the Coral Sea and southwestern Pacific Ocean showing distribution of the species; black dots represent collecting stations, arrow shows type locality. F, Surface sclerites of the polyparium. G, Interior sclerites of the polyparium. H, Stalk surface sclerites. I, Stalk interior sclerites. Scale bar for F-I represents 0.1 mm. — FicurE 1. Photographs of holotypes. A, Minabea acronocephala (8 mm in height). B, M. aldersladei (38 mm in height). C, M. cosmarioides (42 mm in height). D, M. goslineri (9 mm in height). E, M. kosiensis (26 mm in height). PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 Ficure 3. Minabea acronocephala Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.10 mm. B, 0.10 mm. C, 0.07 mm. D, 0.10 mm. E, 0.06 mm. F, 0.08 mm. G, 0.06 mm. H, 0.09 mm. WILLIAMS: SOFT CORAL GENUS MINABEA gastric cavities of autozooids and siphonozooids. C, Detail of a single autozooid surrounded by 10 siphonozooids, length of figure represents 1.6 mm. D, Map of the western Pacific Ocean showing distribution of the species; black dots represent collecttig stations, arrow shows type locality. E, Surface sclerites of the polyparium. F, Interior sclerites of the polyparium. G, Stalk surface sclerites. H, Stalk interior sclerites. Scale bar for E-H represents 0.1 mm. and C. Duke, 11 June 1972, 10 colonies, NIM C-1902; 1 colony, NTM C-5451. Snake Reef, 14°20’S, 145°10’E, 2-5 m, P. Alderslade, 14 December 1990, 5 colonies, NTM C-10427; 10-12 m, J. Hooper, 14 December 1990, 2 colonies, NTM C-10449; 1 colony, NTM C-10448. Flinders Reef, 17°30’S, 148°10’E, 20-25 m, Z. Dinesen, NTM C-4083. Off Cairns, 16°50’S, 145°45’E, 7-10 m, D. Schubot, November 1978, 10 colonies NTM C-901. Moore Reef, 17°00’S, 146°15’E, P. Al- derslade, 1979, 1 colony NTM C-1904. Western Australia: edge of Clerke Reef, Rowley Shoals, 17°23'S, 119°23’E, 19 m, J. Hooper, 18 July 1987, 1 colony, NTM C-5889. BELAU. Palau Islands: Ngemelis Islands, Great Reef, Bai- lechesengel Island; 7°08’N, 134°29’E, 4.6 m, D. Faulkner, 28 August 1973, 5 colonies, USNM 58498. D1aGnosis. — Colonies elongate-digitiform, ta- pering distally, often curved. Stalk restricted to proximal most one-sixth to one-third of colony. Sclerites elongate barrels and eight-radiates 0.04— 0.12 mm long. Color either uniform yellow or uniform red-orange. DESCRIPTION. — The 55 colonies that were ex- amined vary in length from 19 to 69 mm and 7 to 18 mm in width. The colonies are elongate digitiform, often somewhat curved. Rarely, col- onies may be bifurcated forming two distal lobes. The colonies taper gradually distally forming a PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 WILLIAMS: SOFT CORAL GENUS MINABEA rounded distal end, the holdfast and stalk being the widest portion of the colonies. The polyps are evenly distributed throughout the entire sur- face of the polyparium. In all but one specimen, the polyps are retracted. The partially extended autozooids from this specimen are approxi- mately 1.5 mm long by 0.8 mm wide. The re- tracted autozooids measure approximately 0.8- 1.0 mm in diameter and are surrounded by si- phonozooids that form minute pores approxi- mately 0.03-0.10 mm in diameter. The stalk forms the proximal most one-third to one-sixth of the colony, while the polyparium comprises the distal most two-thirds to five-sixths. The sclerites of the colonies are all elongate barrels and eight-radiates, 0.04—-0.12 mm in length. The median waists of the sclerites vary from 0.005 to 0.010 mm in length in barrels to 0.012-0.025 mm long in radiates. The surface and interior coenenchyme of the polyparium contain sclerites that are 0.04-0.11 mm in length. Sclerites from the surface of the stalk are 0.05—0.11 mm long, while those of the interior are 0.06-0.12 mm long. Two color varieties are known: colonies from the Madang region of the Bismarck Sea are either entirely lemon-yellow or uniformly red- dish-orange, while colonies from Australia are either a uniform, rich golden-yellow or entirely orange to deep red-orange. All sclerites are col- ored orange or yellow, none are colorless. Exsert autozooids are white. DISTRIBUTION.—The Madang region of the Bismarck Sea (Papua New Guinea), Rowley Shoals (northwestern Australia), the Great Bar- rier Reef (Queensland, Australia), and Belau; 2- 25 meters in depth. ETYMOLOGY.— This species is named for Phil Alderslade of the Northern Territory Museum, Darwin, Australia, a friend, colleague, and en- thusiastic student of the Octocorallia, who gen- erously supplied numerous colonies of this spe- cies for study. REMARKS. — Unlike other species of Minabea, M. aldersladei has this unique combination of characters: an elongate, often curved finger-like growth form that tapers distally, polyps cover over two-thirds of the surface of the colony, scle- — FiGure 5. rites of elongate barrels and eight-radiates, and colony color that is either entirely yellow or uni- form red-orange. Some colonies of M. aldersladei collected off Cairns (Great Barrier Reef) resemble colonies of M. goslineri in that they are yellow in color and low-conical in growth form. The sclerite com- plement of the Cairns material contains many barrels in addition to radiates, which is charac- teristic of M. aldersladei. On the other hand, the sclerites of M. goslineri are exclusively radiates. The sclerites of M. acronocephala and M. al- dersladei are very similar in appearance. In ad- dition, the two species exhibit sympatric distri- bution at Madang, Papua New Guinea. In fact, colonies of both species may occur in the same caves or overhangs on the outer wall of the Sek Island to Pig Island barrier reef. However, these species differ significantly in three respects. Mi- nabea aconocephala is short and dome-like or button-shaped; the polyps are confined to the distal-most extremity of the colony, resulting in a conspicuous and prominent stalk that makes up the bulk of the colony, and the colonies usu- ally contain both red and yellow sclerites. In M. aldersladei, on the other hand, the colonies are usually long and finger-like; the polyps cover the distal two-thirds or more of the colony resulting in a short and less conspicuous stalk, and the colonies are always either uniformly yellow or red-orange. Minabea aldersladei bears a superficial resem- blance to M. indica (Thomson and Henderson, 1905). The two species are differentiated below under REMARKS for M. indica. Minabea cosmarioides sp. nov. (Figs. 1C, 6, 7) Tyre MATERIAL. — Holotype: SAM-H4260, dredged from a bottom of coarse sand and stones, off Mbotyi (31°33,2’S, 29°51,9'E), SOUTH AFRICA. Indian Ocean: Transkei, 250 meters depth, 4 July 1986, steel dredge, R. N. Kilburn and D. Herbert on board RV Meiring Naude. OTHER MATERIAL.—SAM-H4087, | colony cut longitudi- nally into two halves, same data as holotype. D1acnosis. — Colonies robust digitiform, near- ly cylindrical, slightly tapering to a broadly rounded apex. Polyparium forming the distal two- Minabea aldersladei Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.08 mm. B, 0.08 mm. C, 0.09 mm. D, 0.06 mm. E, 0.07 mm. F, 0.07 mm. G, 0.07 mm. H, 0.08 mm. I, 0.08 mm. J, 0.07 mm. K, 0.06 mm. L, 0.05 mm. 10 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 INDIAN TRANSKEI »” OCEAN 32° LONDON Ficure 6. Minabea cosmarioides A, A single colony, 34 mm in height. B, Longitudinal section through colony in A showing gastric cavities of autozooids and siphonozooids. C, Detail of a single autozooid surrounded by 16 siphonozooids, length of figure represents 2.5 mm. D, Map of southeastern Africa showing distribution of the species; black dot represents type locality and sole collecting station. E, Surface sclerites of the polyparium. F, Interior sclerites of the polyparium. G, Stalk surface sclerites. H, Stalk interior sclerites. Scale bar for E-H represents 0.1 mm. thirds or three-quarters of the colonies. Sclerites DESCRIPTION. — The two colonies examined are of two distinct kinds: smaller eight-radiates0.05— 41 mm long by 15-28 mm wide and 34 mm long 0.11 mm long and large rotund barrels 0.14-0.19 by 14-23 mm wide. The holotype is the larger mm long. Color dull brownish-orange. of the two specimens. The colonies are digitiform _ FicurE 7. Minabea cosmarioides Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.13 mm. B, 0.05 mm. C, 0.15 mm. D, 0.06 mm. E, 0.15 mm. F, 0.05 mm. G, 0.14 mm. H, 0.06 mm. I, 0.05 mm. J, 0.06 mm. WILLIAMS: SOFT CORAL GENUS MINABEA 12 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 and robust, cylindrical, and the polyparium ta- pers distally only slightly if at all. The distal ex- tremity is broadly rounded. The proximal third to a quarter of the colonial length represents the stalk while the polyparium comprises the distal two-thirds to three-fourths of the colony length. The stalk broadens proximally to form the spreading holdfast. The polyps are evenly dis- tributed over the entire surface of the polypari- um. The numerous siphonozooids measure ap- proximately 0.2 mm in diameter and surround the retracted autozooids, which are approxi- mately 1.6 mm in diameter. In the smaller col- ony, five autozooids of the apex region were pre- served extended and measure 5 mm long by 1.5 mm wide. The autozooids of the holotype were all preserved completely retracted. The sclerites of the surface coenenchyme are of two distinct types: smaller eight-radiates 0.05-0.07 mm long and large rotund barrels 0.14—0.19 mm in length. All sclerites have very short waists 0.005—0.012 mm in length. Radiates of the surface of the poly- parium vary in length from 0.05 to 0.07 mm, while the barrels from the same area are 0.14— 0.19 mm long. Sclerites of the interior of the polyparium are large barrels, 0.17-0.19 mm long. Radiates of stalk surface are 0.05—0.11 mm in length, while the barrels are 0.17—0.19 mm long. Sclerites of the interior of the stalk are also large barrels and measure 0.15—0.18 mm in length. Color of the colonies is predominantly a dull brownish-orange. The polyparium may be whit- ish with brownish-orange retracted autozooids and siphonozooids, resulting in a blotchy ap- pearance, while the stalk is uniformly brownish- orange. Extended autozooids are a uniform grey- ish-white. The brownish-orange coloration is restricted to sclerites of the surface of the coe- nenchyme, as the interior of the colony is a uni- form yellowish-white and the sclerites are col- orless. DISTRIBUTION.—This species is known only from the type locality off the central coast of Transkei, South Africa in the western Indian Ocean; 250 meters in depth. EtyMoLocy.—The specific epithet is derived from the chlorophyte genus Cosmarium and the Greek suffix oides, denoting likeness of form, in reference to the shape of the larger sclerites, which superficially resemble the body shape of various species of this genus of placoderm desmid. ReMARKS.—Minabea cosmarioides is differ- entiated from other species of the genus by the occurrence of two very distinct types of sclerites in the surface coenenchyme (small eight radiates and large rotund barrels), and only large rotund barrels in the interior of the colony. Minabea goslineri sp. nov. (Figs. 1D, 8, 9) Type MATERIAL.— Holotype: CASIZ 078425, on protected shady sides of dead coral pinnacles, shoreward of Anae Island (13°20'S, 144°40'E), MARIANA ISLANDS. Guam: Agat Bay off Nimitz Beach, 1-3 meters depth, 5 December 1990, Scuba, T. M. Gosliner. Paratypes: CASIZ 078426, 3 colonies (one of which is cut longitudinally into two halves), same data as ho- lotype. OTHER MATERIAL. —CASIZ 078427, 27 colonies, same data as holotype. DriaGnosis.—Colonies low-conical in shape, wider than long. Stalk short and inconspicuous. Polyps cover entire surface of conical or rounded polyparium. Sclerites six- or eight-radiates and double stars (0.05—0.12 mm long), with elongate median waists. Colony color uniform yellow. DESCRIPTION.—The 31 colonies examined range from 3 to 10 mm in height and from 6 to 14 mm in diameter. Colonies are short, often broader than long, dome-shaped to slightly con- ical. The stalk is very short and often inconspic- uous as the polyparium arises very close to the holdfast. The length of the stalk is usually con- siderably less than that of the polyparium. The polyparium often tapers to a central apex at the distal terminus. The polyps cover the entire sur- face of the conical or rounded polyparium, which represents most of the bulk of the colony. In preserved and retracted specimens, the siphon- ozooids are seen as minute pores approximately 0.8 mm in diameter, which surround the larger autozooids that are approximately 1.0 mm in diameter. The siphonozooids superficially may be very inconspicuous or not evident due to the extreme retraction in preserved material. Scle- rites are six- or eight-radiates and double stars (0.05—0.12 mm long) with relatively robust tu- berculation, and elongate median waists (0.010- 0.025 mm long). The sclerites are densely dis- tributed in the surface as well as in the interior of the colonies. Sclerites of the surface of the polyparium vary in length from 0.06 to 0.12 mm, while the interior of the polyparium has sclerites that are 0.09-0.12 mm in length. Sclerites of the surface of the stalk are 0.05—0.10 mm long, while those of the stalk interior vary from 0.08 to 0.10 mm in length. Color of the colonies is lemon yellow throughout. WILLIAMS: SOFT CORAL GENUS MINABEA 13 en a a ses dasbaechn iy, 13°30’ 132157 - 144°45' Ficure 8. Minabea goslineri A, A single colony, 7.2 mm in height. B, Longitudinal section through colony in A showing gastric cavities of autozooids and siphonozooids. C, Detail of a single autozooid surrounded by four siphonozooids, length of figure represents 1.6 mm. D, Fifteen colonies showing variability of colony shape, scale bar represents 12 mm. E, Map of Guam, Northern Mariana Islands, showing distribution of the species; black dot represents type locality and only collecting station. F, Surface sclerites of the polyparium. G, Interior sclerites of the polyparium. H, Stalk surface sclerites. I, Stalk interior sclerites. Scale bar for F-I represents 0.1 mm. DIsTRIBUTION.—Minabea goslineri is known discoverer, Dr. Terrence M. Gosliner of the Cal- only from the type locality (Guam, Mariana Is- ifornia Academy of Sciences, a long-time friend lands, in the western Pacific; 1-3 metersin depth). and colleague. Colonies are found on protected, shady sides of REMARKS. — Minabea goslineriis differentiated overhangs or vertical surfaces in shallow water from other members of the genus by its low con- of areas with coral heads or coral reef. ical growth form, in which the diameter of the ETYMOLOGY.—This species is named for its colonies is generally greater than the total height, PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 Ficure 9. Minabea goslineri Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.06 mm. B, 0.07 mm. C, 0.07 mm. D, 0.08 mm. E, 0.07 mm. F, 0.07 mm. G, 0.07 mm. H, 0.06 mm. an inconspicuous stalk, and sclerites that are s1x- Bellonella indica Thomson and Henderson, 1905:274, pl. 6, ° 6 = i yn Sy 9) or eight-radiates (many tending toward double fig. 5. Non Bayer, 1974: 261. Non Faulkner and Chesher, : , fi ; 1979:267, pl. 22. stars) with conspicuously elongate median waists. Type MATERIAL.— Holotype: BMNH Reg. No. 1933. 3. 13. 206 (labeled as Nidalia indica), SRI LANKA. South of Galle: deep water, Sir. J. A. Thomson Collection, one specimen in which part of the polyparium has been cut way longitudinally Minabea indica (Thomson and Henderson, 1905), comb. nov. (Figs. 10, 11) and is missing. WILLIAMS: SOFT CORAL GENUS MINABEA 15 INDIAN OCEAN Ficure 10. Minabea indica (Thomson and Henderson, 1905). A, The holotype, 20 mm in height. B, Longitudinal section through part of the polyparium showing two autozooids and two smaller siphonozooids, length of figure represents 5.5 mm. C, Map of southern India and Sri Lanka showing the type locality (black dot). D, Sclerites from the surface and interior of the polyparium. E, Sclerites from the surface and interior of the stalk. Scale bar for D-E represents 0.1 mm. Diacnosis.— Colony digitiform, cylindrical in shape. Polyparium restricted to the distal-most 40-50% of the colony. Sclerites robust barrels, six- seven- or eight-radiates, and double stars (0.045—-0.11 mm). Color red-orange with yellow retracted polyps. DESCRIPTION. — The preserved colony is 20 mm in length, although Thomson and Henderson (1905) recorded it as 24 mm long. The polypar- ium is 6 mm in width, while the stalk is 9 mm wide at its base. The polyps are restricted to the distal-most two-fifths to one-half of the total col- ony length. The autozooids have bright yellow sclerites in their bases but these do not form permanent calyces. The colony is tightly con- tracted and consequently the siphonozooids are not visible on the surface. These can only be observed in the longitudinal section of the col- ony. The sclerites are densely distributed throughout the surface and interior of the colony. The sclerites of the polyparium are highly vari- able in type, 0.045-0.11 mm in length, and com- posed of robust barrels, six-radiates, seven-ra- diates, and eight-radiates, with some radiates PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 FicurE 11. Minabea indica (Thomson and Henderson, 1905). Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.06 mm. B, 0.08 mm. C, 0.07 mm. D, 0.05 mm. E, 0.08 mm. F, 0.08 mm. G, 0.05 mm. H, 0.08 mm. WILLIAMS: SOFT CORAL GENUS MINABEA approaching double stars with elongated median waists. The stalk contains sclerites that are main- ly robust barrels with some radiates. Colony col- or is dichromatic, crimson-vermillion with yel- low retracted polyps. DISTRIBUTION.—The species is known only from the type locality, southwestern Sri Lanka, central Indian Ocean, in deep water (specific depth and latitude/longitude not given). REMARKS. — Bellonella indica can be allocated to the genus Minabea on the basis of having dimorphic polyps and sclerites that are mostly radiates and barrels (which are not illustrated in the original description). The absence of per- manent calyces and presence of siphonozooids precludes the acceptance of Bellonella as the val- id generic designation. Thomson and Henderson (1905:274) make the following ambiguous state- ment regarding the presence of autozooids and siphonozooids: “Smaller forms occur among the larger, but there is no evidence of dimorphism of zooids.” The severely contracted nature of the specimen makes seeing the siphonozooids ex- tremely difficult. The yellow/red dichromatic coloration and the restriction of the polyps to less than half of the colony length, as well as the species’ occurrence in deep water, serve to distinguish it from both M. robusta and M. aldersladei (the two species that it seems to resemble most closely). Minabea robusta contains tuberculate spheroids, while M. aldersladei and M. indica do not. Minabea indica can be differentiated from M. aldersladei as fol- lows: In M. aldersladei the polyparium com- prises 66-83% of the colony; the sclerites are elongated barrels or eight radiates; the colony color is monochromatic (uniform yellow or red- orange); and the species is restricted to shallow water (2-25 m). In contrast, M. indica has the polyparium restricted to 40-50% of the colony; the sclerites are variable (robust barrels, six-, sev- en-, eight radiates, and double stars); the color is dichromatic (red-orange and yellow); and the species is known only from deep water. Minabea kosiensis sp. nov. (Figs. 1E, 12, 13) Type MATERIAL.— Holotype: SAM-H4261, dredged from a bottom of small rocks, Kosi River mouth (26°55,2’S, 32°56,1’E), SOUTH AFRICA. Zululand: Indian Ocean, 370 meters depth, 7 June 1987, steel dredge, G. C. Williams on board RV Meiring Naude. Paratype: One colony cut longitudinally into two halves, SAM-H4262, same data as holotype. 17 OTHER MATERIAL.— One colony, SAM-H4076, same data as holotype. DiaGnosis.—Colonies digitiform, relatively short, cylindrical in shape. Sclerites eight-radi- ates, thorny spindles and intermediate forms (0.06—0.24 mm long). Color white with salmon- pink retracted autozooids and siphonozooids. DESCRIPTION.—The three colonies examined are 20, 26, and 30 mm in length, all with an average width of 9 mm. The colonies are digi- tiform, relatively short, cylindrical in shape as the distal region does not taper to any appreciable degree and the distal terminus of a particular colony is broadly rounded. The distal-most three- quarters to one-half of the colonies are occupied by the polyparium, while the proximal-most one- quarter to one-half comprises the stalk. Proxi- mally the stalk produces a somewhat spreading holdfast. The polyps are evenly distributed over the surface of the polyparium. The retracted au- tozooids are approximately 1.2 mm in diameter and are surrounded by minute siphonozooids that measure approximately 0.12-0.16 mm in di- ameter. Several autozooids that were preserved extended in two of the colonies measure ap- proximately 2.5 mm in length by 1.5 mm in width. Sclerites are eight-radiates and thorny spindles as well as forms intermediate between these two (0.06-0.24 mm in length). Some of the spindles may be slightly club-shaped and have conspicuous, sharply pointed tubercles. The sclerites from the surface coenenchyme of the polyparium are 0.06—-0.23 mm in length. Scler- ites from the surface of the stalk are 0.06-0.20 mm long. Sclerites from the interior of the poly- parium measure 0.10—0.24 mm in length. In the interior of the stalk, the sclerites are 0.12-0.15 mm in length. The stalk sclerites are predomi- nantly thorny radiates or intermediate forms with very few elongate spindles present, while the polyparium contains both spindles and radiates in approximately equal numbers. Color of the colony is white with light salmon-pink, retracted autozooids and siphonozooids. The extended au- tozooids are uniform white. The spindles and sclerites of intermediate form are colorless, while the smaller radiates are reddish in color. The interior of the colonies is white due to the col- orless nature of the sclerites, while the pinkish coloration is restricted to the surface coenen- chyme of the polyparium, which contains red- dish radiates. 18 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 FiGure 12. Minabea kosiensis A, A single colony, 20 mm in height. B, Longitudinal section through colony in A showing gastric cavities of autozooids and siphonozooids. C, Detail of a single autozooid surrounded by 28 siphonozooids, length of figure represents 2.4 mm. D, Map of southeastern Africa showing distribution of the species; black dot represents type locality and only collecting station. E, Surface sclerites of the polyparium. F, Interior sclerites of the polyparium. G, Stalk surface sclerites. H, Stalk interior sclerites. Scale bar for E-H represents 0.1 mm. DISTRIBUTION.— This species is known only EtyMoLoGy.—The specific epithet is derived from the type locality at Kosi Bay (border of from the name of the type locality, Kosi Bay (near South Africa and Mozambique), 370 meters in the border of Mozambique and Zululand), and depth. the Latin suffix -ensis, meaning belonging to. WILLIAMS: SOFT CORAL GENUS MINABEA FiGure 13. REMARKS.—Minabea kosiensis is distin- guished from other Minabea species by the oc- currence of short and slender-digitiform/cylin- drical growth form, and sclerites of eight-radiates, elongate spindles, and intermediate forms in the polyparium with a predominance of thorny ra- diates and intermediate forms in the stalk. Red- dish coloration is restricted to radiates in the surface coenenchyme of the polyparium. This species is here recorded as occurring at the great- est depth of any Minabea species. Minabea ozakii Utinomi, 1957 Minabea ozakii Utinomi, 1957:139. Utinomi and Imahara, 1976:205. Alderslade, 1985:113. Minabea kosiensis Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.12 mm. B, 0.06 mm. C, 0.05 mm. D, 0.15 mm. E, 0.23 mm. F, 0.10 mm. REMARKS. — This is the type species for the ge- nus Minabea. Type material was not available to me for study. An assessment of the species is presented below (see Discuss1on— Historical Account). DISTRIBUTION. — Honshu, Japan. Minabea phalloides (Benham, 1928) Anthomastus phalloides Benham, 1928:79. Minabea phalloides Utinomi and Imahara, 1976:211. REMARKS.— Type material for this species was not available to me for examination. An assess- ment is presented below (see Discuss1on— His- torical Account). 20 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 TABLE 1. COMPARATIVE CHARACTERS FOR THE GENUS Min4BeA UTINOMI, 1957. Colony shape Distribution Species and length Sclerites Color and depth Minabea acrono- cephala Minabea alder- sladei Minabea cosma- rioides Minabea gosli- neri Minabea indica Minabea kosien- SiS Minabea ozakii dome-shaped: low hemispherical (4-14 mm) digitiform: elongate ta- pering (19-69 mm) digitiform: robust cy- lindrical (34-41 mm) dome-shaped: low con- ical (3-10 mm) digitiform: elongate cy- lindrical (20-24 mm) digitiform: short cylin- drical (20-30 mm) digitiform: elongate ta- pering (<45 mm) barrels and eight-radi- ates (0.05—0.13 mm) Elongate barrels and eight-radiates (0.04— 0.12 mm) eight-radiates and ro- tund barrels (0.05— 0.19 mm) double stars, six-radi- ates and eight-radi- ates (0.05—0.12 mm) robust barrels, six-, seven-, eight-radi- ates, double stars (0.045-0.11 mm) eight-radiates, thorny spindles and inter- mediates (0.06-0.24 mm) eight-radiates, thorny spindles, slender yellow or orange, often with red retracted polyps yellow or red-orange dull brownish-orange yellow crimson-vermillion with yellow retracted polyps white with salmon- pink retracted polyps dull orange or cinna- mon New Guinea and Fiji (5-21 m) New Guinea, NW Australia, Great Barrier Reef (2-25 m) Transkei (South Afri- ca) (250 m) Guam (1-3 m) Sri Lanka (deep water) Natal (South Africa) (370 m) Japan (250-270 m) rods (0.037-0.26 mm) Minabea phal- digitiform: elongate ta- radiates and thorny pale-yellowish New Zealand (depth loides pering (47 mm) spindles (0.05-0.25 unknown) mm) Minabea robusta __ digitiform: elongate ta- pering (57-170 mm) mm) DISTRIBUTION. — New Zealand. Minabea robusta Utinomi and Imahara, 1976 (Figs. 14, 15) Minabea robusta Utinomi and Imahara, 1976:206; Figs. 1-3; pl. 1. Alderslade, 1985:113, Fig. 5b. MATERIAL EXAMINED.— Paratype: MSM-INV-75-048, one partial colony in three pieces, colony originally attached to a boulder, JAPAN. Suruga Bay: Uchiura coast, Wakamatsu-zaki (approximately 34°45’N, 138°30’E), 11 June 1975, 30 meters in depth. DESCRIPTION. — The three portions of the para- type examined are from the polyparium: the dis- tal tip of the colony (25 mm in length by 10-12 mm in width), a middle portion (45 mm long by 15-20 mm in width), and a basal portion (25 mm in length by 30-33 mm in width). None of the material is representative of the stalk. The distal and middle portions are longitudinally sec- radiates, robust barrels and tuberculate spheroids (0.04—0.11 light yellow-orange Japan (25-30 m) with yellow or or- ange retracted polyps tioned with only one half from each section pres- ent. The gastric cavities of the autozooids form parallel straight-sided tubes. The siphonozooids are numerous and prominent, pustulate, ap- proximately 0.5 mm in diameter, and surround the retracted autozooids, which are 1.5—2.0 mm in diameter. The sclerites are variable and in- clude radiates (predominantly eight-radiates), tuberculate spheroids, and robust barrels (0.04— 0.11 mm in length). The tuberculation of sclerites from the interior is somewhat more coarsely thorny than that of the surface sclerites. DISTRIBUTION. —Sagami Bay and Suruga Bay, Honshu, Japan; 25-30 meters in depth. REMARKS.—The other Japanese species, Mi- nabea ozakii, differs from M. robusta by having slender rods in the interior. Minabea robusta is differentiated from other digitiform members of the genus by the occurrence of tuberculate sphe- roids and robust barrels in addition to radiates. WILLIAMS: SOFT CORAL GENUS MINABEA 21 Ficure 14. Minabea robusta. A, Reconstruction of a colony from fragmented paratype (MSM-INV-75-048) and adapted from Utinomi and Imahara, 1976:pl. 1 (Fig. 3), 170 mm in height. B, Longitudinal section of distal terminus of paratype showing gastric cavities of autozooids and siphonozooids; portion shown is 25 mm in height. C, Detail of a single autozooid surrounded by five siphonozooids, length of figure represents 3.0 mm. D, Map of Japan showing collecting stations for the species (arrow indicates type locality). E, Polyparium surface sclerites. F, Polyparium interior sclerites. Scale bar for E, F represents 0.1 mm. DISCUSSION HistoricAL AccouNT.—The genus Minabea was established by Utinomi (1957) for M. ozakii from near Minabe, eastern entrance to the Inland Sea, Honshu, Japan at 250-270 meters in depth. According to Utinomi, the species is character- ized by having an elongate and tapering digiti- form shape, in addition to radiates and spindles in the surface of the polyparium and stalk, with the interior of the polyparium containing slender rods. A recent request for loan of type material of M. ozakii was not acknowledged, and thus a detailed comparison with recently collected ma- terial representing the new taxa was not possible. A second species, M. robusta, was subsequent- ly described in 1976 by Utinomi and Imahara for material collected in Sagami and Suruga bays, Honshu, Japan at 25-30 meters depth. This spe- cies, also elongate and tapering digitiform, was differentiated by the possession of short-waisted capstans (six- or eight-radiates), and robust bar- rels and tuberculate spheroids in the surface as PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 Ficure 15. Minabea robusta Scanning electron micrographs of sclerites. Measurements refer to sclerite length. A, 0.07 mm. B, 0.06 mm. C, 0.08 mm. D, 0.08 mm. E, 0.08 mm. F, 0.06 mm. G, 0.06 mm. H, 0.06 mm. WILLIAMS: SOFT CORAL GENUS MINABEA well as in the interior of the colony. An emended generic diagnosis was included to accommodate both species. A paratype of M. robusta (MSM- INV-75-048) was recently acquired on loan and a comparison was subsequently made with re- cently collected material. Anthomastus phalloides Benham, 1928 was described presumably from the region of the Foveaux Strait, south of the South Island of New Zealand, at an unrecorded depth. Like the two Japanese species, it is elongate digitiform, and the colony gradually tapers distally. The polyp- iferous region contains radiates and thorny spin- dles in contrast to the surface of the stalk which contains only radiates, while the interior also contains spindles in addition to radiates. Tixier- Durivault (1970) described Anthomastus agilis Tixier-Durivault, 1970 from New Caledonia (depth unrecorded). The colonial growth form is clavate and the sclerites are spindles and wart clubs. Neither the New Zealand nor the New Caledonia species can be allocated to the genus Anthomastus Verrill, 1878, which is character- ized by a distinctly capitate growth form, a re- | duced number of greatly enlarged autozooids, and the occurrence of large thorny spindles or rods in the interior of the colony. Utinomi and Imahara (1976) transferred these two species to the genus Minabea and considered them as prob- able synonyms. Based solely on the description provided by Benham, I agree that Anthomastus phalloides should be allocated to Minabea. In contrast, Tixier-Durivault’s species presents an- other set of circumstances. According to P. N. Alderslade (pers. comm.), the holotype of An- thomastus agilis has interior spindles up to 2.2 mm in length, which clearly indicates that it is not a species of Minabea. The taxon is in the process of being redescribed by Alderslade, along with two other species, as members of a new genus of Alcyoniidae. Bellonella indica Thomson and Henderson, 1905, a single specimen described from “deep water” off of southwestern Sri Lanka in the In- dian Ocean, can be allocated to the genus Mina- bea on the basis of digitiform growth habit, pres- ence of dimorphic polyps, sclerites as capstans and related forms, and absence of permanent calyces. Taking into account the five new species de- scribed in the present work, a total of at least nine worldwide species can therefore be recog- 23 nized for the genus Minabea. These taxa are dif- ferentiated in Table 1. SIMILAR GENERA.— Two other alcyoniid gen- era from the Indian Ocean are superficially sim- ilar to Minabea. Acrophytum Hickson, 1900 and Verseveldtia Williams, 1990 both have dimor- phic polyps, relatively small and numerous au- tozooids, and are unbranched with a growth form and complement of sclerites that may resemble Minabea. Acrophytum claviger Hickson, 1900, the only species known in the genus, is distributed along the South African coast between Cape St. Francis (Cape Province) and Port Durnford (Natal). Ac- rophytum is distinguished by its elongate digi- tiform shape, colorless sclerites that are predom- inantly clubs and wart clubs, and a marked scarcity of sclerites in the interior of the colony. The colonies are relatively rigid and firm due to the presence of very dense mesoglea. Verseveldtia, on the other hand, is distinctly capitate with the polyps restricted to the distal portion of the enlarged capitulum; the auto- zooids contain small platelets; many if not all the colonial sclerites are colored, being predomi- nantly radiates, double stars, barrels and tuber- culate spheroids; and they are densely distributed in the interior of the colony. The two described species of Verseveldtia, also from South Africa, are distributed between East London (Cape Province) and Durban (Natal). An undescribed species of Verseveldtia has recently been recog- nized from off western Australia (P. N. Alder- slade, pers. comm.). Anthomastus granulosus Kiikenthal, 1910, originally described from Ja- pan and later described from subsequently col- lected material by Utinomi (1960), has sclerites that closely resemble the barrels and capstans of various species of Minabea and Verseveldtia, and in addition lacks the sparsely spinose rods in the interior of the colonies that is characteristic of the genus Anthomastus. Regarding the markedly capitate nature of Kiikenthal’s and Utinomi’s specimens, it is possible that the species belongs to the genus Verseveldtia (Williams 1990). Thomson (1921) applied the name Anthomastus granulosus to a specimen from South Africa that is probably identifiable as Verseveldtia buccini- forme Williams, 1990. Minabea differs from Acrophytum and Ver- seveldtia by the combination of digitiform growth form, sclerites that are barrels, radiates, or spin- 24 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 1 3500 km INDIAN OCEAN Ficure 16. Map of the Indo-West Pacific showing distribution of the genus Minabea. Inset of southern Africa shows the overlapping distributions of the genera Acrophytum (1), Minabea (2), and Verseveldtia (3). Genus Acrophytum Minabea Verseveldtia TABLE 2. Colony shape in mature colonies digitiform: finger- like digitiform: dome- shaped to fin- gerlike Capitate with nar- row and elon- gate stalk Meso-_ Siphono- glea zooids dense _ few, sparse thin many, dense thin many, dense Polypary Interior and stalk sclerites sclerites very colorless tubercu- sparse lated clubs dense colored barrels, radiates, double heads, spindles and rods dense colored barrels, double stars, ra- diates, and tu- berculate sphe- roids COMPARATIVE CHARACTERS FOR THREE SIMILAR GENERA OF DIMORPHIC SOFT CORALS. Polyp sclerites absent absent numerous small platelets Distribution and depth South Africa (30-146 m) Indo-West Pa- cific (1-370 m) South Africa and Western Australia (50-97 m) WILLIAMS: SOFT CORAL GENUS MINABEA dles, which are for the most part colored and distributed throughout the colony. The geographic ranges of all three genera over- lap along the east coast of South Africa between Mbotyi (Transkei) and Durban (Natal) (Fig. 16, inset). The genera are differentiated in Table 2. ACKNOWLEDGMENTS This paper represents Christensen Research Institute publication number 21. I wish to thank Director Matthew Jebb and the staff of the Chris- tensen Research Institute, Madang, Papua New Guinea, for their enthusiastic support of field research undertaken during November of 1990. CRI provided a research fellowship for funding of field work during this period. Matthew Jebb translated the abstract into Pidgin English. I am grateful to Terrence M. Gosliner, director of re- search at the California Academy of Sciences, San Francisco, for the collection of some mate- rial; Phil Alderslade of the Northern Territory Museum, Darwin, for kindly providing loans of Australian material; Katsumi Suzuki of the Ma- rine Science Museum, Tokai University, Shi- zuoka, Japan, for loan of the paratype of M. ro- busta; Charles Hussey, Invertebrate Curation Group, The Natural History Museum (London) for loan of the holotype of M. indica; and Tim- othy E. Coffer of the National Museum of Nat- ural History, Smithsonian Institution, Washing- ton, D.C., for loan of material from Palau. I am indebted to Richard Kilburn and David Herbert of the Natal Museum, Pietermaritzburg, for their help during dredging cruises of the RV Meiring Naude in 1986 and 1987; Michelle van der Merwe and Liz Hoenson of the South African Museum, Cape Town, for curatorial assistance; and Lisa Borok and Terrence Gosliner (California Acad- emy of Sciences) for assistance in the preparation of scanning electron micrographs. I express my gratitude to Frederick M. Bayer, of the National Museum of Natural History, Smithsonian Insti- tution, Washington, D.C., for sharing his knowl- edge of the genus. SUMMARY (PIDGIN) Sotpela toksave long dispela stori: wanpela nu- pela toksave bilong stretim stori bilong wanpela kain malumalu-koral lain (Minabea Utinomi, 1957, family Alcyoniidae) i kamap, bikos mipela 25 bin kisim faivpela nupela kain long solwara na- baut long Pacific na Papua Niugini. Nau tasol i gat nainpela kain samting long dispela lain. Dis- pela lain i kamap long ol rip i stap insait long Afrika, na i go long Fiji, na Japan, na New Zea- land, sampela kain i stap long rip i no daun, na sampela 1 stap long ples i daunbilo long 370 me- ters. Sampela toktok long skelim dispela lain wantaim ol kandere lain long narapela hap i ka- map. LITERATURE CITED ALDERSLADE, P. N. 1985. Redescription of Acrophytum cla- viger (Coelenterata: Octocorallia). The Beagle 2:105-113. Bayer, F. M. 1974. A new species of Trichogorgia and re- cords of two other octocorals new to the Palau Islands. Mi- cronesica 10:257-271. . 1981. Key to the genera of Octocorallia exclusive of Pennatulacea (Coelenterata: Anthozoa), with diagnoses of new taxa. Proc. Biol. Soc. Wash. 94:902-947. BENHAM, W.B. 1928. On some alcyonarians from New Zea- land waters. Trans. Proc. New Zealand Inst. 59:67-84. FAULKNER, D. AND R. CHESHER. 1979. Living corals. Clark- son N. Potter, Inc., New York. 311 pp. Hickson, S. J. 1900. The Alcyonaria and Hydrocorallinae of the Cape of Good Hope. Marine Invest. S. Africa 1:67- 96. KUKENTHAL, W. 1910. Zur Kenntnis der Gattung Antho- mastus Verr. In Beitrage zur Naturgeschichte Ostasiens. Suppl.-Bd. 1(9)1-16. F. Doflein, ed. Abhandl. math.-phys. Klasse K. Bayer. Akad. Wiss. THomson, J. A. AND W. D. HENDERSON. 1905. Report on the Alcyonaria collected by Professor Herdman, at Ceylon, in 1902. Jn Report to the Government of Ceylon on the Pearl Oyster Fisheries of the Gulf of Manaar Part 3, 20:269- 328. Tuomson, J.S. 1921. South African Alcyonacea. Trans. Roy. Soc. South Africa 9:149-175. Trxier-DurIvauLT, A. 1970. Les Octocoralliaires de Nou- velle-Calédonie. Expedition Francaise sur les Récifs coral- liens de la Nouvelle-Calédonie 4:171-—350. 1987. Sous-Classe des Octocoralliaires. In Traité de Zoologie— Anatomie, Systématique, Biologie 3(3):3-185. P.- P. Grasse, ed. Masson, Paris. Utinomi, H. 1957. Minabea ozakii n. gen. et n. sp., a new remarkable alcyonarian type with dimorphic polyps. J. Fac. Sci. Hokkaido Univ. (Zool.) 13(1-4):139-146. . 1960. Noteworthy octocorals collected off the south- west coast of Kii Peninsula, middle Japan. Publ. Seto Marine Biol. Lab. 8(1):1-26. Urtinomi, H. AND Y. IMAHARA. 1976. A new second species of dimorphic alcyonacean octocoral Minabea from the bays of Sagami and Suruga, with the emendation of generic di- agnosis. Publ. Seto Marine Biol. Lab. 23(3/5):205-212. Wiiuiams, G. C. 1990. A new genus of dimorphic soft coral from the south-western fringe of the Indo-Pacific (Octocoral- lia: Aleyoniidae). J. Zool. 221:21-35. © CALIFORNIA ACADEMY OF SCIENCES, 1992 Golden Gate Park San Francisco, California 94118 eal Laboratory i PROCEEDINGS ~~ LIBRARY OF THE : | f | 6 17199 CALIFORNIA ACADEMY OF SCIENCES ~™ | Woods Hole, Vol. 48, No. 2, pp. 27-95, 37 figs., 7 tables. ‘e GRENADIERS (PISCES, GADIFORMES) OF THE NAZCA AND SALA Y GOMEZ RIDGES, SOUTHEASTERN PACIFIC By Yuri I. Sazonov Department of Ichthyology, Zoological Museum, Moscow Lomonosoy State University, Herzen Street 6, Moscow 103009, Russia and Tomio Iwamoto Department of Ichthyology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118, U.S.A. Asstract: Twenty-five species of grenadiers are recorded from the Nazca and Sala y Gomez ridges in the southeastern Pacific. New species include: Caelorinchus immaculatus, C. multifasciatus, C. nazcaensis, C. spilonotus, Hymenocephalus neglectissimus, H. semipellucidus, Kuronezumia pallida, Ventrifossa macrodon, V. teres, and V. obtusirostris. Caelorinchus immaculatus is very similar to C. karrerae from the southeastern Atlantic and Indian Ocean. The H. striatissimus complex is examined using new data. Hymenocephalus semipellucidus and H. neglectissimus appear to be part of this complex. The subspecies H. s. hachijoensis from Japan is elevated to full species status. Kuronezumia, formerly considered a subgenus of Nezumia, is redefined and elevated to generic status to include K. pallida, K. bubonis, K. leonis, K. macronema, K. dara, and two undescribed species. Despite proximity of the ridges to the mainland coast of Peru, relationships of the associated fauna are to the west, particularly the western Pacific and Hawaiian Islands. Of the 25 species from these ridges, eight are definitely known from the vicinity of the Hawaiian Islands: Caelorinchus spilonotus, Cetonurus crassiceps,* Coryphaenoides paradoxus,* H. striatulus, Malacocephalus laevis,* Mataeocephalus acipenserinus, Nezumia propinqua,* and Pseudocetonurus septifer. Three other species whose identifications are undetermined may be part of, or have close counterparts in, the Hawaiian fauna: Gadomus sp. cf. melan- opterus, Hymenocephalus sp. cf. aterrimus, and Trachonurus villosus*? The four species marked with an asterisk are broadly distributed through the Pacific, Indian, and Atlantic oceans. Malacocephalus laevis is known from the continental slopes of southern California and on seamounts off Baja California, but nowhere else along the Pacific coast of Central and South America. Caelorinchus immaculatus is also recorded from mainland South America; Nezumia conyergens is questionably represented by a specimen from the Sala y Gomez Ridge. Received May 1, 1991. Accepted December 1, 1991. CONTENTS DESCRIPTION OF NAZCA AND SALA Y GOMEZ | RATT eT cM AT sre ta pe pepe Direed ESET Ne, 32 PERS IMOACT ie)! Slt ent hese yh bh el 27. Key To NSG SpEciEs OF GRENADIERS.............. 52 [GRO WCTION.<< 6 Writ .ci mer ty eee ol Sr D8 BATHYGADIDABS 3 .cxjeert oa! } be cepee dt t. 36 MATERIALS AND METHODS ow ecceeeeeesee eee 28 1. Gadomus sp. cf. melanopterus Gil- CLASSIFICATION OF GRENADIERS .... 30 ent LOO Fe ees nrrcenl tori ehbS betes ot 36 [27] 28 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 IMACROURIDAR.A 229) nt 38 Subfamily Macrouroidinae 38 2. Macrouroides inflaticeps (Smith and Radcliffe, 1912) et weewter, <0 5» 38 3. Squalogadus modificatus Gilbert and EMG OS pO iG haan eee eb 39 Subfamily Macrourinae 2 ee 39 4. Cetonurus crassiceps (Ginther, jal) eee ete eRe |e 40 5. Caelorinchus immaculatus n. sp... 41 6. Caelorinchus multifasciatus n. sp... 44 7. Caelorinchus nazcaensis n. Ssp............. 46 8. Caelorinchus spilonotus n. sp... 49 9. Coryphaenoides paradoxus (Smith and Radelitte: VOD) ee aces 52 10. Hymenocephalus sp. cf. aterrimus Eloi Se en 54 11. Hymenocephalus gracilis Gilbert and BESS 79119 (0 arene 35) 12. Hymenocephalus neglectissimus n. <0 ee aad ar a 56 13. Hymenocephalus semipellucidus n. SPeo ee ee ee” 60 14. Hymenocephalus striatulus Gilbert, OOS ee ke 63 15. Kuronezumia pallida n. sp. 65 16. Malacocephalus laevis (Lowe, 1843) 68 17. Mataeocephalus acipenserinus (Gil- bertand! Cramer Usman 70 18. Nezumia convergens (Garman, BSS 9) eh A AOR, 9 Be eles 71 19. Nezumia propinqua (Gilbert and Cramer: gS Oi),) see tee, 2 20. Pseudocetonurus septifer Sazonov and Shceherbache ver 1982 aes beats es TS 21. Trachonurus villosus (Giinther, | ke A) nee = A iy ar 77 22. Ventrifossa johnboborum Iwamoto, LOS 2 da.20 sqemone Ss » mere: dee eet eS. coal 78 23. Ventrifossa macrodon N. sp. ....- 81 24. Ventrifossa obtusirostris n. sp. 84 25. WCnUOSSA LEneS 1). SP joe 86 BIOGEOGRAPHICAL CONSIDERATIONS 88 ACKNOWLEDGMENTS ose eceessssceseceeeneeeeeeeeeeeeneeeneee 91 IRESUMEN) 2 <8) re ne ee 91 eT TERATURE CRED = 2.62.2 ee 92 INTRODUCTION The research vessel Ichthyandr of the Sevas- topol Bureau of Underwater Researches of the former USSR Ministry of Fisheries surveyed the Nazca and Sala y Gomez oceanic ridges (Fig. 1) off the southwestern coast of Peru on its fifth cruise between April and November 1979. Pre- liminary studies (Parin et al. 1981) of the ma- terial collected during that cruise suggested the existence of an extremely diverse ichthyofauna on these ridges. The species composition of the macrourid fishes, or grenadiers, differed mark- edly from that known along the nearby conti- nental margins and, in fact, appeared to be more similar to those from the Indo-Pacific region, not the eastern Pacific. Subsequently, between 1979 and 1987, several other Soviet expeditions investigated the ridges and collected many grenadiers from depths of about 300-800 m, with a few trawls below 1,000 m. In preparing the section on the grenadiers for the preliminary report (Parin et al. 1981), Sa- zonov recognized 11 species, eight of which he considered new or of uncertain taxonomic status. Additionally, Sazonov and Shcherbachev (1982) described a new genus and species (Pseudoceto- nurus septifer) from the Sala y Gomez Ridge. Because of the obvious Indo-West Pacific af- finities of the grenadiers, it was necessary to ex- amine many type specimens deposited in U.S. museums, particularly the U.S. National Mu- seum of Natural History (USNM) and the Cal- ifornia Academy of Sciences (CAS). The idea of studying the fauna on a cooperative basis was initially proposed by N. V. Parin during a Work- shop on Cold-Water Fishes held in Orono, Maine, in 1979. The study was begun in 1981 and a preliminary manuscript was drafted in 1988 after Iwamoto spent 3.5 months in the Laboratory of Oceanic Ichthyofauna, Russian Academy of Sci- ences, Moscow. This paper presents the results of the cooperative effort. Twenty-five species of grenadiers are herewith recorded from the Nazca and Sala y Gomez ridg- es, ten of which are described as new. Two others may represent undescribed taxa, but the avail- able material does not justify their formal rec- ognition. Considering that for the most part only the shallowest levels of these oceanic ridges have been sampled, the faunal list can be expected to grow, especially in bathybenthic and abyssal forms. MATERIALS AND METHODS Almost all specimens reported here were col- lected by Soviet oceanographic and fishery re- search vessels. Most collections were made be- tween 1979 and 1987, although a few specimens found at the Zoological Institute in St. Petersburg SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 29 an Ambrosia I. mc i) a ° oa, o a o 0 o _fam in Sala y Gomez | S) /' Easter I. rn foo = —— _ Figure 1. Map of southeastern Pacific showing location of the Nazca and Sala y Gomez ridges. Inset shows general location of larger map. Depth contours in fathoms (solid lines represent 2,000 fathoms or about 3,658 m; broken lines represent 1,000 fathoms or about 1,829 m). Based in part on a chart by Mammerickx et al. 1975. 30 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 (ZIN) and the Zoological Museum, Moscow State University (ZMMGU) were procured in 1973 and 1975 during cruises of the Hercules and As- tronomer of the Pacific Fisheries and Oceanog- raphy Research Institute, Vladivostok (TINRO). Most of the specimens used for this report are deposited in the Institute of Oceanology, Russian Academy of Sciences, Moscow (IOAN) and ZMMGU, with representatives also deposited in CAS and ZIN. A list of Russian vessels and sta- tion data are provided in Table 1. Institutional abbreviations follow Leviton et al. (1985) as amended by Leviton and Gibbs (1988). Abbreviations and methods of taking measurements and counts are the same as those described in a previous paper (Iwamoto and Sa- zonov 1988). The abbreviation NSG is used throughout to refer to the combined Nazca and Sala y Gomez ridges. The following abbrevia- tions are also used: cr.—cruise; fm(s)— fathom(s); HL—head length; spec.—specimen; sta.—sta- tion; TL—total length; tr.—trawl. The descriptions of species are based mainly on the materials examined from NSG and usu- ally do not include specimens from outside the area, even though such additional materials have been examined when available. These peripheral specimens are usually considered in the RE- MARKS AND COMPARISONS sections. Several species have been adequately de- scribed and illustrated in recent literature. For these, a brief diagnosis or a short diagnostic de- scription is provided. For other species, more extensive descriptions are given, but synonymies are kept to a minimum; those included are lim- ited to recent references and a few older ones with good descriptions or illustrations. As this is not meant to be a comprehensive revision of the groups treated, generic accounts are for the most part very brief, and synonyms are omitted. CLASSIFICATION OF GRENADIERS There has been a recent surge of interest in the higher classification of gadiform fishes, which culminated in a workshop on the systematics of the group in 1986 (see Cohen 1989). The work- shop brought together disparate workers in the group to air their views on its classification and phylogeny and search for a modicum of agree- ment. For the so-called macrouroid fishes, or grenadiers, there was none. Howes (1988, 1989), in a radical departure, removed the bathygadine TABLE 1. Soviet research vessels and trawl stations on the Nazca and Sala y Gomez ridges at which grenadiers were col- lected. Hercules (1973) tr. 40: Nazca Ridge, 19°30.4’S, 80°11.2'W; 950 m; bot. trawl; 20.X.1973. Hercules (1975) sta. 68: Sala y Gomez Ridge, 25°18.8’S, 93°34.2'W; 610 m; bot. trawl; 1.XI.1975. sta. 70: Sala y Gomez Ridge, 25°01'S, 91°18’W; 610-620 m; bot. trawl; 2.X1.1975. sta. 74: Sala y Gomez Ridge, 24°57.5'S, 88°27.1'W; 550-630 m; bot. trawl; 5.XI.1975. sta. 110: between Nazca and Sala y Gomez ridges, 25°35.0’S, 85°13’W; 1,600-2,000 m; midwater trawl; 3.XII.1975. Astronomer (1975) tr. without no.: Nazca Ridge, 23°30.5’S, 81°45’W; 300-330 m; bot. trawl; 29.VI.1975. sta. 104: Sala y Gomez Ridge, 25°00’S, 88°30’ W; 550 m; bot. trawl; 24.VII.1975. tr. without no.: Sala y Gomez Ridge, 25°02'S, 88°35’'W; 550 m; bot. trawl; 24. VII.1975. Ichthyandr cr. 5 (1979) tr. 1: Nazca Ridge, 19°40’S, 80°18'W; 980 m; 26.VIII.1979. tr. 13: Nazca Ridge, 21°30’S, 81°42’'W; 330 m; 5.1X.1979. tr. 14: Nazca Ridge, 21°24’S, 81°41'W; 330-340 m; 6.1X.1979. tr. 15: Nazca Ridge, 21°27’S, 81°41’W; 330 m; 7.IX.1979. tr. 17: Nazca Ridge, 21°41'S, 81°41'W; 310-330 m; 9.I1X.1979. tr. 40: Sala y Gomez Ridge, 25°52’S, 86°47’'W; 440 m; 24.1X.1979. tr. 44: Nazca Ridge, 21°32’S, 81°38’W; 330 m; 30.1X.1979. tr. 50: Sala y Gomez Ridge, 25°07’S, 99°37'W; 450-480 m; 24.X.1979. tr. 51: Sala y Gomez Ridge, 25°07'S, 99°39'W; 360-400 m; 24.X.1979. tr. 52: Sala y Gomez Ridge, 25°02'S, 99°26’W; 770 m; 24.X.1979. tr. 53: Sala y Gomez Ridge, 25°02’S, 88°32’W; 540 m; 29.X.1979. tr. 54: Sala y Gomez Ridge, 25°01'S, 88°27'W; 535-575 m; 30.X.1979. tr. 55: Sala y Gomez Ridge, 25°00’S, 88°31’W; 345-540 m; 30.X.1979. tr. 56: Sala y Gomez Ridge, 25°46'S, 86°33'W; 410 m; 31.X.1979. tr. 57: Sala y Gomez Ridge, 25°42’S, 86°32'W; 420 m; 31.X.1979. Ichthyandr cr. 6 (1980) tr. 1: Nazca Ridge, 21°27'S, 81°40’W; 340 m; 5.VIII.1980. tr. 2: Nazca Ridge, 21°08'S, 81°39’W; 320 m; 5.VIII.1980. tr. 3: Nazca Ridge, 21°28’S, 81°39’W; 340 m; 6.VIII.1980. tr. 10: Nazca Ridge, 21°27'S, 81°41'W; 320-340 m; 9. VIII.1980. tr. 33: Nazca Ridge, 21°28'S, 81°41’W; 330 m; 21.1X.1980. tr. 56: Sala y Gomez Ridge, 25°45'S, 86°35’W; 430 m; 2.X.1980. Odissey cr. 2 (1981-82) tr. 11: Nazca Ridge, 22°11’S, 81°21’W; 235 m; bot. trawl; 23.XII.1981. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 31 TasLe 1. Continued. tr. 14: Nazca Ridge, 22°11.6’S, 81°18.8’W; 230 m; bot. trawl; 25.XII.1981. tr. 15: Nazca Ridge, 22°12'S, 81°19’W; 230-250 m; bot. trawl; 25.XII.1981. Akademik Kurchatov cr. 34 (1983) sta. 3594: Nazca Ridge, 21°28.8’S, 81°42.0’'W; 340 m; Sigs- bee trawl; 25.XI.1982. Professor Mesiatzev cr. 13 (1983) tr. 1: Sala y Gomez Ridge, 24°58.6'S, 88°28’ W; 565-555 m; 1.1X.1983. tr. 2: Sala y Gomez Ridge, 24°59.8’S, 88°25'W; 550-560 m; 1.1X.1983. tr. 4: Sala y Gomez Ridge, 25°45’'S, 86°37'W; 410-420 m; 2.1X.1983. tr. 7: Sala y Gomez Ridge, 25°41'S, 85°31'W; 285-300 m; 3.1X.1983. tr. 10: Sala y Gomez Ridge, 25°34.4’S, 85°20.7'W; 1,070- 1,100 m; 4.1X.1983. tr. 14: Sala y Gomez Ridge, 25°52.7'S, 84°34.2'W; 1,050 m; 5.1X.1983. tr. 17: between Sala y Gomez and Nazca ridges, 23°38.8’S, 85°27.5'W; 1,220-1,230 m; 7.1X.1983. tr. 23: Nazca Ridge, 22°08’S, 81°19'W; 235-250 m; 12.1X.1983. tr. 25: Nazca Ridge, 22°04’S, 81°17'W; 235-225 m; 12.1X.1983. tr. 31: Nazca Ridge, 22°06’S, 81°17'W; 14.1X.1983. tr. 35: Nazca Ridge, 22°07’'S, 81°18'W; 14.1X.1983. its 362 Nazca Ridge,.22°07'S, 81°18’ W: 235—230 m: 14.1X.1983. tr. 39: Nazca Ridge, 22°06’S, 81°16’W; 230-240 m; 16.1X.1983. tr. 43: Nazca Ridge, 20°46’S, 80°52’W; 18.1X.1983. tr. 44: Nazca Ridge, 20°44.9’S, 80°52.3’'W; 340-780 m; 18.1X.1983. tr. 94: Nazca Ridge, 19°35.5’S, 80°15.7’'W; 940-960 m; 10.X1.1983. tr. 109: Nazca Ridge, 22°06’S, 81°17’'W; 225-240 m; 15.X1.1983. tr. 120: Nazca Ridge, 22°05'S, 81°17’W; 230 m; 18.X1.1983. Professor Mesiatzev cr. 15 (1984) tr. 49: Sala y Gomez Ridge, 25°01.7'S, 88°26.9'W; 525-530 m; 10.X.1984. tr. 50: Sala y Gomez Ridge, 25°57’S, 88°31'W; 565 m; 11.X.1984. tr. 52: Sala y Gomez Ridge, 25°45’S, 88°36'W:; 400 m; 11.X.1984. tr. 53: Sala y Gomez Ridge, 25°44’S, 86°36’W; 390-385 m; 11.X.1984. tr. 54: Sala y Gomez Ridge, 25°48’S, 86°15'W; 304 m; 12.X.1984. Professor Shtokman cr. 18 (1987) sta. 1826: Nazca Ridge, 20°44.8’S, 80°53.7'W; 300-0 m over bot. depth 330-340, IKMWT; 17.IV.1987. sta. 1828/35: Nazca Ridge, 20°48.8’S, 80°53.5’'W; 730-720 m; traps; 18.I1V.1987. 225-240 m; 340-325 m; 320-325 m; TABLE 1. Continued. sta. 1845: Nazca Ridge, 21°24.0’'S, 81°38.6’'W; 330 m; bot. trawl; 19.1V.1987. sta. 1851: Nazca Ridge, 21°23.2’S, 81°38.3’W; 330-350 m; bot. trawl; 19-20.1V.1987. sta. 1856: Nazca Ridge, 21°43.9’S, 81°04.9’'W; 900-920 m over bot. depth 1,270-1,200 m; IKMWT; 20.IV.1987. sta. 1864/76: Nazca Ridge, 22°04.1'S, 81°16.7'W; 530-500 m; traps; 21.1V.1987. sta. 1867: Nazca Ridge, 22°06.1'S, 81°17.0'W; 225-247 m; bot. trawl; 21.1V.1987. sta. 1873: Nazca Ridge, 22°07.1'S, 81°16'W; 235 m; bot. trawl; 22.I1V.1987. sta. 1879/1882: Nazca Ridge, 23°26.4’S, 83°20.1'W; 505 m; traps; 23.1V.1987. sta. 1887: between Sala y Gomez and Nazca ridges, 24°41'S, 85°25.6'W; 50-300 m over bot. depth 408->2,000 m; IKMWT; 24.IV.1987. sta. 1925: Sala y Gomez Ridge, 25°32.4’S, 85°29.5'W; 50- 300 m over bot. depth 1,200->2,000 m; IKMWT; 27.1V.1987. sta. 1938: Sala y Gomez Ridge, 25°42.4’S, 86°35.3’W; 380 m; Sigsbee trawl; 28.1V.1987. sta. 1940: Sala y Gomez Ridge, 25°41.0'S, 86°35.9'W; 380 m:; bot. trawl; 28.I1V.1987. sta. 1941: Sala y Gomez Ridge, 25°48.6'S, 86°34.1'W; 410- 385 m; 28.1V.1987. sta. 1949: Sala y Gomez Ridge, 25°37.5’'S, 86°29.4'W; 200- 0 m over bot. depth 390 m; IKMWT; 29.IV.1987. sta. 1955/62: Sala y Gomez Ridge, 24°57.7'S, 88°36.1'W; 560 m; traps; 29-30.1V.1987. sta. 1956/61: Sala y Gomez Ridge, 24°54.0'S, 88°32.0'W; 580 m; traps; 29-30.IV.1987. sta. 1964: Sala y Gomez Ridge, 24°56.3’S, 88°32.6'W; 580- 564 m; bot. trawl; 30.1V.1987. sta. 1965: Sala y Gomez Ridge, 24°58.5'S, 88°29.3'W; 562- 545 m; bot. trawl; 30.1V.1987. sta. 1970: Sala y Gomez Ridge, 25°34.2’S, 89°09.1’'W; 540- 560 m; Sigsbee trawl; 1.V.1987. sta. 1971: Sala y Gomez Ridge, 25°30.4'S, 89°10.3’W; 570- 580 m; bot. trawl; 1.V.1987. sta. 1976: Sala y Gomez Ridge, 25°33.6’'S, 89°11.9'W; 563- 590 m; 1.V.1987. sta. 1977: Sala y Gomez Ridge, 25°09.9’S, 90°18.7'W; 545- 600 m: bot. trawl; 1-2.V.1987. sta. 1996: Sala y Gomez Ridge, 25°08.2’S, 99°25'W; 750- 800 m, 5.V.1987. sta. 2018: Sala y Gomez Ridge, 25°07.9’S, 99°26.8'W; 730- 790 m; 7.V.1987. sta. 2019: Sala y Gomez Ridge, 25°05.7’S, 99°27.7'W; 750 m; Sigsbee trawl; 7.V.1987. sta. 2023: Sala y Gomez Ridge, 25°07'S, 99°40'W; 439-500 m; bot. trawl; 8.V.1987. and trachyrincine grenadiers from the suborder Macrouroidei and included them in the suborder Gadoidei, raising both to family status (see also Howes and Crimmen 1990). Okamura (1989) 32 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 maintained the original view of the Macrouroidi- dae as a distinct family in the Macrouroidei and also raised the Trachyrincinae to family status, but he kept them and the Bathygadinae in the Macrouroidei, and also added to this suborder the enigmatic Euclichthyidae. Iwamoto (1989), Markle (1989), and Nolf and Steurbaut (1989a, b) maintained more traditional views of Tra- chyrincinae, Macrouroidinae, Bathygadinae, and Macrourinae as subfamilies of Macrouridae. We recognize that there are apt to be changes in the higher classification of what we now con- sider as macrouroid fishes, and that the tradi- tional view of the family Macrouridae being monophyletic will probably not stand. There will undoubtedly be a sorting-out period before a consensus is reached based on additional re- search. For the purposes of this paper, we find it convenient to follow a conservative classification and treat the major groups as subfamilies, with the exception of the Bathygadidae, and call them all grenadiers. DESCRIPTION OF NAZCA AND SALA Y GOMEZ RIDGES One of the most notable geologic features of the oceanic basin of the southeastern Pacific is a long, deep, and broken submarine range that ex- tends westward from the Peru-Chile Trench in the vicinity of Nazca, Peru, and traverses the basin to connect with the East Pacific Rise around longitudes 110°-115°W (Fig. 1). It is one of the world’s major mountain ranges, but it is exposed at the surface only on San Ambrosio, San Felix, Sala y Gomez, and Easter islands (Norris 1961). The Nazca Ridge forms the easternmost part of this range, beginning at the western rim of the Peru-Chile Trench at a latitude of about 15°S and extending in a southwesterly direction for about 600 nautical miles. The northeastern half of the ridge arises from a bottom depth of more than 4,500 m to reach upwards to within 2,200- 2,900 m of the surface. Only a few pinnacles in this part of the ridge are taller. This contrasts with the higher southeastern half, where a con- siderable portion lies above the 2,000-m isobath. Dominating this half is an elongated promi- nence, offset from the main axis of the ridge, that rises to within 320 m of the sea surface. The central part of the range lies to the east of Sala y Gomez Island and forms a well-defined ridge that rises 2,400-3,000 m above the sur- rounding sea floor (Norris 1961). This ridge, called the Sala y Gomez Ridge (Fisher and Norris 1960), extends more than 1,000 nautical miles to the east. It is punctuated by many promi- nences that rise to within 1,500 m of the surface. Several peaks that lie between longitudes 97°W and 101°W come to within the upper 500 m of the surface. Between the Nazca and Sala y Gomez ridges and to the south of the Nazca Ridge are a cluster of high seamounts that rise close to the surface. San Felix and San Ambrosio islands constitute the tips of the easternmost peaks of this cluster. The seamounts in this region were extensively investigated by Russian vessels. Key To NSG Species OF GRENADIERS This key is presented as a convenient guide to the identification of grenadiers from the Nazca and Sala y Gomez ridges. It should be used only as a guide, however, and preliminary identifi- cations should be confirmed with the descrip- tions. It can be confidently used only for adult specimens, as juveniles undergo ontogenetic changes that render many of the key characters useless. Because its coverage is geographically limited, some character states that key out NSG genera will not agree with the broader range of states given in the diagnoses for genera (e.g., pel- vic rays in Caelorinchus is given as seven in the key but six or seven in the diagnosis because six is found only in one western Pacific species). Fig- ures for the key are mainly from FAO Species Identification Sheets. For a more comprehensive key to the genera and species of grenadiers, the reader is referred to the FAO Species Catalogue on Gadiformes of the world (Cohen et al. 1990). la. One dorsal fin, no portion elevated, spi- nous rays absent; pelvics small or absent; head enormous, globose; eyes 10 or more in HL (Macrouroidinae) (Fig. 2a) 2 1b. Two dorsal fins, the first elevated, first two rays spinous; pelvics small to large; head moderate to large; eyes 5 or less in Hl (Pig: 2b) icc nt. oe 3 2a. Pelvic fins absent = ae Ie As Macrouroides inflaticeps (p. 38) Small pelvics with 5 or 6 rays es _ Squalogadus modificatus (p. 39) 3a. Mouth large, terminal (Fig. 2c); first (out- ermost) gill slit unrestricted; gill rakers 2b. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 33 FiGureE 2. 3b. 4a. 4b. Sa. Sb. 6a. (a) Squalogadus, (b) Nezumia, (c) Gadomus. on first arch long and numerous (Fig. 3a); second dorsal begins close behind first dorsal, its rays longer than those of anal; scales without spinules .. (Bathygadidae) .. Eee: Gadomus sp. cf. melanopterus (p. 36) Mouth small to large, usually subinferior (Fig. 2b); first gill slit restricted by mem- branes connecting upper and lower por- tions of first arch to gill cover; gill rakers tubercular or tablike (Fig. 3b); second dorsal separated by a distinct gap from first dorsal, its rays usually shorter than those of anal; scales beset with spinules in all but a few species (Macrourinae) .. 4 Branchiostegal rays-6: 222 S65 Yeer 2) Branchiostepal rays! 7 ee 2B 9 Snout blunt, scarcely protruding; ridge of stout scales, if present, not continuous to preopercle, and ends in a blunt or rounded tip (Fig. 4a); rakers present on both sides of first arch; pelvic rays 9-11 (Garely( 9) SLO rar Paaees Bee Re: FO. Coryphaenoides paradoxus (p. 52) Snout sharply and distinctly pointed; a continuous ridge of stout scales from tip of snout to angle of preopercle, the ridge ending in a sharp point (Fig. 4b); rakers absent on outer side of first gill arch; pel- wie rays 7 (Caelorinchus).t2es 0. 6 Light organ long, the anterior end just Ficure 3. Outer gill arches of (a) Gadomus and (b) Ne- zumia. behind isthmus (Fig. 5a); prominent body markines presente =< et C7] 6b. Light organ short, the anterior end not extending to a line between pelvic fin origins (Fig. 5b); no body markings _... 8 7a. Aseries of about 7 saddlemarks on body; first dorsal not black tipped ee TP Caelorinchus multifasciatus (p. 44) Figure 4. Diagrammatic illustrations of (a) Coryphae- noides paradoxus and (b) Caelorinchus comparing snout shape and suborbital ridge. Figure (a) redrawn from Iwamoto and Sazonov (1988, fig. 28). 34 FiGure 5. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Diagrammatic ventral views of Caelorinchus showing (a) long light organ extending from anus forward to large fossa on chest and (b) short light organ, scarcely apparent, anterior to anus. 7b. 8a. 8b. 9a. 9b. 10a. Not more than 2 or 3 saddlelike marks on body; first dorsal fin black tipped .... a _ Caelorinchus spilonotus (p. 49) Snout rather short, much less than orbit diameter; anterolateral edge of snout not supported by bone .. 2 Weare tit GC aelorinchus n nazcaensis is (D. 46) Snout long, much longer than orbit; an- terolateral edge of snout completely sup- POLrLEG Dy DONC cee. TR ce al pause... Caelorinchus immaculatus (p. 41) Ventral striae, consisting of parallel lines of black over a silvery ground, over much of isthmus, shoulder girdle, and ventral aspects of abdomen; ventral light organ with two lenslike bodies, one immedi- ately before anus and connected to an- other on the chest by a black midventral line (Fig. 6). (Hymenocephalus) ........ 10 No ventral striae; light organ not as above bi Ai. tues Three on NY 1a dbwn..2 Teninn ails. Nokes 14 Body terete, head about as deep as wide; small serrations along leading edge of second spinous ray of first dorsal fin; gill rakers on lower limb of outer arch fewer than 15 . Hymenocephalus gracilis (p. 55) Ficure 6. Diagrammatic lateral (a) and ventral (b) views of Hymenocephalus showing extent of ventral striations and location of vent and light organs. 10b. lla. 1 1b. 12a. 2b: 13a. Body laterally compressed, head deeper than wide; second spinous ray of first dorsal smooth; gill rakers on lower limb of outer arch more than|S 2a 1 Pelvic rays 8; orbits 42-55% HL _. 12 Pelvic rays 13-15; orbits 21-43% HL 13 Many pigment cells scattered ventral and posterior to black blotch on dorsum of trunk and part of caudal region; pectoral rays 114-119 (usually 116-117); interor- bital width 16.5—22.1, suborbital width 8.5-11.8, snout length 20.9-27.0% of HL ee Hymenocephalus semipellucidus (p. 60) Black blotch on dorsum with sharp out- lines, surrounded by singular pigment cells with few (if present) above and be- hind blotch; pectoral rays 112-114 (rarely i115), usually 112-113; interorbital width 20.8-27.4, suborbital width 5.2-9.5, snout length 16.4—23.8% of HL Av Hymenocephalus neglectissimus (p. 56) Orbits 21-23% HL; pelvic rays 13; bar- bel. absent... «3.355 1 ae Hymenocephalus sp. cf. aterrimus (p. 54) SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 35 b FiGcure 7. Diagrammatic ventral views of (a) Cetonurus and (b) Nezumia showing positions of pelvic and anal fins relative to periproct region. Arrow in (b) points to small fossa between pelvic fin bases. 15d: 14a. 14b. 15a. 15b. 16a. Orbits 35-42% HL; pelvic rays 14-15; barbel rudimentary or obsolescent sxc AYES Hymenocephalus striatulus (p. 63) Anus and urogenital opening surrounded by a broad naked margin, the entire re- gion (periproct) closely abutting or nar- rowly separated from anal fin origin, the anus closer to anal origin than to pelvic insertions; no accessory fossa of light or- gan anterior to anus (Fig. 7a) 15 Periproct moderate to small, broadly separated from anal fin origin by several scale rows, anus usually closer to pelvic insertions than to anal origin; usually a small fossa anterior to periproct (Fig. 7b) epee ctletest Preseli vy oeriha wh OM meorresett Puree et 17 Spinous ray of first dorsal smooth; pelvic origin below midbase of first dorsal fin, far behind vertical through pectoral or- Trachonurus villosus (p. 77) Spinous ray of first dorsal serrated; pel- vic origin below or anterior to origin of first dorsal fin, about on same vertical as pPectoralierigins ms Ue Ae 16 Head stout, somewhat depressed; snout long and pointed; no enlarged scales along dorsal and anal fins Ficure 8. Diagrammatic lateral (a) and dorsal (b) views of Cetonurus. Arrow in (b) points to enlarged scales along second dorsal fin. 16b. vas 17b. 18a. 18b. 19a. Head soft, inflated; snout broadly round- ed (Fig. 8a); a series of enlarged spiny scales along dorsal and anal fins (Fig. 8b) Hades _ Cetonurus crassiceps (p. 39) Lower j jaw ‘teeth large, widely spaced in 1 row; second spinous dorsal ray smooth; scales on branchiostegal membrane; an- terior fossa of light organ bean shaped (Figs. 9) e1sl< Malacocephalus laevis (p. 68) Lower jaw teeth small, closely spaced, in 1 or more rows; spinous dorsal ray ser- rated or smooth; no scales on branchios- tegal membrane, or if scaled, dorsal ray serrated; anterior fossa usually tear-drop Shaped deere et Peers ALT os Win Andee che 18 Head notably large and broad, preoper- cle and suborbital bones deep and large; interorbital width 33-44%, suborbital width 19-26%, orbit-preop. 53-64% HL : ._ Pseudocetonurus septifer (p. 75) Head n more normally proportioned for a macrourid; interorbital <31%, subor- bital <21%, orbit-preop. <47% HL ... 19 Snout distinctly pointed and tipped with a spiny bifid tubercle; a suborbital shelf 36 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Ficure 9. Diagrammatic lateral (a) and ventral (b) views of Malacocephalus. Arrows point to scales on branchiostegal rays. formed of two rows of coarse, modified scales; snout partially or extensively na- ked ventrally; mouth small, upper jaw length 26-38% HL (Nezumia) 20 Snout bluntly rounded to pointed, tipped with a single, rather small tubercle or terminal tubercle absent; suborbital shelf of more than two rows or no modified scales; snout almost entirely scaled ven- trally; mouth moderate to large, upper jaw 35-55% HL .. A Pelvic rays 14-1 8: inner er gill rakers of first arch 9-12 Nezumia propinqua (p. 72) Pelvic rays 11; inner gill rakers of first arch 6-10 ....... Nezumia convergens (p. 71) Pelvic rays 11; inner gill rakers of first arch 8-9; suborbital about 20% HL _... _ Kuronezumia pallida (p. 65) Pelvic rays s 8-10: inner gill rakers of first arch 12-19; suborbital 8—14% HL (Ven- irifossa) nat Ait Gren a emer ty PEM 22 Spinous second ray of first dorsal fin smooth _ Ventrifossa macrodon (p. 81) Spinous second ray of first dorsal fin ser- Tated OM LOO ital setatey re} 23 19b. 20a. 20b. 2 ae 21b. 22a. 22D: 23a. Barbel long, much greater than orbit di- ameter, about one-half HL fied Sic Ventrifossa obtusirostris (p. 84) Barbel moderate to short, 20) and long. Branchiostegal rays 7, upper 2 on epihyal. Scales small, bearing few to many fine erect spinules. One dorsal fin, no portion elevated, spinous rays absent; pelvics ab- sent. REMARKS.—Only the single species known. Macrouroides inflaticeps Smith and Radcliffe, 1912 Macrouroides inflaticeps Smith and Radcliffe in Radcliffe, 1912: 138, pl. 31, fig. 2 (Philippines, near Batan I., Lagonoy Gulf, Luzon, 13°23'S, 124°00'30’”W, 408 fms [746 m], Albatross sta. 5450). Norman 1939:48 (1 spec., near Maldives, Indian Ocean); Marshall 1964:92; 1973:516; Marshall and Taning 1966:1-5, pl. 1 (1 spec., South Atlantic off St. Helena); Parin et al. 1981:11 (1 spec., se. Pacific on Nazca Ridge); Shcher- bachev and Piotrovskiy 1982:45—-47 (10 spec., Indian Ocean); Arai in Uyeno et al. 1983:209 (12 spec., Suriname and French Guiana). Parin 1990:17 (listed from area between Nazca and Sala y Gomez ridges). SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 39 D1acnosis.— As for genus. DISTRIBUTION. — Philippines; Indian Ocean (including off SW tip of Australia); E and W trop- ical Atlantic; and E Pacific on Nazca Ridge. Mid- water to bottom in bathyal and abyssal depths (747-4,000 m) (Shcherbachev and Piotrovskiy 1982). SizeE.—To more than 48 cm. REMARKS.—Only one specimen of this dis- tinctive species was taken off NSG. That 475+ mm individual, the largest known, was reported by Parin et al. (1981). Radcliffe (1912) and Mar- shall and Taning (1966) provide good descrip- tions and illustrations; Shcherbachev and Pio- trovskiy (1982) provide a distribution map plotting all known records of the species, includ- ing theirs from the Saya de Malha Bank and Naturaliste Plateau in the Indian Ocean. MATERIAL EXAMINED. —ZMMGU 17683 (158 mm HL, 475+ mm TL); Nazca Ridge, 1,600-2,000 m; Hercules; midwater trawl 110. Squalogadus Gilbert and Hubbs, 1916 Type species: Squalogadus modificatus Gilbert and Hubbs, 1916, by original designation. Diacnosis.— As for Macrouroides, but having a small pelvic fin of five or six rays. REMARKS. — Only the single species known. Squalogadus modificatus Gilbert and Hubbs, 1916 (Figure 2a) Squalogadus modificatus Gilbert and Hubbs, 1916:156, pl. 8, fig. 2 (off Kyushu, Japan; 32°32'N, 132°23'’W; 720 fms [1317 m]; Albatross sta. 4956). Golovan’ 1978:222 (2 spec., West Africa, 1,450-1,550). Pakhorukov 1981:26 (3 spec., Whale Ridge; 850—1,550 m). Shcherbachev and Piotrovskiy 1982: 47 (17 spec., Indian Ocean; Pacific over Lord Howe Rise, 800-1,740 m). Shiobara 1982:143-146 (2 spec., Suruga Bay, Japan). Sawada in Amaoka et al. 1983:192 (1 spec., Iwate Pref., Japan; about 39°N). Parin 1990: 17 (listed from area between Nazca and Sala y Gomez ridges). Squalogadus intermedius Grey, 1959:330, fig. 53 (Gulf of Mex- ico, 29°07'N, 87°54’ W; 600 fms [1,097 m]; Oregon sta. 1426) (See Marshall [1973] for additional earlier references.) Dracnosis.—As for genus. SizE.—To more than 41 cm. DISTRIBUTION. — Worldwide in tropical seas, and in temperate western North Pacific and North Atlantic, but not yet recorded from central and eastern North Pacific and western South Atlan- tic. Shcherbachev and Piotrovskiy (1982) pro- vide a distribution map, to which may be added our Nazca specimen, the first record from the eastern South Pacific. They state that the species is restricted to bathyal depths of 600-1,740 m, ‘“‘where it may be found together with M. infla- ticeps.” REMARKS. — Our two specimens had six pelvic fin rays on each fin compared with the five gen- erally reported in this species. Squalogadus mod- ificatus is occasionally taken in great numbers in the north-central Gulf of Mexico. In an unpub- lished report of cruise 14 of the R/V Oregon IT (issued by the National Marine Fisheries Service, Pascagoula Fisheries Laboratory, Pascagoula, Mississippi), during a three-week period in Jan- uary 1970, ““Squalogadus intermedius [=Squalo- gadus modificatus] ... was taken frequently, catches ranging up to 219 pounds [about 100 kg] per 4-hour tow with a 70-foot [21.3 m] trawl.” Deepwater trawling during that cruise was con- ducted at 500 to 1,000 fathoms [914—1,829 m]. MATERIAL EXAMINED.—ZMMGU 17684 (152 mm HL, 459+ mm TL) and CAS 67408 (154 HL, 410 TL); Nazca Ridge, 1,050 m; Prof. Mesiatzev cr. 13, tr. 14. SUBFAMILY MACROURINAE The largest subfamily in the Macrouridae with more than 250 species, generally allocated to 30 or more genera. A diverse assemblage that may not be monophyletic. Size at maturity ranges from 12 cm in some Hymenocephalus to more than 160 cm in Albatrossia pectoralis (Gilbert, 1892). Species predominantly benthopelagic, but a few strictly bathypelagic species known. Distribution worldwide in depths from about 200 m to more than 6,000 m. Cetonurus Giinther, 1887 Type species: Coryphaenoides crassiceps Giinther, 1878, by monotypy. D1aGcnosis.—Macrourine grenadiers with 7 branchiostegal rays. Head large, broad; interor- bital width much greater than orbit diameter; suborbital width more than orbit diameter; snout high, broad. Mouth rather small, subterminal; armed with small teeth in 1-3 irregular rows. Barbel very small to rudimentary. Head and body essentially completely covered with small scales beset with fine erect spinules; a series of enlarged scales along each side of dorsal fins and anteriorly along anal fin. Lateral line without grooved scales; lateral line course marked by a series of widely 40 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 spaced black papillae (free neuromasts). First dorsal with a serrated spinous ray; V. 8-10. Anus within a broad naked area (periproct) that abuts anal fin origin. Swim bladder with 2 retia mirabi- lia. DISTRIBUTION. — Warm waters of the Atlantic, Pacific, and Indian oceans. Usual capture depths range from about 1,000-1,800 m, but C. globi- ceps recorded from more than 4,200 m. REMARKS. — Only two broadly distributed and closely related benthopelagic species. Cetonurus crassiceps (Giinther, 1878) Coryphaenoides crassiceps Giinther, 1878:25 (n. of Kermadec Is., Challenger sta. 170 and 171; 520 and 650 fm). Coryphaenoides (Cetonurus) crassiceps: Giinther 1887:143, pl. Bie Cetonurus crassiceps: Marshall 1973:613 (in key); Pakhorukov 1976:327 (4 spec., Rio Grande Rise); 1981:25 (Walvis Ridge); Sazonov and Shcherbachev 1985:180, fig. 1b, 2, 3 (diagnosis, distribution). Parin 1990:15 (listed from Nazca and area between Nazca and Sala y Gomez ridges). Driacnosis. — Orbits, 4.3—5.7 in HL (4.3-4.6 in NSG specimens); interorbital width about one- half HL; barbel very short to rudimentary, less than 3.5% HL; upper jaw extends to below pos- terior half of orbit. COUNTS AND MEASUREMENTS.—1D. II,8—10; 1P. 116-119; V. 10-11; GR-I (outer/inner) (4)7- 9/1 + 1 + (10-11), GR-II (0-1) + 1 + (10-12) /1 + 1 + (10-11); scales 1D. 12-14, mid.-1D. 6-11, 2D. 10-13; caeca 10-11 (3 spec.). Measurements in percent HL (8 spec., smallest excluded): postrostral 68.2—72.5; snout 35.4-42.7; preoral 32.3-38.0; orbit 21.7—23.0; interorbital 46.6—53.6; suborbital 23.7—30; postorbital 50.7- 58.9; orbit-preop. 50.3—58.2; upper jaw 31-37.1; gill slit 9.4-12.9; pre-1D. 99-106; pre-V. 95-127; pre-A. 118-139; body depth 73-95; V.-A. 23- 31; height 1D. 39-46 (3 spec.); base 1D. 17.3- 21.7; 1D.-2D. 25.6—32.8; 1P. 41-47 (2 spec.); V. 26-42 (4 spec.). SizE.—To more than 44 cm. DISTRIBUTION. — Cetonurus crassiceps has been reported from the subtropical South Atlantic, central tropical Atlantic, off Kermadec Islands, and off Hawaii (see Sazonov and Shcherbachev 1985: fig. 3). Records from the Nazca and Sala y Gomez ridges are the first in the southeastern Pacific, but are not unexpected. Surprisingly, however, the species is yet to be found in the Indian Ocean. COMPARISONS.—Sazonov and Shcherbachev (1985) readily differentiated C. crassiceps from its congener C. globiceps (Vaillant, 1888) and showed the circumglobal distribution of the lat- ter species. Their records show a common oc- currence of the two species only on the Sierra Leone Rise in the eastern Atlantic. The NSG specimens of C. crassiceps show no apparent dif- ference from representatives from other locali- ties. MATERIAL EXAMINED.—(9 spec.) Sala y Gomez Ridge: ZMMGU 17688(5:29-68 mm HL, 112+-314+ mm TL) and CAS 75971 (3:41-73.5 HL, 183-315+ TL); 1,070-1,100 m; Prof. Mesiatzev cr. 13, tr. 10. Nazca Ridge: ZMMGU 17689 (1, 62 HL, 320+ TL); 940-960 m; Prof. Mesiatzev cr. 13, tr. 94. Caelorinchus Giorna, 1809 Type species: Lepidoleprus caelorhincus Risso, 1810, by sub- sequent designation. DiAGnNosis.—Macrourine grenadiers with branchiostegal rays 6. Snout blunt and barely protruding to greatly elongated and sharply pointed. Orbit large, oval to elliptical, the long axis horizontal (or nearly so). Mouth small and inferior to moderately large (to about 40% HL). Outer gill slit greatly restricted; no gill rakers on outer side of first arch. Ridges of head often stout and sharply spined. A stout suborbital ridge formed of modified sales running from tip of snout posteriorly onto preopercle, ending in a sharp point. Scales covered with spinules, which vary widely in size, shape, arrangement, and den- sity. Spinous first dorsal ray smooth (a few ter- minal denticles in a few species). Pelvic fin rays almost always 7 (6 in one species). Anus at or near anal fin origin. Ventral light organ variously developed; elongated with two large dermal win- dows at each end in some, to almost rudimentary in others; lens absent. Abdominal vertebrae 1 1- 12. Swim bladder oval to bilobed anteriorly; retia mirabilia usually 4, but number variable in a few species (as many as 9-11 in C. canus). DISTRIBUTION. — Worldwide in tropical to temperate seas, but absent in high polar latitudes. Two species found south of Antarctic Conver- gence. Species most numerous in tropical waters. REMARKS.— We follow Eschmeyer (1990:70) in the use of the spelling of the generic name. The dipthong ae as originally used by Giorna (1809) was changed to oe by later workers on the SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 41 Figure 10. Caelorinchus immaculatus new species; holotype, ZMMGU 17692 (69 mm HL, 268+ mm TL) captured on Sala y Gomez Ridge in a trap set at 560 m, Prof. Shtokman cr. 18, sta. 1955/62, 29-30 April 1987. Photograph by Sergei Dudarev. assumption that the name Coelorinchus was de- rived from ‘hollow snout.’ The name Caelorin- chus, meaning burin snout, alludes to the loz- enge-shaped snout tip of the type species, C. caelorhincus (note: a burin is an engraving tool, sometimes having a lozenge- or diamond-shaped tip). The genus is a distinctive and diverse group with more than 100 species, although only about three-quarters of that number are named. Its re- lationship is closest to Macrourus (Okamura 1970a, b; Iwamoto and Sazonov 1988). Three of the four species recorded from NSG are most closely similar to species of the Indo-West Pa- cific. The fourth (C. nazcaensis) appears similar to C. aconcagua Iwamoto, 1978 from the con- tinental slopes of the eastern Pacific. Caelorinchus immaculatus new species (Figures 10, 11a, 12) Coelorinchus innotabilis: Iwamoto 1978:329-332 (in part; 1 spec., Chile).—Parin et al. 1981:11. ““Coelorinchus sp. nova 2 Sazonov et Iwamoto”: Parin 1990: 15 (listed from Nazca and Sala y Gomez ridges). Parin et al. 1990:42 (stomach contents). Kotlyar and Parin 1990:104, fig. 3v (otolith). DrAGnosis. — Snout sharp, slender, longer than orbits, anterolateral margins completely sup- ported by bone. Subopercle forms a narrow pointed triangle. Upper jaw teeth in broad short bands that fall well short of posterior angle of jaws; upper jaw about one-fourth HL. Underside of head scaleless except for a small patch postero- ventrally on preopercle. Scales with short, greatly reclined spinules in parallel rows; scales below 2D. (7-8.5) + 1 + (11-13). Anterior rays of sec- ond dorsal almost as well developed as opposite members of anal fin; 1D.—2D interspace less than base of first dorsal. No bold markings on body. Anus separated by 2-3 scale rows from anal fin; ventral light organ small with short pedicel. COUNTS AND MEASUREMENTS (data for holo- type in square brackets).—D. II,10 [II,10] + 102- 113, A. 99-116, 1P. 118-121 [i119]; GR-I (outer/ inner) 0/(1-2) + (0-1) + (6-7) 8-9 total [1 + 1 + 7], GR-II 0 + 1 + (5-7), 6-8 total/(1-2) + (6-8) 8-10 total [1 + 0 + 8]; scales below 1D. 7-9 [9.5], below 2D. (7-10) + 1 + (10.5-15) [9.5 + 1 + 10] (18.5-24 total scales in diagonal row to A.), below mid-1D. 5.5-8 [5.5], lat.1. 39-48 [41]. Total length of specimens examined 137-315 mm, HL 37.9-70 mm; the following in percent HL: postrostral 58-65 [61.7]; snout 37—43 [39.1]; preoral 35-45 [40.7]; orbit 28-32 [31.0]; inter- orbital 19-23 [19.3]; suborbital 12-15 [13.6]; postorbital 27-33 [30.9]; orbit-preop. 30-35 [31.7]; upper jaw 22-26 [23.9]; gill slit 8—1 1 [10.9]; barbel 5.6-13.5 [12.3]; body depth 40-56 [55]; pre-A. 133-151 [151]; V.-A. 31-45 [36]; height 1D. 36-57 [47]; 1D.-2D. 5-15 [10.4]; 1P. 37-45 [45]; V. 29-49 [37]. DESCRIPTION. — Head 3.8-4.2 in TL, about as wide as deep. Trunk and tail moderately com- pressed laterally. Anus slightly removed from anal origin a distance equal to about 0.5 diameter of pupil. Head ridges rather strong, sharp, and gen- erally narrow; head divided into dorsal and ven- tral parts by suborbital ridge. Snout sharply pointed; median and lateral processes of nasal bones connected and forming a complete bridge across anterolateral edges of snout. Interopercle completely hidden behind preopercle; ventral tip of subopercle produced into a flexible, pointed tab, its tip slightly exposed beyond posteroven- tral margin of preopercle. Gill membranes broadly connected to isthmus without a free pos- terior fold. Gill slits restricted, the outermost about as long as pupil diameter. Uppermost (epi- 42 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 FiGure 11. Scales from dorsum below interspace between dorsal fins of: (a) Caelorinchus immaculatus; (b) C. innotabilis; (c) C. nazcaensis, and (d) C. spilonotus. Drawn by Tomio Iwamoto. Scale bars equal 1.0 mm. branchial) gill rakers padlike; lower rakers tu- bercular and armed with very small spines. Chin barbel thin, short, less than diameter of pupil. Scales on body (Fig. 1 1a) of moderate size and covered with short, reclined spinules, each aligned in close longitudinal rows (9-12 rows in 48 mm HL specimen, 12-13 ina 58 mm HL specimen); outer rows on field slightly divergent from mid- dle rows. Posteriormost spinules extend beyond scale margin. Middle spinule row not enlarged. Scales on head generally with more divergent spinule rows; those dorsally between head ridges mostly with small, reduced spinules or spinules absent. A small scutelike scale at posterior end of occipital sensory canal and before lateral line; no supraoccipital ridge scale. Terminal snout scute somewhat arrowhead-shaped in dorsal view, dorsoventrally flattened, and armed with several longitudinal rows of small spinules, the rows diverging posteriorly from the anterior apex. Broad areas around nostrils and behind leading anterolateral margin of snout naked. Underside of head, including mandible, completely and smoothly naked except for a narrow file of thin scales on posterior end of preopercle. Head pores not prominent. Light organ small, externally manifested in some specimens by a blackish median streak be- fore anus; streak lacking in others. Luminescent gland housed within body wall and in a flattened, black, oval structure connected to rectum. Entire length of gland about 0.5-1.2 pupil diameter. Intestine with multiple loops and somewhat similar to that of C. smithi Gilbert and Hubbs, 1920, as illustrated by Okamura (1970b: fig. 65B). Pyloric caeca of three specimens 9, 9, and 10, short and unbranched. Gas bladder with 4 large retia and gas glands. Fins generally small; first dorsal shorter than postrostral length of head, its spinous second ray scarcely longer than adjacent segmented rays. Second dorsal close behind first dorsal, their in- terspace less than length of base of latter; ante- riormost rays of second dorsal relatively long for a macrourine grenadier, slightly shorter but dis- tinctly slimmer than anterior rays of anal fin. Outer pelvic fin ray somewhat prolonged beyond inner rays, reaching to anteriormost | to 3 anal rays; inner rays extend, at most, only to anus. Color in alcohol generally light brown to some- what swarthy on head, but abdomen and oper- culum dark bluish to blackish; scale pockets prominently outlined. Ventral surfaces of head grayish; in smaller specimens peppered with small melanophores. Anterior rim of orbits and sep- tum between anterior and posterior nostrils black. Fins all blackish or dusky. First dorsal in smaller specimens blackish basally, but pale distally; the distinction less obvious in large Prof. Mesiatzev specimens. Oral, branchial, and peritoneal lin- ings black. Gill arches blackish, but rakers and filaments pallid. We found two color variants. One variant, confined to specimens trawled on the Sala y Gomez Ridge, is very pale with almost unpig- mented pectoral fins and small melanophores scattered over naked areas of the head. The other variant, represented by all specimens from the Nazca Ridge and those taken in traps on the Sala y Gomez Ridge, have darkly pigmented pectorals and large melanophores densely covering naked areas of the head. The darkly pigmented head areas contrast strongly with the light adjacent scaled and ridged areas. The holotype and two paratypes (ZMMGU 18116 and 17693 [213 mm TL)) attributed to the dark variant, however, have somewhat intermediate pigmentation patterns. (A 220+ mm paratype from ZMMGU 17643 is a more typical dark variant.) SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 43 Foop.—Parin et al. (1990:42) found a high percentage of polychaet worms (Polinoidae and Tomopteridae) and copepods (the harpacticoid Cervicornia sp. and pelagic calanoids Clausocal- anus sp., Chirundia streetse, Bradyidius sp., Ae- tideus sp., Pseudocalanus sp., and Oncaea conif- era) in the alimentary canal of 20 individuals examined. Also present were shrimps, isopods, amphipods (Gammaroidea, including Oedice- rotidae), foraminiferans, and ophiurans (Ophi- ura). SizE.—To about 32 cm. DIsTRIBUTION.— Nazca and Sala y Gomez ridges and off central Chile in 340-780 m, its primary depth distribution probably around 400- 550 m. EtyMoLocy.—From the Latin, immaculatus, not spotted; in reference to the lack of a distinc- tive color pattern in the species, in contrast to two other NSG species, C. spilonotus and C. mul- tifasciatus, which have prominent blotches on the body. COMPARISONS AND REMARKS.— Caelorinchus immaculatus 1s close in all diagnostic characters to C. karrerae Trunov, 1984, a species originally described from material collected in the south- eastern Atlantic off South-West Africa (Namib- ia), Valdivia Seamount, Discovery Seamount, and also reported from the southwestern Indian Ocean. (Other specimens of C. karrerae, depos- ited in CAS, IOAN, ZMMGU, and ZIN, have been examined by us from the Saya de Malha Bank, Broken Ridge [West Australian Ridge], and Madagascar Ridge.) Differences between the two species are slight and may represent populational differences, but we think they are sufficient to justify our current treatment. In our comparison of specimens of the two species, C. immaculatus specimens have a somewhat more upturned snout, the orbit shape is less “‘squared off’ an- teriorly than in C. karrerae, the count of diagonal scale rows below the second dorsal is slightly higher (18.5—24 vs. 16-20 in C. karrerae), the snout is shorter (37-43% HL vs. 42-46%), the posterior nostril is somewhat smaller (6.9-9.1% HL vs. 7.7—11.5%), and the upper jaw is longer (22-26% HL vs. 19-23%). Furthermore, there are two or three rows of non-spinulated scales (rarely with rudimentary ridges) lateral to the median nasal ridge in C. immaculatus, whereas in C. karrerae, there is a single row of large scales bearing 4-8 well-developed subparallel ridges with numerous small reclined spinules. Orbit diameter (mm) Orbit to angle of preopercle (mm) Ficure 12. Scatter diagram comparing relationship of dis- tance orbit to angle of preopercle with orbit diameter in Cae- lorinchus innotabilis (plus signs) and C. immaculatus (circles). Diagonal line represents 1:1 ratio of the two measurements. The new species is also closely similar to the New Zealand and Australian species C. innota- bilis McCulloch, 1907, with which the single Chilean specimen was identified by Iwamoto (1978). It differs from that species, however, in having (1) an overall blackish first dorsal (C. in- notabilis has a pale first dorsal base and blackish tips); (2) no scales on underside of head above and lateral to mouth (C. innotabilis has patches of small scales there; (3) a shorter trunk and tail (HL about 3.7-4.2 in TL cf. about 4.2-5.5 TL in innotabilis); (4) a more prolonged preopercle (distance orbit to angle of preopercle less in C. innotabilis, see Fig. 12); (5) a longer upper jaw (22-26% HL cf. 19-21%); (6) a longer postorbital distance (27-29% HL cf. 25-26%); (7) usually more pectoral rays (i1 8-121 cf. 115-119); (8) more scale rows below 1 D. and 2D; and (9) a somewhat deeper body (about 50% HL vs. less than 50%). The two color variants of C. immaculatus present a bit of a problem. Differences between the variants are not restricted to color, but also include some morphometric features. The dark variants have on average a shorter snout and shorter preoral length, and they are somewhat larger (all >59 mm, cf. <68 mm in pale vari- ants.). The different capture localities and cap- ture gear suggest some geographical and ecolog- ical separation. The differences may be related 44 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Ficure 13. Gomez Ridge in 439-500 m. Photograph by Sergei Dudarev. to ontogenetic changes in body shape and hab- itat. Certainly, the proportional-measurement differences are interrelated and attributable to the longer snout of the light variant. More spec- imens and study are necessary before we can firmly establish the taxonomic status of the two variants. MATERIAL EXAMINED. —(dark variants noted with an aster- isk) Holotype: ZMMGU 17692* (69 mm HL, 268+ mm TL); Sala y Gomez Ridge, 560 m; Prof. Shtokman cr. 18, sta. 1955/ 62. Paratypes (85 spec. from 15 stations). CHILE. CAS 38314 (50.6 HL, 215 TL); 32°17’S, 71°39.5'W; 580 m; Anton Bruun cr. 18A, sta. 702. Sala y Gomez Ridge: ZMMGU 17690 (2: 47.5-50 HL, 180-224 TL) and CAS 50895 (37.9 HL, 161+ TL); 535-574 m; Ichthyanar cr. 5. tr. 54. ZMMGU 17691 (39 HL, 177 TL); 25°02’S, 88°35'W; 550 m; Astronomer no trawl no. ZMMGU 17693* (2: 59.3-59.5 HL, 213+-220 TL); 580 m; Prof. Shtokman sta. 1956/61. ZMMGU 17694 (26: 27.4— 63.3 HL, 117-249 TL); 580-564 m; Prof. Shtokman cr. 18, sta. 1964. ZMMGU 17695 (13: 40.5-51.8 HL, 162-220 TL); 562-545 m:; Prof. Shtokmancr. 18, sta. 1965. ZMMGU 17698 (61 HL, 243 TL); 550-630 m; Hercules tr. 74. ZMMGU 18110 Caelorinchus multifasciatus new species; holotype, ZMMGU 18117 (32.7 mm HL, 135 mm TL), from Sala y (3: 53.5-57.5 HL, 214+-231+ TL); 540 m; Ichthyandr cr. 5, tr. 53. ZMMGU 18111 (37.7 HL, 151 TL) and CAS 50896 (2: 39.4-39.9 HL, 137+-139+ TL); 420 m; Ichthyandr cr. 5, tr. 57. ZMMGU 18112 (48.2 HL, 180+ TL); 550-560 m; Prof. Mesiatzev cr. 13, tr. 2. ZMMGU 18113 (18: 38-50.7 HL, 135- 214 TL); 563-590 m; Prof. Shtokmancr. 18, sta. 1976. ZMMGU 18114 (32 HL, 147 TL); 545-600 m; Prof: Shtokman cr. 18, sta. 1977. Nazca Ridge: ZMMGU 17696* (5: 64-70 HL, 254— 315 TL) and CAS 75975 (4: 64-68 HL, 245+-307 TL); 340- 780 m; Prof. Mesiatzev cr. 13, tr. 44. ZMMGU 17697* (69.5 HL, 310 TL); 940-960 m; Prof. Mesiatzev cr. 13, tr. 94. ZMMGU 18115* (62.7 HL, 282 TL); 730-720 m; Prof. Shtok- mancr. 18, sta. 1828/35. ZMMGU 18116* (63.6 HL, 261 TL); 505 m; Prof. Shtokman cr. 18, sta. 1879/82. Caelorinchus multifasciatus new species (Figures 13, 14) “‘Coelorinchus sp. nova 4 Sazonov et Iwamoto”: Parin 1990: 16 (listed from Sala y Gomez Ridge). D1AGnosis.— Body terete and shallow, width across pectoral bases about equal to greatest depth. Anterolateral margins of snout incom- pletely supported by bone. Underside of head Ficure 14. Lateral view (a) and dorsal view (b) of holotype, ZMMGU 18117 (32.7 mm HL) of Caelorinchus multifasciatus new species. Drawn by Tomio Iwamoto. Scale bar equals 25 mm. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 45 completely naked except for a few scales over- lapping anterior edge of snout. Spinules on body scales in 5—6 discrete parallel rows, those on scales atop head between lateral occipital ridges in a single ridgelike longitudinal row on each scale. Second dorsal fin close behind first dorsal, its rays about as high as opposites of anal fin. A series of about 7 dark broad bands or saddles on body. A long light organ extending from anus almost to isthmus, the opposite ends expanded, each recessed in a shallow fossa. COUNTS AND MEASUREMENTS.—D. II,10 + 85+; 1P. 117-118; GR-I (inner) 0 + 1 + 5, GR- II (outer/inner) 0 + 1 + 6; scales below 1D. 6, mid=1): 5.5, 2D. 6.5, lat:1. 35. Measurements in mm, percent HL in paren- theses: postrostral 18.3 (56.0); snout 14.6 (44.6); preoral 14 (42.8); internasal 6.5 (19.9); interor- bital 6.9 (21.1); orbit 8.2 (25.1); suborbital 4.4 (13.5); postorbital 9.7 (29.7); orbit-preop. 9.5 (29.1); upper jaw 7 (21.4); barbel 3.5 (10.7); gill slit about 3.0 (9.2); pre-D. 36 (110); pre-A. 51.3 (157); V.-A. 15.6 (48); body depth 15 (46); 1D.- 2D. 4.3 (13.1); base 1D. 6.8 (20.8); height 1D. 15.5 (47); 1P. 14 (43); V. 14 (43). DESCRIPTION. — Body and head long, slender, terete; width over opercles about equal to body depth, width over pectoral base slightly less. Snout sharply pointed, a slender scute at tip; profile of snout viewed from above a slender convex cone. A prominent bony suborbital ridge, strengthened with thickened coarsely spined scales, runs from snout tip to preopercle, its posterior tip pointed, the ridge dividing the head into upper and lower parts. Orbits about 4 in head, shorter than snout, about 1.2 in postorbital length of head. Inter- orbital space flat, width slightly less than orbit diameter. Mouth small, inferior, rictus restricted: upper jaw less than orbit; barbel rather thick at base but tapers rapidly distally. Subopercle forms a narrow short tip posteroventrally. Branchiostegal membranes broadly attached to isthmus; gill openings strongly restricted, ex- tend anteriorly only to level of posterior edge of preopercle. Gill rakers few, tubercular; outer gill slit restricted, length about equal to barbel length. Scales of body with needlelike spinules aligned in 5 or 6 parallel rows, spinules in middle row slightly if at all larger than those in adjacent rows. Scales atop head somewhat elongated, with spi- nules erect and aligned in single, longitudinal, sharp, ridgelike rows. Dorsally on each side of median nasal ridge and behind leading edge of snout with scattered small scales, each covered with few erect spinules, and extensive naked ar- eas. A squarish naked area medially at end of occipital region and a few boundary areas on head with naked margins. Whether these naked areas become more extensively scaled in larger specimens is unknown. Ridges of head with modified scales; those on suborbital in 2 rows, heavily thickened and beset with stout spinules. Tip of snout with a slender sharp scute. Under- side of head entirely naked, although some scales overlap anterolateral snout edges onto ventral surface; whether these become more extensive or better defined in larger specimens is uncertain. Teeth in both jaws small and scarcely visible above thick gum papillae; those in upper jaw in short broad band that falls well short of posterior extent of rictus; those in lower jaw in a long tapered band roughly half width of premaxillary band. Fins rather small; first dorsal short, its second spinous ray slender with 3 fine needlelike spines near distal tip; second dorsal well developed and about equal in height to anal fin. Interspace be- tween dorsals short, much less than length base of first dorsal. Outer pelvic ray slightly thickened and prolonged, extending just past vent. Light organ large and prominently developed along ventral midline of abdomen and chest. Posterior end abuts anus, which lies immediately before anal fin; anterior end in a shallow fossa on chest, just behind isthmus and before pelvic fin bases, the two ends connected by a broad black median line. Pigment pattern highly distinctive, marked most prominently by a series of about seven broad saddles or bands, the anteriormost one at ante- rior end of nape and separated (by a narrow pale area that runs from hind margin of gill cover over nape) from a second band that extends over 9 or 10 scale rows to near the posterior border of the first dorsal. A faint diagonal pale strip leads to another more diffuse band that subtends only about 6 scale rows and ends under the 2nd or 3rd ray of second dorsal. A much darker band (9-10 scales wide) begins below the Sth or 6th ray of second dorsal. Broad pale areas separate the remaining bands, which are dark and subtend 8-9 scale rows each; the last band blending im- perceptibly into darkened tail tip. Abdomen dark, somewhat bluish, covered with large melano- phores having silvery middle. Interorbital region with a dark narrow band; another across lateral 46 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Ficure 15. occipital ridges at posterior margin of naked area at head of nape. Dorsal aspects of head gray, underside pale, virtually white except for faint punctations over gill membranes, gular region, posteroventrally on preopercle, and along as- cending process of premaxilla. Anterior parts of mouth white, but inner walls dark. Lips and bar- bel white. Long spinous ray of first dorsal black on basal one-third or so, remainder dusky; fin somewhat darker distally and anteriorly. Other fins pale to sparsely flecked with small melano- phores. Lining of gill cavity, the gill arches, and rakers heavily punctate. SizE.— Probably a small species, attaining at least 14 cm. DISTRIBUTION.—Known only from a single small specimen taken on the Sala y Gomez Ridge in 439 m. ETyMOLoGy.—From the Latin mul/to, many, and fasciatus, with bands, for the distinctive col- or pattern on the new species. COMPARISONS AND REMARKS.—The new spe- cies is one of the more striking members of the genus, with its slender, terete body, sharply pointed snout, and multiple bold saddle marks. It appears to be most closely related to C. cin- gulatus and C. spilonotus, the three being equally characterized by the combination of long light organ extending almost to the isthmus; antero- lateral margin of snout incompletely supported by bone; underside of head naked; long, slender, sharply pointed snout; spinules on scales nee- dlelike and arranged in relatively few parallel rows; rays of second dorsal about equal in height to those of anal; orbit about four in HL, longer Caelorinchus nazcaensis new species; paratype, CAS 50949 (70 mm HL, 280 mm TL), from Nazca Ridge in 340 m. Photograph by Susan Middleton. than upper jaw; and saddle marks on body. Cae- lorinchus multifasciatus is easily distinguished from the other two species by its more numerous bands and more terete body; its median nasal process not dark as in the others; its first dorsal lacking a black tip; its snout somewhat longer and more slender; and its spinules on scales atop head in single, longitudinal, ridgelike rows. The species appears to be unavailable to trawl gear because of its rocky-bottom habitat; only a single small specimen was captured in a torn bottom trawl. The holotype was captured on one of the westernmost seamounts sampled during the 18th cruise of the Prof. Shtokman in spring of 1987. No other Caelorinchus was collected there, but specimens of Mataeocephalus were captured at depths of 739-763 m in the area surrounded by rocks near the peak of the sea- mount. MATERIAL EXAMINED. — Holotype: ZMMGU 18117 (32.7 mm HL, 135 mm TL); Sala y Gomez Ridge, 439-500 m; Prof. Shtokman cr. 18, sta. 2023. Caelorinchus nazcaensis new species (Figures llc, 15, 16) Coelorinchus sp.: Parin et al. 1981:11 (brief description, 12 spec. Nazca Ridge, here reported). “‘Coelorinchus sp. nova 1 Sazonov et Iwamoto”: Parin 1990: 15 (listed from Nazca Ridge). Parin et al. 1990:42 (stomach contents). Kotlyar and Parin 1990:104, fig. 3a (otolith). D1acnosis. —Snout shorter than orbit, antero- lateral margin not supported by bone, terminal scute sharply pointed; orbit 2.5-3.0 in HL, 0.9- 1.2 into postorbital length. Free end of suboper- cle variably prolonged into a flap. Mouth rather SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 47 Ficure 16. Caelorinchus nazcaensis new species; paratype, CAS 50898 (77 mm HL, 322 mm TL), from Nazca Ridge in 340 m. Scale bar for (a) and (b) equals 25 mm; that for (c) equals 1.0 mm. Drawn by Amy Pertschuk. large for genus and unrestricted at posterior an- gle, upper jaw about one-third HL. Upper jaw teeth in narrow bands extending to posterior an- gle of mouth. GR-I about 12-14 total. Underside of head (including lower jaws) naked except for small patch of scales above end of mandible. Anus separated by 1-3 scale rows from anal fin; light organ very small, length slightly more than length posterior nostril; no pedicel. COUNTS AND MEASUREMENTS. —(data for ho- lotype in square brackets). D. II,9-11 [II,10] + 118-129; 1P. 119-123 [120/121]; A. about 123- 138; GR-I (outer/inner) 0/(1-3) + 1 + (9-11) (total 12-14), GR-II (1-2) + (0-1) + (9-11) (total 11-13) [1 + 1 + 10)/(1-3) + (O-1) + (8-11) (total 11-13) [3 + 1 + 10]; scales below 1D. 7- 10 [8], below mid-1D. 5.5-7 [5], below 2D. 6.5- 8 [6], lat.1. 35-43 [42]; pyloric caeca 30-44. Head length 46-94 [67.7] mm; other dimen- sions given as percent HL: snout 28.2-32.1 [31.3]; preoral 23.6—29.2 [29.0]; orbit 33.8-39.7 [37.4]; interorbital 18.9-24.5 [21.9]; suborbital 10.4- 13.3 [11.8]; postorbital 34.9-41.4 [33.5]; orbit- preop. 34.2-40.7 [35.6]; upper jaw 30.0-37.8 [33.8]; gill slit 14.6-22.3 [18.9]; barbel 10.9-18.1 [17.1]; body depth 58-79 [67]; pre-D. 100-111 [107]; pre-V. 104-116 [106]; pre-A. 130-156 [139]; V.-A. 32-51 [37]; height 1D. 44-56 [52]; 1D. base 23:2—28.0 [25.1]; 1D.-2D. 27.8-52.1 [38.7]; 1P. 38-50 [51]; V. 29-42 [39]. DESCRIPTION. — Head large, about 4 in total length, broad, greatest width slightly less than greatest depth of head. Orbits huge, more than one-third HL, much longer than snout, 0.9-1.2 postorbital length, about 0.95-1.2 of length from orbit to angle of preopercle. Posteroventral end of subopercle variously developed into a short flap directed ventrally or posteroventrally. In- teropercle completely hidden behind preopercle. Snout rather short, sharply pointed, tipped with a stout, narrowly conical scale. Median and lat- eral processes of nasal bones separated by a broad gap, leaving anterolateral margin of snout un- supported by bone. Mouth of moderate size, up- per jaw about one-third HL; maxilla extends to below middle of orbits. Dentition consists of small, conical, slightly recurved teeth in moderately wide bands, the mandibular band narrower than that of premax- illary. Premaxillary band about 6-7 teeth rows wide, tapering posteriorly to a few teeth wide and extending along most of mouth gape, but falling just short of end of rictus. Mandibular band sim- ilarly shaped but longer than premaxillary band, extending beyond end of rictus. Squamation characterized by rather large scales covering most of body and dorsal surfaces of head, but almost entirely absent on ventral head surfaces (a few small scales ventrally near junc- tion of infraorbital and preopercular ridges in 48 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 some specimens). A narrow half-moon-shaped naked area on each side of leading edge of snout. Gill and nasal membranes naked. Suborbital shelf formed of two rows of stout, spiny, scutelike scales, the lower margin forming a distinct ridge that extends from snout to just short of pre- opercle angle. Angle formed by dorsal and ven- tral planes of suborbital region obtuse. Median nasal ridge low, formed by a row of about 8 broad scales, each covered with about 12 low spinule ridges that radiate from an anteromedian focus. Two rows of smaller scales on either side of me- dian nasal ridge, these separated laterally by a shallow naked groove from a row of scales along mesial side of supranarial ridges. Surpanarial ridges incompletely scaled in smaller paratypes. A narrow spiny ridge one scale wide separating orbit from nostril region. Head ridges other than those of suborbital not endowed with markedly heavy, enlarged, and spiny scutelike scales. No supraoccipital ridge. Scales of trunk below in- terspace of dorsal fins beset with 5-8 slightly di- vergent, ridgelike rows of short, sawtooth-shaped spinules (Fig. 11c). As with most species of ma- crourids, spinule rows on scales probably be- come more numerous with size. A small scaleless black fossa of light organ im- mediately preceding anus (fossa may be covered by scales in fully scaled specimens, but in our study specimens, scales missing from area). Anus separated from anal fin by 1-3 scale rows. Swim bladder in 72.2 mm HL male about 40 mm long, broader anteriorly, with two short horns directed anteriorly; closely attached to ribs on either side of vertebral column. Drumming mus- cles well developed in ventrolateral position be- hind each horn. A long oval window on dorsal surface. Five short broad retia, each terminating in a small globular gas gland (but a smaller, 70 mm HL male, CAS 50949, had 4 retia and gas glands). Swim bladder in a female of 88.9 mm HL similar in shape, but drumming muscles ru- dimentary and 4 retia. Fins all short; height first dorsal much less than postrostral length of head; pectorals slightly more than postorbital length of head and barely ex- tending to level of vent; pelvic rays fall well short of that level except for outermost ray, which just reaches to vent in some specimens. Color in alcohol overall swarthy to grayish brown, paler ventrally on head, but anteroventral margin of snout markedly darker than more pos- terior areas; opercular region darkish; abdomen bluish. Fins generally dusky, outer ray of pelvic fins whitish. Branchial, oral, and peritoneal lin- ings bluish to brownish black. Lips and barbel pale. Foop.—Parin et al. (1990:42) recorded mostly pelagic food items in the stomachs of 10 speci- mens examined. These included the shrimp gen- era Gennada, Sergestes, Plesionika, the mysid Paralophogaster glaber, the copepod Pleuro- mamma, squid, and fish (Chauliodus sp., Lam- panyctus sp., Myctophidae gen., sp.). SizE.—To about 40 cm. DIsTRIBUTION.— Known only from the type- specimens taken in 225-530 m on the Nazca Ridge. EtyMoLoGy.— Derived from the type locality, the Nazca Ridge. COMPARISONS AND REMARKS.—Caelorinchus nazcaensis is closely similar to C. aconcagua Iwamoto, 1978, from the Pacific coast of Chile, sharing with that species a rather deep head (deeper than broad); relatively large mouth (al- most one-third HL); essentially completely na- ked underside of head (note exception in new species); naked areas on either side of leading edge of snout; small light organ without pedicel immediately before anus; rather wide outer gill slit (more than one-half orbit diameter); numer- ous gill rakers (12 or more total on first arch) compared with other members of genus; snout much shorter than orbit diameter; gill mem- branes narrowly joined across isthmus and form- ing a deep “V”; and similar-shaped opercular bones. The two species are readily differentiated by (1) orbit diameter (somewhat larger in C. aconcagua, 37.4—43.5% HL cf. 33.8-39.7%); (2) barbel length (shorter in C. aconcagua, 6.2-10.8% HL cf. 10.9-18.1%); (3) pectoral fin length (51.9- 66.3% HL in C. aconcagua cf. 37.8-51.1%); (4) dark bluish color of trunk (extends around entire body in C. aconcagua, but restricted ventrally on belly below level of pectoral fin origin in new species); (5) oral cavity (white in C. aconcagua, blackish throughout in new species); (6) pyloric caeca (more than 30 in new species cf. 16-20 in C. aconcagua); (7) scale spinulation; and (8) head covering (thinner, and scales dorsally on head not as dense as in C. aconcagua). Caelorinchus nazcaensis also resembles C. fas- ciatus (Giinther, 1878), sharing with that and related southern hemipshere species (including C. aspercephalus Waite, 1911, C. australis (Rich- ardson, 1839), C. biclinozonalis Arai and Mc- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 49 FiGureE 17. Millan, 1982, C. bollonsi McCann and Mc- Knight, 1980, C. cookianus McCann and McKnight, 1980, and C. mirus (McCulloch, 1926) the following features: (1) snout short, rather blunt; (2) orbits huge, much longer than snout, about equal to or greater than postorbital length; (3) anterolateral margin of snout not supported by bone; (4) light organ rather small, not ex- tending to level of bases of pelvics; (5) postero- ventral angle of subopercle rounded to variably produced, but not into a long narrow point. The new species differs from C. aspercephalus, C. australis, and C. biclinozonalis in lacking scales on the underside of the head (except for a few isolated ones in some individuals). Features dif- ferentiating C. nazcaensis from C. mirus, C. fas- ciatus, C. cookianus, and C. bollonsi include: lon- ger premaxillary tooth bands (not extending to end of rictus in others); mouth opening not no- tably restricted as in other species; first dorsal fin shorter than postrostral length of head (equal to or longer than in others); pectoral and pelvic fins shorter; spinules on body scales broadly saw- tooth-shaped (cf. smaller, finer, greatly reclined and imbricate); and terminal snout scute sharply pointed (blunt in others). MATERIAL EXAMINED.—(all from Nazca Ridge) Holotype: ZMMGU 18094 (67.7 mm HL, 288 mm TL); 235-250 m; Prof. Mesiatzev cr. 12, tr. 23. Paratypes (33 spec.): CAS 50949 (2:65-70 HL, 270-280 TL); 340 m; Ichthyandr cr. 6, tr. 1. CAS 50948 (54 HL, 240 TL); 320 m, Ichthyandr cr. 6, tr. 2. CAS 50898 (77 HL, 322 TL); 340 m; Ichthyandr cr. 6, tr. 3. CAS 50897 (84 HL, 350 TL) and ZMMGU 17699 (8:46-89 HL, 197-383 TL); 330 m; Ich- thyanadr cr. 5, tr. 13. ZMMGU 17700 (59.5 HL, 242 TL); 330- 340 m; Ichthyandr cr. 5, tr. 14. ZMMGU 17701 (3:60-93 HL, 270-402 TL) and CAS 50899 (69 HL, 265+ TL); 330 m; Ichthyanar cr. 5, tr. 15. ZMMGU 17702 (2:69.5-75 HL, 270+- 307 TL); 310-330 m; Ichthyandr cr. 5, tr. 17. CAS 50900 (2: 70-76 HL, 290-305 TL); 330 m; Ichthyandr cr. 5, tr. 33. ZMMGU 17703 (6:71-82.5 HL, 289-357 TL); 330 m; Prof. Shtokman cr. 18, sta. 1845. ZMMGU 17704 (4:53.5-86 HL, 225-370 TL); 330-350 m; Prof. Shtokman cr. 18, sta. 1851. ZMMGU 17705 (82 HL, 310 TL); 530-500 m; Prof. Shtokman cr. 18, sta. 1864/76. Caelorinchus spilonotus new species; paratype, ZMMGU 18124 (52.8 mm HL), from Sala y Gomez Ridge in 545-600 m, Prof. Shtokman cr. 18, sta. 1977, 1-2 May 1987. Photograph by Sergei Dudarev. Non-type material (47 spec.): ZMMGU 18093 (2:62.2-70 HL, 239+-295+ TL); 300-330 m; Astronomer trawl without no. (23°30.5’S, 81°45’W). ZMMGU 18095 (78.4 HL, 336+ TL); 235-225 m; Prof. Mesiatzevcr. 13, tr.25. ZMMGU 18096 (59 HL, 255 TL); 340-325 m; Prof. Mesiatzey cr. 13, tr. 35. ZMMGU 18097 (72.7 HL, 329 TL); 235-230 m; Prof. Mesia- zey cr. 13, tr. 36. ZMMGU 18098 (70.5 HL, 294+ TL); 230- 240 m; Prof. Mesiatzev cr. 13, tr. 29. ZMMGU 18099 (2: 58- 59.5 HL, 247-253 TL); 320-325 m; Prof. Mesiatzev cr. 13, tr. 43. ZMMGU 18100 (3:63-80.5 HL, 267+-324+ TL); 340- 780 m; Prof. Mesiatzey cr. 13, tr. 44. ZMMGU 18101 (68.5 HL, 286 TL); 225-240 m; Prof. Mesiatzev cr. 13, tr. 109. ZMMGU 18102 (91 HL, 399 TL); 225-240 m; Prof. Mesiatzev cr. 13, tr. 31. ZMMGU 18103 (2:72-74.5 HL, 295-315 TL); 230 m; Prof. Mesiatzev cr. 13, tr. 120. ZMMGU 18104 (81.8 HL, 342 TL); 320-340 m; Ichthyandr cr. 6, tr. 10. ZMMGU 18105 (72.3 HL, 278+ TL); 230-250 m; Odissey cr. 2, tr. 15. ZMMGU 18106 (2:65.5—71.3 HL, 273-297 TL); 235 m; Odis- sey cr. 2, tr. 11. ZMMGU 18107 (80 HL, 348 TL); 230 m; Odissey cr. 2, tr. 14. ZMMGU 18108 (75 HL, 320 TL); 225- 247 m; Prof. Shtokman cr. 18, sta. 1867. ZMMGU 18109 (5: 68-87.5 HL, 292-366 TL); 235 m; Prof. Shtokman cr. 18, sta. 1873. IOAN uncat. (21:61-92 HL, 252+-400 TL); 340 m; Akademik Kurchatov cr. 34, sta. 3594. Caelorinchus spilonotus new species (Figures 11d, 17, 18) Coelorinchus cingulatus: Parin et al. 1981:11 (12 spec. from Sala y Gomez Ridge). “‘Coelorinchus sp. nova 3 Sazonov et Iwamoto”: Parin 1990: 15 (listed from Sala y Gomez Ridge). Kotlyar and Parin 1990:104, fig. 3b (otolith). DiAGNosis.—Snout acutely pointed, tipped with a narrow, pointed scute; median nasal bone blackish; anterolateral margins of snout not sup- ported by bone; ventral surfaces of snout entirely naked. Subopercle posteroventrally produced into a short triangular flap. Mouth small, restricted at posterior angle; upper jaw 20-26% of HL. Up- per jaw teeth in broad short bands that fall short of posterior angle of mouth. Scales on body with slender, conical, reclined, imbricate spinules ar- ranged in 5-10 parallel to slightly divergent rows (more rows in larger specimens). Anterior rays of second dorsal fin as long as opposite members of anal fin. Interspace between dorsal fins short, 50 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 FiGure 18. Drawn by Tomio Iwamoto. Scale bar equals 25 mm. less than base of first dorsal. Two prominent sad- dle marks, one immediately before first dorsal fin connecting to broad pectoral blotch, second saddle separated from first by a distance about equal to snout length. First dorsal fin black tipped. Anus immediately before anal fin. Light organ extends forward between pelvic fins and onto middle of chest, both ends expanded into a tear- drop shape and lying within a shallow fossa. COUNTS AND MEASUREMENTS (from 26 speci- mens; mean, X, in parentheses, holotype in square brackets).— 1D. II,8—10 (II,9.35) [II,9]; 1P. 117- 122 (119.73, n = 52) [i119]; GR-I (outer/inner) total 6-8 [0/0 + 1 + 5], GR-II total 5-7 [0 + 0 + 5/1 + 1 + 5]/total 6-8; scales below 1D. 6.5— 8 [7.5], below 2D. 5.5-7.5 [7], below mid-1D. 4.5-5.5 [5], lat.1. 37-44 [43]. Total length of specimens examined +96-206 mm [184+ mm], HL 29.5-60 mm [60 mm]. The following in percent HL: postrostral 49.7-60.5 (55.5) [54.5]; snout 41.8-51.4 (46.1) [46.7]; preoral 35.4—44.8 (39.1) [40.5]; internasal 17—22 [19.3]; orbit 21.2—27.9 (24.7) [24.2]; interorbital 18.0-24.1 (20.6) [21.0]; suborbital 10.8-14.8 (12.5) [12.8]; orbit-preop. 25.3—32.8 (29.8) [29.2]; upper jaw 20.3-26.0 (24.0) [24.2]; gill slit 8.6- 13 [13.3]; barbel 4.8-9.0 (6.9) [7.8]; body depth 37.6-47.5 (41.7) [44.2]; light organ 47.3-60.7 (51.7) [53.3]; pre-D. 102-115 (107) [107]; pre- V. 100-115 (107) [106]; pre-A. 139-161 (150); V.-A. 38-55.8 (46.6) [46.7]; height 1D. 34-49 (41.3) [44]; 1D. base 13.8-—21.6 (18.7) [17.8]; 1D- 2D. 4.2-13.3 (9.2) [9.5]; 1P. 37-45 (40) [38]; V. 27-39 (33) [33]. DESCRIPTION. — Head about as deep as wide. Snout sharply pointed; anterolateral margins not supported by bone; a broad gap between median and lateral processes of nasal bone. Head ridges LLY My Ue K< Ul <— Caelorinchus spilonotus new species; paratype, CAS 50950 (50.3 mm HL), from Sala y Gomez Ridge in 440 m. distinct but not especially strong, except for sharp suborbital ridge, which divides ventral and dor- sal portions of head. Supraoccipital ridge weakly developed. Interopercle completely hidden be- hind preopercle; subopercle developed postero- ventrally into a short triangular flap. Gill mem- branes broadly attached to isthmus, without a free posterior fold. Gill slits somewhat restricted, outer slit slightly longer than pupil diameter. Gill rakers low, epibranchial rakers shaped like flat- tened disks; remainder tubercular. Chin barbel short, almost rudimentary, usually shorter than length of posterior nostril. Scales of body (Fig. 11d) moderate sized, cov- ered with parallel to slightly divergent rows of small, sharp, conical, imbricate spinules, the pos- teriormost tips extending beyond margin of scale. About 5 or 6 spinule rows in smallest specimens examined (about 30 mm HL); 9 or 10 in largest specimens (about 60 mm HL). Scales on head generally like those of body, but with shorter, more erect spinules in fewer, more divergent rows. Nostril membranes completely naked, but sub- orbital shelf below and behind nostrils complete- ly scaled. Narrow strips on either side of median nasal ridge naked. Terminal snout scute sharp, rather small, narrow, and weak (tip broken off in several specimens); protrudes directly forward or slightly upwards; upcurve of tip more highly developed in smaller specimens. Light organ large, long, extending from anus to chest just behind isthmus. Luminescent glands as described and illustrated for the subgenus Quincuncia by Okamura (1970b:figs. 43, 81). Al- imentary canal about like that illustrated for Caelorinchus longissimus by Okamura (1970b: fig. 64D). Extent of anteriormost bends could not be accurately determined because all specimens SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 51 had everted stomachs that distorted bending pat- tern. Pyloric caeca short, thick; 9-12 in 9 spec- imens. Swim bladder oval, blunter end anteriad, with 4 or 5 (in 5 specimens) short retia, each tipped with a broad, flat gas gland. Eggs in the large ovaries of CAS 50951 (58.8 mm HL) dis- tinct, the largest about 1.0 mm in diameter. Fins generally small, height first dorsal about as long as snout, second spinous ray not pro- duced. Pectoral fin short, about equal to snout length. Outer pelvic ray about equal to or slightly longer than postorbital length of head, falls short of anal fin. Second dorsal relatively well devel- oped for genus, height of rays anteriorly about equal to opposite members of anal fin. Interspace between dorsal fins short, less than length base of first dorsal. Color in alcohol light brown to tawny overall, paler ventrally on head and on tail; darker ven- trally on trunk. Saddles and blotches darker brown. Two prominent saddles, one before first dorsal fin, the second posteriorly about one snout length behind first and originating at position of 9th ray of second dorsal fin. The first saddle blending in larger specimens with large blotch above pectoral fin base. This pectoral blotch 5 or 6 scale rows wide, posterior saddle 6 or 7 rows wide. A faint blotch in larger specimens between dorsal-fin interspace and lateral line. Underside of head entirely lacking pigmentation except for fine scattered punctations posteriorly. Median process of nasal bone blackish or dusky. Mouth pale except for scattered punctations posteriorly on roof of mouth and in gullet. Gill chamber and gill membranes pale with scattered punctations; gill arches and filaments pale. Peritoneal mem- brane blackish; stomach pale to dusky near esophagus. Fin rays whitish except for black- tipped first dorsal, and uppermost pectoral ray often darker. Foop.— One female (58.8 mm HL, CAS 50951) had a large squid beak in its otherwise empty stomach; all other specimens had everted stom- achs. S1zE.—A small species, maximum length slightly more than 20 cm. DISTRIBUTION. — Known only from the Sala y Gomez Ridge and Hawaiian Islands in 330- 600 m. ETyMoLoGy.—From the Greek, spilos (spot), and notos (back), in reference to the peculiar col- oration of this species. COMPARISONS. — The species is most similar to, and may be conspecific with, C. cingulatus Gil- bert and Hubbs, 1920, a species known only from the two small type-specimens taken in the Phil- ippines and one 232 mm specimen recently re- corded from the Okinawa Trough by Okamura (in Okamura and Kitajima 1984:229, 366, fig. 161). The holotype (USNM 78221, 40.2 mm HL) and small paratype (41.4 mm HL, 148+ mm TL; USNM 78233) of C. cingulatus were ex- amined for comparison. Most meristic and mor- phometric features of the two species are similar, but their pigmentation patterns differ. (Pigmen- tation in the paratype is almost totally lost from long preservation; thus, color-pattern compari- sons were made from the original description and from the description and illustration by Oka- mura in Okamura and Kitajima 1984.) Caelorin- chus cingulatus lacks the large blotch above the pectoral fin base that is so prominent in C. spi- lonotus. Furthermore, the first saddle mark be- gins on the nape in front of the first dorsal fin and extends ventrally to join the pectoral blotch. None of this saddle lies below the soft rays of the first dorsal fin, as described and illustrated by Okamura for C. cingulatus. The second saddle mark does not extend below the lateral line in the new species, but does so in C. cingulatus, and the second saddle begins under the 7th—10th rays of the second dorsal in the new species, but under the sixth in C. cingulatus. Gilbert and Hubbs (1920:483) describe brown streaks that “‘radiate backward from the eye... the upper one, more conspicuous, extends hor- izontally backward, just below the postorbital scaly ridge, to the upper angle of the branchial aperture, where it is continuous with the dusky opercular blotch.”’ This upper streak is promi- nent in Okamura’s (in Okamura and Kitajima 1984, fig. 161, p. 228) photograph of his speci- men but completely lacking in our specimens of C. spilonotus. In C. cingulatus a diagonal streak extends anteriorly and ventrally from the second saddle mark, joining the darker abdominal re- gion. This streak is also absent in C. spilonotus. The new species may have fewer gill rakers on the inner series of the second arch (7 or 8 cf. 9 in C. cingulatus) and fewer pyloric caeca 9-12 (cf. about 14 in the holotype and 15 in the para- type of C. cingulatus, although Gilbert and Hubbs [1920:482] reported 21 for the paratype). With so few specimens of C. cingulatus, these enu- merated characters cannot be adequately as- sessed to determine their value as species char- 52 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 acters. The possibility exists that they may simply reflect individual or geographic variation within one species. The three Hawaiian specimens of C. spilonotus have noticeably longer, more attenuated snouts, with the terminal scute narrow, long, and sharply pointed. Naked areas laterodorsally behind the leading edges of the snout are more extensive than in the NSG specimens. Punctations on the underside of the head are lacking in the NSG specimens but noticeably present in the Hawai- ian specimens (and in C. cingulatus). Head col- oration of the Hawaiian specimens is slightly darker than that of the NSG specimens and fol- lows more closely the color description of C. cin- gulatus. We do not consider these differences as sufficient to recognize the Hawaiian population as distinct. The elongation of the snout is rem- iniscent of the situation in C. caribbaeus (Goode and Bean, 1885), where the snout length is re- markably variable within a single population in the northern Gulf of Mexico (see Marshall and Iwamoto 1973). Among other Pacific members of the genus, the new species and C. cingulatus share many important characters with C. gladius Gilbert and Cramer, 1897 from the Hawaiian Islands. The chief differences are: C. gladius lacks saddle marks but has a prominent black blotch over the pec- toral base; the snout is sharp and narrow, with the terminal scute remarkably long and devoid of spinules dorsally. The second spinous ray of the first dorsal fin has a few minute denticula- tions distally along the anterior margin in the two C. gladius specimens examined (47 and 66 mm HL, CAS-SU 8517), and surprisingly, also in the Hawaiian specimens of C. spilonotus. Cer- tain proportions differ notably as a result of the extremely elongated snout of C. gladius (e.g., the snout length is considerably greater than the postrostral length of the head in C. gladius but is much shorter than that measure in C. spilono- tus from the NSG, and about equal in the Ha- walian representatives). MATERIAL EXAMINED. — Holotype: ZMMGU 18125.(60 mm HL, 187+ mm TL); Sala y Gomez Ridge, 545-600 m; Prof. Shtokman cr. 18, sta. 1977. Paratypes (82 spec.). Sala y Gomez Ridge: CAS 50950 (50.3 mm HL, 182 mm TL); 440 m; Ichthyandr cr. 5, sta. 40. CAS 50952 (2:30.6-38.4 HL, 108+-115+ TL); 480 m; Ichthyandr cr. 5, sta. 50. ZMMGU 17706 (4:42.5-56.5 HL, 138+-204+ TL); 540 m; Ichthyandr cr. 5, tr. 53. ZMMGU 17707 (6:29.5- 58.5 HL, 96+-201 TL) and CAS 50951 (6:43.5-58.8 HL, 131+-202 TL); 410 m; Ichthyandr cr. 5, tr. 56. CAS 50953 (56.2 HL, 192+ TL); 420 m; Ichthyandr cr. 5, tr. 57. ZMMGU 17708 (12:30.5-43.6 HL, 116-162 TL); 380 m; Prof. Shtok- man cr. 18, sta. 1940. ZMMGU 17709 (5:35-52 HL, 133-188 TL); 410-385 m; Prof. Shtokman cr. 18, sta. 1941. ZMMGU 18123 (15:42.7-52.3 HL, 130+-172 TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. ZMMGU 18124 (30:25-58 HL, 98-198 TL); 545-600 m; Prof: Shtokman cr. 18, sta. 1977. Hawaiian Islands: LACM 45409-1 (3:52.7-58.2 HL, 176+- 181+ TL); Townsend Cromwell tr. 57; Mar-Apr 1972. Non-type material (7 spec.). Sala y Gomea Ridge: ZMMGU 18118 (30.5 HL, 97+ TL); 360-400 m; Ichthyandr cr. 5, tr. 51. ZMMGU 18119 (2:48.5-56.5 HL, 162+-182+ TL); 535- 575 m; Ichthyandr cr. 5, tr. 53. ZMMGU 18120 (44.5 HL, 157 TL); 550 m; Astronomer trawl without no. (25°02’S, 88°35'W). ZMMGU 18121 (48.8 HL, 174 TL); 565 m; Prof. Mesiatzev cr. 15, tr. 50. ZMMGU 18122 (2:44.7-54 HL, 138+ - 170+ TL); 400 m; Prof. Mesiatzev cr. 15, tr. 52. Coryphaenoides Gunnerus, 1765 Type species: Coryphaenoides rupestris Gunnerus, 1765, by monotypy. DiaGcnosis.—Macrourine grenadiers with 6 branchiostegal rays. Snout shape variable but never greatly produced beyond mouth. Subor- bital ridge variously and usually weakly devel- oped, not connected to preopercular ridge, which in turn is never sharply angular at its posterior end. Dentition variable among species, arranged in broad bands to 1 or 2 rows, but teeth never few and fanglike. Barbel present. Second spinous ray of first dorsal finely serrated along leading edge, although serration occasionally obsolete in large individuals of some species. Gill rakers tu- bercular to short and tablike, those of outer series of first arch rudimentary. Anus immediately in advance of anal fin or slightly anterior to it; no light organ. Pyloric caeca simple, unbranched, usually fewer than 20. Retia and gas glands 4-7. (Adapted from Iwamoto and Sazonov 1988.) Coryphaenoides paradoxus (Smith and Radcliffe, 1912) Macrourus paradoxus Smith and Radcliffe in Radcliffe 1912: 115-116, pl. 25, fig. 1 (off eastern Palawan, Philippines, 2,021 m). Coryphaenoides (Nematonurus) paradoxus: Gilbert and Hubbs 1916:143. ?Nematonurus macrocephalus Maul, 1951:17 (type locality Madeira). ?Coryphaenoides macrocephalus: Marshall and Iwamoto 1973: 575-578 (North Atlantic). Coryphaenoides sp.: Parin et al. 1981:11 (Nazca Ridge). Coryphaenoides paradoxus: Wilson et al. 1985:1,243-1,254 (Darwin Guyot, central North Pacific, about 1,600 m). Iwa- moto and Sazonov 1988:72-75, fig. 3c, 24, 28 (se. Pacific). Parin 1990:16 (listed from Nazca ridge and area between Nazca and Sala y Gomez ridges). SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 53 Diacnosis.—A large species (120+ cm) in subgenus Coryphaenoides with 1D. II,9-11, 1P. 116-121, V. 9-11 (rarely 8). Snout low, scarcely protruding. Orbits relatively small, about 1.5 into snout, 1.8—1.9 into interorbital space, 5.0—6.0 into HL. Interopercle barely exposed as a broad fleshy naked tab. Suborbital region vertical; no strong sharp ridges on head. Mouth large, almost ter- minal, upper jaw extends to below posterior orbit margin. Teeth prominent, an outer enlarged se- ries on premaxillary behind which a narrow vil- liform band; mandibular teeth in about 3 irreg- ular series near symphysis, becoming uniserial posteriorly. Head fully scaled except lips, gill membranes, interopercle. (After Iwamoto and Sazonov 1988.) Size.—A large species attaining more than 120 cm. DIsTRIBUTION. — Broadly distributed in the Pa- cific (Kermadecs, Hawaiian Islands, Nazca Ridge), Indian Ocean [our data], and Atlantic (if C. macrocephalus is a synonym). REMARKS.—The two specimens from NSG were fully described by Iwamoto and Sazonov (1988). In the eastern Pacific, the species is likely to be confused only with C. bulbiceps, which at- tains a similar large size. The stronger teeth, lon- ger barbel, smaller orbit, lower snout, and higher pelvic fin ray count readily distinguish C. par- adoxus from C. bulbiceps. Although we (Iwa- moto and Sazonov 1988:75) previously alluded to the similarity of C. paradoxus and C. rudis Giinther, 1878, we now strongly suspect that the two are the same. We have examined the type specimens of C. paradoxus and C. rudis and found no characters that would suggest specific differ- ence. The scale spinules in the C. rudis lectotype appeared longer and more densely placed, but these features of the spinules tend to be highly variable in most grenadiers. If this synonymy proves correct, C. rudis has priority. We are re- luctant at this point, however, to make this syn- onymy because of the lack of comparative ma- terial from near the type locality. As far as we can determine, only the lectotype of C. rudis is known (the paralectotype is a species of Nezu- mia). McCann and McKnight’s (1980:35) record of the species from off New Zealand stems from misidentification of a specimen representing a perhaps-undescribed species. Brauer’s (1906:246) record from the western Indian Ocean is based on a specimen (ZMB 17641) representing an un- described species of Kuronezumia. MATERIAL EXAMINED. —ZMMGU 16528 (155 HL, 765 TL); Nazca Ridge, 980 m; Jchthyandr cr. 5, sta. 1. ZMMGU 16527 (166 HL, 820 TL); Sala y Gomez Ridge, 1,070-1,100 m; Prof. Mesiatzey cr. 13, tr. 10. Hymenocephalus Giglioli, 1884 Type species: Hymenocephalus italicus Giglioli, in Giglioli and Issel, 1884, by original designation (also monotypic). DraGnosis.— Branchiostegal rays 7. Head bones weak, often paper-thin. Head covering membranous. Ventral striae on abdomen, chest, and shoulder girdle; light organ with two lens- like bodies, one before the anus, the second on the chest anterior to pelvic fin bases, both con- nected externally by a long median-ventral line. DISTRIBUTION. —The genus is primarily trop- ical to warm temperate in distribution and gen- erally absent in high-latitude waters. With five species, the genus (sensu lato, including Spico- macrurus Okamura, 1970a and Hymenogadus Gilbert and Hubbs, 1920) is surprisingly well represented on the Sala y Gomez and Nazca ridg- es. Four species are found off the Hawaiian Is- lands. The genus is absent, however, along the entire mainland Pacific coasts of the Americas (although a single, apparent stray of undeter- mined species was reported by Iwamoto 1979: 140). In contrast, the genus is well represented in the western Pacific. Gilbert and Hubbs (1920) re- corded seven species and three subspecies from the Malay Archipelago (to include their terms “Philippine Islands and East Indies’); Okamura (in Okamura and Kitajima 1984) recorded five species from the Okinawa Trough, and s1x spe- cies and two subspecies (Okamura 1970a) from waters off Japan. The number of species in the Indian Ocean is as yet undetermined, but prob- ably ranges between two and four in the western and central parts, and more from the Malay Ar- chipelago and Australia. No Hymenocephalus species is known from New Zealand; only two have been reported from Australia, although rep- resentatives of other species are deposited in IOAN, ZMMGU, and AMS from warm-water regions of that continent. The Atlantic has only four representatives. Some of the species of Hymenocephalus have wide distributions; others appear to be rather narrowly confined. Hymenocephalus gracilis is an example of a widely distributed species, hav- ing first been described from off the Philippines and subsequently recorded from Japan (Oka- 54 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 FiGureE 19. TL) from off Sala y Gomez Ridge in 580-564 m. Drawn by Y. I. Sazonov. mura 1970a), the Atlantic (Marshall and Iwa- moto 1973), the western Indian Ocean (Iwamoto 1982), and now the southeastern Pacific. The Hawaiian species H. aterrimus has also been re- corded from the western Atlantic (Marshall and Iwamoto 1973) and western Indian Ocean (Shcherbachev 1987); a related form is here re- ported from NSG. Hymenocephalus striatulus, originally described from Hawaii, is now known from Soviet collections from the NSG, and its closest relative appears to be H. billsamorum Marshall and Iwamoto, 1973, from the Gulf of Mexico and Caribbean Sea. Hymenocephalus longiceps Smith and Radcliffe, 1912 and H. stria- tissimus Jordan and Gilbert, 1904, are wide- spread in the western Pacific, from Japan south to Indonesia; the latter is represented by four geographic subspecies, and its closest relative ap- pears to be the new species from NSG. Most of the remaining species have relatively confined distributions, although available collections are still inadequate to circumscribe their limits with certainty. REMARKS. —A distinctive group of about 22 or (a) Hymenocephalus sp. cf. aterrimus. (b) H. neglectissimus, holotype ZMMGU 18219 (18.5 mm HL, 123+ mm more small species, few attaining lengths of more than 25 cm. Hymenocephalus sp. cf. aterrimus Gilbert, 1905? (Figure 19a) ?Hymenocephalus aterrimus Gilbert, 1905:666, pl. 93 (Holo- type: USNM 51649; off Kauai, Hawaii, 385-500 fms; A/- batross sta. 3989). Parin 1990:16 (listed from Nazca Ridge and area between Nazca and Sala y Gomez ridges). D1aGcnosis.—A species of subgenus Papyro- cephalus, with orbit diameter 21-23% HL; in- terorbital width 24-35% HL; outer gill slit 27— 30% HL. Barbel absent. Pelvic fin rays 13. Color mostly blackish. COUNTS AND MEASUREMENTS.—1D. II,9-10; 1P. 113-115; GR-I (outer/inner) 17-19/4 + 17- 19; GR-II 3 + 17-18/3 + 15-16. Total length 84-136 mm; HL 26.5-35 mm. Following di- mensions in percent HL (3 spec.): postrostral 71.4—75.5; snout 26.4—34.3; suborbital 20-21.9; postorbital 48.5-51; orbit-preop. 52.8-55.4; up- per jaw 45.4—48.7; body depth 69-76; light organ 59-68; pre-1D. 104; pre-V. 103-110; pre-A. 147- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 55 160; V.-A. 50-59; base 1D. 31.4-37.9; 1D.-2D. 32.1-40.8; 1P. 70; V. 60. DESCRIPTION. — Trunk and tail strongly com- pressed laterally, dorsal and ventral profiles taper rapidly behind first dorsal fin for about one HL distance. As typical for subgenus, head large and deep, bones generally paper-thin, scale bones over occipital region large and thin. Orbits small, more than 4 into head. Snout high, median nasal ridge notably elevated. Subopercle deep; propercle large, forming most of gill cover, outer edges crenulated; opercle commensurately small; a deep subopercular notch. Preopercular ridges forming a large triangular process at posteroventral cor- ner. Teeth in moderately wide bands in both jaws, the individual teeth uniformly short and stoutly conical with rather blunt tips. Scales all missing, but scale pockets large. Anterior lens of light organ small, round, dif- ficult to see; posterior lens slightly larger, oval, within a broad, teardrop-shaped, black, naked area immediately before anus, as typical of ge- nus. Black line connecting the two lenses poorly defined. Ventral striae generally as described for others of genus, but not as well defined and much more difficult to see. Pectoral and pelvic fins about on same vertical; first dorsal origin slightly behind that; anal fin begins midway below long interspace between dorsal fins. Pectoral and pelvic fins rather long, both extending slightly past anal origin. Color overall black or swarthy, darkest over trunk and ventral aspects of fish; tail mostly dark brown. A broad midlateral band of heavier me- lanophores beginning above midlateral portion of trunk, extending posteriorly onto and even- tually completely including tail. Paired and first dorsal fins with dark rays and pale interradial membranes; anal fin pale but bases of rays black. Floor of buccal cavity below tongue forming a black triangle; tongue, however, pale dorsally with only a few scattered large melanophores on ven- tral surface. Roof of mouth with splotches of black. SizeE.—To at least 130 mm. DIsTRIBUTION.— Known only from the Nazca and Sala y Gomez ridges. COMPARISONS AND REMARKS.—The NSG specimens are similar in many characters to H. aterrimus and H. papyraceus. Using the key pro- vided by Gilbert and Hubbs (1920:520), our four specimens key out to H. aterrimus, and most meristic and morphometric features agree with the original and Marshall and Iwamoto’s (1973: 607) descriptions of that species. However, com- parison of the four with Hawaiian paratypes (CAS-SU 8513) of H. aterrimus, four Atlantic representatives (CAS 14514), and others from the Atlantic and Indian oceans deposited in IOAN and ZMMGU suggests that the head is too deep and narrow, and the median nasal ridge is too high to be that species. The poor condition of the specimens, however, might account for the apparent differences. We also compared our ma- terial with a 32.4 mm HL specimen of H. papyra- ceus (ZMMGU 18258) taken in the East China Sea off southern Japan. Among other characters, H. papyraceus differs in having a smaller barbel, fewer pelvic fin rays (11), larger orbit, narrower suborbital, and longer upper jaw. With more and better specimens, it should be possible to better compare and quantify these apparent differences, but for now, it seems best to leave the identity as questionably H. aterrimus. MATERIAL EXAMINED.—ZMMGU 17729 (2:33-34.3 mm HL, 125-131+ mm TL) and CAS 75792 (34 HL, 134 TL); Sala y Gomez Ridge, 1,070—1,100 m; Prof. Mesiatzev cr. 13, tr. 10. ZMMGU 18128 (26.5 HL, 84+ TL); Nazca Ridge, 340-780 m; Prof. Mesiatzey cr. 13. Hymenocephalus gracilis Gilbert and Hubbs, 1920 (Figure 20a) Hymenocephalus gracilis Gilbert and Hubbs, 1920:522, fig. 31 (Holotype: USNM 78227, Albatross sta. 5292, 162 fms, off Luzon, Philippines). Marshall and Iwamoto 1973:602, fig. 31 (N. Atlantic spec.). Parin 1990:16 (listed from Nazca and Sala y Gomez ridges). Hymenogadus gracilis: Okamura 1970a:61, pl. 17, fig. 27 (Ja- pan); 1984:201, 359, fig. 143 (E. China Sea). DIAGNOosTIC DESCRIPTION. — Body low, terete, gradually tapering posteriorly; tail laterally com- pressed. Head low, its length 18-22% TL; ridges low and poorly developed. Snout low, rather long, 1.3-1.7 in orbit (usually 1.3-1.4, relatively short- er in juveniles), pointed and protruding well be- yond mouth. Orbit oval, 29-41% HL, upper margin reaching upper profile of head. Maxillary extends to vertical through posterior margin of orbit. Nasal and medial rostral ridges conspic- uous; suborbital ridge developed, dividing sub- orbital region into upper and lower portions, the former very narrow. COUNTS AND MEASUREMENTS. —(30 NSG spec.) 1D. II,9-11; 1P. 115-118; V. 8 (1 spec. with 8/7); 56 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Ficure 20. Hymenocephalus species from Sala y Gomez Ridge, collected on the 18th cruise of Prof. Shtokman. (a) Hy- menocephalus gracilis, ZMMGU 17740 (22 mm HL), sta. 1941, in 410-385 m, 28 April 1987. (b) Hymenocephalus striatulus, ZMMGU 17715 (27.8 mm HL), sta. 1970, in 540-560 m, 1 May 1987. Photographs by Sergei Dudarev. GR-I (outer/inner) 7-11/3-4 + 14-16 (17-18 total), GR-II 14-18 total. Total length 58+-125 mm; HL 11.8-24 mm. Following in percent HL: postrostral 73-82; snout 22.6-27.9; preoral 11.1-17.3; orbit 29.2—40.7; interorbital 14.7—22.6; suborbital 7.5—10.6; post- orbital 28.8-36.5; orbit-preop. 32.6—43.3; upper jaw 42.3-49.6; gill slit 20.8-29.8; barbel 23.7- 32.4; body depth 47-67; light organ 54-70; pre- 1D. 98-113; pre-V. 93-108; pre-A. 143-163; V.- A. 49-66; 1D. height 47-77 (6 spec.); base 1D. 31-45; 1D.-2D. 30-87; 1P. 35-56; V. 44-55. SizE.—To 130 mm. DISTRIBUTION. —South China Sea off Luzon, Japan, East China Sea, southeastern Pacific, tropical western North Atlantic, eastern Atlantic off Morocco, and Indian Ocean off Zanzibar (our unpublished data). COMPARISONS AND REMARKS.—This species seems to have a preference for island areas. The capture of three specimens in midwater trawls fished well off bottom suggests occasional forays into the bathypelagic realm. Hymenocephalus tenuis Gilbert and Hubbs, 1917, is a closely re- lated species known only from the immature ho- lotype collected in the Hawaiian Islands. Differ- ences between the two species are slight (see Gilbert and Hubbs 1920:520) and should be re- evaluated when more specimens of H. tenuis be- come available. MATERIAL EXAMINED. —(39 spec., 7 sta.) Sala y Gomez Ridge (eastern part): ZMMGU 17735 (20.7 mm HL, 98+ mm TL); 390-385 m; Prof. Mesiatzev cr. 15, tr. 53. ZMMGU 17736 (10.9 HL, 56+ TL); 300-0 m over bot. depth >2,000 m; IKMT; Prof. Shtokman cr. 18, sta. 1887. ZMMGU 17737 (about 6.2 HL, 28 TL); 300-0 m over bot. depth >2,000 m; IKMT; Prof. Shtokman cr. 18, sta. 1925. ZMMGU 17738 (2: 11.8-16.8 HL, 58+-93+ TL); 380 m; Sigsbee trawl; Prof. Shtokman cr. 18, sta. 1938. ZMMGU 17739 (10:19-23.2 HL, 80+-120 TL) and CAS uncat. (10:11.3—23 HL, 64-124 TL); 380 m; Prof. Shtokman cr. 18, sta. 1940. ZMMGU 17740 (8: 18-24 HL, 84-120 TL) and CAS uncat. (5:19-23 HL, 97-116 TL); 410-385 m; Prof. Shtokman cr. 18, sta. 1941. IOAN uncat. (12.0 HL, 58.5 TL); 200-0 m over bot. depth >2,000 m; IKMT; Prof. Shtokman cr. 18, sta. 1949. Hymenocephalus neglectissimus new species (Figures 19b, 21) No literature applies to this species. DiaGnosis.—Orbits large, greatest diameter about 44-55% HL; interorbital 20.8—27.4%; bar- bel thin, 15-28%; GR-I (lower limb, inner series, including raker at angle) 13-17, rarely 18. Body depth over anal origin 47-58% HL. Pectoral rays 112-114 (rarely 115); V. 8. Black blotch on dor- sum with clear-cut outlines; few isolated or no pigment cells above its posterior projection, few or none behind it, and few small widely spaced dots below (on sides of abdominal region). COUNTS AND MEASUREMENTS. —(see also D1- AGNOsIS) (data on holotype in square brackets) 1D. II,7-8 (rarely 9) (x = II,7.9, n = SO) [II,8]; SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES >) 9 A wb A oP ® D IL at Li ate 7 F =F ‘wal gee e = 6 rep + = L 9 mm tt ¢ L Sie ou e° Z Oo e @ O oe ad 0 EP oe® - ® eyo Peo ® e e on 8 8 iit dis Be] [Ea] (igs 5p Aes ! aT | 1 a a 18 21 24 27 30 33 HEAD LENGTH (mm) Ficure 21. Scatter diagram showing relationship of interorbital width to head length in H. semipellucidus (solid circles), H. neglectissimus (squares), H. s. striatissimus (plus signs), H. s. aeger (circled plus signs), and H. hachijoensis (triangles). Data for Hymenocephalus s. aeger from Gilbert and Hubbs (1920), for H. hachijoensis from Okamura (1970a). 1P. 112-114 (rarely 115) (% = 12.8, n = 92); GR-I (outer) 11-15 (* = 12.9, n = 51) [14]; GR-I (inner) (2-3) + (0-1) + (13-17, rarely 18), total 16-21 (X = 18.2, n = 51) [3 + 0 + 15]; GR-II (outer) (2, rarely 3) + (1, rarely 0) + (13-16), total 16-19 (* = 17.4, n = 51) [2 + 1 + 15); GR-II (inner) (2-3) + (0-1) + (13-15), total 16- 19 (x = 17.1, n = 51) [3 + 1 + 14]; caeca 10- He (X—= 10.3, n = 12). Total length 78 +-125 mm; HL 12.5-19.2 mm (X = 16.5, n = 29) [18.5]. The following in per- cent of HL: postrostral 77.9-89.6 (x = 85.4, n = 28) [86.5]; snout 16.4—23.8 (% = 19.6, n = 27) [17.3]; preoral 10.4—-18.9 (X = 14.8, n = 15) [13.0]; postorbital (greatest length) 37.5-43 (x = 40.8, n = 8) [37.8], (least length) 26.7—38.6 (X = 33.8, n = 28); suborbital 5.2-9.5 (%* = 6.9, n = 29) [7.0]; orbit-preop. 28.1-37.3 (% = 32.3, n = 28) [28.6]; orbit (max.) 44.8-54.8 (x = 49.9, n = 29) [49.7], (horizontal) 44.6-49.7 [49.7]; upper jaw 53.3-58.8 (x = 56.1, n = 28) [55.7]; interorbital 20.8-27.4 (X = 23.9, n = 26) [21.1]; gill slit 31.1- 36.9 (x = 33.3, n = 16) [32.4]; barbel 15.3-28.0 (x = 21.5, n = 28) [18.9]; pre-A. 146-166 (x = 155, n = 28) [162]; pre-V. 92-113 (* = 105, n = 28) [109]; V.-A. 47.5-68 (% = 587, n = 28) [60.5]; body depth 66-88 (x = 77, n = 46) [78], depth over A. origin 47-58 (X = 54, n = 24) [55]; light organ 65-89 (x = 80, n = 24) [76.8]; height 1D. 75-92 (x = 83, n = 9); 1D.-2D. 60.1-90.4 (xX = 72.4, n = 26) [73.5]; length 1P. 63-92 (x = 77, n = 24) [80]; length V. 71 (n = 1). DESCRIPTION. — Head 6.5-—7.6 into TL (21 specimens); greatest body depth usually some- what less than postorbital length of head, the trunk tapering more rapidly to origin of anal fin 58 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 than beyond in adults, but tapering gradually in smaller fishes. Head moderately broad, its great- est width notably less than depth of head over midorbit. Orbit circular (rarely greatest diameter oblique); both maximum and horizontal diam- eters of orbit exceed maximum postrostral length of head. Snout short, rather low, bluntly round- ed, not produced beyond anterior tip of pre- maxillary. Dentition consists of small, conical, slightly recurved teeth in 2-4 irregular rows on premax- illary and 1-3 on dentary (one row on short pos- teriormost part of tooth band adjacent to coro- noid process). Luminescent organ as illustrated for H. s. stria- tissimus by Okamura (1970b, fig. 74A), but ex- ternally both lenses seem somewhat larger and lack pigmentation on surface. Alimentary canal short and simple with only 2 bends (see Okamura 1970b, fig. 62A). Pyloric caeca short, 10-11. Scales on all specimens now lost. Fins weak, often broken, especially pelvics. Pectoral and pelvic (in one specimen only) fins slightly produced beyond anal fin origin, extend- ing to about 4th to 8th anal ray. Spinous second ray of first dorsal fin with a short threadlike pro- longation; its overall length about equal to post- rostral length. Second dorsal rudimentary over almost entire length, its origin above 13th to 16th anal ray. Color superficially resembles that of H. semi- pellucidus (see below) or H. striatissimus (as il- lustrated by Okamura 1984), but with numerous constant differences. Blackish blotch on dorsum darker and with clear-cut outlines (see Fig. 19b); its posterior projection rather short with isolated (if any) small pigment cells above, very few be- hind (posterior end of projection appears some- what eroded), and few pigment cells below pro- jection (some in short midlateral series); pigment rarely passes from trunk to tail beyond vertical of 5th to 7th anal fin ray. Tail appears transparent over almost entire length. Dorsalmost 5-9 cer- atobranchial rakers on second gill arch blackish; those more ventral whitish; lateral rakers on first arch all unpigmented. Stomach light to dark gray. SizE.—To at least 130 mm. DISTRIBUTION. — Known only from the Sala y Gomez Ridge in 525-600 m. ETYMOLoGy.—Latin neglectus (neglected or unnoticed) is used in the superlative, neglectis- simus, to reflect the late discovery of the species. Hymenocephalus semipellucidus and H. neglec- tissimus were first collected together in 1983, but the latter was not recognized as being distinct from the former until 1990. COMPARISONS AND REMARKS. — Hymenoceph- alus neglectissimus is very similar to H. semipel- lucidus from NSG and to the species in the H. striatissimus complex from the tropical western Pacific. It is partially sympatric with H. semipel- lucidus, the two having been collected together at four localities bounded by latitudes 25°00'S and 25°30’'S and longitudes 88°30’W and 89°00'W. Superficially, the two NSG species of Hymeno- cephalus may be easily and faultlessly separated by their color patterns. Hymenocephalus neglec- tissimus differs from members of H. semipelluci- dus and also the western Pacific H. striatissimus complex in having a more contrasting outline of the dark dorsal blotch, with a rather short pos- terior projection that is somewhat eroded cau- dally and surrounded by few or no isolated pig- ment cells. In contrast, H. semipellucidus and H. striatissimus have a dorsal blotch with eroded outlines surrounded by numerous, relatively large pigment cells and a relatively long posterior pro- jection to the blotch that extends well back (pos- teriorly) onto the tail. Hymenocephalus neglec- tissimus also differs in having unpigmented lateral gill rakers on the first arch. It further differs from H. semipellucidus in having fewer pectoral fin rays (112-114 vs. 114-118) (Table 3), a somewhat shorter snout, wider interorbital (Fig. 21), nar- rower suborbital, slightly longer upper jaw, and longer barbel, but considerable overlap is seen in these mensural features. Body depth in H. neglectissimus appears to decrease more rapidly posteriorly than it does in H. semipellucidus; the difference between the maximum body depth in, and the depth at anal origin of, 23 specimens varied from 1 1.2—33.1% (x = 22.7) in the former, compared with 5.5—27.9% (x = 16.7) in 18 spec- imens of the latter. In this respect, H. neglectis- simus more closely compares with H. striatissi- mus (14-29%, usually 22-26%). The new species differs from H. striatissimus (including subspecies striatissimus, aeger, and torvus, but excluding H. hachijoensis) in having a somewhat shorter snout, larger orbit, shorter postorbital, and narrower interorbital (see Fig. 21). Gill rakers are also fewer in H. neglectissi- mus, but this character shows considerable over- lap (see Table 3). Hymenocephalus neglectissimus is a smaller SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 59 TABLE 3. Selected counts of pectoral fin rays and gill rakers of the Hymenocephalus striatissimus complex. Pectoral fin rays 12 13 14 15 16 17 18 19 x SD n H. semipellucidus — - 2 13 46 29 U 1 16.30 0.899 98 H. neglectissimus 21 41 29 1 _ - — - 13.11 0.763 92 H. striatissimus: sp. Tasman Sea - 1 1 _ — _ = = 13.50 = 2 Timor and Coral seas 2 3 _ 1 — _ — _ 13.00 1.095 6 Molucca Sea 1 1 4 1 — = = _ 13.71 0.951 7 Sulu Sea 1 4 6 9 — _ = _ 14.15 0.933 20 S. China Sea 1 1 _ _ _ = = 13.50 — 2 Japan 3 1 _ » Z — — - 13.88 1.808 8 Inner gill rakers, first arch, lower limb 13 14 15 16 i7/ 18 19 Xx SD n H. semipellucidus 1 16 25 9 _ _ _ 14.82 0.740 51 H. neglectissimus _ 6 27 16 1? 1? = 15.29 0.782 51 H. striatissimus: sp. Tasman Sea _ _ _ 1 _ oa — 16.00 _ 1 Timor and Coral seas _ _ _ 1 — 1 2 18.00 _ 4 Molucca Sea — - — 2 3 _ = 16.60 = 5 Sulu Sea - - _ — 7 3 2 17.58 0.793 12 S. China Sea _ — — _ 1 _ _ 17.00 — 1 Japan _ _ — _ - 2 1 18.22 _ 3 Total gill rakers, second arch 15 16 il7/ 18 19 20 21 22 23 Xs SD n H. semipellucidus 5) 19 17 8 1 1 — a — 16.69 1.049 51 H. neglectissimus 1 2 24 21 3 — _ - _ WES) 1.647 51 H. striatissimus: sp. Tasman Sea — _ 1 _ _ — _ — _ 17.00 - 1 Timor and Coral seas - — _ 1? — — 1 2 1 21.20 _ 5 Molucca Sea = = = 3 2 _ — _ _ 18.40 - 3 Sulu Sea _ — a _ 2 1 6 2 1 20.92 - 12 S. China Sea = _ _ _ = _ 1 = ~ 21.00 _ 1 Japan — — = _ — — 3 _ = 21.00 = 5 Total inner gill rakers, second arch 15 16 17 18 19 20 21 22 X SD n H. semipellucidus 6 20 19 6 a _ _ — 16.49 0.857 51 H. neglectissimus — 13 7) 13 3 — _ _ ile 0.864 51 H. striatissimus: sp. Tasman Sea = 1 — _ — o = _ 16.00 _ l Timor and Coral seas = _ — _ 1 -_ 2» _ 20.33 = 3) Molucca Sea — 2 1 2 a — - - 17.10 5 Sulu Sea _ _ 1 1 6 22 1 - 19.09 1.044 11 S. China Sea — — _ — — 1 — — 20.00 — 1 Japan — = _ _ — 1 1 1 21.00 _ 3 species (maxima at about 130 mm TL and 20 ‘1984, and our material). It may well be the small- mm HL) than H. semipellucidus (130-160 mm __ est macrourid species. TL and 20-26.5 mm HL) and H. striatissimus MATERIAL EXAMINED. —(all from Sala y Gomez Ridge) Ho- (150-180 mm TL, 25-28 mm HL; data from jotype: ZMMGU 18219 (18.5 mm HL, 123+ mm TL); 580- Okamura et al. 1982, Okamura and Kitajima 564 m:; Prof. Shtokman cr. 18, sta. 1964. 60 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 = Ficure 22. Hymenocephalus semipellucidus new species; paratype, CAS 75978 (26 mm HL), from Sala y Gomez Ridge in 565-555 m, Prof. Mesiatzey cr. 13, tr. 1, 1 Sept. 1983. Drawn by Tomio Iwamoto. Scale bar equals 10 mm. Paratypes (marked with an asterisk *) and other material (270 specimens from 8 localities). ZMMGU 18220 (30* + 94: 15.5-19.6 HL, 60+-130 TL); data as for holotype. ZMMGU 17724 (55:12.6-18.2 HL, 90-128 TL); 562-545 m; Prof. Shtokman cr. 18, sta. 1965. ZMMGU 17726 (25:14.2-18.5 HL, 85-115 TL); 545-600 m; Prof. Shtokman cr. 18, sta. 1977. ZMMGU 18140 (20* + 20:12.5-18.7 HL, 78+-125 TL); 525- 530 m; Prof. Mesiatzev cr. 15, tr. 49. ZMMGU 18217 (17.5 HL, 99+ TL) and CAS 75976 (16.3 HL, 116+ TL); 565-555 m; Prof. Mesiatzev cr. 13, tr. 1. ZMMGU 18218 (7:16.5-19 HL, 86+-115 TL); 550-560 m; Prof: Mesiatzey cr. 13, tr. 2. ZMMGU 18221 (27:14.2-18.7 HL, 80+-123 TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. Hymenocephalus semipellucidus new species (Figures 21-23) **Hymenocephalus sp. nova Sazonov et Iwamoto”: Parin 1990: 17 (listed from Sala y Gomez Ridge). Hymenocephalus striatissimus (non Jordan and Gilbert, 1904): Parin et al. 1990:42 (stomach contents). Kotlyar and Parin 1990:106 (otolith). DiAGnosis.— Orbits large, greatest diameter about 43-51% HL; interorbital 16.5—22.1% HL; barbel thin, 13-22% HL; GR-I (lower limb, inner series) 14-16 (rarely 13). Body depth over anal fin origin 48-64. Pectoral rays 114-118 (rarely i119); V. 8. COUNTS AND MEASUREMENTS. — (holotype data in square brackets) 1D. II,7-9 (x = 8.1, n = 51) [11,8]; 1P. 114-119 (x =116.5, n = 98) [117]; GR-I (outer) 9-15 (x = 11.6, m= 51) [13]; GR-I (inner) (2-4) + (0-1) + (13-15), total 16-20 (x = 17.9, n=51)[3 + 0 + 14]; GR-II (outer) (1-4) + (O- 1) + (12-16), total 15-19 (x = 16.7, n = 51) [2 + 1 + 13]; GR-II (inner) (2-3) + (0-1) + (12- 15), total 15-18 (x = 16.5,n = 51) [2 + 1 + 13]; caeca 9-13 (x = 11.4, n = 17). Total length 102+-165 mm; HL 17.7-26.5 mm. The following in percent of HL: postrostral 76.9-86.8 (x = 81.9, n = 25) [79.6]; snout 20.9- 27.0 (x = 24.6, n = 25) [24.9]; preoral 11.8-17.4 (x = 14.1, n = 21) [15.1]; orbit max. 42.9-50.5 (x = 45.7, n = 26) [49.0]; orbit horiz. 37.0-48.2; interorbital 16.5-22.1 (x = 19.6, n = 25) [20.4]; postorbital (greatest) 38.3-45.1 (x = 42.0, n = 24) [40.8]; postorbital (least) 33.1-44.9 (x = 37.9, n = 24) [37.1]; suborbital 8.5-11.8; orbit-preop. 30.2-39.6 (x = 35.5, n = 25) [35.9]; upper jaw 49.2-56.8 (x = 52.6, n = 25) [52.2]; gill slit 26.7- 34.5 (x = 31.4, n = 16) [31.4]; barbel 13.0-—22.0 (x = 17.7, n = 26) [15.5]; body depth (max.) 61- 85 (x = 72, n = 26) [66]; depth over A. orig. 48.2-63.9 (x = 54.9, n = 18) [53.1]; pre-A. 152- 173 (x = 162, n = 26) [165]; pre-V. 97-111 (x = 103, n = 26) [97]; V.-A. 53.6-81.8 (x = 69.0, n = 26) [80.8]; height 1D. 66-96 (x = 77, n= 17) [78]; 1D.-2D. 67.6-88.9 (x = 77.5, n = 24) [71.8]; 1P. 58-78 (x = 70, n = 25) [76.7]; V. 62- 78 (x = 71, n = 22) [67]; light organ 71.9-92.6 (x = 80.8, n = 18) [77.6]. DESCRIPTION.—Head 6-7 into TL; greatest body depth usually somewhat less than postros- tral length of head, the trunk tapering gradually to and beyond origin of anal fin. Head rather broad, its greatest width slightly less than depth of head over midorbit. Greatest diameter of orbit oblique and about equal to or somewhat more than postorbital length of head (taken from orbit margin to uppermost posterior angle of opercle). Snout rather low, bluntly rounded, not produced beyond anterior tip of premaxillary. Dentition consists of small, conical, slightly recurved teeth in 1-3 irregular rows in both jaws in most specimens (a few individuals have pre- maxillary bands 4 or 5 teeth wide). Luminescent organ as illustrated for H. s. stria- tissimus by Okamura (1970b, fig. 74A). Alimen- tary canal short and simple, with only 2 bends (see Okamura 1970b, fig. 62A). Pyloric caeca short, thick; 9-13 in 17 specimens. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 61 Few scales remain, almost all of these lacking spinules. A few isolated scales with short conical spinules as illustrated for genus by Okamura (1970b, fig. 2). Fins rather weak, tips of first dorsal broken off in all specimens; pectoral and pelvic fins extend posteriad to level of 4th to 9th anal fin ray. Sec- ond dorsal rudimentary over almost entire length, its origin above 14th to 16th anal ray. Color description given by Gilbert and Hubbs (1920:533-534) for H. striatissimus aeger accu- rate for new species as well—this and other char- acters suggest close affinity of the two species. Black blotch on dorsum with eroded outlines, surrounded by large pigment cells with numerous branched projections; posterior projection of blotch variable in length (extends either to level of anal-fin origin or to vertical of origin of 13th— 18th anal ray). Pigment cells surrounding blotch numerous below and behind, fewer (rarely ab- sent) above. Arrangement of these cells varies greatly: from irregularly scattered to grouped along myosepta (more densely along those going dorsad and posteriad above lateral line, and ven- trad and anteriad below blotch). Some cells may be arranged in a line going posteriad along mid- lateral myosepta. Dark pigment attains level of 9th to 25th anal ray. Dorsalmost 9 or so cera- tobranchial rakers on second gill arch blackish, lateral rakers on first arch also blackish (some- times pale on anteriormost few); those more ven- tral pale. Stomach blackish. Hymenocephalus semipellucidus and Russian specimens from Japan of H. striatissimus do not show the silvery sheen and black striations as prominently as in A/batross specimens of latter species, but differences probably an artifact of preservation. Alcohol-fixed specimens retain sil- very appearance, whereas those fixed in formalin lose silvery reflections. Thick translucent skin covers abdominal region of NSG specimens, but careful teasing and stripping away of epidermis reveals striations underneath. Black melano- phore pattern over body of new species promi- nent (see Fig. 22). Similar dark pigment patterns may have been present in H. striatissimus spec- imens examined, but specimens currently faded and show only faint traces of such patterns. Foop.—Parin et al. (1990:42) found the spe- cies to feed predominantly on copepods (includ- ing Pleuaromamma sp., Gaussia scotti, Phyllopus mutatus, Chirundina streetsi, Arietellus sp., and Aetideidae sp.). Benthic polychaets (including Polinoidae) and the gonostomatid fish Cyclo- thone were also important. S1zE.—To at least 165 mm. DISTRIBUTION. — Known only from the Sala y Gomez Ridge in 550-800 m. ETyMo.Locy.—From Latin semi, half, and pel- lucidus, clear or transparent, in reference to the partially transparent head covering, and the translucent caudal region where the vertebra can be seen in fresh specimens. COMPARISONS AND REMARKS.—The new spe- cies keys out to H. striatissimus using Gilbert and Hubbs’s (1920:520) key, but it differs from that species in a number of important features. The gill-raker counts, notably the inner series in the first (outermost) arch, show the best sepa- ration of the two species (see Table 3) in our material. Note from the table that gill raker counts in our H. striatissimus specimens suggest a clinal difference in populations of that species, with the Molucca Sea specimens, representing H. s. aeger, completely separated from the typical subspecies H. s. striatissimus of Japan, and the Sulu Sea and South China Sea specimens intermediate in their counts. South China Sea specimens were consid- ered by Gilbert and Hubbs (1920:531) to be an intergrade between H. s. torvus and H. s. stria- tissimus. In gill-raker counts, ZMMGU specimens of H. striatissimus from the Coral and Timor seas ap- pear closest to H. striatissimus specimens from the Sulu Sea. A small specimen (ZMMGU 18260) from the Tasman Sea, however, is almost indis- tinguishable from H. semipellucidus and H. ne- glectissimus, differing only in having a distinct color pattern similar to that of H. striatissimus and slightly different morphometry from that of H. semipellucidus. Pectoral ray counts in the new species tend to be higher than those of H. striatissimus, but there is considerable overlap in this feature (Table 3). The suborbital region appeared to be narrower in the new species, and a scatter diagram plotting measurements of the least width against the greatest diameter of the orbit (Fig. 23) showed a good separation between the Malay Archipelago specimens of H. striatissimus on the one hand, and Japan specimens of that specimens, and H. semipellucidus specimens on the other hand. The separation of the Japanese specimens of H. stria- tissimus from other specimens of that species lends additional support for continued recogni- tion of subspecies. A more extensive study may 62 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 SUBORBITAL WIDTH (mm) ORBIT DIAMETER (mm) FiGure 23. Scatter diagram showing relationship of suborbital width to orbit diameter in Hymenocephalus semipellucidus (circles), H. s. striatissimus from Japan (triangles), and H. striatissimus aeger from the Malay Archipelago (plus signs). necessitate elevation of two or more of the sub- species to species level. The interorbital width is distinctly narrower in H. semipellucidus, ranging 16.5—22% HL, as compared with about 28% HL in Japanese spec- imens of H. striatissimus, and 28-39% in Malay Archipelago specimens. Although there is con- siderable overlap in this character between H. semipellucidus and H. neglectissimus, the scatter diagram (Fig. 21) comparing the interorbital width of five populations of Hymenocephalus shows a distinct separation between the two spe- cies. The new species also appears to have a more gradually tapered body than does H. striatissi- mus. This is best reflected in a comparison of the body depth below the origin of the first dorsal fin (greatest body depth) and the depth over the anal fin origin. In the new species, the first mea- surement ranged about 61-85% HL, the second 48-64% HL, with a difference in each specimen of 5.5—27.9% between the two proportions. In contrast, the H. striatissimus specimens had body depths of 55-81% HL, and 37-56% HL, with a difference of 14-29% (most examples at 22-26%). Differences between H. semipellucidus and H. neglectissimus were discussed under the descrip- tion of the latter species. Hymenocephalus semi- pellucidus differs from the three subspecies of H. striatissimus (H. s. stratissimus, H. s. aeger, and H. s. torvus) more than each differs one from another. The fourth subspecies, H. s. hachijoen- sis, from Japanese waters, may represent a sep- arate species intermediate (as Okamura 1970a suggested) between H. striatissimus and H. lon- giceps. We have examined three specimens of H. s. hachijoensis (ZMMGU 18243 and 18244 from the North-West (=~Emperor) Ridge and ZMMGU uncat. from the Kyushu-Palau Ridge) and found that they differ from the three other subspecies of H. striatissimus (as well as the two NSG spe- cies) in color pattern and barbel length, which would seem sufficient to elevate H. hachijoensis to species level. MATERIAL EXAMINED. —(103 spec. from 7 sta.) Sala y Gomez Ridge: Holotype: ZMMGU 18129 (24.5 mm HL, 152 mm TL); 750 m; Prof. Shtokman cr. 18, sta. 2019. Paratypes (40 spec.): ZMMGU 17723 (7:20.0—24.8 HL, 138— 162 TL); 580-564 m; Prof. Shtokmancr. 18, sta. 1964. ZMMGU 17725 (14:17.7-23.3 HL, 106+-150 TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. ZMMGU 17728 (3:22.0-22.8 HL, 104+ -138 TL); data as for holotype. ZMMGU 17721 (4:19.5- 23 HL, 91+-149 TL) and CAS 75978 (2:20.0-26.5 HL, 137- 165 TL); 565-555 m; Prof. Mesiatzev cr. 13, tr. 1. ZMMGU 17722 (6:18-21 HL, 105+-132+ TL) and CAS 75977 (4:22- 23 HL, 108+-142+ TL); 550-560 m; Prof. Mesiatzev cr. 13, lime Other material: ZMMGU 17727 (53:19.5-—26.5 HL, 96+- 164 TL); 730-790 m; Prof. Shtokmancr. 18, sta. 2018. ZMMGU 18205 (9:22.5-26 HL, 121-184 TL); 750-800 m; Prof. Shtok- man cr. 18, sta. 1996. Specimens of Hymenocephalus striatissimus used for com- parisons. —Coral Sea: ZMMGU 18259 (2:23.7—26.5 HL, 125+ - 147 TL); 18°48'S, 149°58’E; 660 m; Lyra tr. 20; 14 Jun. 1968. Timor Sea: ZMMGU uncat. (3: ca. 22—25.2 HL, 110+-146 TL); 9900'S, 130°38.8’E; 445-520 m; Akademik Berg tr. 553. Tasman Sea: ZMMGU 18260 (questionably identified as H. striatissimus) (15.1 HL, 97.5 TL), 34°17.8'S, 171°30.9’E; 670- 630 m; Dmitry Mendeleev cr. 16, sta. 1265; 5 Jan. 1976. JA- PAN. CAS-SU 8549 (paratype of H. striatissimus: 25 mm HL, 110+ mm TL); Suruga Bay; A/batross (no other data). Molucca Sea: CAS-SU 25463 (8 paratypes of H. s. aeger: 11.2—26.5 HL, 69-146 TL); 00°15'N, 127°24'36”’E: 545 m; Albatross sta. 5621; 28 Nov. 1909. CAS 57180 (3:25.1-28.1 HL, 160-178 TL); Suruga Bay, off Heda [Heta]; shrimp trawl; 18 Feb. 1969. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 63 PHILIPPINES. CAS-SU 25620 (18.8 HL, 111 TL); near Jolo, 6°02'55’N, 120°53’E; 186 fms [340 m]; Albatross sta. 5173, 5 Mar. 1908. CAS uncat. (14.7 HL, 93 TL); off n. Luzon, 18°29'45”N, 121°39’E; 150 fms [274 m]; Albatross sta. 5328, 19 Nov. 1908. CAS-SU 25464 (9.9-13.5 HL, 50-88 TL); be- tween Jolo and Tawi Tawi, 5°48'12”N, 120°30'48”E; 224 fms [410 m]; Albatross sta. 5563, 21 Sep. 1909. Specimens of H. hachijoensis used for comparisons: Em- peror Seamounts: ZMMGU 18243 (32.7 HL, 205 TL); ca. 32°N, 173°E; depth unknown; Mys Unony tr. 86; Sep. 1979. ZMMGU 18244 (26.2 HL, 163+ TL); 41°904’N, 170°32’E; 1,050-1,060 m; Mys Unony tr. 103; 22 Sep. 1979. Kyushu- Palau Ridge. ZMMGU uncat. (ca. 29.5 HL, 160+ TL); no exact data on catch; Prof. Deryugin; 1971. Hymenocephalus striatulus Gilbert, 1905 (Figure 20b) Hymenocephalus striatulus Gilbert, 1905:665-666, fig. 259 (type-locality Hawaii, off SW coast Oahu; 192-352 fathoms [351-644 mJ). Parin 1990:16 (listed from Sala y Gomez Ridge). Parin et al. 1990:41-42 (stomach contents). Kotlyar and Parin 1990:104, fig. 3d (otolith). Hymenocephalus sp.: Parin et al. 1981:11-12 (67 specimens from Sala y Gomez Ridge). DriaGnosis.—A species of subgenus Hymeno- cephalus, with snout rather low, pointed, pro- jecting beyond mouth, about as long as inter- orbital or longer; orbit diameter 35-43% HL, much greater than interorbital width; suborbital narrow, 9—15% HL; barbel rudimentary (1.5 or more into least suborbital width) or obsolete. Pectoral rays 114-120; V. 14-15. CouUNTS AND MEASUREMENTS.—(from more than 100 Sala y Gomez specimens; mean values in square brackets) 1D. II,8—-11 [x = II,9.5; 1P. 114-120 [116.9]; V. 14-15 [14.5; GR-I [outer/in- ner] 17-23 [20.2]/(4-6) + (21-25) [5.0 + 1 + 22.5], GR-II (3-5) + 1 + (20-24) [4.4 + 1 + 22.0]/(3-5) + 1 + (19-23) [4.2 + 1 + 20.9]. Total length 103-180; HL 28.5-34.5 mm. The following in percent of HL: postrostral 73.3-81.2 [76.9]; snout 23.2-30.5 [27.5; orbit 34.7-42.8 [39.1]; interorbital 21.2—29 [24.9]; postorbital 33— 42 [37.0]; orbit-preop. 34.2-40.6 [37.2]; subor- bital 9.1-14.9 [10.9]; upper jaw 47.2—58.8 [51.6]; barbel 3.1—12; gill slit 29.2—37 [33.3]; pre-D. 95- 106 [102]; pre-A. 138-152 [146]; pre-V. 95-113 [102]; V.-A. 43.8-59.1 [50.8]; 1D. base 28.6- 36.8 [32.3]; 1D. height 52-75 [63]; 1D.-2D. 35.3- 73.4 [55.2]; 1P. 48-61 [53]; V. 55-84 [70]; light organ 51.5-62.5 [56.7]; body depth 58-70 [65]. DESCRIPTION.— Body slender, subcylindrical, greatest width over pectoral bases about three- fourths greatest depth; the trunk tapering grad- ually posteriorly to tail tip. Head low and broad, greatest width about equal to or more than its depth. Head bones somewhat stouter than most others of genus, and mucous cavities less devel- oped—these conditions somewhat intermediate between those of subgenus Hymenogadus and Hymenocephalus. Head covering transparent. Orbit large, its greatest diameter on a diagonal and slightly shorter than postorbital length. In- terorbital space moderately broad, although sharply narrowed forward of midorbit level. Pre- opercle ridges form a large triangular process at posteroventral corner. Snout moderately pointed and protruding beyond large mouth. Upper jaws extend posteriorly to hind edge of orbit. Teeth in moderately wide bands in both jaws, the individual teeth uniformly short and stoutly conical, with rather blunt tips. Small round anterior lens of luminescent organ in middle of chest, but difficult to discern; pos- terior lens slightly larger, oval, within a broad teardrop-shaped black naked area immediately before anus, as typical for genus. Black line con- necting the two lenses poorly defined. Ventral striae not especially prominent and less extensive than in H. striatissimus. Pyloric caeca 11 or 12, short (less than orbit diameter or interorbital width) and thick. Two slender retia terminate in 2 flattened half-moon- shaped gas glands. First dorsal fin long-based, its height some- what less than postrostral length; second dorsal scarcely developed except near tail tip; pectoral fins slender, extending to about level of anal or- igin; pelvic fins moderate in size, the outermost ray slightly prolonged, extending to about Sth to 8th anal ray. Color in alcohol overall grayish or blackish except over transparent head covering. Abdo- men from base of pectoral fins ventrally to pel- vics and back to vent dark violet to blue; chest and bases of paired fins black. Scattered mela- nophores of different sizes cover all of trunk and tail, the melanophores generally larger and more widely spaced on ventral half of body. A broad, dense midlateral band of melanophores begins about midlateral portion of trunk and extends posteriorly onto, and eventually completely in- cludes, tail. Anterior edge of nape and leading edge of snout black. Large scattered melano- phores cover most of suborbital and lower por- tion of preopercle, all of opercle, and parietal region. Jaws anteriorly, and gular and branchios- tegal membranes black. Floor of mouth below tongue forming a black triangle; tongue, how- 64 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 ever, pale dorsally with only a few scattered large melanophores on ventral surface; roof of mouth with splotches of black. Gill filaments pale; arch- es and rakers lightly peppered with small mela- nophores or pale. Paired and first dorsal fins with dark rays and pale interradial membranes; anal fin pale, but bases marked by small black dots. Foop.—Parin et al. (1990:42) found H. stria- tulus to feed predominantly on pelagic organ- isms, especially copepods (Pleuromamma sp., Oncaea conifera, Xanthocalanus sp., Aetideidae spp.), but also small fish, chaetognaths, gam- marids (Lysianassidae), shrimp (Bentheogenne- ma pasithea), and mysids (Paralophogaster gla- ber). Benthic polychaets (including the family Polinoidae) were also important food items. SizeE.—A small species, attaining about 180 mm ie DIsTRIBUTION.—Known only from the Ha- waiian Islands and the Sala y Gomez Ridge in depths of about 350 to 640 m. COMPARISONS AND REMARKS. —Gilbert’s (1905) original description and illustration of the species are excellent; they should be referred to for ad- ditional details. Specimens from the Sala y Go- mez Ridge show one notable difference from those collected off Hawaii; they have a rudimentary but distinct mental barbel, the length of which goes about 1.5 times into the least suborbital width, in contrast to the almost obsolete barbel in Hawaiian specimens. We do not feel this single difference is sufficient to recognize the species as distinct from H. striatulus, especially knowing that barbel length in species of Hymenocephalus can vary widely. A notable example is the wide- ranging species H. striatissimus of Japan, South China Sea, Philippines, and East Indies. Gilbert and Hubbs (1920:527) recognized three subspe- cies of H. striatissimus, and Okamura (1970a: 50) described a fourth, based in part on geo- graphical differences in barbel development. Per- haps the Sala y Gomez specimens should also be recognized as a subspecies of the Hawaiian H. striatulus, but we choose not to recognize them as such at this time. Gilbert and Hubbs’s (1920:521) key to the sub- genera and species of Hymenocephalus can be used to distinguish H. striatulus from all other known species of the genus except H. billsamo- rum Marshall and Iwamoto, 1973, from the trop- ical western Atlantic. The two species share in common a relatively slender body, low pointed snout, high pelvic fin ray counts (13-15), and a rudimentary or obsolescent barbel. Hymeno- cephalus striatulus differs from H. billsamorum in having larger orbits, narrower suborbital and interorbital regions, and slightly more pectoral fin rays (116-118 vs. 114-115). MATERIAL EXAMINED. — Sala y Gomez Ridge: ZMMGU 17711 (33 spec.:24-34.5 mm HL, 117+-150+ mm TL); 540 m; Ichthyanar cr. 5, tr. 53. ZMMGU 17712 (34:25-37 HL; 112+- 180+ TL); 345-540 m; Ichthyandr cr. 5, tr. 55. ZMMGU 17713 (275:17.8-34 HL, 90-170 TL); 580-564 m; Prof. Shtok- man cr. 18, sta. 1964. ZMMGU 17714 (185:19.2-36.5 HL, 91-164 TL); 562-545 m; Prof. Shtokman cr. 18, sta. 1965. ZMMGU 17715 (2:27.8-29.3 HL, 116+-118+ TL); 540-560 m; Prof. Shtokman cr. 18, sta. 1970. ZMMGU 17716 (4:23.5- 30.8 HL, 80+-144 TL); 570-580 m; Prof. Shtokman cr. 18, sta. 1971. ZMMGU 17717 (37:18.8-32.5 HL, 95-154 TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. ZMMGU 17718 (100:15.3-33.5 HL, 75-135 TL); 545-800 m; Prof. Shtokman cr. 18, sta. 1977. ZMMGU 17719 (3:30.6-32.2 HL, 125-142 TL); 750-800 m; Prof. Shtokman cr. 18, sta. 1996. ZMMGU 17720 (24:16-35 HL, 62.5+-165 TL); 730-790 m; Prof. Shtokman cr. 18, sta. 2018. ZMMGU 18212 (7:28-34.5 HL, 103+-180 TL); 565-555 m; Prof. Mesiatzev cr. 13, tr. 1. ZMMGU 18213 (10:29-34 HL, 128+-185 TL); 550-560 m; Prof. Mesiatzey cr. 13, tr. 2. ZMMGU 18214 (2:30.7-32 HL, 122+-141 TL); 410-420 m; Prof. Mesiatzey cr. 13, tr. 4. ZMMGU 18215 (13:20-32.3 HL, 100-158 TL); 530 m; Prof. Mesiatzev cr. 15, tr. 49. ZMMGU 18216 (2:28.9-32.5 HL, 125-134+ TL); 400 m; Prof. Mesiatzey cr. 15, tr. 52. Kuronezumia Iwamoto Kuronezumia Iwamoto, 1974 [as subgenus of Nezumia Jordan, 1904]. Type species: Nezumia (Kuronezumia) bubonis twamoto, 1974. D1aGnosis.— Macrourines with 7 branchios- tegal rays. Body and head compressed and deep, greatest depth below origin of first dorsal fin about 90-110% HL. Snout rounded in profile, almost entirely covered (except narrow median ventral and ventralmost margin) with small uniform, finely spinulated scales. Suborbital region ver- tical, without an angular midlateral ridge, the region covered with small scales without a row of enlarged, scutelike scales. Mouth moderate in size, upper jaw extends posteriorly to below an- terior half of orbit, length 30-44% HL. Dentition in broad villiform bands in both jaws; lower jaw band broad and short; outer series on premax- illary slightly enlarged. Gill rakers 8-11 (total) on outer side of second arch. Small patches of scales on branchiostegal membranes in most spe- cies. Body scales rather small, adherent, densely covered with long slender spinules; transverse ridges in most adults (but variously absent in some). Anus far removed from anal fin, closer to pelvic bases. Anterior dermal window of light SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 65 Ficure 24. Kuronezumia pallida new species; holotype, ZMMGU 17730, 98.5 mm HL, from Sala y Gomez Ridge in 550 m. Photograph by Susan Middleton. organ usually small, situated between pelvic fin bases and separated from anus by a broad scaly area, which is greatly swollen in two species. Py- loric caeca about 35-40 except K. /eonis, which has 15-18. Color overall light gray or brown to swarthy, fins dusky to blackish, naked mem- branes blackish to dark gray. Abdominal verte- brae 12-13. REMARKS.—The genus, here elevated from subgeneric status, includes K. bubonis, K. pallida, K. dara (Gilbert and Hubbs, 1916), K. leonis (Barnard, 1925), K. macronema (Smith and Rad- cliffe, 1912), and two undescribed species. The definition of the genus has had to be expanded to accommodate K. Jeonis and one of the un- described species. The peculiarly enlarged, bul- bous swelling housing the light organ in K. bu- bonis was used to diagnose the subgenus Kuronezumia, but K. pallida and K. dara lack this swelling. The combination of other features, especially the head physiognomy and squama- tion, nonetheless serve to unite the seven species and justify recognizing them as representatives ofa distinct taxon. The genus is now known from the tropical western Atlantic, throughout the Pa- cific (NSG, Hawaii, Japan, off southeastern Aus- tralia [AMS specimens provided by J. R. Pax- ton], off New Zealand [specimen of K. /eonis examined in LACM and NMNZ, and K. bubonis in NMNZ], and the Indian Ocean (Madagascar Ridge, West Australian Ridge [Shcherbachev 1987]; Kerguelen Plateau [one juv. of K. leonis, ZMMGU uncat.]). Kuronezumia pallida new species (Figures 24, 25a) Nezumia sp.: Parin et al. 1981:12 (brief descr. of specimens here described as new). “Kuronezumia sp. nova Sazonov et Iwamoto”: Parin 1990:16 (listed from Sala y Gomez Ridge). D1aGcnosis.— No large pores of sensory canals on head. Length upper jaw 34-42 HL, smaller in juveniles. GR-I (inner) 8—9 total; scales below 1D. 11-14; V. 11. No bulblike swelling of light organ; anterior dermal window represented by a small, lenticular scaleless area between midbases of pelvic fins. Color pale brown, except bran- chiostegal membranes and fins darker. COUNTS AND MEASUREMENTS. —(data for ho- lotype followed in parentheses by 79 mm para- type and 25 mm juvenile) 1D. II,11 (11,13, II,11); 2D. about 140+; 1P. 124/125 (125/124, 125/127); V. 11/11 (11/11, 11/11); GR-I [outer/inner] 7 (8, 7/0 +9(0 + 8,1 + 8), GR-T10+7(1 + 8, 1 + 9)//0 + 9, (0 + 9, 1 + 9); scales below 1D. 11 [about 14 on right side] (12-13, 12), below 66 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 2D. 11 [about 14 on right side] (11, 12). Abdom- inal vertebrae 12; anal pterygiophores anterior to first haemal spine 12. The following in percent of HL: postrostral 74.1 (73.4, 64.0); snout 28.4 (28.5, 28.8); preoral 17.3 (15.8, 17.6); orbit 23.9 (25.3, 34.8); inter- orbital 24.4 (26.6, 25.2); orbit-preop. 46.3 (42.4, 34.8); suborbital 19.8 (20.2, 18.0); upper jaw 40.1 (42.4, about 34); barbel 26.4 (23.4, 22.8); gill slit 17.8 (17, —); body depth about 95 (94, 86); pre- D. 120 (110, 111); pre-V. 104 (104, 94); pre-A. 144 (137, 131); pre-vent 119 (114, 114); V.-A. 46.3 (38.6, 37.2); 1D. height—(82, 101); 1D. base 28.9129. 15 31-2)2 1D 2D; 35.0\(38.6, 29.2): 1P- —(61, 68); V.—(73, 101). DESCRIPTION.—A large, deep-bodied species, greatest depth about equal to HL in adults, about 5.5 in TL; head deep, relatively compressed. Snout high, blunt, with vertical anterior profile. Orbits moderate in size, circular, diameter less than snout length, about equal to interorbital width. Mouth large, jaws subterminal; posterior end of maxillary extends to vertical through mid- dle of orbit. Head ridges inconspicuous; infra- orbital region deep, vertical, without modified scutelike scales, about 13 scales wide, 6 or 7 in upper portion. Scaled areas of opercle and sub- opercle together form a deep, inverted triangle. Interopercle narrowly exposed and scaled along ventral and posterior margins. Free margin of preopercle smooth. Gill membranes broadly united (at level of hind border of orbits), almost without free margin behind their connection with isthmus. Mental barbel moderately thick, long, about equal to orbit diameter, tapered to a fine tip. Sensory canals on head not broadened or swol- len; no enlarged open pores. Free neuromasts serially arranged along surfaces of supraorbital, infraorbital, postorbital, and mandibular canals, and on anterior surface of snout. Olfactory cavity rather small (about diameter of pupil), nasal ro- sette occupies small portion of cavity; anterior nostril rounded, much smaller than posterior nostril, which is semi-elliptical and about 2.1 into least suborbital width; internarial mem- brane narrow, with flap length equal to diameter of anterior nostril. Premaxillary teeth in broad, abruptly tapered bands that fall short of rictus (about 5 rows wide anteriorly); outermost series notably enlarged, straight, conical. Dentary teeth band short, broad, 3-5 rows wide; all teeth uniformly small. Lips, interdental spaces, and adjacent portions of oral cavity covered with numerous, branched (at tips), hairlike papillae that almost cover teeth. Similar papillae on anteriormost portion of snout above upper jaw. Head scales densely covered with erect to sub- erect needlelike spinules in 2-10 divergent series; those bordering orbits with 1-2 prolonged crests radiating from orbit (most pronounced on sub- orbital). Spinules on scales more densely placed in paratype, whereas in holotype spinulated sur- faces of adjacent scales separated from each other by smooth interspaces. No scutelike scales at tip of snout and anterior end of nasal bones; scales around tip scarcely stouter than more peripheral ones. Head almost completely scaled except over gular membranes, anterior end of mandible, and along lower part of snout and suborbital im- mediately above upper lip. Scales present over lowermost branchiostegal rays in two patches on each side in holotype; upper patch much smaller than lower one. Body scales (Fig. 25a) small, densely covered with conical retrose spinules forming 7-15 radiating series. Transverse ridges (reticulate pattern) absent in holotype, but this may be size related—in a 56 mm HL paratype of K. bubonis (CAS 27874), transverse ridges were sparse but present around the focus (Fig. 25b). Scales cover proximal part of pelvics between fin rays. Light organ represented by black, scaleless, rather small anterior dermal window situated be- tween midbases of pelvics and separated from periproct by convex scale-covered area. Peri- proct rather narrow, preceded by posterior der- mal window of size about equal to anterior win- dow. Internally, light organ consists of large bulblike black luminous gland separated from body cavity by peritoneum, without muscular tissue over upper surface; smaller reflector situ- ated between luminous gland and pelvic bones immediately above anterior dermal window. In- ner layer of skin between both dermal windows uniform in consistency, becoming thinner ante- riorly, blackish with slight opalescence, without indication of separate lenses. Swim bladder moderately large, oval, with blunt end anteriad. Two gas glands well devel- oped, connected to two relatively long retia. Stomach everted in paratype; number of pyloric caeca not determinable. Origin of first dorsal fin slightly behind, origin of pelvic fins well before, vertical through pec- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 67 toral origins. Spinous second ray of first dorsal scarcely extended beyond longest branched rays; serrations on leading edge relatively small and low. Interspace between dorsals rather short, about one-third longer than base of first dorsal. Pectoral fin moderately long, originating well be- low level of top of gill opening. Pelvic fin with long filamentous outer ray extending to about 8th-10th anal ray. Origin of anal fin somewhat before vertical of posterior end of first dorsal. Color in alcohol: body and head grayish with yellow tinge. Gular membrane light gray, bran- chiostegal membrane blackish. Fins dark gray to blackish, but anal fin blackish on anterior portion only. Oral and branchial cavities unpigmented. SizE.—To about 55 cm TL. DIsTRIBUTION. — Known only from the Sala y Gomez Ridge in 540-800 m. EtyMoLoGy.—From the Latin pallida, light or pale, in reference to the pale coloration of the species, contrasted with the dark-colored K. bu- bonis. COMPARISONS AND REMARKS.— The new spe- cies appears most closely related to K. dara (Gil- bert and Hubbs, 1916) from Japanese waters. Previous to Okamura’s report (in Okamura and Kitajima 1984:217, 363, fig. 153) of two speci- mens (220-318 mm TL) from the Okinawa Trough, the three known specimens of that spe- cies were too small to make comparisons with other species of Kuronezumia. Okamura’s de- scription and excellent color figure of the larger Okinawa specimen show the close similarity of that species to other members of the genus. The similarities lie not just in the physiognomy of the head, but also in the structure of the fins, the squamation, and the overall color. Kuronezumia pallida appears to differ from K. dara in having smaller orbits (24-25% HL cf. 25-36%), longer upper jaw (34-42% HL cf. 30-35%), interorbital space about equal to or slightly more than orbit diameter (cf. about equal to or less than orbit diameter), and barbel about equal to orbit (cf. much shorter than orbit, at least in small K. dara; condition not stated for large specimen by Oka- mura). The 171 mm juvenile (ZMMGU 18066) in our material agrees closely in these propor- tional measurements with K. dara. Allometric growth probably accounts for the differences be- tween juveniles and adults of the species. It thus seems that juveniles of K. dara and K. pallida are indistinguishable by these measurements. The new species differs from K. bubonis and an un- FiGure 25. Scales from below interspace between dorsal fins of (a) Kuronezumia pallida (holotype, ZMMGU 17730, 98.5 mm HL) and (b) K. bubonis (paratype, CAS 27874, 56 mm HL). Scale bars represent 0.5 mm. Drawn by Tomio Iwa- moto. described species from the South China Sea (see Iwamoto 1974) primarily in lacking the bulbous swelling before the periproct and in having a paler coloration. MATERIAL EXAMINED.—Sala y Gomez Ridge: Holotype: ZMMGU 17730 (98.5 mm HL, 548+ mm TL); 550 m; As- tronomer sta. 104. Paratype: ZMMGU 17731 (79 mm HL, 379+ mm TL); 540 m; Ichthyandr cr. 5, tr. 53. Non-type material: ZMMGU 18066 (25 HL, 171 TL); 750- 800 m; Prof. Shtokman cr. 16, sta. 1996. 68 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Malacocephalus Giinther, 1862 Type species: Macrourus laevis Lowe, 1843, by monotypy. D1acnosis.— Macrourines with 7 branchios- tegal rays. Mouth large, usually greater than 45% HL; GR-I (inner series) usually less than 12 total. Teeth large, widely spaced, in 1 row in lower jaw, usually larger posteriorly; in 2 rows to narrow band in premaxillary. Head completely and uni- formly scaled, lacking enlarged, modified scales; suborbital area vertical or nearly so and covered with small, finely spinulated scales; lowermost branchiostegal rays scaled; scales without trans- verse ridges. Anus removed from anal fin and closer to pelvic fin, preceded by 2 naked fossae (dermal windows of light organ), one round to bean-shaped fossa between pelvic fin bases, the other immediately before anus at anterior end of periproct region. Pyloric caeca 50-100, multiply branched. DIsTRIBUTION. — Worldwide in warm to tem- perate seas. REMARKS. — The genus is divided into two sub- genera by some authorities: subgenus Pawnurus Parr, 1946 (two species, with serrated first dorsal fin ray) and subgenus Malacocephalus (two to four species, lacking serrated first dorsal ray). Malacocephalus boretzi Sazonov, 1985, de- scribed from the central North Pacific, has char- acters that lessen the distinction between the two nominal subgenera. Sazonov (1985:17), how- ever, supported their continued recognition and also provided evidence (in the internal structure of the light organ) that showed a close relation- ship of Lucigadus Gilbert and Hubbs, 1920 (which he elevated to full genus) to Malacocepha- lus. A question still remains whether or not there is more than one species in the Malacocephalus laevis complex (see REMARKS in following de- scription). Malacocephalus laevis (Lowe, 1843) Macrourus laevis Lowe, 1843:92 (off Madeira). Malacocephalus laevis: Giinther 1862:397. Parin et al. 1981: 12 (Sala y Gomez Ridge). Golovan’ and Pakhorukov 1987: 73 (Nazca and Sala y Gomez ridges). Parin 1990:16 (Sala y Gomez Ridge). Parin et al. 1990:43 (stomach contents). Kot- lyar and Parin 1990:106 (otolith). Macrurus (Malacocephalus) laevis: Giinther 1887:148. DiaGnosis. — Upper jaw 45-54% HL. Premax- illary teeth in 2 distinct rows, mandibular teeth in 1 row. A smooth spinous second ray of first dorsal; fin rays V. 9 (occasionally 8 or 10); 1P. 116-121. REMARKS. — This widespread species has been more than adequately described and illustrated by numerous recent authors (e.g., Marshall 1973; Okamura 1970a), and the reader is referred to other sources for a complete description. Iwa- moto (1979) reported a single individual from off southern California, the first record of the species from the eastern Pacific. Subsequent to that report, numerous specimens were collected in 1979 on seamounts off the Baja California peninsula by the Japanese fishery research vessel Kaiyo Maru (Eichii Fujii, Tokai Regional Fish- ery Research Laboratory, Tokyo, per. comm. with TI, Dec. 1979) and by Soviet vessels off the Sala y Gomez Ridge (Parin et al. 1981:12). The Kaiyo Maru specimens were examined (by TI) in 1980 at the Far Seas Fishery Research Laboratory in Shimizu, Japan. The Soviet collections consti- tute the primary basis for this account. We identify the eastern Pacific specimens with Malacocephalus laevis, a species originally de- scribed from the North Atlantic. The status of three other nominal Pacific species of the genus has not been completely resolved. The three in- clude M. nipponensis Gilbert and Hubbs, 1916, M. hawaiiensis Gilbert, 1905, and M. luzonensis Gilbert and Hubbs, 1920. Okamura (1970a:69) placed M. nipponensis into the synonymy of M. laevis, but later (Okamura in Okamura et al. 1982: 145, 347-348) recognized M. nipponensis based on its lack of “‘scaled patch on the gular mem- brane and in a few other characters.”’ Iwamoto (1970:411) recognized all three species and pro- vided a key to the genus based on the literature and examination of relatively few specimens. Marshall (1973:650-652) also recognized the three species but noted that “Examination of a good series of individuals may show that nip- ponensis is identical with hawaiiensis.”’ In Table 4, selected counts and measurements of speci- mens of Malacocephalus from three oceans are compared. Although the data were not treated statistically, a casual inspection suggests that they do not support taxonomic recognition of the pop- ulations. (The raw data for this table are depos- ited in the library of the Department of Ichthy- ology, California Academy of Sciences; photocopies are available to any interested party.) SizE.—To at least 52 cm. DISTRIBUTION. — Probably worldwide in tem- perate to tropical waters, but absent on the con- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 69 TasBLE 4. Comparison of selected measurements and counts of Malacocephalus laevis from the Sala y Gomez Ridge, the Atlantic Ocean, and the Indian Ocean. Character Sala y Gomez Indian Ocean Atlantic Ocean TL (mm) 231-458 265-492 326-450 HL (mm) 37.5-79.0 52.5-93.0 49-76 % of HL [no. of specimens] [18] [5] [10] Snout 22.6-29.6 26.6—-28.7 21.4-29.6 Orbit 29.8-36.7 30.8-37.1 30.4-36.1 Interorbital 26.6-34.7 31.7-34.8 27.6-36.1 Orbit-preop. 38.7-45.6 42.6-48.4 40.7-45.7 Postorbital 41.4-45.6 43.8-48.1 40.7-48.5 Suborbital 11.6-14.7 9.9-12.4 11.2-13.0 Upper jaw 47.1-53.5 46.7-52.4 45.9-53.0 Gill slit 24.1-29.3 23.1-29.5 22.4-28.7 Barbel 19.5-32.0 22.6-26.3 22.4-27.0 Pre-A. 117-148 130-140 121-154 Body depth (max.) 74.4-91.6 77.0-84.4 78.5-94.6 Height 1D. 40-69 [7] 62-69 [2] 62-64 [2] 1D.-2D. 40.0-69.0 49.2-61.0 46.1-70.3 Length 1P. 53-56 [3] 51-59 [4] 63-66 [3] Length V 30-38 [5] 32-46 [5] 33-40 [5] Counts 1D. II,10-13 II,10-11 II,10-12 Wee i16-120 il7-119 i17-i21 V. 8-10 8-9 8-9 GR-I (outer) total 8-11 7-9 7-11 GR-I (inner) total 10-13 10-12 11-14 GR-II (outer) total 9-12 9-11 11-14 GR-II (inner) total 9-12 10-11 11-14 tinental slopes of the eastern central and South Pacific. Depth range generally between 300 and 700 m, but often encountered shallower or deep- er. MATERIAL EXAMINED. —(84 spec., 19 sta.) Sala y Gomez Ridge: ZMMGU 18074 (2:63.7-73.3 mm HL, 392+-420+ mm TL); 550 m; Astronomer trawl without no.; 25°02’S, 88°35'W; 24 Jul. 1975. ZMMGU 18075 (6:39.5-78.5 HL, 231+-463+ TL); 345-540 m; Ichthyanadr cr. 5, tr. 55. ZMMGU 18076 (14:41.7- 53.3 HL, 255-322 TL); 410 m; Ichthyandr cr. 5, tr. 56. ZMMGU 18077 (3:37.5-66.8 HL, 230+-403+ TL); 540 m; Ichthyandr cr. 5, tr. 53. ZMMGU 18078 (2:50.5-61 HL, 288+-359+ TL); 535-575 m; Ichthyandr cr. 5, tr. 54. ZMMGU 18079 (72 HL, 440+ TL); 420 m; Ichthyandr cr. 5, tr. 57. CAS uncat. (8:42.5— 79 HL, 248-443 TL); 430 m; Ichthyandr cr. 6, tr. 56. ZMMGU 18080 (4:40-74.5 HL, 257-410 TL); 565-555 m; Prof. Mesi- atzev cr. 13, tr. 1. ZMMGU 18081 (4:61.5-67 HL, 281+ - 443+ TL); 550-560 m; Prof. Mesiatzev cr. 13, tr. 2. ZMMGU 18082 (38 HL, 239 TL); 410-420 m; Prof. Mesiatzev cr. 13, tr. 4. ZMMGU 18083 (2:39-53.5 HL, 238-295 TL); 285-300 m; Prof. Mesiatzev cr. 13, tr. 7. ZMMGU 18084 (3:40-49.5 HL, 242-288 TL); 400 m; Prof. Mesiatzev cr. 15, tr. 52. ZMMGU 18085 (5:37.8-44.1 HL, 239+-264+ TL); 530 m; Prof. Mesiatzev cr. 15, tr. 49. ZMMGU 18086 (43.5 HL, 271 TL); 304 m; Prof. Mesiatzev cr. 15, tr. 54. ZMMGU 18087 (6: 42.8-86.2 HL, 255-475 TL); 410-385 m; Prof. Shtokman cr. 18, sta. 1941. ZMMGU 18088 (5:39.5-88 HL, 214+-463 TL); 580-564 m; Prof. Shtokmancr. 18, sta. 1964. ZMMGU 18089 (5:40.5-84 HL, 237-490 TL); 562-545 m; Prof. Shtokman cr. 18, sta. 1965. ZMMGU 18090 (5:38.3-57.7 HL, 238+-333+ TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. ZMMGU 18092 (7:45.5-100 HL, 237+-480 TL); 545-600 m; Prof. Shtokman cr. 18, sta. 1977. Mataeocephalus Berg, 1898 Type species: Coelocephalus acipenserinus Gilbert and Cra- mer, 1897, by original designation. D1aGnosis.—Macrourines with 7 branchios- tegal rays. Snout long, pointed, somewhat de- pressed, tipped with a two-pronged scute. Mouth inferior, upper jaw less than 30% HL. Teeth in most species in short, broad bands, confined to anterior end of jaws. Outer gill rakers on first arch rudimentary or absent. Spinous ray of first dorsal fin serrated or smooth. Periproct at or close to anal fin origin. DISTRIBUTION. — Tropical waters of Pacific and Indian oceans in upper to mid-slope depths of about 550-1,700 m (according to our unpubl. data). REMARKS.—A small genus, comprising five or more species, in need of revision. See Iwamoto (1979:144) for a discussion of the genus. 70 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Mataeocephalus acipenserinus (Gilbert and Cra- mer, 1897) Coelocephalus acipenserinus Gilbert and Cramer, 1897:422- 423, pl. 42, fig. 1 (Kaiwi Channel, Hawaiian Is.; 572-728 m; Holotype: USNM 47721). Mataeocephalus acipenserinus: Berg, 1898:41. Parin 1990:16 (listed from Sala y Gomez Ridge). Kotlyar and Parin 1990: 106 (otolith). DIAGNosIs. — Teeth in lower jaw confined to a short, broad, lunate patch at tip of jaws, none laterally. Underside of head anterior to lower jaw angle naked; scales below mid-1D. 6—7.5, below 2D. 8 or 9 (rarely 7). Some denticulations on leading edge of dorsal spine; V. 8; 1P. 120-125. Anus at posterior end of a broad, oval, naked area situated close to anal fin origin and far re- moved from pelvic fin insertions. COUNTS AND MEASUREMENTS.—(of 21 Sala y Gomez spec.) (see also DiaGnosis) 1D. II,8-10; V. 8 (7/8 in one spec.); GR-I (outer/inner) 3-5 total/6-8 total, GR-II 5-8 total/6—9 total; scales below 1D. 8-9. Total length 118-250 mm, HL 26.5-54 mm. The following in percent HL: postrostral 58.1- 64.8; snout length 38.5—44.0, width at lateral an- gles 34.1-43.2; preoral 37.8-46.2; orbit 28.9- 34.4; interorbital 19.7—-23.0; postorbital 26.3- 33.0; suborbital 13.9—17.1; orb.-preop. 27.7-35.0; upper jaw 20.0-27.5; barbel 3.5-8.6; gill slit 7.4— 10.4; pre-1D. 105-113; pre-A. 126-140; pre-vent 116-135; pre-V. 96-119; V.-A. 22.6-35; body depth 47-60; 1D. base 17.3-22; 1D. height 50- 63; 1D.-2D. 13.5-—32.2; 1P. 41-57; V. 45-70. DESCRIPTION.—Entire head profile sharply conical in lateral view, the profile continuing smoothly to dorsal fin base dorsally and to pelvic or anal fin bases ventrally; thereafter profile ta- pers to slender, attenuated tail tip. Orbit large, about 1.5 into snout, almost equal to postorbital; sides of snout viewed dorsally gently curved from orbits to bifid scute on snout tip; interorbital about 1.3 into orbit, equal to width between su- pranarial ridges, space slightly concave; mouth “U” shaped, small, inferior, the rictus greatly restricted by lip folds laterally; upper jaw extends to a vertical slightly posterior to midorbit; barbel very small, length much less than large posterior nostril; suborbital forming a rounded triangle at posteroventral corner; interopercle exposed along posteroventral margins. Gill membranes broadly attached to isthmus; gill openings extend only to below preopercular ridge. Gill slits restricted, the first (outermost) slit about 10% of HL; other slits restricted by narrow fold of membrane between lowermost parts of arches. Rakers on outer side of first arch small, scarcely visible as tiny pimples that show no contrast with dark arches; rakers elsewhere short and tubercular. Scales below origin of second dorsal in 51 mm HL specimen with fine needlelike spinules ar- ranged in 5-8 parallel rows, fewer rows in smaller specimens. Underside of head naked forward of jaw angle except for overlapping series of mod- ified scales along leading edge of snout; some scale patches ventrally on preopercle. Suborbital strongly angular in cross section; heavy, modi- fied, scutelike scales in 2 rows on dorsal part leading forward to rather sharp edge of snout. Tip of snout with a pair of spiked, conical tu- bercles, as characteristic of genus. Ridges other than suborbital not strongly supported by heavy scales. A prominent naked groove dorsally be- hind broad anterolateral margins of snout. Jaw teeth all small, in a broad short band in premaxillae, in a short tapered band only at an- terior end of lower jaw. Disposition of teeth as illustrated for M. tenuicauda by Iwamoto (1979, fig. 9). Lengths of first dorsal and pectoral fins about half HL. Serrations on spinous dorsal ray weakly and sparsely developed, but present in all ex- amined specimens. Outer pelvic ray slightly thickened and prolonged, extending past anal or- igin to level of 10th anal ray or beyond in some specimens. Anal fin well developed, origin about on vertical through first dorsal insertion; pectoral and pelvic origins about on same vertical, that of first dorsal well behind. Interspace between dorsals 1-2 times length base of first dorsal. Periproct region broad, the region black, na- ked, and separated by a narrow gap from anal fin. Anus at posterior end of periproct, much closer to anal fin than to pelvic fin insertions. A large lenslike structure forming anterior portion of periproct extending forward towards, but fall- ing short of, pelvic girdle. Photophore length 6.2- 8.5% HL in 12 Sala y Gomez specimens (this compares with 6.7—8.1% in 3 Hawaiian para- types and 3.1-6.7%, usually less than 6%), in 21 specimens of M. microstomus from the Indian Ocean. Pyloric caeca short and thick: 14, 14, and 15 counted in 3 specimens. Overall color variable, but Sala y Gomez spec- imens generally rather light grayish overall, whit- ish on underside of head anterior to lower jaw SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 71 angle, and deep bluish over opercles and abdo- men. Fins clear to light dusky. Rim of nostrils narrowly blackish. Mouth black; gill cavity black dorsally and along gill slits and arches, but pale ventrally and on gill filaments. Size.—Attains at least 25 cm TL. DIsTRIBUTION.—Our Sala y Gomez Ridge specimens were collected at two closely adjacent stations on the westernmost seamount surveyed during the 18th cruise of the Prof. Shtokman. Until this cruise, M. acipenserinus was known only from off Hawaii. Its presence on the Sala y Gomez Ridge was unexpected, although several slope-dwelling Hawaiian species of other fami- lies have been captured on the ridge (Parin et al. 1981:5,and N.V. Parin, pers. comm.), suggesting a closer affinity of the two faunas than previously realized. COMPARISONS AND REMARKS.—A revision of the genus Mataeocephalus is in preparation by one of us (YS) and Y. N. Shcherbachev. Prelim- inary findings show the genus to comprise two species groups—M. adustus group and M. mi- crostomus group—each with several species of uncertain status. Our Sala y Gomez Ridge spec- imens of Mataeocephalus belong to the second group, which includes M. microstomus (Regan 1908), M. nigrescens Smith and Radcliffe, 1912 (probably a synonym of the former species), M. tenuicauda (Garman 1899), and M. acipenseri- nus. The species in this group lack scales on the underside of the snout, and the mandibular teeth are in short lunate patches (as illustrated by Iwa- moto 1979, fig. 9a). Proportional measurements and counts of all species in the M. microstomus group are similar. Differences between species are slight, and the main differentiating character appears to be the degree of development of the light organ. In this regard, our specimens agree most closely with the Hawaiian species, M. acipenserinus, in hav- ing the light organ anteriorly prolonged. In con- trast, M. microstomus has a small subtriangular gland with a tiny anterior dermal window (ADW) just in front of the periproct, and M. tenuicauda has a rudimentary gland and no ADW. Our Sala y Gomez specimens differ slightly from paratypes of M. acipenserinus in that the anterior margin of the ADW in the paratypes is much closer to the insertion of the inner pelvic rays than is the case in our specimens. This re- flects the shorter distance between pelvic and anal fins in the paratypes (22.6-31.9% HL, usu- ally less than 26% cf. usually 26%-34.9% [but 22.6% in one specimen]). Considering the great variability and consequent low taxonomic value of this character, and in the absence of other discernible differences, we recognize the Sala y Gomez populations as M. acipenserinus. MATERIAL EXAMINED. — Sala y Gomez Ridge: ZMMGU 17732 (8:26.5-51.5 mm HL, 117-250 mm TL) and CAS uncat. (5: 40-50 HL, 166+-182+ TL); 750-800 m; Prof. Shtokman cr. 18, sta. 1996. ZMMGU 17733 (8:32.3-54.7 HL, 155-240 TL); 730-790 m; Prof. Shtokman cr. 18, sta. 2018. Hawaiian Is- lands: CAS-SU 3142 (4 paratypes, 44-51.5 HL, 161-177 TL); 21°08'30’N, 157°49’W; 627 m:; Albatross sta. 3470. Nezumia Jordan, 1904 Type species: Nezumia condylura Jordan and Gilbert, 1904, by original designation. DriaGnosis.— Macrourines with 7 branchios- tegal rays. Teeth small, in narrow to broad bands in both jaws; those on premaxillary do not occur past maxillary process. Gill rakers <12 on inner side of first arch in most species. Snout variously naked on ventral surfaces, anteriorly tipped with spiny tubercles in most species. Suborbital shelf with 2 rows of modified scutelike scales forming a prominent ridge. Body scales covered with nee- dlelike to shield-shaped spinules; transverse ridges present. Spinous second ray of first dorsal fin serrated in most species. Anus removed from anal fin origin and situated within an oval-shaped periproct. A small naked fossa of light organ be- tween pelvic fins. DISTRIBUTION. — Worldwide in temperate to tropical seas; most species found at upper to mid- dle continental-slope depths (about 200-1,500 m). Nezumia convergens (Garman, 1899) (Figure 26) Macrurus convergens Garman, 1899:210, pl. 48, fig. 1 (Gulf of Panama, 695-1,020 fm [1,271-1,865 m]; Albatross sta. 3353, 3357, 3393). Nezumia convergens: Parin and Sazonov 1982:86 (5 spec.; Peru). Parin 1990:16 (recorded from area between Nazca and Sala y Gomez ridges). See Iwamoto (1979:171) for synonymy. D1aGnosis.— Body slender, greatest depth 7- 8 in TL in large adults. Upper jaw 26-33% HL; barbel 8—20% HL (usually 1.5-2.0 in orbit); GR-I and GR-II (inner series) (1-2) + (5-8) (usually 5-7 on lower limb). Spinules on body scales con- ical to narrowly lanceolate, aligned in 10-12 slightly convergent longitudinal rows in large 72 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 FiGure 26. Nezumia convergens. From Iwamoto (1979:fig. 15b) adults, middle row often slightly enlarged. Al- most all of ventral surfaces of snout and antero- ventral surface of suborbital naked; mandibles naked anteriorly. Second spinous ray of first dor- sal fin longer than postrostral length of head, its leading edge beset with closely spaced teeth; first dorsal fin uniformly dusky; V. 10-11. Anus in middle half of space between pelvic fin insertion and anal fin origin. Light organ not well devel- oped externally; scaleless fossa not present be- tween pelvic fin bases. COUNTS AND MEASUREMENTS. —(of NSG spec- imen only) 1D. II,10; 1P. 123; V. 11; total GR-I [outer/inner] 9/10, GR-II 10/10; scales below 1D. 11, below mid-1D. about 10, below 2D. 9.5; lat.1. 46. The following in mm, percent HL in paren- theses: snout 14.1 (29.1); preoral 11.6 (23.9); or- bit 14.5 (29.9); interorbital 9.9 (20.4); postorbital 20.8 (42.9); orbit-preop. 18.8 (38.8); suborbital 7.4 (15.3); upper jaw 16.2 (33.4); barbel 7.5 (15.5); gill slit 6.8 (14.0); pre-A. 75 (154); pre-vent 66 (137); 1D.-2D. 14.5 (30); height 1D. 49 (101); 1P. 29 (59); V. 30 (62); body depth 36 (73). SizE.—To about 30 cm. DISTRIBUTION. — Broadly distributed along continental slopes of eastern Pacific, from Gulf of California and northern Mexico south to Chile (lat. 35°S) and also in Galapagos Archipelago. Depth range 600-1,865 m. REMARKS. —See Iwamoto (1979:171, figs. 15b, 18f) for a description and illustration. This fine specimen taken near the Sala y Gomez Ridge agrees well in most respects with specimens of N. convergens from the continental margins, al- though a few characters fall outside the ranges established for the species by Iwamoto (1979). Most notable is the gill raker count, which was high by one raker in the inner series of the first and second arches. Scale spinules also appeared to be more broadly lanceolate. The lateral line scale count over a distance equal to the predorsal length differed by two (46 vs. 36-44), the post- orbital length was slightly longer (43% HL vs. 34-40%), and the height of first dorsal was great- er (101% HL vs. 68-93%). Unfortunately, only the single specimen was collected, thus preclud- ing a more meaningful analysis of these differ- ences. MATERIAL EXAMINED.—ZMMGU 17734 (1:48.5 mm HL, 272+ mm TL); area between the Nazca and Sala y Gomez ridges, 1,050 m; Prof. Mesiatzev cr. 13, tr. 14. Nezumia propinqua (Gilbert and Cramer, 1897) (Figures 27, 28a, b) Macrourus propinquus Gilbert and Cramer, 1897:424, pl. 42, fig. 2 (type-locality Kaiwi Channel, Hawaiian Islands; 642 m). Gilbert 1905:667. Lionurus propinquus: Gilbert and Hubbs 1916:144 (list). Nezumia propinquus: Marshall and Iwamoto 1973:625 (list). Okamura in Okamura et al. 1982:163, 350, color fig. 97 (6 spec., 216-250 TL; Kyushu-Palau Ridge; 695-219 m). Nezumia propinqua: Iwamoto 1983:8; 1986:339 (2 spec.; Mo- zambique; 740 m). Parin 1990:16 (listed from Sala y Gomez Ridge). Kotlyar and Parin 1990:106 (otolith). D1AGnosis. — Body scales covered with slender conical spinules in 5-8 parallel rows (in adults to about 20 cm TL); scales below 2D. 10-13. Mandibles and underside of head to posterior angle of mouth naked. A distinct black tip on first dorsal fin. Length pectoral fin about 60% HL. Pelvic fin rays 14-18; 1 P. rays 118-122 (count data included for specimens from other areas). COUNTS AND MEASUREMENTS. —(from 17 Sala y Gomez specimens) 1D. II,1 1-13; 1P. 119-122; V. 15-17 (rarely 14 or 18); total GR-I (outer/ inner) 7—12/9-12; total GR-II 9-11/8—11; scales below 1D. 12-14, below mid-1D. 8-11, below 2D. 10-13, lat.1. 43-47; caeca 24 (1 spec.). SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES FiGurE 27. Photograph by Sergei Dudarev. Total length 125+-200+ mm; HL 19.3-34 mm. The following in percent of HL: snout 26- 32; preoral 20-29; orbit 31-40; interorbital 22- 28; postorbital (greatest) 41-45, (least) 34-39; orbit-preop. 32—40; suborbital 12—15 (19); upper jaw 31-38; barbel 17-28; gill slit 14-19; pre-A. 131-159; pre-V. 96-121; pre-vent 119-142; V.- A. 36-53; 1D.-2D. (17) 24-57; height 1D. 84—- 108; 1P. 58-69; V. (74) 80-99; body depth 78- 100. DESCRIPTION.— Head about 6 in TL, body depth slightly less than head length; trunk short, length abdominal cavity much less than HL. Snout bluntly pointed, tipped with a broad coarse tubercle; mouth subterminal, small, upper jaw about one-third HL, extends to below hind edge of pupil; barbel short but prominent, about one-half orbit diameter; gill membranes rather narrowly restricted across isthmus with a mod- erately free posterior fold, the gill openings ex- 73 Nezumia propinqua (approximately 30 mm HL) from Sala y Gomez Ridge, collected on Prof. Shtokman cr. 18. tending forward to below hind edge of orbits. Anus within a broad periproct, somewhat closer to anal fin origin than to pelvic origins. A small oval dermal window between pelvic bases. Ab- dominal cavity terminates above 9th or 10th anal ray. First dorsal fin with a small keellike first spine; second spinous ray prolonged, armed along lead- ing edge with sharp and widely spaced serrations, length of spine about equal to HL. Second dorsal fin rudimentary over entire length. Pectoral and pelvic fins rather short; outer ray of pelvics slight- ly prolonged into a filament that extends to end of abdominal cavity. Origin of pelvics below pos- terior margin of opercle and in advance of pec- toral base, which in turn is in advance of first dorsal origin; anal origin below midbase of first dorsal. Scales with about 7-10, more or less parallel rows of small, reclined, needlelike spinules (Fig. Ficure 28. Scales from dorsum below interspace between dorsal fins of: (a) Nezumia propinqua (CAS-SU 8538; Hawaii); (b) N. propinqua (Sala y Gomez Ridge); (c) N. condylura (CAS-SU 22925; Japan). Scale bar represents 0.5 mm. Drawn by Tomio Iwamoto. 74 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 28b). Suborbital shelf formed of a double row of stout scales. Ventral aspects of snout, most of suborbital, and lower jaw completely naked. Ce- phalic lateral-line pores rather prominent along lower margins of snout, suborbital, preopercle, and lower jaws. Teeth in broad bands in both jaws; outer series of upper jaw slightly enlarged. Large papillae densely interspersed between teeth and along mesial side of jaw, giving superficial appearance of mandibular band being much broader than it actually is. A 29-mm HL specimen from CAS 75990 had 24 short, thick pyloric caeca, their lengths about equal to pupil diameter. Ovaries in that speci- men moderately developed, but no eggs yet vis- ible. Largest Sala y Gomez specimen (ZMMGU 17749, 34 mm HL) a mature female with well- developed ovaries containing eggs about 1.0 mm in diameter. Color in alcohol pale yellowish brown overall; abdomen bluish, turning to blackish ventrally on trunk. Fins all pale except distally black on first dorsal. Mouth pale; broad outer margins of gill cavity dark; gill rakers and gular membranes dark or blackish. Underside of head pale. S1zE.—To about 25 cm. DISTRIBUTION. — Nezumia propinqua has been reported from Hawaii, the Kyushu-Palau Ridge, off Vietnam, off western Australia, and off Mo- zambique, in 523-870 m. COMPARISONS AND REMARKS.— Nezumia pro- pinqua and N. condylura are extremely close and may be conspecific. Okamura (in Okamura et al. 1982:350) used the distance anus to anal fin as the primary character to distinguish the two spe- cies. From measurements we have made, that distance ranges from 15-20% HL in N. condylura compared with 12-32% HL (most greater than 20% HL) in N. propinqua. On that basis, the Sala y Gomez specimens would fall into N. condylura. The length of pelvic fins also suggests a separa- tion of the two: 79—100% in N. propinqua vs. 68- 81% in N. condylura. Based on this measure- ment, the Sala y Gomez specimens fall into N. propinqua. Another character that may offer a means of separating the two species is the num- ber of spinule rows on body scales. In specimens of 27-28 mm HL, N. propinqua specimens had 5-8 spinule rows on scales below the interspace of the dorsal fins, whereas N. condylura speci- mens had 8-12 rows (see Fig. 28). No other char- acter we have examined suggests a separation, but a good size-series of fresh, well-preserved specimens of each have not been compared. For now, its seems best to treat the current material as N. propinqua, recognizing, however, that fu- ture workers making a more thorough study may arrive at another conclusion. Nezumia evides Gilbert and Hubbs, 1920, rep- resents another closely related species that can be distinguished by its slightly fewer scale rows below the second dorsal fin (9-10). MATERIAL EXAMINED.—(17 spec.) Sala y Gomez Ridge: ZMMGU 17741 (31.5 HL, 188+ TL); 565-555 m; Prof. Mesi- atzev cr. 13, tr. 1. ZMMGU 17742 (2:22.3-31.5 mm HL, 151+—- 191.5 mm TL) and CAS 75990 (2:29-32 HL, 164+-200+ TL); 550-560 m; Prof. Mesiatzey cr. 13, tr. 2. ZMMGU 17749 (34 HL, 175+ TL); 540-550 m; Prof. Shtokman cr. 18, sta. 1970. ZMMGU 17750 (4:20-29.5 HL, 125+-169+ TL); 750- 800 m; Prof. Shtokman cr. 18, sta. 1996. ZMMGU 18070 (6: 19.3-33 HL, 121+-174.5+ TL); 730-790 m; Prof. Shtokman cr. 18, sta. 2018. ZMMGU 18126 (26 HL, 157+ TL); 580- 564 m; Prof. Shtokman cr. 18, sta. 1964. Pseudocetonurus Sazonov and Shcherbachev, 1982 Type species: Pseudocetonurus septifer Sazonov and Shcher- bachev, 1982, by original designation. D1aGcnosis.— Macrourines with 7 branchios- tegal rays. Head notably large and broad, pre- opercle and suborbital bones deep and large, opercle commensurately small; orbit small, 19- 30% HL, diameter much less than broad inter- orbital; snout high, slightly projecting beyond mouth. Mental barbel small, 10% or less of HL. Gill opening wide, extending forward to below hind end of maxilla; gill membranes loosely and narrowly attached to isthmus. Gill rakers usually 16 or 17 total on inner series of first arch. Teeth small, close-set, in narrow tapered band on pre- maxilla, uniserial on dentary. Scales with nu- merous small, awl-shaped spinules; no reticula- tions on anterior field; lateral line scales absent, a series of dark papillae in its place. Vent about halfway between pelvic fin insertion and anal fin origin (usually closer to pelvic insertion), sur- rounded by a black, oval to teardrop-shaped na- ked area and preceded by a small, round dermal window of light organ between pelvic fin bases. Pyloric caeca short, 22-34. Color black to dark brown overall. DISTRIBUTION. — Known from Hawaii and the Nazca and Sala y Gomez ridges. REMARKS.—Only the single species known. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 75 FiGureE 29. Pseudocetonurus septifer: (a) Juvenile (ZMMGU 18127, 26.7 mm HL), Professor Shtokman sta. 1856, IKMWT in 900-0 m, 20 April 1987. (b) Adult (ZIN uncat., 72 mm HL, 393+ mm TL) from Nazca Ridge in 950 m; Hercules sta. 40. Photograph by Susan Middleton. Pseudocetonurus septifer Sazonov and Shcher- bachev, 1982 (Figure 29) Pseudocetonurus septifer Sazonov and Shcherbachev, 1982:712, fig. 2 (type locality Sala y Gomez Ridge, 850-860 m). Parin 1990:16-17 (listed from NSG). Iwamoto in Cohen et al. 1990, fig. 229 (in key). Dracnosis.— As for the genus. CouUNTS AND MEASUREMENTS.—1D. II,8—12 (usually 9 or 10); 1P. 116-120; V. 9 or 10, rarely 8; GR-I (outer/inner) 7-12 (usually 9-11)/(1-2) + (12-16) (total 13-18, usually 16 or 17), GR- II (1-3) + (12-15) (total 14-18, usually 15 or 16)/(1-3) + (12-15) (total 14-17, usually 15 or 16); scales 1D.-A. 33-39 (n = 18), midbase 1D.- A. 25-34; 2D.-A. 18-24 (n = 17); caeca 22-34 (n = 10). Total length 117+-393+ mm; HL 24-72 mm. The following in percent HL: postrostral 72.3- 79.1; snout 25:1-34.4; preoral 18.8-27.0 (usually 20-25); orbit 18.8—31.2; interorbital 32.8-44.4; postorbital 46.6-58.6; orbit-preop. 52.5-64.4; suborbital 19.3-25.5; upper jaw 40.9-48.5; bar- bel 6.5—10.3; outer gill slit 22.6-30.6; body depth 75.4—94.5; pre-D. 95-111; pre-V. 102-127; pre- vent 109-134; pre-A. 119-146; V.-A. 12-38.1; 1D. height 52-66; 1D. base 22.8—30.2; 1D.-2D. 24-38; 1P. 68-81; V. 41-63. DESCRIPTION. — Head large, 5.0-5.5 into TL, relatively broad and deep, greatest width about 1.5 into greatest depth; trunk compressed, grad- ually tapering into a long straplike tail (the tip often missing). Orbits small, semi-elliptical to oval, its greatest diameter usually oblique. In- terorbital broad, irregularly convex, 2.3-3.0 into HL. Suborbital region deep, almost vertical, con- vex, without division into upper and lower parts, and without stout scutelike scales. Preopercle no- tably large, with posterior margin oblique, more strikingly so in smaller specimens. Subopercle posterior margin forming a deep notch; sub- opercle and interopercle broadly exposed beyond preopercle in holotype and 393 mm specimen 76 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 (ZIN uncat.), but interopercle narrowly exposed in smaller specimens. Free margins of opercular bones crenulated. Snout short, bluntly pointed, and high, forming a pronounced hump in dorsal profile, its length usually greater than orbit di- ameter; anterior profile subvertical. Mouth large, jaws subterminal; posterior end of maxillary ex- tends to vertical through midorbit. Mental barbel very small, almost a rudiment, length less than half orbit diameter. Gill membranes narrowly attached to isthmus below hind end of upper jaws (opercular opening consequently wide). Sensory canals on head large. Free neuromasts on head indistinct and poorly developed. Lateral line canal absent on body, being replaced by small black papillae (free neuromasts?) situated some- what irregularly along normal course of lateral line. Olfactory capsules small; posterior nostril slightly larger than anterior; internarial mem- brane forming a prominent flap equal to diam- eter of anterior nostril in 393 mm ZIN specimen, but flap inconspicuous in smaller specimens. Teeth in jaws small, conical, closely spaced; uniserial on dentary; in narrow band 3-4 irreg- ular rows wide anteriorly on premaxillary, nar- rowing to about 2 rows wide posteriorly, outer series slightly enlarged. Squamation in 393 mm specimen overall forming a smooth, velvety cover without coarse ridges or scutes. Head almost completely scaled, including underside of snout, suborbital region, and mandibular rami. No scaly ridges on head; no scutes at terminal and lateral tips of snout. Scales on posterior margins of opercular bones lacking spinules (or absent?). Scales absent on gular and branchiostegal membranes, over su- pranasal ridge, dorsal edge of orbit, on posterior portion of median nasal ridge, and in rather wide band surrounding postorbital sensory canal. Scales greatly enlarged over postorbital canal and posterior portion of interorbital space to level of anterior margin of pupil, but very small anteri- orly on snout, especially along leading edges, with some scales nonimbricate. Elsewhere on head, scales intermediate and about equal in size. Body scales large, covered with numerous slender, sub- erect, slightly curved spinules in quincunx order. Light organ externally consists of a teardrop- shaped anterior dermal window between mid- base of pelvic fins, connected by a narrow isth- mus to posterior dermal window at anterior end of periproct. Anus much closer to pelvic fin bases than to origin of anal fin. Swim bladder well developed. Pyloric caeca short, unbranched. Ra- diographs of two specimens showed abdominal vertebrae 12; anal pterygiophores before first haemal spine 12. Origin of first dorsal fin opposite that of pelvic fin; pectoral fin situated slightly anterior to these; anal fin origin slightly behind vertical through last ray of first dorsal. Second unbranched ray of first dorsal high, not especially stout, terminating in a short filament; leading edge smooth except for distal one-fifth or so where numerous small, inconspicuous denticles present. Rays of second dorsal moderately developed, but much finer and shorter than opposites of anal fin. Interdorsal space about equal to snout length. Pectoral fins long, more than two-thirds HL, situated well be- low level of top of gill opening; pelvic fin long, outer ray filamentous and extending well past anal fin origin. Color in alcohol overall black to brownish black. Scale pockets on head and body of largest specimens dark blue with black margins; in smaller specimens most dorsal surfaces of head, snout, and suborbital grayish. Opercle, gular and gill membranes, and jaws black; in small speci- mens leading edge of snout and rim of olfactory capsule faintly blackish; oral cavity mostly black, but floor anteriorly whitish; branchial cavity black except pale anteriorly on isthmus and gill fila- ments. All fins black. Holotype generally lighter in color than above description suggests, but this probably a result of fading, which was not so evident in other specimens. SizE.—To more than 39 cm. DIsTRIBUTION.— Nazca and Sala y Gomez ridges and Hawaiian Islands, in 340-950 m. REMARKS.—The collection of many small ad- ditional specimens of this species, originally de- scribed from a single large individual, has shown its similarity to members of the genus Ventri- fossa. This supports moving Pseudocetonurus even farther from Cetonurus, with which it was initially loosely grouped, but ‘“‘without ... in- dicating a close kinship”’ (Sazonov and Shcher- bachev 1982:6). In fact, aside from the lack of a developed lateral line, the expanded interorbital, suborbital and preopercular regions, the rela- tively large head, and the high snout, there is little to suggest that Ventrifossa and Pseudoce- tonurus are much different. The dark color, small eyes, and enlarged sen- sory canals suggest a bathypelagic existence, and the capture of many specimens in a midwater SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES Wd) trawl over the Nazca Ridge lends support to this idea. The swim bladder is rather well developed, nonetheless, and in large adults it is enveloped in a tough tunica externa and filled with spongy tissue. The rete-gas gland complex of a small 150 mm specimen was well developed and consisted of a straplike rete that terminated posteriorly in a tightly appressed pair of flattened, horse-hoof- shaped gas glands. The rete bundle did not ap- pear to be separated into two bundles, although closer inspection may show them to be so, as it is in other members of the tribe Malacocephalini. The discovery of four specimens in good con- dition from the Hawaiian Islands was surprising and suggests the possibility of a much wider dis- tribution for this species. The specimens were dip netted at the surface in 1950 after having been killed by the lava flow from a volcanic erup- tion of Mauna Loa on the Island of Hawaii. Gos- line et al. (1954) and Gosline (1954) reported on the kill and collection of the fishes from this erup- tion. MATERIAL EXAMINED.—Sala y Gomez Ridge: Holotype: ZMMGU P16011 (54 mm HL, 257+ mm TL); 25°20.2’S, 93°35.5'W; 850-860 m. OTHER MATERIAL: Nazca Ridge: ZMMGU 17743 (19:24— 41.5 HL, 93+-212+ TL) and CAS 67409 (10:29.5-38.5 HL, 125+-184+ TL); 340-780 m; IKMWT; Prof. Mesiatzev cr. 13, sta. 44. ZIN uncat. (72 HL, 393+ TL); 950 m; bottom trawl; Hercules tr. 40. ZMMGU 18127 (26.7 HL, 126+ TL); 900-920 m over bottom depth of 1,270-1,200 m; IKMWT; R/V Prof. Shtokman cr. 18, sta. 1856. Hawaiian Islands: LACM 45410-1 (4:39-45 HL, 209-231 TL); lava flow kill, Kona, Ha- waii; Y. Yamaguchi, collector; 7.VI.1950. Trachonurus Giinther, 1887 Type species: Coryphaenoides villosus Giinther, 1877, by monotypy. D1acnosis.—Macrourines with 7 branchios- tegal rays. Teeth in narrow bands in both jaws. Scales bristly, covered with short to moderately long, erect, conical spinules; head entirely scaled, including patches on gular and (in some) bran- chiostegal membranes; scales along second dor- sal and anal fins somewhat enlarged and thick- ened in most species. Second spinous ray of first dorsal fin slender, flexible, without serrations; 6- 8 segmented first dorsal rays; pelvic fin with 6 or 7 rays, its origin below base (usually midbase) of first dorsal. Anus in middle of broad oval periproct spanning most of distance between anal and pelvic fins, a broad, naked anterior extension ending at pelvic fin bases. Abdominal vertebrae 12 or 13. Swim bladder with 2 retia mirabilia. Pyloric caeca 6-14, stubby to moderately long and thick. Color blackish to dark brown. SizE.—To at least 64 cm. DISTRIBUTION. — Worldwide in tropical to warm-temperate waters, but not recorded from continental slopes of eastern Pacific. REMARKS.—The genus constitutes a small group of four nominal species, all of which were synonymized by Marshall (1973) with Tracho- nurus villosus. A closer re-examination of spec- imens from Hawaiian and Atlantic waters may require a re-evaluation of the status of T. sen- tipellis Gilbert and Cramer, 1897, and T. sulcatus (Goode and Bean, 1885). Study (by TI) of old Albatross collections from the Philippines and recent collections from Australia and New Zea- land suggests the possibility of as many as four undescribed species represented. Shcherbachev et al. (1979) reported numerous specimens of 7. villosus from the Indian Ocean; our subsequent examination of this material suggests that some of the specimens may represent undescribed spe- cies. Trachonurus villosus (Giinther, 1877)? Coryphaenoides villosus Giinther, 1877:441 (south of Yeddo [=Tokyo)). Macrurus (Trachonurus) villosus: Giinther 1887:142, pl. 36, fig. B (lectotype, BMNH 1887.12.7.105, here designated; s. of Yeddo, Challenger sta. 232 in 345 fm [631 m]. Paralec- totype, BMNH 1887.12.7.106; s. of Philippine Is., Chal- lenger sta. 214 in 500 fm [914 m]). Trachonurus villosus: Parin 1990:17 (listed from vicinity of Nazca and Sala y Gomez ridges). D1AGNosis.— Orbit about 15-16% HL; sub- orbital width about half orbit diameter. Outer premaxillary teeth slightly enlarged. A well-de- veloped grooved lateral line; scale spinules rel- atively short and stout; diagonal scale rows from hind edge of first dorsal to anal fin more than 25; gular and branchiostegal membranes heavily scaled. Rays 1P. 116-118. Pyloric caeca moder- ately long, not stubby. REMARKS.— The two specimens from the Sala y Gomez Ridge appear quite similar to speci- mens of Trachonurus from the Gulf of Mexico and Caribbean Sea, and some specimens from Australia. A more thorough study is necessary to determine if they are the same. They differ from a syntype of 7. villosus (BMNH 1887.12.7.105; 26 mm HL, 177+ mm TL), taken south of Yeddo (Tokyo) in having a smaller or- 78 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 bit, broader suborbital, longer distance orbit-pre- opercle, longer upper jaw, and more pectoral fin rays. It cannot be discounted that the exception- ally large size of these Sala y Gomez specimens may account for these differences in proportional measurements. A second syntype (BMNH 1887.12.7.106) from south of the Philippines ap- pears to be a different species with a longer upper jaw, longer barbel, longer first gill slit, wider 1D.— 2D. interspace, and larger scales. (This specimen was in poor shape when examined by TI in Oct. 1986.) The features characterizing 7. villosus have not been properly assessed, and the relationships of the species with others of the genus have not been adequately determined. Because two spe- cies are probably represented in the syntypic se- ries of T. villosus, the Yeddo syntype is here des- ignated as the lectotype. It was obviously the one used by Giinther in his original description, which listed only “‘south of Yeddo”’ as the type locality, and it was the specimen figured in his report on the Challenger fishes (Giinther 1887:142, pl. 36B). The Sala y Gomez specimens differ from Ha- waiian specimens of the genus (nominally 7. sen- tipellis) in having a deeper body, a more rounded, less conical snout, more diagonal scale rows be- low the hind margin of first dorsal, more nu- merous pectoral fin rays, smaller orbit, narrower suborbital, and more heavily scaled gular and branchiostegal membranes. MATERIAL EXAMINED. —(2 spec., both Prof. Mesiatzev) Sala y Gomez Ridge: ZMMGU 17744 (1:124 mm HL, 637+ mm TL); 1,070-1,100 m; cr. 13, tr. 10. ZMMGU 17745 (1:65 HL, 3035> me) 1e050imeacr 18s tr 14. Ventrifossa Gilbert and Hubbs, 1920 Type species: Coryphaenoides garmani Jordan and Gilbert, 1904, by original designation. DraGnosis.— Macrourines with 7 branchios- tegal rays. Head smoothly rounded, without coarse, angular ridges, and lacking stout, modi- fied scales; snout without a terminal scutelike scale except in subgenus Sokodara. Mouth mod- erate to large, upper jaw 35-55% HL. Jaw teeth in narrow to moderately broad bands in upper jaw, outer series enlarged; lower jaw teeth small, none enlarged, in | to several irregular series laterally. Gill rakers 12—20 total on inner series of first arch. Branchiostegal membranes naked. Body scales small, densely covered with slender conical or triangular spinules; no transverse ridg- es on anterior field. Spinous second ray of first dorsal finely serrated along leading edge or smooth. Anus removed from anal fin origin and closer to pelvic fin; periproct oval-shaped, ex- tending forward to small fossa between pelvic fin bases. Pyloric caeca about 34-75. SizeE.—To at least 53 cm. DISTRIBUTION. — Tropical waters of the Indo- Pacific, but with an isolated pocket of two species in the tropical western Atlantic (Gulf of Mexico and Caribbean Sea). Species generally confined to upper continental slope depths ranging from about 200 m to more than 1,000 m, but most often captured at 500-800 m. REMARKS.—A medium-sized group with 24 named species and several more known to us, yet to be described. Sazonov (1985) removed the subgenus Lucigadus Gilbert and Hubbs, 1920, with its five species, and elevated it to the status of genus. Iwamoto (1979) has recently treated the group in some detail. Ventrifossa johnboborum Iwamoto, 1982 (Figure 30) Ventrifossa johnboborum Iwamoto, 1982 (in part):55-61, fig. 1 (type locality Bismarck Sea; non-type specimens are V. fusca Okamura, 1982). Shcherbachev et al. 1986:202 (3 spec. from western Indian Ocean on Freda Seamount, 810 m). Shcherbachev 1987:7 (listed from Indian Ocean). Parin 1990: 17 (listed from Sala y Gomez Ridge). Kotlyar and Parin 1990:107, fig. 4a (otolith). Malacocephalus sp.: Parin et al. 1981:13 (1 spec., ZMMGU 17746, here reported, from Ichthyandr cr. 5, tr. 53). DiaGnosis.—A broad-headed species of sub- genus Sokodara (see Iwamoto 1979). Interorbital width 27.8-33.4% HL, slightly less than orbit diameter, which goes about 3 in head; upper jaw 39-45% HL; barbel 13-19% HL. Scales small, 65-71 from lateral line origin over distance equal to pre-1D. length. Dorsal spines weakly serrated in adults; 1D. IJ,10-11 (13); 1P. 121-123; V. 9- 10. COUNTS AND MEASUREMENTS. —(5 specimens; see also DiAGnosis) GR-I (outer/inner) 7—10/12- 14 total; GR-II (outer/inner) 10—13/1 1-14 total; caeca 65-73 (2 spec.); vertebrae 14-16. Total length 280+-391+ mm, HL 66.5-85.1. The following in percent of HL: postrostral 73.8— 76.7; snout 27.8-31.6; preoral 19.2—23.8; orbit 31.4-35.7; suborbital 10.5-—14.0; postorbital 38.3-43.6; orbit-preop. 39.1—42.7; gill slit 20.4— 28.7; pre-1D. 101-110; pre-V. 97-114; pre-A. 127-137; V.-A. 33-49; base 1D. 23.8-31.0; 1D.- 2D. 33.1-45.1. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 79 Ficure 30. Ventrifossa johnboborum. Sala y Gomez Ridge in 540 m, Ichthyandr cr. 5, sta. 53. Drawn by Tomio Iwamoto. Scale bar equals 25 mm. DESCRIPTION. —Greatest body depth slightly more than postrostral length of head, about 1.3- 1.4in HL, more than 6 in TL. Head width more than half its length. Snout broad, width at lateral angles (anterior to nostrils) slightly less than both interorbital width and snout length; moderately protruding beyond large subterminal mouth, length less than orbit diameter. Orbits large, 2.7— 3.2 in HL, about 1.2-1.3 in postorbital length. Suborbital region gently rounded, separated into upper and lower regions by a ridge-line that runs close to orbital rim anteriorly but diverges widely posteriorly; scales on suborbital not modified into heavy tubercular scutes. Preopercle broadly rounded; interopercle exposed along posterior and ventral margins of preopercle. Upper jaw 2.2- 2.4 into HL; maxillary extends to below posterior one-third of orbits; rictus not restricted poste- riorly. Barbel small, slender, about 1.8—2.5 in orbits. Premaxillary dentition in holotype in narrow band 4-5 irregular rows wide near symphysis, narrowing posteriorly. In Sala y Gomez speci- mens, teeth in 2 or 3 irregular rows, outer series slightly enlarged. A wide toothless gap at pre- maxillary symphysis. Mandibular dentition small, in 2 or 3 irregular series that taper to 1 row posteriorly; inner series larger than outer series. Scales all small, densely covered with small, slender, conical, relatively erect spinules ar- ranged in irregularly quincunx or widely diver- gent, V-shaped rows. Scales uniformly cover all of head except gill membranes, lips, and nostril membranes. A row of small scales on anterior orbital rim between posterior nostril and orbit. Scale pockets present on exposed interopercle surfaces, although no scales remain there on ex- amined specimens. No stout spiny ridges or heavily modified scales on head or body, except tip of snout has small, blunt, tubercular median scale (not noticeably enlarged or protruding). Fins unscaled. Pyloric caeca about 65 in holotype, 73 in a male 325+ mm TL. Caeca short, about equal to suborbital width, branched at base. Retia 2, long, slender; gas glands small. Paired drumming mus- cles in 325 mm male large, covering almost half anteroventral surface of swim bladder. Stomach in that specimen contained remains of cepha- lopods, including beaks, eyes, and body parts. Light organ with 2 dermal windows, the anterior one between pelvic bases small and round. First dorsal and paired fins moderately large for genus. Height of first dorsal fin about equal to postrostral length of head; spinous second ray weakly serrated distally, scarcely prolonged be- yond succeeding rays; when laid back, longest rays barely reach or extend slightly beyond origin of second dorsal fin. Base of first dorsal about 1.1-1.6 into space between first dorsal and sec- ond dorsal. Origin of pelvic fin anterior to that of pectoral, which is anterior to that of first dor- sal. Outer pelvic fin ray slightly produced into a filament that extends to about 3rd or 4th anal ray. Origin of anal fin under hind edge of first dorsal. Length of pectoral about equal to length of snout plus orbit. Color in alcohol overall grayish brown to swar- thy, bluish over abdomen; fins black. Gill mem- branes and entire oral cavity black. Gill chamber pallid over epihyal, ceratohyal, anteroventral 80 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 Ficure 31. m, Prof. Shtokman cr. 18, sta. 2018, 7 May 1987: (a) specimen 67.5 mm HL; (b) 69 mm HL. Photograph by Sergei Dudarev. portion of cleithrum, and gill filaments; remain- der dark dusky to blackish. Gill arches and rak- ers, leading edge of snout, and periproct region blackish. Peritoneum and stomach black. Anal rays blackish anteriorly fading to dusky poste- riorly. Barbel overall pallid but base blackish. Lips edged with thin black margin. COMPARISONS AND REMARKS.— Ventrifossa johnboborum is most closely related to V. fusca Okamura, 1982, described from the Kyushu-Pa- lau Ridge, and V. misakia Jordan and Gilbert, 1904, from Japan. Ventrifossa johnboborum can be distinguished from the other two species by its blackish lining of the oral cavity (compared with pallid or white) and pelvic fin ray count (9- 10 cf. 8). It is further distinguished from V. fusca by its larger orbit (2.9-3.1 in HL, cf. 3.4-3.6), shorter interspace between dorsal fins (2.3-3.0 in HL cf. 1.8-2.1), and longer barbel (1.8—2.7 in orbit diameter, cf. 3.8-4.1). The original description of V. johnboborum was based on the holotype and three other spec- imens not designated as type-specimens. These other specimens, one from off Batag Island in the Philippines, the other two from the South China Sea, were sufficiently different from the holotype so as to cast doubt as to their conspecificity, but their differences (V. 8, barbel length 11-12% HL, upper jaw length 36-39% HL) did not justify description of a second species. The specimens otherwise agree closely with V. johnboborum and also V. fusca, especially concerning the pelvic ray count, and proportional measurements of the Ventrifossa macrodon new species; non-type material, ZMMGU 18139, from Sala y Gomez Ridge, in 730-790 barbel and upper jaw. They disagree with V. fusca in having dark mouth linings (which questions the validity of that character in V. johnboborum), but otherwise seem so similar that they are ten- tatively identified with that species. (Dr. Osamu Okamura, BSKU, kindly loaned two paratypes of V. fusca for comparison with southeastern Pa- cific specimens of V. johnboborum.) The subgenus Sokodara is represented by three Pacific species (two in the western Pacific, one in the South Pacific), and two Indian Ocean spe- cies. Ventrifossa misakia is known only from off Japan; V. fusca from the western North Pacific south of Japan; V. johnboborum from the Bis- marck Sea, the western Indian Ocean, and on the Sala y Gomez Ridge; and V. nasuta Smith, 1935, from South Africa and Mozambique. The dis- tributions of V. fusca and V. johnboborum will undoubtedly be extended with more collecting, and it should be interesting to see if and where the two distributions abut or adjoin. SizeE.—To at least 47.5 cm (Shcherbachev et al. 1986). DISTRIBUTION. — Known from the western In- dian Ocean, the Bismarck Sea, and the Sala y Gomez Ridge in 540-810 m, but probably ex- tends across the intervening areas on seamounts and oceanic ridges of the Indian Ocean and South Pacific Ocean. MATERIAL EXAMINED.— Bismarck Sea: Holotype: AMS 1.15602-002 (85.1 mm HL, 360+ mm TL). Sala y Gomez Ridge (5 spec.): ZMMGU 17746 (70.3 HL, 340+ TL); 540 m; Ichthyandr cr. 5, tr. 53. ZMMGU 17747 (2:66.5-70.7 HL, SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 81 FiGure 32. Scale bar equals 25 mm. 280+-325+ TL); 610-620 m; Hercules tr. 70. ZMMGU 17748 (2:72.5-78.5 HL, 361+-391+ TL); 750-800 m; Prof: Shtok- man cr. 18, sta. 1996. Ventrifossa macrodon new species (Figures 31-33) Ventrifossa sp. 1: Parin et al. 1981:12 (counts and measure- ments of 24 specimens here described as new). Ventrifossa sp. nova 1 Sazonov and Iwamoto”: Parin 1990:17 (listed from Sala y Gomez Ridge). Parin et al. 1990:42 (stom- ach contents). Kotlyar and Parin 1990:106, fig. 4b (otolith). Diacnosis.—A species of subgenus Ventrifos- sa with moderately deep body (depth 76-87% HL). Interorbital 20-24% HL, much less than orbit diameter, which goes about 3 in HL; upper jaw 46-50% HL; barbel less than snout length, 23-27% HL; GR-I 17-19 total (inner). No en- larged spinules on body scales. Second spinous ray of first dorsal fin not serrated; 1D. II,9-11; 1P. 119-126; V. 9-10. Leading edge of snout, su- pranarial ridges, and median nasal ridge black- ish; first dorsal fin blackish, without distinct black blotch. COUNTS AND MEASUREMENTS. —(24 speci- mens; data for holotype in square brackets) 1D. II,9-11 [11,9]; 1P. 119-126 (x = 122.07; SD = 2.129) [123/124]; V. 9-10 (11) [10]; GR-I (outer/ inner) 8-14 [11]/(2-3) + (0-1) + (13-16) (16— 19 total) [3 + 14], GR-II (1-3) + (0-1) + (12- 15) (15-18 total) [3 + 14]/(2-3) + (0-1) + (12- 15) (15-18 total) [3 + 14]; scales below midbase 1D. 7.5-10 [8]; below 2D. 8.5—11 [10]; caeca 50 in 54.5 mm HL paratype; vertebrae 12 (5 spec.). Total length 157+-—363+ [350] mm, HL 26.7- 72 [67.7] mm. The following in percent HL: post- rostral 73.7-80 [77.1]; snout 23.2-28.9 [25.6]; preoral 6.5—10.5 [7.8]; orbit 29.2-43.0 [30.7]; in- Ventrifossa macrodon new species; paratype, CAS 51796, from Nazca Ridge in 540 m. Drawn by Amy Pertschuk. terorbital 19.6-26.2 [21.7]; suborbital 9.4—13.2 [11.4]; postorbital 39.0-46.7 [41.9]; orbit-preop. 38.0-46.6 [44.2]; upper jaw 46.2-54.8 [51.7]; gill slit 22.5-29.1 [22.2]; barbel 22.7-33.9 [29.2]; body depth 73-88 [87]; pre-D. 98-118 [110]; pre- V. 100-123 [105]; pre-A. 127-152 [137]; V.-A. 26-43 [38]; height 1D. 46-77 [—]; 1D. base 21.8- 29.1 [27.0]; 1D.-2D. 48.6-76.8 [66.2]; 1P. 49- 61 [53.5]; V. 30-46 [32]. DESCRIPTION. — Head and trunk compressed, relatively deep; greatest width of head about equal to width of trunk over pectoral fin bases, slightly more than postorbital length, about three-fifths body depth. Greatest body depth more than 6 in TL, 1.1-1.3 in HL. Snout low, blunt, scarcely protruding beyond the large mouth; distance be- tween supranarial ridges less than (1.1—1.3 into) least interorbital width, 1.4-1.8 into orbit di- ameter. Orbits large, longer than snout, about 3 (2.8-3.4) in HL, but about 2.3-2.7 in juveniles with HL less than 35 mm. Interorbital space flat to slightly convex, width slightly more than that across supranarial ridges. Ridges on head not well developed; rounded or smoothly angular where present. Suborbital region narrow, gently rounded; a ridge line apparent but not demar- cated by stout scales (as in species of Nezumia). Preopercle deep and broadly lobate; posterior border slightly inclined forward and with a shal- low inflection. A relatively deep notch formed along posterior margin of subopercle. Interoper- cle exposed along its posterior and ventral bor- ders. Strongly toothed jaws unrestricted at angles by lip folds; extend to below level of posterior third of orbits. Barbel long, slender, slightly lon- ger than interorbital width. Gill membranes of opposite sides narrowly united under posterior third of orbits, forming a narrow free fold. 82 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 BARBEL LENGTH (mm) 40 50 60 70 80 HEAD LENGTH (mm) Ficure 33. Comparison of barbel lengths in three species of the Ventrifossa atherodon species complex. Ventrifossa macrodon (circles), V. atherodon (squares), V. sp. (South China Sea) (triangles). Premaxillary dentition consists of moderately wide band of small teeth (4-6 rows wide) flanked by a row of widely spaced, enlarged, conical teeth with arrowhead-shaped tips. Mandibular teeth smaller than enlarged premaxillary teeth and ar- ranged in basically 1 row anteriorly, but 2 irreg- ular rows laterally (small but prominent teeth with arrowhead-shaped tips form a continuous regularly spaced series interspersed laterally and posteriorly with | to 3 much smaller teeth that are not exposed above the gum papillae). Thin, finely spinulated scales uniformly cover all of head except over nostril membranes, lips, fins, and gill membranes. Head scales densely covered with short, conical, erect spinules (none enlarged) giving a shagreenlike feel to surface. No scutelike scales on ridges; no enlarged ter- minal scute at snout tip. Scales over suborbital shelf small, generally similar to other scales of head. Body scales rather deciduous; those along base of second dorsal fin lack spinules or only partially covered with spinules; those more ven- trad densely covered with minute, fine, conical, reclined spinules, arrayed in roughly quincunx pattern; 50-70 spinules per scale on larger of these scales in specimens 63.3—70.5 mm HL. Spi- nules of head and body scales flecked with me- lanophores. Fins unscaled except at base. Vent region as for most others of genus (see Marshall 1973:655, fig. 49; Okamura 1970a:74, text-fig. 32). Light organ and alimentary canal similar to that illustrated by Okamura (1970a, text-fig. 34). A small, ill-defined posterior dermal window immediately before anus, connected by a narrow isthmus of black skin to a small oval anterior dermal window situated in a shallow fossa between bases of pelvic fins. Pyloric caeca numerous, thin walled, short (length less than pupil diameter), branched at bases. Ovaries of 63.6 mm HL female (CAS 51795) and 63.3 mm HL female (CAS 51796) large, but eggs small, 0.3-0.4 mm diameter. First dorsal fin relatively low, height less than postrostral length of head, about equal to length pectoral fin; first ray of first dorsal fin rudimen- tary, closely appressed to spinous second ray, which is slender and without serrations; second dorsal fin poorly developed over entire length, begins far posterior to first dorsal. Color in alcohol somewhat swarthy overall; SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 83 abdominal region bluish to violet. Ridges of snout along leading edge, over medial nasal ridge, and over supranarial ridges marked with prominent black edges. A diffuse transverse band extends across each parietal ridge at posterior end of su- praoccipital region. Behind level of orbits, pa- rietal region blackish. Suborbital region darker dorsally than ventrally. An irregularly triangular area of black pigmentation extends posteroven- trally from orbit to preopercle. A blackish hor- izontal streak on ventral aspect of eye just below cornea in some but not in others. Lips, gular membrane, and gill membranes blackish; barbel blackish at base, grayish to pale distally. Papillae of gums blackish; oral valve on roof of mouth edged with black; most linings of mouth whitish; pharynx gray. Gill filaments, gill arches, and hy- oid region of gill cavity pale, remainder of gill chamber blackish. Peritoneal lining finely pep- pered with small melanophores ventrally, dusky to blackish dorsally; stomach pale. Fins blackish; anal fin paler posteriorly. Foop.—Stomachs of two individuals (CAS 51795 and 51796) contained unidentifiable re- mains of a fish and a shrimplike crustacean. Pa- rin et al. (1990:42) reported the mysid Paralo- phogaster glaber in the diet of 31% of their 13 specimens; a galatheid crab, Munida sp., in 15%. Penaeids, copepods (Pleromammama sp.), am- phipods (Eurisidae), and the fish genus Cyclo- thone were also eaten. S1zE.—To about 40 cm. DIsTRIBUTION. — Known only from the Sala y Gomez Ridge in 345-770 m. ETyMoOLoGy.— The name macrodon (latinized Greek macros, long, and odus, tooth) refers to the presence of enlarged teeth in the outer row of the premaxillary, a not uncommon character in this genus. The name, however, also reflects the close relationship of the new species to V. macroptera Okamura, 1982, and V. atherodon (Gilbert and Cramer, 1897) by the combination of elements forming each name. COMPARISONS AND REMARKS.— The new spe- cies is most closely related to V. atherodon, V. macroptera, and an undescribed species from the South China Sea. The four can be distinguished from each other by characters given in the key that follows. In addition, the new species has a somewhat longer barbel than the other three: 23- 34% HL cf. 12-24% in V. atherodon, 18-24.5% in V. macroptera, and 17-22% in the South Chi- na Sea species (see also Fig. 33). Key to the species in the Ventrifossa atherodon complex la. Scales below second dorsal fin with dis- tincily-enlarged(spinules sa nt pte ce oer tee species (South China Sea) 1b. No enlarged spinules on scales 2 2a. Pectoral fin 1.3-1.5 in HL; orbits 25- 32. )90sEUE 2 seek, fk eee eee Bicfies Patel macroptera (Kyushu-Palau Ridge) 2b. Pectoral fin 1.7—2.0 in HL; orbits 20-43% 3a. Length barbel 23-34% HL; outer gill slit 22-29% HL (usually 22-26%) wuss... Macrodon (Sala y Gomez Ridge) 3b. Length barbel 12-24% HL; outer gill slit 26-32% HL (usually 26-28%) Sa ees atherodon (Hawaiian Islands) The four species constitute a well-defined spe- cies group in the genus Ventrifossa distinguished by a non-serrated spinous dorsal ray, low blunt snout, high gill raker counts, rather large mouth, and 9-10 pelvic fin rays. The subgeneric name Atherodus Gilbert and Hubbs, 1920, is available for this group. Although a count of V. 9 is com- mon among species of Ventrifossa, a count of 10 is normally encountered only in this species group and in V. ctenomelas (usually 9-10), a species endemic to the Hawaiian Islands where it is sym- patric with V. atherodon. Material Examined.—(138 spec.) all from Sala y Gomez Ridge: Holotype: ZMMGU 18132 (67.7 mm HL, 350 mm TL); 580- 564 m; Prof. Shtokman cr. 18, sta. 1964. Paratypes (39 spec.): ZMMGU 17751 (2:38-55 HL, 192- 277+ TL); 770 m; Ichthyanadr cr. 5, tr. 52. CAS 51796 (2:63- 69 HL, 290+-305 TL) and ZMMGU 17752 (55 HL, 307+ TL); 540 m; Ichthyanadr cr. 5, tr. 53. ZMMGU 17753 (9:30.5- 56 HL, 169+-307 TL) and CAS 51794 (2:48-60 HL, 265- 290 TL); 535-575 m; Ichthyanar cr. 5, tr. 54. ZMMGU 17754 (9:36.3-69 HL, 201+-367+ TL) and CAS 51795 (3:38-70.5 HL, 210+-353+ TL); 345-540 m; Ichthyandr cr. 5, tr. 55. ZMMGU 17755 (2:59-72 HL, 325+-350+ TL); 550 m; As- tronomer tr. 104. ZMMGU 18133 (9:54-69 HL, 245+-350 TL); same data as for holotype. Other material: ZMMGU 18071 (3:26.7-50 HL, 157-260 TL); 565-555 m; Prof. Mesiatzey cr. 13, tr. 1. ZMMGU 18072 (4:34.3-53 HL, 177-271 TL); 550-560 m; Prof. Mesiatzeyv cr. 13, tr. 2. ZMMGU 18073 (2:39.5-46.3 HL, 184+-246 TL); 530 m; Prof. Mesiatzev cr. 15, tr. 49. ZMMGU 18134 (11:32- 76 HL, 168-404 TL); 562-545 m; Prof. Shtokman cr. 18, sta. 1965. ZMMGU 18135 (33.7 HL, 181 TL); 570-580 m; Prof. Shtokman cr. 18, sta. 1971. ZMMGU 18136 (13:29-55 HL, 160-260+ TL); 563-590 m; Prof. Shtokman cr. 18, sta. 1976. ZMMGU 18137 (36:29-61 HL, 132+-288+ TL); 545-600 m; Prof. Shtokman cr. 18, sta. 1977. ZMMGU 18138 (11: 29.5-73 HL, 169-383 TL); 750-600 m; Prof. Shtokman cr. 84 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 FIGURE 34. Ventrifossa obtusirostris new species; holotype, ZMMGU 18130 (49 mm HL), from Sala y Gomez Ridge in 739- 760 m, Prof. Shtokman cr. 18, sta. 1996. Photograph by Sergei Dudarev. 18, sta. 1996. ZMMGU 18139 (17:26.3-73 HL, 131+-359 TL); 730-790 m; Prof. Shtokman cr. 18, sta. 2018. Ventrifossa obtusirostris new species (Figure 34, 35) “Ventrifossa sp. nova. 3 Sazonov and Iwamoto”: Parin 1990: 17 (listed from Sala y Gomez Ridge). D1aGnosis.—A species of subgenus Ventrifos- sa with snout blunt and rounded, scarcely pro- duced beyond mouth, its length 26.5-28.3% HL; upper jaw 47.7-52% HL; barbel long, 137-166% of orbit diameter, 48-54% of HL; pectoral fins long, 72-78% HL. Color dark overall, first dorsal blackish distally, paler near base; anterior por- tion of anal fin with black margin. COUNTS AND MEASUREMENTS. —(from 5 spec- imens; holotype data in square brackets) 1 D. II,11 [11]; 1P.121-123 [121]; V. 9-10 [9/10]; GR-I (out- er/inner) 8-10 [9]/(1-2) + (12-14) [2 + 13] (14— 15 total), GR-II 2 + 12/(1-2) + (11-13) [2 + 11] (11-14 total); scales below 1D. 10-13 [11], below midbase 1D. 7.5—9.5 [9.5], below 2D. 9- 10 [9]. Total length 249+-294 mm, HL 44.0-51.5 mm. The following in percent HL: postrostral 74.8-78.6 [78.3]; snout 26.5—28.1 [27.2]; preoral 11.7-17.3; interorbital 25.7-27.4 [26.5]; orbit 33.5-36.3 [33.5]; suborbital 13.5-14.3 [13.5]; postorbital 42.7-46.3 [46.3]; orb.-preop. 40.9- 45.5 [44.6]; upper jaw 47.7—52.0 [51.1]; barbel 47.7-54.3 [54.3]; gill slit 26.3-30.4 [30.4]; pre- 1D. 112-122 [123]; pre-A. 128-133 [133]; pre- vent 111-121 [119]; pre-V. 102-110 [111]; V.- A. 26-31 [27.8]; body depth 80-100 [100]; 1D.- 2D. 39-54 [54]; 1D. base 30-36 [32.6]; 1D. height 77-89 [80]; 1P. 72-78 [78]; V. 39-45 [45]. DESCRIPTION. — Body deep, greatest depth al- FiGure 35. 760 m; Prof. Shtokman cr. 18, sta. 1996. Drawn by Tomio Iwamoto. Scale bar equals 25 mm. Ventrifossa obtusirostris new species; paratype, ZMMGU 17756 (51.5 mm HL), Sala y Gomez Ridge in 739- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 85 most equal to HL, tapering abruptly behind ab- domen before leveling off to long tail. Head 5.5- 6.0 into TL, width slightly more than postorbital length of head. Snout blunt, low, scarcely pro- truding beyond mouth, its length about equal to its width across lateral angles, about equal to orbit diameter. Orbits large, about 1.4 into post- orbital length, uppermost rim barely entering dorsal profile of head. Interorbital essentially flat, width less than orbit. Suborbital region almost vertical, gently curved in cross-section, the ridge scarcely discernible in well-preserved specimens, the dorsal shelf uniformly wide. Preopercle mar- gin and ridge broadly rounded, not formed into a lobe; interopercle broad, exposed along pos- teroventral margins. Upper jaw about 2 in HL, maxillary extends to below posterior one-third of orbits; rictus extends to below midorbit. Bar- bel notably long, moderately thick, extending be- yond junction of gill membranes, its length more than postorbital, about 2 in HL. Teeth all small, short, conical, rather bluntly pointed, in a broad band in premaxillary with a slightly enlarged outer series. Mandibular teeth similarly small, barely perceptible among gum papillae, in a long narrow band. Scales small, uniformly covering all of body and head except fins, gill membranes, lips, nasal membranes. Scales thin, covered with short, fine, black, partially reclined spinules, rather sparsely and irregularly arranged in an irregularly quin- cunx pattern. Abdominal area anterior to anal fin short, dis- tance pelvic fin base to anal fin origin much less than distance isthmus to pelvic fin base. Vent close between pelvic fins, slightly behind their insertions; periproct region with anterior exten- sion to small teardrop-shaped dermal window situated anterior to pelvic fin origin. First dorsal fin moderate in height, slightly more than postrostral length of head, its second spinous ray finely serrated along leading edge, rather flexible and weak, its tip scarcely pro- longed; second dorsal fin rudimentary for almost entire length; anal fin well developed. Pectoral fin extends to end of abdomen; pelvic fin rather small, outer ray thin, short, extending past base of third or fourth anal fin ray. Origin of pelvics far forward, below hind margin of operculum; pectoral fin origin slightly behind that; first dorsal origin about on same vertical as pectoral or slightly behind, anal fin origin below or forward of hind edge of first dorsal. Color in alcohol dark brownish dorsally, paler on body below midlateral line; bluish to dark gray or blackish over abdomen and head. Lead- ing edge of snout blackish, but inconspicuous because of dark ground color. Pectoral and pelvic fins black; first dorsal overall blackish, but some- what paler along base; anal fin pale except an- terior rays blackish distally. Barbel starkly con- trasting in white. Oral cavity completely pale or white except for thin dark edge of dorsal oral valve and gums-of lower jaws; lips black. Gill cavity black along inner wall and peripherally on outer wall; pale in remaining areas; gill arches, rakers, and filaments pale. S1zE.—To at least 30 cm. DIsTRIBUTION.—Known only from a single collection on the Sala y Gomez Ridge in 750- 800 m. EryMo_Locy.—From the Latin, obtusus, blunt, and rostrum, snout, in reference to the short blunt snout of the species as compared with others of the genus. The name is to be treated as a feminine adjective. COMPARISONS AND REMARKS. — It is somewhat perplexing that this species was collected only once in the NSG region, despite relatively ex- tensive trawling at appropriate depths. Whether this scarcity of collections can be attributed to unavailability of populations to the sampling gear, insufficient collecting within their primary range, or a general scarcity of individuals is not known. Relationships of V. obtusirostris appear to be closest to three western tropical Pacific species, all of which have a relatively long barbel (about one-half HL): V. saikaiensis Okamura, 1984, V. longebarbata Okamura, 1982, and V. rhipido- dorsalis Okamura, 1984. The first species differs from V. obtusirostris in the following: somewhat fewer pelvic rays (8-9), a broad area of spinule- less scales behind the first dorsal, a slightly short- er pectoral fin (61-72% HL), and a slightly short- er barbel (115-133% of orbit). Members of the last two species can be distinguished by their shorter pectoral fin (less than 70% HL). In ad- dition, V. Jongebarbata has a somewhat shorter postorbital length (39-41%) and narrower inter- orbital (23-24% HL). The long pectoral fin char- acteristic of V. obtusirostris is matched only in V. macroptera, but that species differs in its members having a shorter barbel (20-25% HL, less than orbit diameter), more gill rakers (GR-I 16-18 total inner), a smaller orbit (18-31% HL), and broader interorbital (30-32% HL). 86 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. N FiGure 36. m, Prof. Shtokman cr. 18, sta. 2018, 7 May 1987. Photograph by Sergei Dudarev. MATERIAL EXAMINED. —(all from Sala y Gomez Ridge) Ho- lotype: ZMMGU 18130 (46.0 mm HL, 275+ mm TL); Sala y Gomez Ridge, 750-800 m; Prof. Shtokman cr. 18, sta. 1996, trawl 29. Paratypes (same locality as holotype): ZMMGU 17756 (3: 44.5-51.5 HL, 258+-294 TL) and CAS uncat. (44.0 HL, 249+ Tie): Ventrifossa teres new species (Figures 36, 37) Ventrifossa sp. 2: Parin et al. 1981:12 (counts and measure- ments of 4 specimens from Sala y Gomez Ridge). “Ventrifossa sp. nova 2 Sazonov et Iwamoto”: Parin 1990:17 (listed from Sala y Gomez Ridge). Kotlyar and Parin 1990: 107, fig. 4b (otolith). D1aGnosis.—A species of subgenus Ventrifos- sa with slender body (greatest depth 57-75% HL). Interorbital width 18-24% HL, much less than orbit diameter, which goes about 3 in HL; upper jaw 40-45.5% HL; barbel less than snout length, 19-27% HL. First dorsal fin spine weakly ser- rated; 1D. II,9-10; V. 8-9; GR-I 13-15 total (in- ner). Leading edge of snout blackish; no black median nasal streak; first dorsal blackish, with- out distinct black blotch. COUNTS AND MEASUREMENTS. —(from 38 spec.; holotype data in square brackets) 1D. II,(8) 9- 10 [11,10]; 1P. 120-122 (rarely 119 or 123) [123]; V. 8-9 [9]; GR-I (outer/inner) 7—12 [10]/(1-2) + (O-1) + (10-13) [1 + 1 + 12] (12-15 total), GR- II (1-2) + (0-1) + (9-12) [2 + 11] (11-14 total)/ (1-2) + (0-1) + (9-12) [1 + 1 + 12] (11-14 total); caeca 34-36 (4 spec.); abdominal verte- brae 11. Total length 131-239 mm, HL 23.8-52.2. The following in percent HL: postrostral 72-78 [75.7]; snout 26.4—-31.8 [26.8]; preoral 1 1.2—17.6 [13.6]; orbit 31.6-37.3 [31.4]; interorbital 18.5—23.9 [18.1]; suborbital 7.2—12.3 [9.2]; postorbital 37.0- 41.4 [38.3]; orbit-preop. 32.9-38.5 [36.8]; upper jaw 38.7-44.7 [42.1]; gill slit 18.8-25.9 [21.1]; Ventrifossa teres new species; holotype, ZMMGU 18131 (52.2 mm HL), from Sala y Gomez Ridge in 730-790 barbel 19.4—26.7 [27.4]; body depth 57-75 [71]; pre-D. 103-112 [111]; pre-V. 99-122 [107]; pre- A. 126-159 [144]; V.-A. 27-49 [42]; 1D. height 47-68 [49]; 1D. base 20.5-30.2 [27.8]; 1D.-2D. 48-80 [63]; 1P. 49-61 [54]; V. 28-37 [34]. DESCRIPTION.—A_ slender-bodied species of Ventrifossa, width across pectoral fin bases about three-fourths body depth; greatest head width three-fourths its depth; greatest body depth about 1.3-1.8 into HL, 6.4—8.7 into TL. Nape and first dorsal fin base low. Snout low, conical in lateral profile; relatively blunt and narrow in dorsal view, width across anterolateral angles slightly greater than interorbital width; length about 1.2 into or- bit. Orbits large, about 3 in head, about 1.3 into postorbital length. Interorbital space probably relatively flat in live specimens, but slightly con- cave in examined preserved specimens. Subor- bital region traversed by a low, rounded ridge, the upper and lower surfaces of region meeting at an obtuse angle (in cross-section); upper shelf area not markedly narrowed anteriorly. Pre- opercle broadly lobate; interopercle broadly ex- posed posteriorly beyond preopercle. Upper jaw long, about 2.2—2.6 into HL; maxillary extends to below posterior one-third of orbits; rictus (mouth cleft) not restricted posteriorly. Barbel slender, moderately long, about equal to inter- orbital width. Dentition of premaxillary in a moderately wide band; inner teeth very small, bordered by an out- er series of slightly enlarged, widely spaced, con- ical, recurved teeth. Mandibular teeth all small, in 2 irregular series. Tips of outer premaxillary teeth and all mandibular teeth shaped like ar- rowheads. Scales uniformly cover all of head except for nostril membranes, lips, gill membranes, and a narrow median triangular area above upper lip. Head scales thin, densely covered with short, SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 87 FiGureE 37. Ichthyandr cr. 5, trawl 55. Drawn by Amy Pertschuk. Scale bar under head equals 10 mm, that under scale equals 1.0 mm. conical, erect spinules, none enlarged. No scute- like scales on ridges; no enlarged terminal scute at snout tip; scales over suborbital shelf not coarse or scutelike. Most body scales missing on type specimens; those remaining dorsally below sec- ond dorsal fin covered with fine, conical, reclined spinules, arrayed in roughly quincunx pattern; 24-26 spinules per scale on larger of these scales. Spinules of head and body scales flecked with melanophores. Fins unscaled except at base. The 43.5 mm HL specimen from CAS 51797 had large ovaries filled with eggs of different sizes, some measuring as much as 1.4 mm in diameter. Pyloric caeca relatively few for a Ventrifossa (ac- curate number could not be determined), length slightly less than interorbital diameter. Vent- periproct region much like that of other members of genus; anterior dermal window of light organ small, circular, slightly anterior to pelvic bases. First dorsal fin relatively short, length less than postrostral length of head; first spinous ray close- ly appressed to second spinous ray and not pro- Ventrifossa teres new species; paratype, ZMMGU 17759 (39.5 mm HL), Sala y Gomez Ridge in 354-540 m, truding through integument. Second spinous ray weakly denticulate, the short denticles confined to distal half. Paired fins moderate in size, neither pelvics nor pectorals much longer than postor- bital length of head. Color in alcohol overall pale with widely spaced dots on ventral half of trunk and tail; abdomen bluish. Branchiostegal and gular membranes blackish on exposed surfaces. Lips edged with thin black margin. Diffuse blackish blotches over posterior part of posttemporal and ventrally on preopercle at junction with infraorbitals. Barbel completely white to base. Edges of mandibular ramus black. Leading edge of snout posteriorly to anterior end of suborbital ridge blackish; su- pranarial ridges blackish, this coloration extend- ing onto tops of orbital rim. Outer edge of gill cover with a diffuse blackish margin; upper edge of opercle with a thin black margin; a diffuse horizontal black streak slightly below upper edge of opercle. Oral cavity completely pallid; bran- chial cavity generally pallid with scattered punc- 88 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 tations and blackish outer margins. Gill fila- ments pale; arches and rakers dusky. Paired fins black; first dorsal blackish with no pronounced blotches, but base pale; anal and second dorsal pale. SizeE.—A small species, attains about 24 cm. DISTRIBUTION. — Known only from the Sala y Gomez Ridge, in 345-610 m. ETyYMOLOGyY.—From Latin, teres, terete, in ref- erence to the slender, cylindrical body. COMPARISONS AND REMARKS.— Ventrifossa teres resembles V. mucocephalus Marshall, 1973, from the tropical western Atlantic in its low head profile, relatively terete body, and small chin bar- bel. The two species differ in the following: (1) V. teres usually has V. 9, rarely 8, cf. V. 8, seldom 7 or 9 in V. mucocephalus; (2) 1D. 1,9 or 10 in V. teres cf. 1D. II,11-12 (rarely II, 10); (3) inter- orbital 18.5-—23.9% of HL in V. teres cf. 23-28% HL; (4) orbit to preopercle distance 32.9-38.5% HL cf. 40.5-46.0%; (5) black margin along su- pranarial ridges in V. teres, but unmarked in V. mucocephalus; and (6) V. teres is a smaller spe- cies attaining about 24 cm TL, cf. more than 41 cm in V. mucocephalus. Ventrifossa teres also resembles V. nasuta (Smith, 1964) from southern Africa in its low head profile and relatively terete shape, but the two differ markedly in the latter having a stout terminal snout scute and lacking distinctive head markings. Several morphomet- ric and meristic features also distinguish the two. MATERIAL EXAMINED.—(39 spec., all from Sala y Gomez Ridge) Holotype: ZMMGU 18131 (mature 2, 52.2 mm HL, 237 mm TL); 730-790 m; Prof. Shtokman cr. 28, sta. 2018. Paratypes: CAS 51797 (2:39.0-43.5 HL, 174-184 TL): 770 m; Ichthyanar cr. 5, tr. 52. ZMMGU 17758 (38 HL, 188 TL); 535-575 m; Ichthyandr cr. 5, tr. 544. ZMMGU 17759 (39.5 HL, 196 TL); 345-540 m; Ichthyandr cr. 5, tr. 55. ZMMGU 17760 (2:41-48.5 HL, 196+-226+ TL); 610 m; Hercules tr. 68. ZMMGU 17761 (13:36.5-47.5 HL, 157+-214+ TL); 750- 800 m; Prof. Shtokman cr. 18, sta. 1996. ZMMGU 18067 (4: 23.8-49.2 HL, 131-239 TL); 562-545 m; Prof. Shtokman cr. 18, sta. 1965. CAS uncat. (5:41-45 HL, 184+-202+ TL) and ZMMGU 18068 (8:38-46.5 HL, 172.5+-207+ TL); same data as for holotype. ZMMGU 18069 (2:31.1-32.7 HL, 143+-143+ TL); 545-600 m; Prof. Shtokman cr. 18, sta. 1977. BIOGEOGRAPHICAL CONSIDERATIONS Table 5 lists the 25 species of macrourids known from NSG and compares their distribu- tions in various parts of the world’s oceans. Only two species are known from the continental slope of South America: Caelorinchus immaculatus and Nezumia convergens, and the identification of the latter from a single specimen is as yet tentative. Caelorinchus immaculatus is a representative of a southern temperate fauna, as its relationships appear closest to C. karrerae of the southeastern Atlantic and southern part of the Indian Ocean, from Africa to Australia. None of the NSG spe- cies are common to the tropical eastern Pacific fauna, although the widespread Malacocephalus laevis is also known from the northeastern Pacific in slope waters of southern California and on seamounts off California and Mexico. Eight of the 25 species are widespread tropical or subtropical taxa: Macrouroides inflaticeps, Squalogadus modificatus, Cetonurus crassiceps, Coryphaenoides paradoxus, Hymenocephalus gracilis, Malacocephalus laevis, Nezumia propin- qua, and Trachonurus villosus (but unresolved taxonomic problems exist with Trachonurus). Eight species are endemic to NSG and four (possibly five) others (Caelorinchus spilonotus, Hymenocephalus striatulus, Mataeocephalus acipenserinus, Pseudocetonurus septifer, and pos- sibly Gadomus sp. cf. melanopterus) are known only from NSG and the Hawaiian Islands. The ties to the Hawaiian fauna are strong, with 11 species (three are questionably included here) common to the two areas, the most for any re- gion. The western tropical Pacific and Indian Ocean also show a high affinity, with eight and nine species in common, although all but Ven- trifossa johnboborum are widespread species. Those areas are followed closely by the Atlantic Ocean, with seven species in common with the NSG. Thus, despite the proximity of the Nazca Ridge to the mainland coast of Peru, the relationship of the fauna of the Nazca and Sala y Gomez ridges is clearly to the west, in the tropical central and western Pacific. This is in line with the dis- tributions of other deep-sea fish and invertebrate taxa known from this area (see Parin 1990; Parin et al. 1981; Nesis 1990) and also the shore fauna of many eastern Pacific islands (Rosenblatt et al. 1972; Springer 1982). Wilson et al. (1985) dis- cuss this relationship in the fauna of seamounts of the central North Pacific. The 32% endemism of the grenadier fauna (8 of 25 spp.) is somewhat lower than the 46% endemism (65 of 141 spp.) recorded by Parin (1990) for the entire NSG fish fauna. That so many of the grenadiers repre- sented are circumglobal species accounts for the lower percentage of endemic species. It can be expected that more drags made below 1,500 m SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES TABLE 5. 89 List of the grenadiers of the Nazca and Sala y Gomez ridges, comparing their distributions in different areas. Species denoted with an asterisk (*) may have still-unresolved taxonomic problems and may involve more than one species. A question mark (?) following a plus sign (+) indicates the possible occurrence of the species in the region. Abbreviations: A— Atlantic Ocean; E—endemic to Nazca and Sala y Gomez ridges; ESP—continental eastern South Pacific, HI— Hawaiian Island region; IO—Indian Ocean; Nazca— Nazca Ridge; SyG = Sala y Gomez Ridge; WP—western tropical Pacific, WSP—Australian- New Zealand region (western South Pacific). SyG . Macrouroides inflaticeps . Squalogadus modificatus . Cetonurus crassiceps . Caelorinchus nazcaensis . C. spilonotus C. immaculatus . C. multifasciatus . Coryphaenoides paradoxus* . Hymenocephalus gracilis . H. neglectissimus . H. semipellucidus . H. striatulus . H. sp. cf. aterrimus* . Kuronezumia pallida . Malacocephalus laevis* . Mataeocephalus acipenserinus . Nezumia propinqua . N. sp. cf. convergens . Pseudocetonurus septifer . Trachonurus villosus . Ventrifossa johnboborum . V. macrodon . V. teres . V. obtusirostris ONDWAPWNH Totals . Gadomus sp. cf. melanopterus* HI WP WSP ESP IO A + s°H — — - = = — _ + — = + = = - + - _ + + ar 3° = + — - + + + _ - - — = + — - + _ - + = = = = = = + + + +? - ari +? + - + — — + 3 + — = = = = = + = = = = = = + + _ - _ - + aie - _ -_ ata aFY + — = = — = = 3 + + + — iF + + + - - = = = at 3 + - = ar = = = = = +? = = + 3 _ - - — = seh +? 35 +? - +? +? + = + — = + = + — = — = — a + = = = — = + = = — — = 21 11 8 4 2 9 7 will result in the capture of other species that will further reduce the percentage of endemics, be- cause abyssal grenadiers are more broadly dis- tributed than their shallow-water counterparts, as a general rule. The relative richness of the two ridges is clearly skewed towards the Sala y Gomez Ridge, with 21 species of grenadiers represented, compared with only eight on the Nazca Ridge (see Table 5). Nezumia sp. cf. convergens was taken only in the region between the two ridges; Macrouroides inflaticeps was taken in the intermediate area and on the Nazca Ridge; Cetonurus crassiceps and Hymenocephalus sp. cf. aterrimus were taken on both ridges and in the intervening area. Parin (1990) considered the Nazca fauna a depauperate one containing a few peculiar elements that make it distinctive from that of the Sala y Gomez. Nesis (1990), using cephalopod data, confirmed the existence of a zoogeographical boundary on the southwestern side of the Nazca Ridge at about 84°-85°W. Comparison of the grenadiers of the NSG and Hawaiian Islands shows some interesting par- allels and contrasts (Table 6). The Hawaiian Is- land region has a slightly greater number of re- corded species (29) than NSG (25), and the number of endemics is notably higher (15, or 52%, compared with 9 and 36%). This high en- demism of the Hawaiian fauna is reflected in the few species shared in common with the western tropical Pacific—the number of NSG species in this category is twice that of the Hawaiian Islands (8 vs. 4). Strangely, however, the number of In- dian Ocean and Altantic Ocean species shared in common with NSG on the one hand, and the Hawaiian Islands on the other hand, are com- parable. If we exclude five wide-ranging (and problematic) species (Cetonurus crassiceps, Cor- yphaenoides paradoxus, Hymenocephalus ater- rimus, Malacocephalus laevis, and Trachonurus villosus), the contrasts become even more inter- esting. In the Hawaiian fauna, Kuronezumia bu- bonis and Mesobius berryi remain as the only 90 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 TABLE 6. List of the grenadiers of the Hawaiian Islands region, comparing their distributions in different areas. Species denoted with an asterisk (*) may have still-unresolved taxonomic problems and may involve more than one species. A question mark (?) following a plus sign (+) indicates the possible occurrence of the species in the region. Abbreviations: A— Atlantic Ocean; E—endemic to Hawaiian Islands region; ESP—continental eastern South Pacific; IO—Indian Ocean; Nazca— Nazca Ridge; SyG—Sala y Gomez Ridge; WP—western tropical Pacific, WSP—Australian-New Zealand region (western South Pacific). E Nazca . Gadomus melanopterus* = . Bathygadus bowersi + — . Bathygadus micronemus* af . Cetonurus crassiceps _ . Caelorinchus aratrum + = . C. spilonotus = . C. doryssus + - . C. gladius + . Coryphaenoides paradoxus* — + . C. longicirrhus* - = . Hymenocephalus striatulus — . H. antraeus + — . H. tenuis Ar . H. aterrimus* — +? . Kuronezumia bubonis _ _ . Malacocephalus laevis* _ = . Mataeocephalus acipenserinus _ — . Mesobius berryi _ _ . Nezumia propinqua — . N. burragei* . N. ectenes* oa . N. holocentra + — + + eS Ss Ss Fs es OwomaonnanfkwnrndkK OO WANDA UN HRWN NY N— . N. obliquata . N. hebetata* 25. Pseudocetonurus septifer - 26. Sphagemacrurus gibber + _ 27. Trachonurus villosus* _ _ 28. Ventrifossa atherodon + — 29. V. ctenomelas Totals 15 5 a) RW two species also found in the Indian and Atlantic oceans. Of the NSG fauna, only Hymenocepha- lus gracilis and Squalogadus modificatus remain. That the two faunas are closely related is shown in the number of species they share exclusively: Gadomus melanopterus, Caelorinchus spilono- tus, Hymenocephalus striatulus, Mataeocephalus acipenserinus, and Pseudocetonurus septifer. It thus appears that the two faunas share a host of wide-ranging species as well as several unique ones, but each has been isolated long enough to develop a large percentage of endemic elements. Okamura and Yatou (in Okamura et al. 1982) reported on the grenadiers from the Kyushu-Pa- lau Ridge and Tosa Bay, collected by the Fish- eries Agency of the Government of Japan (Table 7). The Kyushu-Palau Ridge extends in a south- easterly direction from the island of Kyushu to SyG WP WSP ESP @) A aia - — - - — + - + — — 3 = = — = — + ate “E? = ts Pi — = — +? <= + = — — = = +? - - - au +? _ - +f = 3 +f + = + - + +f + = = = = — = = = — ar ar a + = sae - + = = a = — + + 4 — + af 11 4 5 0 8 7] the Palau Islands at a latitude of approximately 6°N. The ridge area investigated was confined to approximately 25°-30°N. Of the 25 species re- corded, 15 were confined to the Kyushu-Palau Ridge, and 10 others were taken on the conti- nental slope off Tosa Bay (Shikoku Island, Japan) and other continental margins. Although the coverage area and collection depths were much more limited in the Japanese surveys than in the Soviet surveys of the NSG, some interesting comparisons can be made. There were three species that were possibly the same in the two areas: Malacocephalus nipponensis (?=M. laevis), Nezumia propinqua, and Tracho- nurus villosus (2). Six of the 15 species (40%) were endemic to the Kyushu-Palau Ridge, and these were equally divided among Caelorinchus and Ventrifossa. The two genera were represented in SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 91 the NSG fauna by four and three endemic spe- cies. The Kyushu-Palau Ridge, although being relatively closely adjacent to an exceeding rich faunal region (Japan and the Philippines), none- theless has a distinctive fauna, which appears to be less diverse than the faunas of the Hawaiian Archipelago and the Nazca and Sala y Gomez ridges. The propensity for the genera Caelorin- chus and Ventrifossa to have the most species in these ridge areas is notable, although the thought is tempered with the knowledge that Caelorin- chus is by far the most diverse genus of grena- diers, with more than 100 species. ACKNOWLEDGMENTS We thank the many people who helped in var- ious ways in the preparation of this paper. Fore- most, we thank Nikolai V. Parin, head of the Laboratory of Oceanic Ichthyofauna, for initi- ating this cooperative study, for his encourage- ment and advice, and for his help with materials, facilities, personnel, and other things too nu- merous to mention. Yuri N. Shcherbachev (IOAN) contributed enormously with his advice, knowledge of the group and collections, and as- sistance with manuscript review, references, data gathering, and facilities arrangements (the last especially for TI). Many people were extremely hospitable and helpful to TI during his visits to their institutions. These include, but are not lim- ited to: E. Karmovskya, Y. N. Shcherbachev, V. Tchuvasov, and others of IOAN; I. A. Verighina and others of ZMMGU; A. Andriashev, V. V. Barsukov (deceased), A. V. Neyelov, and others of ZIN; S. Jewett, L. R. Parenti, V. G. Springer, R. P. Vari, (USNM); D. M. Cohen, R. Feeney, R. J. Lavenberg, J. Seigel, and C. C. Swift (LACM); J. E. Randall and A. Sugimoto (BPBM); K. Hartell (MCZ); O. Crimmen, G. Howes, A. Wheeler, and (BMNH); O. Okamura (BSKU). Technical assistance at CAS was provided by D. G. Catania, J. Fong, and P. M. Sonoda; T. Moritz and B. Edgar provided references; S. Tatro trans- lated the abstract into Spanish. Photographs of specimens were provided by S. K. Middleton (CAS) and S. Dudarev (IOAN). Many individ- uals were involved in collecting specimens, but we especially acknowledge the efforts of G. A. Golovan’, N. P. Pakhorukov (both from the Se- vastopol Laboratory), and A. N. Kotlyar IOAN, formerly with VINRO). Although we began this study many years ago, its completion was aided TaBLe 7. Grenadiers of the Kyushu-Palau Ridge and Tosa Bay (after Okamura et al. 1982). Species found on the Kyushu-Palau Ridge (species endemic to ridge marked with an asterisk*): 1. Gadomus colletti Jordan and Gilbert, 1904 2. Caelorinchus gilberti Jordan and Hubbs, 1925 . C. hexafasciatus Okamura, 1982* . C. longicephalus Okamura, 1982* . C. matsubarai Okamura, 1982* . Hymenocephalus longiceps Smith and Radcliffe, 1912 H. striatissimus Jordan and Gilbert, 1904 . H. lethonemus Jordan and Gilbert, 1904 9. Malacocephalus nipponensis Gilbert and Hubbs, 1916 10. Nezumia propinqua (Gilbert and Cramer, 1897) 11. Trachonurus villosus (Ginther, 1877) 12. V. fusca Okamura, 1982* 13. V. japonica (Matsubara, 1943) 14. V. longibarbata Okamura, 1982* 15. V. macroptera Okamura, 1982* Tosa Bay species not found on Kyushu-Palau Ridge: 1. Caelorinchus anatirostris Jordan and Gilbert, 1904 2. C. smithi Gilbert and Hubbs, 1916 3. C. japonicus (Temminck and Schlegel, 1884) 4. Coryphaenoides nasutus Giinther, 1877 5. C. marginatus Steindachner and Déderlein, 1884 6. Kuronezumia dara (Gilbert and Hubbs, 1916) 7. Nezumia condylura Jordan and Gilbert, 1904 8. N. proxima (Smith and Radcliffe, 1912) 9. Ventrifossa garmani (Jordan and Gilbert, 1904) 0. V. nigrodorsalis Gilbert and Hubbs, 1916 by a 1988 visit to the USSR by TI, sponsored by the National Academy of Sciences/National Research Council through its Soviet and East European Program. Director G. E. Schweitzer and Program Associate C. M. Turzak are thanked for their assistance. The CAS In-House Research Fund provided support to TI for museum visits. RESUMEN Hay registrado veinticinco especies de grana- deros desde las cordilleras submarinas de Nazca y Sala y Gomez en el sur del Pacifico Oriente. Las especies nuevas incluyen: Caelorinchus im- maculatus, C. multifasciatus, C. nazcaensis, C. spilonotus, Hymenocephalus neglectissimus, H. semipellucidus, Kuronezumia pallida, Ventrifos- sa macrodon, V. teres, y V. obtusirostris. Cae- lorinchus immaculatus es muy semejante a C. karrerae desde el sur el Atlantico Oriente y el mar de las Indias. El complejo de H. striatissimus esta examinado usando informaciones nuevas. Hymenocephalus semipellucidus y H. neglectis- simus parecen pertenecer a este complejo. El sub- 92 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 espécie de H. s. hachijoensis de Japon esta ele- vado a la categoria del espécie pleno. Kuronezumia, considerado antes como subgé- nero de Nezumia, esta redefinado y elevado a la categoria de género incluyendo K. pallida, K. bu- bonis, K. leonis, K. macronema, K. dara, y dos especies no descritos. A despecho de la prox- imidad de las cordilleras submarinas a la cuesta del Peru, los afines de la fauna asociada estan hacia del oeste, particularmente al Pacifico Oc- cidental y las Islas Hawaii. Ocho de los veinti- cinco especies de estas cordilleras submarinas eran definitivamente conocidos de la proximi- dad de las Islas Hawaii: Caelorinchus spilonotus, Cetonurus crassiceps,* Coryphaenoides paradox- us,* H. striatulus, Malacocephalus laevis,* Ma- taeocephalus acipenserinus, Nezumia_ propin- qua,* y Pseudocetonurus septifer. Tres otros especies cuyos identificaciones no son determi- nados pueden ser parte o tener afines en la fauna de las Islas Hawaii: Gadomus sp. cf. melanop- terus, Hymenocephalus sp. cf. aterrimus, and Trachonurus villosus*? Los cuatro espécies mar- cados con asterisco estan distribuidos amplia- mente por los oceanos Pacifico, Indio y Atlan- tico. Malacocephalus laevis se conoce de los taludes continentales en el sur de California y por las montanas submarinas cerca de Baja Cal- ifornia, pero nunca se los encuentra por las cues- tas de América Central y América del Sur. Cae- lorinchus immaculatus también se nota de América del Sur; Nezumia convergens esta re- presentada discutiblemente en un espécimen tn- ico de la cordillera submarina de Sala y Gomez. LITERATURE CITED AMAOKA, K., K. NAKAYA, H. ARAYA, AND T. YASUI, EDS. 1983. Fishes from the North-eastern Sea of Japan and the Okhotsk Sea off Hokkaido. The intensive research of unexploited fishery resources on continental slopes. Japan Fish. Resource Conser. Assoc., Tokyo. 371 pp. Aral, T. AND P. J. MCMILLAN. 1982. A new macrourid fish, Coelorinchus biclinozonalis from New Zealand, and rede- scription of C. australis from Australia. Japan. J. Ichthyol. 29(2):115-126. BARNARD, K.H. 1925. Descriptions of new species of marine fishes from S. Africa. Ann. Mag. Nat. Hist., Ser. 9, 15(87): 498-504. BerG, C. 1898. Substitucion de nombres genéricos. Comm. Mus. Nac. Buenos Aires 1:41-43. Brauer, A. 1906. Die Tiefsee-Fische. I. Systematischer Teil. Wissenschaftliche Ergebnisse der deutschen Tiefsee-Expe- dition auf dem Dampfer “Valdivia,” 1898-1899, 15:1-432, text-figs. 1-176, pls. 1-18. Couen, D. M., ED. 1989. Papers on the systematics of gad- iform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. CoHEN, D. M., T. INADA, T. IWAMOTO, AND N. SCIALABBA. 1990. FAO species catalogue. Gadiform fishes of the world (Order Gadiformes). An annotated and illustrated catalogue of cods, hakes, grenadiers and other gadiform fishes known to date. FAO Fish. Synop. no. 125, Vol. 10. 442 pp. EscHMEYER, W. N. 1990. Catalog of the genera of Recent fishes. Calif. Acad. Sci., San Francisco. 697 pp. FisHER, R. L. AND R. M. Norris. 1960. Bathymetry and geology of Sala y Gomez, southeast Pacific. Geol. Soc. Amer. Bull. 71:497-5S02. Garman, S. 1899. Reports on an exploration off the west coasts of Mexico, Central and South America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the U.S. Fish Commission steamer ‘“‘Albatross” during 1891. Lieut.-Commander Z. L. Tanner, U.S.N., commanding. Part 26. The fishes. Mem. Mus. Comp. Zool. Harv. Coll. 24:1- 431, pls. 1-84, A-N. GiGuiou!, E.H. 1884. Pelagos, saggi sulla vita et sui prodotti del mare. Genova. 433 pp. GiLBeRT, C.H. 1892. Scientific results of explorations by the U.S. Fish Commission steamer “Albatross.” No. 22. De- scriptions of thirty-four new species of fishes collected in 1888 and 1889, principally among the Santa Barbara Islands and in the Gulf of California. Proc. U.S. Natl. Mus. 14:539- 566. 1905. The deep-sea fishes of the Hawaiian Islands. Pp. 575-713, figs. 230-276, pls. 66-101 in Jordan, D. S. and B. W. Evermann (eds.), The aquatic resources of the Ha- waiian Islands. Bull. U.S. Fish Comm. 1903, 22(pt. 2, sect. 2). GtLBerT, C. H. AND F. CRAMER. 1897. Report on the fishes dredged in deep water near the Hawaiian Islands, with de- scriptions and figures of twenty-three new species. Proc. U.S. Natl. Mus. 19(1114):403-435, pls. 36-48. GiLBerT, C. H. AND C. L. Husss. 1916. Report on the Jap- anese macrouroid fishes collected by the United States fish- eries steamer ‘“‘Albatross” in 1906, with a synopsis of the genera. Proc. U.S. Natl. Mus. 51:135-214, pls. 8-11. 1917. Description of Hymenocephalus tenuis, a new macrouroid fish from the Hawaiian Islands. Proc. U.S. Natl. Mus. 54(2231):173-175. . 1920. The macrourid fishes of the Philippine Islands and the East Indies. U.S. Natl. Mus. Bull. 100, 1(pt. 7):369- 588. Giccurist, J.D. F. AND C. von Bonpe. 1924. Deep-sea fishes procured by the S. S. “Pickle.” Part II. Rep. Fish. Mar. Biol. Surv. S. Afr. 3(1922):1-24, pls. 1-6. Giorna, M. E. 1809. Mémoire sur des poissons d’espéces nouvelles et de genres nouveau. Mém. Acad. Sci. Turin (1805— 1808)16(1):177-180. Goober, G. B. AND T. H. BEAN. 1885. Descriptions of new fishes obtained by the United States Fish Commission main- ly from deep water off the Atlantic and Gulf coasts. Proc. U.S. Natl. Mus. 8(37-38):589-605. Go.ovan’, G. A. 1978. Composition and distribution of the ichthyofauna of the continental slope of north-western Af- rica. [In Russian, with English summary.] Trans. Inst. Oceanol. Acad. Sci. USSR 111:195-258. Gotovan’, G. A. AND N. P. PAKHoRUKOv. 1987. Distribu- tion and behavior of fishes on the Naska and Sala y Gomez submarine ranges. Voprosy Ikhtiologii (3):369-376. Gos.ine, W. A. 1954. Fishes killed by the 1950 eruption of Mauna Loa. II. Brotulidae. Pac. Sci. 8(1):68-83. SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 93 Gos.ing, W. A., V. E. Brock, H. L. Moore, AND Y. YaA- MAGUCcHI. 1954. Fishes killed by the 1950 eruption of Mau- na Loa. I. The origin and nature of the collections. Pac. Sci. 8(1):23-27. Grey, M. 1959. Deep sea fishes from the Gulf of Mexico with the description of a new species Squalogadus inter- medius (Macrouroididae). Field. Zool. 39(29):323-346. Gunnerus, J. E. 1765. Efterretning om berglaxen, en rar norsk fishk, som kunde kaldes: Coryphaenoides rupestris. K. Norske Videnskabers Selskab Skrifter Trondhjem, 3(4):50- 58, pl. 3, fig. 1, 2. Guntuer, A. 1862. Catalogue of the fishes of the British Museum 4:1-534. 1877. Preliminary notes on new fishes collected in Japan during the expedition of H.M.S. ‘Challenger.’ Ann. Mag. Nat. Hist., ser. 4, 20:433-446. . 1878. Preliminary notices of deep-sea fishes collected during the voyage of H.M.S. ‘Challenger.’ Ann. Mag. Nat. Hist., ser 5, 2:23-28. 1887. Report on the deep-sea fishes collected by H.M.S. Challenger during the years 1873-76. Rep. Sci. Res. Challenger 22(Zool.)(pt. 1)[text]:1-—335; (pt. 2)[plates]:pls. 1- 13% Howes, G. J. 1988. The cranial muscles and ligaments of macrouroid fishes (Teleostei: Gadiformes); functional, eco- logical and phylogenetic inferences. Bull. Brit. Mus. (Nat. Hist.), Zool. Ser., 54(1):1-62. . 1989. Phylogenetic relationships of macrouroid and gadoid fishes based on cranial myology and arthrology. Pp. 113-128 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. Howes, G. J., AND O. A. CrimMen. 1990. A review of the Bathygadidae (Teleostei: Gadiformes). Bull. Br. Mus. Nat. His. (Zool.) 56(2):155-203. Iwamoto, T. 1970. The R/V Pillsbury Deep-Sea Biological Expedition to the Gulf of Guinea, 1964-65. 19. Macrourid fishes of the Gulf of Guinea. Stud. Trop. Oceanogr. (4)(pt.2): 316-431. 1974. Nezumia (Kuronezumia) bubonis, a new sub- genus and species of grenadier (Macrouridae: Pisces) from Hawaii and the western North Atlantic. Proc. Calif. Acad. Sci., ser. 4, 39(22):507-516, figs. 1-3. 1978. Eastern Pacific macrourids of the genus Coe- lorinchus Giorna (Pisces: Gadiformes), with description of a new species from Chile. Proc. Calif. Acad. Sci., ser. 4, 41(12):307-337, figs. 1-20. 1979. Eastern Pacific macrourine grenadiers with seven branchiostegal rays (Pisces: Macrouridae). Proc. Calif. Acad. Sci. 42(5):135-179, figs. 1-18. . 1982. Ventrifossa johnboborum, a new grenadier from the western Pacific (Macrouridae: Pisces). Austr. Zool. 21 (pt. 1):55-61. . 1983. Family No. 93: Macrouridae. Pp. 330-341 in Smith, M. M. and P. C. Heemstra (eds.), Smiths’ sea fishes. Macmillan South Africa, Johannesburg. 1989. Phylogeny of grenadiers (suborder Macrou- roidei): another interpretation. Pp. 159-173 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles Co., Sci. Ser. 32. 262 pp. Iwamoto, T. AND Y. I. Sazonov. 1988. A review of the southeastern Pacific Coryphaenoides (sensu lato) (Pisces, Gadiformes, Macrouridae). Proc. Calif. Acad. Sci. 45(3):35- 82, figs. 1-9. Jorpan, D. S. 1904. See Jordan and Gilbert 1904. Jorpan, D. S. AND C. H. GitBert. 1904. Macrouridae. Pp. 602-621, in Jordan, D. S. and E. C. Starks. List of fishes dredged by the steamer Albatross off the coast of Japan in the summer of 1900, with descriptions of new species and a review of the Japanese Macrouridae. Bull. U.S. Fish Comm. 22(1902):577-630, pls. 1-8. Kot yar, A. N. AND N.N. Parin. 1990. Otoliths and age of bottom and near-bottom fishes from the Nazca and Sala y Gomez Submarine Ridges. [In Russian with English sum- mary.] Tr. Inst. Okeanol. Akad. Nauka SSSR 125:97-126. Leviton, A. E. AND R. H. Gras, Jr. 1988. Standards in herpetology and ichthyology: standard symbolic codes for institutional resource collections in herpetology and ichthy- ology. Supplement No. 1: additions and corrections. Copeia 1988(1):280-282. Leviton, A. E., R. H. Grsss, Jr., E. HEAL, AND C. E. Dawson. 1985. Standards in herpetology and ichthyology: part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia 1985(3):802- 832. Lowe, R. T. 1843. Notices of fishes newly observed or dis- covered in Madeira during the years 1840, 1841 and 1842. Proc. Zool. Soc. Lond. 11:81-95. MAMMERICKx, J., R. N. ANDERSON, H. W. MENARD, AND S. M. SmitH. 1975. Morphology and tectonic evolution of the east-central Pacific. Geol. Soc. Amer. Bull. 86:111-118. Markie, D. F. 1989. Aspects of character homology and phylogeny of the Gadiformes. Pp. 59-88 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. MarsHa_t, N. B. 1964. Bathypelagic macrourid fishes. Co- peia 1964(1):86-93. MarsuHa._t, N. B. 1973. Family Macrouridae. Pp. 496-665, in Cohen, D. M. (ed.) Fishes of the western North Atlantic. Mem. Sears Found. Mar. Res. (1)(pt.6). MarsHatt, N. B. AND T. lwAMoto. 1973. Genus Coelorhyn- chus Giorna 1809, pp. 538-563, Genus Hymenocephalus Giglioli 1882, pp. 601-612, and Genus Nezumia Jordan 1904, pp. 624-649, in Cohen, D. M. (ed.) Fishes of the western North Atlantic. Mem. Sears Found. Mar. Res. (1)(pt.6). MarsuHa_t, N. B. AND A. V. TANING. 1966. The bathypelagic macrourid fish, Macrouroides inflaticeps Smith and Rad- cliffe. Dana-Rep. 69. 6 pp., | pl. Maut, G. E. 1951. Monographia dos peixes do Museu Mu- nicipal do Funchal. Familia Macrouridae e Merlucciidae. Bol. Mus. Munic. Funchal 5(12):5—55. McCann, C. AND D. G. McKnicut. 1980. The marine fauna of New Zealand: macrourid fishes (Pisces: Gadida). New Zealand Oceanogr. Inst. Mem. 61:1-91. McCuttocn, A. R. 1907. The results of deep sea investi- gation in the Tasman Sea. II. The expedition of the “Woy Woy.” 1. Fishes and crustaceans from eight hundred fath- oms. Rec. Aust. Mus. Sydney 6:345-355, pls. 63-65. 1926. Report on some fishes obtained by the F.1.S. “Endeavour” on the coasts of Queensland, New South Wales, Victoria, Tasmania, South and South-western Australia. Part V. Biogical results of the fishing experiments carried on by the F.I.S. “Endeavour” 1909-1914, 4(pt. 5):157-216, pls. 43-56. Nesis, K. N. 1990. Bottom and near-bottom cephalopods of the Nazca and Sala y Gomez Submarine Ridges (Pacific Ocean). [In Russian with English summary.] Tr. Inst. Okean- ol. Akad. Nauka SSSR 125:178-191. No tr, D. AND E. STEURBAUT. 1989a. Evidence from otoliths 94 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 2 for establishing relationships between gadiforms and other groups. Pp. 37-45 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. 1989b. Importance and restrictions of the otolith- based fossil record of gadiform and ophidiiform fishes. Pp. 47-58 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. Norman, J. R. 1939. Fishes. Sci. Rep. John Murray Exped. 1933-1934, 7(Zool.)(1):1-116. Norris, R. M. 1961. Sala y Gomez—lonely landfall. Pac. Disc. 13(6)(Nov.—Dec.):20-25. OKAMuRA, O. 1970a. Fauna Japonica. Macrourina (Pisces). Academic Press, Tokyo. 216 pp., 64 pls. 1970b. Studies on the macrouroid fishes of Japan. Morphology, ecology and phylogeny. Rep. Usa Mar. Biol. Sta. 17(10-2):1-179. 1982. See Okamura et al. 1982. 1984. See Okamura et al. 1984. 1989. Relationships of the suborder Macrouroidei and related groups, with comments on Merlucciidae and Steindachneria. Pp. 129-142 in Cohen, D. M. (ed.), Papers on the systematics of gadiform fishes. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 32. 262 pp. OKAMURA, O., K. AMAOKA, AND F. MITANI, EDs. 1982. Fishes of the Kyushu-Palau Ridge and Tosa Bay. The intensive research of unexploited fishery resources on continental slopes. Japan Fish. Resource Conserv. Assoc., Tokyo. 435 pp. OKAMURA, O. AND T. KitasiMA, EDS. 1984. Fishes of the Okinawa Trough and the adjacent waters. I. The intensive research of unexploited fishery resources on continental slopes. Japan Fish. Resource Conserv. Assoc., Tokyo. 414 pp. PAKHORUKOV, N. P. 1976. Preliminary list of the bathyal bottom fishes of the Rio Grande Rise. [In Russian.] Biology and distribution of the tropical deep-sea fishes. Trans. P.P. Shirshov Inst. Oceanol. 104:318-331. 1981. Deep sea bottom fishes of the Whale Ridge and adjacent area. [In Russian.] Pp. 19-31 in Parin, N. V. (ed.), Fishes of the open ocean. Shirshov Inst. Oceanol. Acad. Sci. USSR, Moscow. 120 pp. Parin, N. V. 1990. Preliminary review of fish fauna of the Nazca and Sala y Gomez Submarine Ridges (southern East Pacific Ocean). [In Russian with English summary.] Tr. Inst. Okeanol. Akad. Nauka SSSR 125:6-36. Parin, N. V., G. A. GoLovan’, N. P. PAKHoRuKovy, Y. I. SAZONOV, AND Y. N. SHCHERBACHEV. 1981. Fishes from the Nazca and Sala-y-Gomez underwater ridges collected in cruise of R/V “‘Ikthiandr.” [In Russian.] Pp. 5-18 in Parin, N. V. (ed.), Fishes of the open ocean. Shirshoy Inst. Oceanol. Acad. Sci. USSR, Moscow. 120 pp. Parin, N. V., T. A. GORELOVA, AND O. D. BoropuLtna. 1990. Feeding and trophic relationships of fishes inhabiting the Nazca and Sala y Gomez Submarine Ridges. [In Russian with English summary.] Tr. Inst. Okeanol. Akad. Nauka SSSR 125:37-57. Parin, N. V. AND Y. I. SAzONov. 1982. Deep-sea fishes col- lected during a cruise of R/V “Professor Mesyatzev” along the Peruvian coasts in 1972. [In Russian.] Pp. 79-91 in Parin, N. V. (ed.), Insufficiently studied fishes of the open ocean. Shirshov Inst. Oceanol. Acad. Sci. USSR, Moscow. 140 pp. Parr, A. E. 1946. The Macrouridae of the western North Atlantic and Central American seas. Bull. Bing. Oceanogr. Coll. 10(1):1-99. RapcutFre, L. 1912. Descriptions of a new family, two new genera, and twenty-nine new species of anacanthine fishes from the Philippine Islands and contiguous waters. Proc. U.S. Natl. Mus. 43:105-140, pls. 22-31. ReGANn, C. T. 1903. On the systematic position and classi- fication of the gadoid or anacanthine fishes. Ann. Mag. Nat. Hist., ser. 7, 11(65):459-466. 1908. The Percy Sladen Trust Expedition to the In- dian Ocean in 1905 under the leadership of Mr. J. Stanley Gardiner. No. XIV.—Report on the marine fishes collected by Mr. J. Stanley Gardiner in the Indian Ocean. Pp. 217— 255, pls. 23-32. RICHARDSON, J. 1939. Account of a collection of fish from Port Arthur, Van Diemen’s Land. Trans. Zool. Soc. Lond. 7:95-100. Risso, A. 1810. Ichthyologie de Nice, ou histoire naturelle des poissons du département de Alpes-Maritimes. Paris, xxxvi + 388 pp., 11 pls. ROSENBLATT, R. H., J. E. McCosker, AND I. RuBINOFF. 1972. Indo-West Pacific fishes from the Gulf of Chiriqui, Panama. Contrib. Sci. Nat. Hist. Mus. Los Angeles County 234:1-18. Sawapa, Y. 1983. See Amaoka et al. 1983. Sazonov, Y. I. 1985. Two new macrourid species (Gadi- formes) from northern Pacific seamounts. [In Russian with English summary.] Voprosy Ikhtiologii (5):719-727. Sazonov, Y. I. AND Y. N. SHCHERBACHEV. 1982. A prelim- inary review of grenadiers related to the genus Cefonurus Ginther (Gadiformes, Macrouridae). Descriptions of new taxa related to the genera Cetonurus Ginther and Kumba Marshall. [In Russian, with English summary.] Voprosy Ikh- tiologii 22 (5):707—721, figs. 1-4. [also Engl. transl., J. Ich- thyol. 22(5):1-15.] 1985. Preliminary list of macrourids related to the genus Cetonurus Giinther (Gadiformes, Macrouridae). II. Genus Cetonurus Giinther. Comparative taxonomical anal- ysis of recognized groups. [In Russian.] Voprosy Ikhtiologii 26 (2):179-195. SHCHERBACHEV, Y. N. 1987. Preliminary list of thalasso- bathyal fishes of the tropical and subtropical waters of the Indian Ocean. [In Russian.] Voprosy Ikhtiologii (1):3-11. {English version in J. Ichthyol. 27 (2):37-46.] SHCHERBACHEV, Y. N., N. V. PARrin, N. P. PAKHORUKOV, AND A. C. Piotrovskty. 1986. Mesobenthic and mesobentho- pelagic fishes from submarine rises in the western Indian Ocean. [In Russian.] Tr. Inst. Okeanol., Akad. Nauk SSSR 121:195-214. SHCHERBACHEV, Y. N. AND A. S. Prorrovskry. 1982. On the bathymetric and geographical distribution of the species of the subfamily Macrouroidinae (Gadiformes, Macrouridae). [In Russian, with English summary.] Bull. Moscow Soc. Nat- ural., Biol. Ser. 87(5):45-48. SHCHERBACHEY, Y.N., Y. I. SAZONOV, AND A. S. PIOTROVSKIY. 1979. Occurrence of Trachonurus villosus (Giinther) and Mesobius (Hubbs et Iwamoto) (Macrouridae, Osteichthyes) in the Indian Ocean. [In Russian.] Voprosy Ikhtiologi 19: 20-28, figs. 1-2. [English version in J. Ichthyol. 19(1):16- 23: SHIOBARA, Y. 1982. A deep-sea macrouroid fish, Squaloga- dus modificatus from the inshore areas of Suruga Bay, Japan. {In Japanese, with English abstract.) Ann. Rep. Mar. Sci. Mus., Tokai Univ. (10):143-146. Smit, H. M. AND L. RApcuiFFeE. 1912. [See Radcliffe, 1912.] SmitH, J. L.B. 1935. New and little known fishes from South Africa. Rec. Albany Mus. 4:169-—235, pls. 18-23. SPRINGER, V.G. 1982. Pacific plate biogeography, with spe- SAZONOV AND IWAMOTO: GRENADIERS OF THE NAZCA AND SALA Y GOMEZ RIDGES 5) cial reference to shorefishes. Smithsonian Contrib. Zool. (367). 182 pp. Trunov, I. A. 1984. Characteristics of some species of the genus Coelorinchus Giorna (Osteichthyes, Macrouridae) from the Southeast Atlantic. [In Russian.] Voprosy Ikhtiologii 24(3): 355-367. Uyeno, T., K. MAtsuurA, AND E. Fusi, Eps. 1983. Fishes trawled off Suriname and French Guiana. Japan Marine Fishery Resource Research Center: Tokyo. 519 pp. VAILLANT, L. L. 1888. Expéditions scientifiques du Travail- leur et du Talisman pendant les années 1880, 1881, 1882, 1883. Poissons. Masson: Paris. Pp. 1-406, pls. 1-28. Waite, E. R. 1911. Scientific results of the New Zealand Government Trawling Expedition, 1907. Pisces—Part II. Rec. Canterbury Mus. 1(3):157—272, pls. xxii—Ivii. Witson, R. R., Jr., K. L. SmiryH, AND R. H. ROSENBLATT. 1985. Megafauna associated with bathyal seamounts in the central North Pacific Ocean. Deep-Sea Res. 32(10):1,213- 1,254. © CALIFORNIA ACADEMY OF SCIENCES, 1992 Golden Gate Park San Francisco, California 94118 2914 aa a3 7 m & im | Me apret® «4 UU 49 Pode a oe o@ f we « = Yras FO és fem Pru@ Webs = ‘ad —i Se Fe pe 1S Lo) 7 _ ® > i @ } , Sry dig vie: Woes Cegad _ - e 1 hove owe » (o@se) ei (ite > = Se 7 bate 4 & : i“ - geey (2 . oki 5 ish Ary Lif) calkg ¥ . Gew emvi mic y Be n eet ee PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 3, pp. 97-108, 9 figs., 1 table. November 2, 1992 SYNOPSIS OF THE GRENADIER GENUS KURONEZUMIA (PISCES: GADIFORMES: MACROURIDAE), WITH DESCRIPTION OF A NEW SPECIES Yuri N. Shcherbachev P. P. Shirshoy Institute of Oceanology, Academy of Sciences, 23 Krasikova Street, Moscow 117218, Russia, Yuri I. Sazonov Department of Ichthyology, Zoological Museum, Moscow Lomonosov LIBRARY State University, Herzen Street 6, Moscow 103009, Russia, and noy 4 91992 Tomio Iwamoto Department of Ichthyology, California Academy of Sciences, Wood Ss H ole, Ma Ss Golden Gate Park, San Francisco, California 94118 _ Oo rR es Eno ars Asstract: Kuronezumia paepkei n. sp. is described from a specimen collected by the Valdivia in 1899 off the coast of Kenya in 748 m. The specimen, originally recorded by Brauer (1906) as Coryphaenoides rudis, differs from other members of the genus in having larger body scales (8 below second dorsal origin vs. 10- 15) and more gill rakers (12 on the inner series of first arch vs. 8-12, and 11 on outer series of second arch ys. 7-10). A key is provided for the seven species in the genus. New distribution records include K. leonis (Barnard) from southern Australia and New Zealand, and K. bubonis from the South China Sea. Received December 6, 1991. Accepted April 1, 1992. INTRODUCTION While studying Russian collections of Indian Ocean Coryphaenoides (see Iwamoto and Shcherbachev 1991), we examined the specimen that Brauer (1906) reported as ““?Macrurus (Ma- crurus) rudis Giinther,”’ collected during the Val- divia expedition of 1898-99 off the eastern coast of Africa. We were surprised to discover that the specimen is not Coryphaenoides rudis (Ginther, 1878) or even a member of that genus, which is characterized by a count of six branchiostegal rays and the anus situated immediately before the anal fin origin. Instead, it represents a mem- ber of the group of genera with seven branchios- tegal rays and the anus about midway between the pelvic and anal fins. The relatively vertical suborbital region and the numerous small scales on the suborbital shelf identify the specimen as an undescribed species of the genus Kuronezu- mia, although including the new species neces- sitated redefining the genus, which has been done (Sazonov and Iwamoto 1992). The purpose of the present paper is to describe [97] Marine Biological Laboratory EEE - ee Ne 98 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 the new species, provide a key to the species of 4a. Scales below 2D. 8; gill rakers on inner Kuronezumia, and report additional distribu- tional and descriptive information. METHODS Methods for taking counts and making mea- surements follow Iwamoto (1970) and Iwamoto and Sazonov (1988). Abbreviations for institu- tions follow Leviton et al. (1985) and Leviton and Gibbs (1988). Kuronezumia Iwamoto, 1974 Kuronezumia Iwamoto, 1974:509 (as subgenus of Nezumia). Type species by monotypy, Nezumia (Kuronezumia) bubonis Iwamoto, 1974. For a diagnosis of the genus, see Sazonov and Iwamoto (1992). REMARKS. — We recognize seven species of Ku- ronezumia: K. bubonis lwamoto, 1974; K. dara (Gilbert and Hubbs, 1916), K. leonis (Barnard, 1925), K. macronema (Smith and Radcliffe, 1912), K. paepkei (here described), K. pallida Sa- zonov and Iwamoto, 1992, and an undescribed species from the South China Sea (see Iwamoto 1974). The limits of the genus have had to be expanded to accommodate K. /eonis and K. paep- kei because of the enlarged terminal snout scute in the former and the large body scales of the latter. Nonetheless, based on a number of shared characters, the seven species form a distinctive group, whose relationships are, however, still un- clear. The phylogeny of grenadiers with seven branchiostegal rays have yet to be adequately determined. Until that is done, the many genera described within the past two decades must await critical evaluation as to their validity as mono- phyletic units. KEY TO THE SPECIES OF Kuronezumia la. A large scaly tubercular swelling anterior to anus, between pelvic fins 0... 2 1b. Area anterior to anus not swollen nor with Scaly tubercle Meee Sane ah tmak aees C8 3 ZartiVA Tea aod OCT Wet aie Ta K. bubonis 2b. V. 8 (not further considered in this paper) 3a. A large spiny scute at tip of snout; V. 8 or 9 (carelyal@) wih 2u te Belt eS K. leonis 3b. No scute at snout tip; V. 11-13 side of first arch, 12.2 5 aa K. paepkei 4b. Scales below 2D. 10 or more; gill rakers on inner side of first arch fewer than 11 Sa. 1D. Il,11-13; 1P. 124-127; upper jaw 40- AD Ny ee ln K. pallida 5b. 1D. II,8-10; 1P. 116-125; upper jaw 33- 36% AL 2.2. 6 6a. 1D. black overall... eee K. dara 6b. 1D. dark basally, paler distally 0. tte ORIEL SPECIES ACCOUNTS Kuronezumia bubonis Iwamoto, 1974 (Figs. 1, 2) Nezumia (Kuronezumia) bubonis lwamoto, 1974:509-515, Figs. 1-3 (Caribbean Sea, Gulf of Mexico, Hawaii). Nezumia bubonis: Iwamoto 1979:163 (listed). Arai in Uyeno et al. 1983:220, fig. (1 spec., 551 mm TL: Suriname/French Guiana). Iwamoto in Cohen et al. 1990:251, fig. 570 (in key). Okamura in Amaoka et al. 1990:189 (1 spec., 348 mm TL; New Zealand). Kuronezumia bubonis: Shcherbachev 1987:41 (9 spec., Mad- agascar and West Australian ranges, Indian Ocean; 841- 1,200 m). Sazonov and Iwamoto 1992:67, fig. 25b, (scales illus.). COUNTS AND MEASUREMENTS (see Table 1). DISTRIBUTION (Fig. 9).—This widely distrib- uted species, originally described from the trop- ical western North Atlantic and the Hawaiian Islands, has subsequently been recorded from New Zealand (Okamura in Amaoka et al. 1990) and the Indian Ocean (Shcherbachev 1987). The species is newly recorded from the Andaman Sea from one specimen recently found in ZMMGU. REMARKS. — Two small specimens (CAS 77315) from Hawaii appear to be this species, but they lack the prominent tubercular swelling of the light organ of the species and appear to have a more extensive naked region on the underside of the snout. The tubercular light organ may represent a size-related characteristic (if these are in fact K. bubonis, they are the smallest specimens known). The relatively poor condition of the two may account for the nakedness on the underside of snout—the scales may have been sloughed off. SPECIMENS EXAMINED.—PACIFIC OCEAN. Hawaiian Islands: CAS 77315 (22.3-23.8 HL, 90+-147 TL); off Kaiwi, 21°08.8’N, 157°42.2'W; 585-640 m; 41-ft otter trawl; Townsend Crom- well cr. 36. sta. 31; 6.V.1968. South China Sea: SHCHERBACHEV, SAZONOV AND IWAMOTO: GRENADIER GENUS KURONEZUMIA 99 Ficure 1. Kuronezumia bubonis, from off Hawaii. Scale bar equals 25 mm. (After Iwamoto 1974.) Ficure 2. Scanning electron micrographs of body scale of Kuronezumia bubonis: (a) from dorsum below interspace between first and second dorsal fins; (b) enlargement of spinules of same scale. Scale bar in (a) equals 1.5 mm; in (b) equals 0.25 mm. 100 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 TaBLe 1. Ranges of selected measurements and counts of seven species of Kuronezumia. Figures in parentheses denote exceptions outside range of others in species. The last column presents data for the undescribed South China Sea species reported by Iwamoto (1974). bubonis dara leonis* macronema __ paepkei pallida sp. n= 12 n=2 n= 20 n=3 n=1 n=2 n=1 Measurements: aie 360-730 111-253 195-495 208-343 270 379-548 347 HL 54-124 20-51 43-84 33-51 59 79-99 64 in percent of head length Snout 26-32 28-30 27-34 27-28 28 28-29 30 Preoral 14-18 15-16 18-24 20 19 16-17 16 Internasal 17-20 18-19 18-24 _ 21 _ — Interorbital 23-26 25 22-30 23-25 24 24-27 29 Orbit 23-31 26-36 27-35 27-30 29 24-25 25 Suborbital 18-20 18-19 16-18 18 18 20 18 Postorbital 49-51 36-43 (36)40-44 — 50 — 48 Orbit-preop. 40-41 38-40 31-39 37-40 3 42-46 41 Upper jaw 35-44 30-35 29-35 34-36 34 40-42 41 Barbel 20-29 20-25 18-26 20-23 18 23-26 22 Gill slit 18-20 16-18 12-16(19) 13-17 14 18 19 Counts: 1D. II, 11-12 II, 9-10 II, 19-10(8) II,8-10 II, 11 II, 11(13) II, 10 iP 121-124 i21-123 119-126 116-122 121-122 124-125 123 W, (9)11—13(14) 10-12 (7)8-10 (9)11 10-11 11 8 GR-I (outer) 6-8 7-8 (7,8)9-11 - 8 7-8 8 GR-I (inner) 8-10 10 7-12 — 12 8-9 10 GR-II (outer) 9-10 9 8-11 10 11 7-9 9 GR-II (inner) 9-10 10-11 9-12 _ 9 9 — Scales 1D. 19-21 12-15 17-20 11-14 13 11-14 17-18 Scales 12-16 9-10 9-12(14, 15) — 7 — — mid-1D. Scales 2D. 12-15 10-11(12) (11)12-14(15) 10-12 8 11-14 16 Scales lat.1. 44-61 42-44 44-53 47 34-37 _ 50 * Two smallest specimens of K. /eonis excluded beceuse of substantial size-related differences in morphometry. ZMMGU P-17658 (65 mm HL, 360+ mm TL); off Viet Nam, 11°38’N, 109°48’E; 900-1,300 m; Odissey tr. 50; 20.[X.1984. INDIAN OCEAN. Madagascar Ridge: ZMMGU P-17659 (57.5 HL, 234+ TL); 33°30’S, 44°00’S; 940-1,110 m; Zvez- da Kryma cr. 6, tr. 20; 7.VII.1976. ZMMGU P-17660 (94 HL, 446+ TL); 32°44.3’S, 45°02.6’E; 1,240 m; Zvezda Kryma cr. 6, tr. 56; 23. VII.1976. CAS 66502 (47.1 HL, 235+ TL); Walters Shoals, 33°25'18”S, 43°37'30"E; 1,000-1,030 m; 29-m otter trawl; Vityaz’cr. 17, sta. 2721; 17.XII.1988. ZMMGU uncat. (255+ TL); 33°21’S, 44°29’E; 928-940 m; Fiolent cr. 11, tr. 15; 3.1V.1979. West Australian Ridge (Broken Ridge): ZMMGU P-17/662,(60 HE, 256-5 TE)31530'S. 95°04.3/E: 1,080-1,150 m; Zvezda Kryma cr. 6, tr. 131. 18.1X. 1976. ZMMGU P-17663 (124 HL, 745+ TL); 30°54’S, 93°49.0’E; 1,060-1,200 m; Zvezda Kryma cr. 6, tr. 187; 3.X.1976. ZMMGU P-17667 (54.5 HL, 334+ TL); 31903’S, 93°08.3’E; 1,049-1,064 m; Prof. Mesiatzeyv cr. 7, tr. 26; 3.1V.1979. ZMMGU P-17664 (49 HL, 304+ TL); 31°10'S, 93°57'E; 1,047-1,080 m; Prof. Mesi- atzev cr. 7, tr. 27; 3.1V.1979. ZMMGU P-17665 (103 HL, 600+ TL); 31°40.7'S, 95°30’E; 1,050- 840 m; Prof. Mesiatzev cr. 7, tr. 33; 4.1V.1979. ZMMGU P-17666 (49.3 HL, 280+ TL); 30°59’S, 93°35'E: 1,050 m: Fiolent crore. 4.VIII.1977. Kuronezumia dara (Gilbert and Hubbs, 1916) Lionurus darus Gilbert and Hubbs, 1916:197, pl. 10, fig. 1 (Suruga Bay; 355 m). Nezumia darus: Okamura 1970:101-103 (descr. after Gilbert and Hubbs 1916). Iwamoto 1974:507 (compared with K. bubonis). Okamura in Okamura and Kitajima 1984:217, 363, fig. 153 (2 spec., 220-318 mm TL; Okinawa Trough; 560-692 m). Okamura in Okamura et al. 1982:161, 349, fig. 95 (2 spec., 130-144 mm TL; Tosa Bay, 605 m). Nezumia dara: Iwamoto 1979:158 (compared with N. /iolepis). Iwamoto in Cohen et al. 1990:251, fig. 571 (in key). SHCHERBACHEV, SAZONOV AND IWAMOTO: GRENADIER GENUS KURONEZUMIA 101 Ficure 3.—Kuronezumia leonis, syntype from off Cape Point, South Africa. Scale bar equals 25 mm. (From Iwamoto in Smith and Heemstra 1986.) Kuronezumia dara: Sazonov and Iwamoto 1992:65 (compared with K. pallida). DISTRIBUTION (Fig. 9).— Kuronezumia dara has been recorded only from southern Japan and the Okinawa Trough in 355-692 m. REMARKS. — We did not find representatives of this species in the collections in Moscow or CAS. Kuronezumia leonis (Barnard, 1925) (Figs. 3, 4) Lionurus leonis Barnard, 1925a:503 (Cape Point; 640 m); 1925b: 349 (additional descr.). Macruroplus leonis: Smith 1949:134 (brief descr.). Nezumia leonis: Krefft 1968:36 (Uruguay). Karrer 1973:219 (SE Atlantic). Iwamoto in Smith and Heemstra 1986:338, fig. 93.21 (brief descr.; illus.). Iwamoto in Cohen et al. 1990: 250, fig. 567 (in key). Nezumia sp. A: Okamura in Amaoka et al. 1990:191, fig. (2 spec., New Zealand). COUNTS AND MEASUREMENTS (see also Table 1).—GR-I (outer/inner series) 7—11/(1-3)+(6- 9)(7-12 total); GR-II (1-2)+(6-10)(8-11 total)/ (1-2)+(7-11)(9-12 total); pyloric caeca 14-18 (3 spec.). The following in percent of head length: rictus 21.8-27.2; ID.-2D. interspace 36-79; height 1D. 72-92; length 1P. 50-65; length V. 41-47; post. nostril 4.3-8.6; body depth 75-109. DISTRIBUTION (Fig. 9).—Kuronezumia leonis was previously known from a few specimens col- lected off southern Africa and the southwestern Atlantic. The species has in fact a surprisingly broad distribution in the southern hemisphere, with collections now known from southern Aus- tralia, New Zealand, and the Kerguelen under- water ridge in the southern Indian Ocean. REMARKS. — The relatively pointed snout tipped with a large terminal scute in K. /eonis is unique in the genus, as other members have a broadly rounded snout lacking modified scute-like scales. SPECIMENS EXAMINED.—SOUTH ATLAN- TIC. Argentina: ISH uncat. (47.6 HL, 313+ TL); Argentina (slope); Walther Herwig sta. 927; 28.1X.78. Rio Grande Rise: ZMMGU P-13707 G22HEA 326-- TL); 31°20.258; 35°31. 22We 280 m:; Prof. Mesiatzev cr. 2, tr. 15; 8.VII.1974. Dis- covery Seamount: ZMMGU P-17657 (52.5 HL, Ficure 4. Scanning electron micrographs of body scale of Kuronezumia leonis from dorsum below interspace between first and second dorsal fins. Scale bar equals 1.0 mm. 102 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 323+ TL); 48°03.6’S, 8°14.5’E; 960-970 m; Ev- rika; 22.111.1981. Namibia: ZIN uncat. (60.0 HL, 325+ TL); no other data. South Africa: SAM 12637 (lectotype, here designated, 43.4 mm HL, 195+ mm TL); Cape Point, NE by E % E, 8 miles [13.1 km], 91 fm [166.4 m]; Pieter Faure sta. 17555; shrimp trawl; 27.VIII.1903. BMNH 1927.12.6.19 (paralectotype, 45.6 HL, 240+ TL); off Cape Point. CAS 76476 (64.3 HL, 433+ TL); Agulhas Bank, 36°41.6'S, 20°17’E; Fiolent cr. 5, tr. 114; 15.X1.1974. ZMMGU P-17656 (69 HL, 409+ TL); Agulhas Bank, 35°53.5'S, 19°20.6’E; 950-1,000 m; Fiolent cr. 4, tr. 140; 4.14.1974. INDIAN OCEAN. Mozambique: CAS 76475 (2: 22.8-23.9 HL, 111+-130+ TL); off Maputo; 25°26'30’S, 35°12'00”E; 715-700 m; Vityaz’ cr. 17, sta. 2630; 23.XI.1988. Kerguelen Ridge: ZMMGU P-160130(25-0' HE, 120+-TE); 56°51.8'S, 70°50.5’E; 945-955 m; Fiolent cr. 10; 6.1.1978. PACIFIC OCEAN. Australia: AMS 1.20068-011 (4:57.4-77.0 HL, 335-490+ TL); N.S.W., E of Broken Bay; 33°27'S, 152°09'E; 886— 895 m; F.R.V. Kapala sta. K77-23-13; bottom trawl; 8.XII.1977. AMS 1.24462-004 (49.0 HL); N.S.W., off Cape Hawke; 32°06’S, 153°08’E; 461 m; F.R.V. Kapala sta. K83-25-02; bottom trawl; 2.X1.1983. AMS I.24157-007 (84.2 HL, 495+ TL); Victoria, off Cape Everard; 149°47’E, 3821948. 19971201 S<*m: « FER.VE t Kapala: 27.1X.1983. AMS I.20098-005 (4:65.1-69.1 HL, 355+-—410+ TL); N.S.W., E of Broken Bay; 33°34'S. 152°O01'E: 904 m: FUR.V. Kapala: 6.XII.1977. New Zealand: LACM 11447-3 (55.8 HE 320-- DL): 47°22. WS) 147552E 2 OSs meee tanin sta. 1981; 24.11.1967. Other specimens not examined in detail: NMV A.5885 (480 TL); Australia, S.A., W of Robe; 37°11.7'S, 138°05.5’E; 1,030-1,070 m. NMV A.3404 (383+ TL); Australia, Vic., off Portland; 38°38'S, 141°904’E; 990-1,100 m. ZIN uncat. (400 TL); Agulhas [South Africa]. Kuronezumia macronema (Smith and Radcliffe, 1912) (Figs. 5, 6) Macrourus macronemus Smith and Radcliffe, in Radcliffe 1912: 115, pl. 24, fig. 3 Jolo Sea, near Cagayan Is., 340 fm). Lionurus (Nezumia) macronemus: Gilbert and Hubbs 1916: 145 (list). Ventrifossa (Ventrifossa) macronemus: Gilbert and Hubbs 1920: 545-546. Ventrifossa macronema: Herre 1953:174 (list). Makushok 1967: 218 (distr.). Marshall 1973:654 (list). Nezumia macronema: twamoto in Cohen et al. 1990:251, fig. 573 (in key). COUNTS AND MEASUREMENTS (see Table 1).— The following in percent HL: 1D.—2D. interspace 22-32; height 1D. 98; length 1P. 52-59; length V. 75-95; body depth 77-87. DISTRIBUTION (Fig. 9).—The species has not been collected, so far as we know, since the orig- inal material, which consisted of three specimens captured in beam trawls at three localities in the Philippines in 622-803 m. REMARKS.— Kuronezumia macronema has been a perplexing species that eluded proper placement for more than three-quarters of a cen- tury. After its initial description, Gilbert and Hubbs (1920) placed it in their new genus Ven- trifossa and considered it to be closest to V. misa- kia (Jordan and Gilbert, 1904). Marshall (1973) followed Gilbert and Hubbs. Iwamoto (1974:513) included the species in Nezumia, and considered it, along with Nezumia burragei (Gilbert, 1905) and N. hebetata (Gilbert, 1905), to “‘fall in a loose group near N. darus in their resemblance to the species of subgenus Kuronezumia.” He contin- ued this idea in 1979 in discussing relationships of Nezumia liolepis and again in 1990 (in Cohen et al. 1990), where he placed the species in his Nezumia “Group C” along with K. bubonis and K. dara. It was not until we realized that Macru- roplus leonis belonged in Kuronezumia that we also considered transferring Macrourus macro- nemus to that genus. Doing so forced comparison with other members of Kuronezumia and quickly established how similar K. macronema was to the others, especially K. dara, which it closely resembles in counts and proportional measure- ments. The original illustration of the species (Rad- cliffe 1912: pl. 24, fig. 3) erroneously depicts a fish with a relatively large mouth, the upper jaw extending to below the hind third of the orbit. In fact, however, the upper jaw fails to reach a vertical through the middle of the orbit. The cor- rect position is shown in our illustration of a paratype (Fig. 5). Iwamoto (1974) reported that K. macronema has lanceolate scale spinules, but closer exami- nation of scales from CAS-SU 25233 using scan- ning electron micrographs has shown the spi- nules to be long, slender, and needlelike, and not flattened into lanceolate shapes. SPECIMENS EXAMINED (all PHILIPPINES).— Holotype: USNM 72931 (51 mm HL, 343 mm TL), Jolo Sea near Cagayan Is., 9°37'05’N, 121°12'37”E; 622 m; Albatross sta. 5424. Para- types: USNM 135346 (48.5 HL, 257 TL); be- SHCHERBACHEV, SAZONOV AND IWAMOTO: GRENADIER GENUS KURONEZUMIA 103 Ficure 5. Kuronezumia macronema, USNM 135346, from Philippines, between Siquijor and Bohol islands. (a) Lateral view; (b) dorsal view of head; (c) ventral view of abdomen. Scale bar equals 2.5 mm. tween Siquijor and Bohol islands; 803 m; A/- batross sta. 5528; CAS-SU 25233 (32.5 HL, 208 TL); off SW Luzon; 622 m; Albatross sta. 5114. Kuronezumia paepkei n. sp. (Figs. 7, 8) ?Macrurus (Macrurus) rudis (non Giinther, 1878): Brauer 1906: 264. MATERIAL. — Holotype: ZMB 17641 (58.5 mm HL, about 27 cm TL, with large pseudocaudal); Indian Ocean off Kenya; 03°07’S, 40°45'E; 748 m; Valdivia sta. 249. DiaGcnosis.—A species of Kuronezumia with 12 mesial gill rakers on first arch; eight scales below origin of second dorsal fin; no bulblike swelling of light organ, anterior dermal window represented by a small scaleless area between midbases of pelvic fins; pores of sensory canals on head small but present; color gray-brown, fins dark. COUNTS AND MEASUREMENTS (see Table 1).— FiGurE 6. Scanning electron micrographs of body scale of Kuronezumia macronema (CAS-SU 25233, 32.5 mm HL) from dorsum below interspace between first and second dorsal fins. Scale bar equals 0.6 mm. 104 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 b (ei Ficure 7. Kuronezumia paepkei n. sp. Holotype (ZMB 17641) from off Kenya in 748 m. (a) Lateral view; (b) dorsal view of head; (c) ventral view of abdomen. Scale bar equals 25 mm. GR-I (outer/inner series) 8/2+10,GR-II 1+10/ rictus 28.3; pre-D. 112; pre-A. 154; pre-V. 123; 1+8; 13 precaudal vertebrae; 11 anal pterygio- snout to vent 141; V.-A. 34.8; 1D. base 29.3; phores anterior to first haemal arch. 1D.-2D. interspace 34.1; height 1D. 95; length The following in percent of HL: postrostral 75; —1P. 63; length V. 79; body depth 96. il ob fe , cee) Saas 2 tienen Ficure 8. Scanning electron micrographs of body scale of Kuronezumia paepkei (holotype, ZMB 17641): (a) from dorsum below interspace between first and second dorsal fins; (b) enlargement of spinules of same scale. Scale bar in (a) equals 0.75 mm, in (b) equals 0.3 mm. SHCHERBACHEV, SAZONOV AND IWAMOTO: GRENADIER GENUS KURONEZUMIA DESCRIPTION. —A medium-sized, deep-bodied species of Kuronezumia, greatest depth about equal to HL or about five in TL (pseudocaudal present). Head moderately compressed and deep, its depth notably less than greatest body depth. Orbit almost circular, its diameter a little more than snout length or about one-fifth more than interorbital width. Snout relatively high with subvertical anterior profile, its tip rounded but not truncated, triangular in shape in dorsal view (Fig. 7b). Mouth small, jaws subterminal; pos- terior end of maxillary extends to about vertical through anterior margin of pupil. Head ridges inconspicuous; infraorbital region moderately deep, subvertical, slightly convex, without mod- ified or enlarged scutelike scales, about 10 scales wide, 5 or 6 on upper portion. Scaled parts of opercle and subopercle forming a deep, inverted triangle. Interopercle narrowly exposed and scaled only at posteroventral margin (scales in small oval patch). Free margin of preopercle smooth. Mental barbel moderately thick, long, equal to preoral length, but conspicuously shorter than orbit diameter, tapering to a fine tip. Gill mem- branes broadly united (at level of middle of oper- cle), almost without free margin behind their connection with isthmus. No scutelike scales at snout tip nor at anterior end of nasal bones; scales around tip of snout slightly stouter than those surrounding, but not raised into tubercles. Scales bordering orbit with- out crests and indistinct from adjacent scales. Head almost completely scaled except over gular and branchiostegal membranes, anterior part of mandibles, and a median swath on lower surface of snout below tip. A narrow, barely perceptible strip of naked skin (bearing sensory pores) along lower margin of infraorbital bones above upper jaw. Body scales (Fig. 8) large (as reflected in the 7 scale rows below the midbase of the first dorsal and 8 below the origin of the second dorsal fin) for the genus, densely covered with long, some- what curved, retrorse, needlelike spinules in ir- regularly quincunx order. Posteriormost spinules extend well beyond margin of scale. Transverse ridges on anterior field in a limited region along anterior edge of spinule field. Scales cover prox- imal part of pelvics between rays. One row of scales just below lowermost rays of pectoral fin with few, low, radiating crests without spinules. Premaxillary teeth short, stout, conical, in broad band (6-7 rows of teeth over most of band) gradually tapering posteriorly. Teeth of outer row scarcely but distinctly enlarged. Teeth on lower 105 jaw in semilunar patch, broad at symphysis (up to 7 rows) but rapidly tapering posteriorly. In- terdental spaces and adjacent parts of oral cavity covered with numerous short papillae; lips and anteriormost portion of snout just above upper jaw also covered with similar papillae, but which are shorter and more widely set. Onigin of first dorsal fin slightly behind vertical through origin of pelvics, and aligned with origin of lowermost pectoral rays. Spinous second ray of first dorsal slightly less than length of head; serrations on leading edge numerous, with 22 low but acute teeth. Interdorsal space short, little more than base of first dorsal fin. Pectoral fins mod- erately long; their origin well below level of top of gill openings. Pelvics with long filamentous ray extending to about 13th anal ray. Origin of anal fin somewhat behind vertical of posterior end of first dorsal. Sensory canals on head not broadened or swol- len; open pores present in all canals, small and situated at tips of relatively low, tubular projec- tions. Free neuromasts on head indistinguish- able. Olfactory cavity small, about equal to pupil diameter, anterior nostril rounded, much smaller than posterior nostril, which is sem1-elliptical and 1.8 times into least suborbital width. Internarial membrane narrow, with flap length equal to half diameter of anterior nostril. Light organ represented by a black scaleless anterior dermal window situated between mid- bases of pelvic fins and separated from periproct by convex, scale-covered area (Fig. 7c). Periproct very narrow. Most internal organs lost (probably a result of earlier dissection). Coloration in alcohol: body gray-brown, rather light, branchiostegal membranes light brown; oral and branchial cavities light with yellowish tinge, outlines of the latter brown; all fins (including posterior part of second dorsal) dark brown (anal fin rather pale basally, but black distally). Brauer (1906) reported coloration of specimen as light brownish; fins, vertical surfaces, and opercles blackish. EtyMoLoGcy.— Named for Dr. Hans-Joachim Paepke, Curator of Fishes at the Museum fiir Naturkunde der Humboldt-universitat zu Ber- lin. DISTRIBUTION (Fig. 9).— Known only from the type locality on the continental slope of east Af- rica off Kenya in 748 m. REMARKS AND COMPARISONS. — Kuronezumia paepkei has larger body scales and more gill rak- ers (medially on the outer arch) than do other members of the genus (see Table 1), but the spe- 106 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 . bubonis . dara . leonis . macronema . paepkei . pallida : SY aia | he 8 y Ficure 9. Distributions of six species of Kuronezumia. cies agrees with the generic concept of Sazonov and Iwamoto (1992) in having a rounded snout profile, a smoothly rounded suborbital region covered with uniformly small scales and lacking a stout, angular, longitudinal ridge formed of en- larged scutelike scales, and in having broad, short teeth bands in both jaws. In addition to its high gill-raker count and large scales (and commensurate lower scale row counts), the new species differs from: K. bubonis and the undescribed South China Sea species in lacking a bulbous swelling of the luminous organ and in having many different proportions (e.g., shorter upper jaw, shorter barbel, longer pelvic fin, narrower outer gill slit); K. pallida and K. dara in having fewer pectoral fin rays (22-23 total vs. 24-26); and different proportions (larger orbits, narrower suborbital, shorter barbel, nar- rower outer gill slit); K. /eonis in lacking a prom- inently enlarged terminal snout scute, having smaller teeth, and having more extensive median naked area on underside of snout, the nakedness extending dorsally to the base of terminal snout scute; and K. macronema in having more exten- sive naked areas on underside of snout. Kuronezumia pallida Sazonov and Iwamoto, 1992 Nezumia sp.: Parin et al. 1981:12 (brief descr., spec. from Sala y Gomez ridge). Kuronezumia sp.: Parin 1990:16 (listed from Sala y Gomez ridge). Kuronezumia pallida Sazonov and Iwamoto, 1992:65-67, fig. 24, 25a (2 spec., Sala y Gomez Ridge, 540-800 m). COUNTS AND MEASUREMENTS (see Table 1). DISTRIBUTION. — Sala y Gomez Ridge in south- eastern Pacific, 540-800 m. REMARKS. — We have no additional specimens to report. The species is very similar to K. dara, and juveniles of the two species were indistin- guishable. Adults, however, can be differentiated by characters given in the key and by their small- er orbit diameter (24-25% HL vs. 26-36%). ACKNOWLEDGMENTS We thank Dr. Hans-Joachim Paepke (ZMB) for loan of the specimen that initiated this study. The following curators and their staffs are thanked for their assistance in various ways: J. R. Paxton and M. McGrouther (AMS); A. Wheeler, N. G. SHCHERBACHEV, SAZONOV AND IWAMOTO: GRENADIER GENUS KURONEZUMIA 107 Howes, O. Crimmen (BMNH); N. V. Parin (IOAN); M. Stehmann and A. Post (ISH); M. F. Gomon (NMV); A. Hulley (SAM); V. G. Spring- er, R. A. Vari and others (USNM); A. Neyelov, V. I. Barsukov (deceased), and others (ZIN)); I. A. Verighina (ZMMGU). The SEM photographs were taken by L. Borok and D. Ubick (CAS); D. Catania (CAS) made radiographs. Iwamoto’s ex- tended visit to Russia in 1988 was supported by the National Academy of Sciences/National Re- search Council through its Soviet-East European Scientific Exchange Program. Visits by Iwamoto to BMNH and ISH were made possible by W. Fischer of FAO. LITERATURE CITED AmAoKA, K., K. MAtsuurA, T. INADA, M. TAKEDA, H. Ha- TANAKA, AND K. OKADA. 1990. Fishes collected by the R/V Shinkai Maru around New Zealand. Japan Mar. Fish. Re- source Res. Center, Tokyo. 410 pp. BARNARD, K. H. 1925a. Descriptions of new species of ma- rine fishes from South Africa. Ann. Mag. Nat. Hist., ser. 9, 15(87):498-504. . 1925b. A monograph of the marine fishes of South Africa. Pt. 1. Ann. S. Afr. Mus. 21:1-418. Brauer, A. 1906. Die Tiefsee-Fische. I. Systematischer Teil. Wissenschaftliche Ergebnisse der deutschen Tiefsee-Expe- dition auf dem Dampfer “Valdivia,” 1898-1899, 15:1-432, text-figs. 1-176, pls. 1-18. Couen, D. M., T. INADA, T. IWAMOTO, AND N. SCIALABBA. 1990. FAO species catalogue. Gadiform fishes of the world (Order Gadiformes): an annotated and illustrated catalogue of cods, hakes, grenadiers and other gadiform fishes known to date. FAO Fish. Synop. no. 125, vol. 10. 442 pp. GiLBerT, C. H. 1905. The deep-sea fishes of the Hawaiian Islands. Pp. 575-713, figs. 230-276, pls. 66-101 in The aquatic resources of the Hawaiian Islands. D. S. Jordan and B. W. Evermann, eds. Bull. U.S. Fish Comm. 1903, 22(pt. 2, sect. 2). GILBerT, C. H. AND C. L. Husss. 1916. Report on the Jap- anese macrouroid fishes collected by the United States fish- eries steamer ‘“‘Albatross” in 1906, with a synopsis of the genera. Proc. U.S. Natl. Mus. 51:135-214, pls. 8-11. . 1920. The macrourid fishes of the Philippine Islands and the East Indies. U.S. Natl. Mus. Bull. 100, 1 (pt. 7):369- 588. GUnTHER, A. 1878. Preliminary notices of deep-sea fishes collected during the voyage of H.M.S. ‘Challenger.’ Ann. Mag. Nat. Hist., ser. 5, 2:23-28. Herre, ALBERT W. 1953. Check list of Philippine fishes. U.S. Fish & Wildl. Serv. Res. Rep. 20. 977 pp. Iwamoto, T. 1970. The R/V Pillsbury Deep-sea Biological Expedition to the Gulf of Guinea, 1964-65. 19. Macrourid fishes of the Gulf of Guinea. Stud. Trop. Oceanogr. Miami (4)(pt. 2):316-431. 1974. Nezumia (Kuronezumia) bubonis, a new sub- genus and species of grenadier (Macrouridae: Pisces) from Hawaii and the western North Atlantic. Proc. Calif. Acad. Sci., ser. 4, 39(22):507-516. 1979. Eastern Pacific macrourine grenadiers with seven branchiostegal rays (Pisces: Macrouridae). Proc. Calif. Acad. Sci. 42(5):135-179, figs. 1-18. Iwamoto, T. AND Y. I. SAzonov. 1988. A review of the southeastern Pacific Coryphaenoides (sensu lato) (Pisces, Gadiformes, Macrouridae). Proc. Calif. Acad. Sci. 45(3):35— 82. Iwamoto, T. AND Y. N. SHCHERBACHEV. 1991. Macrourid fishes of the subgenus Chalinura, genus Coryphaenoides, from the Indian Ocean. Proc. Calif. Acad. Sci. 47(7):207-233. JORDAN, D. S. AND C. H. GitBert. 1904. Macrouridae. Pp. 602-621 in List of fishes dredged by the steamer A/batross off the coast of Japan in the summer of 1900, with descrip- tions of new species and a review of the Japanese Macrouri- dae. D. S. Jordan and E. C. Starks, eds. Bull. U.S. Fish Comm. (1902) 22:577-630, pls. 1-8. Karrer, C. 1973. Uber Fische aus dem Siidostatlantik. Mitt. Zool. Mus. Berlin 49(1):191—257. Krerrt, G. 1968. Neue und erstmalig nachgewiesene Knor- pelfische aus dem Archibenthal des Siidwestatlantiks, einschliesslich einer Diskussion einiger Etmopterus-Arten siidlicher Meere. Arch. Fischereiwiss. 19(1):1-42. Leviton, A. E. AND R. H. Grsss, Jr. 1988. Standards in herpetology and ichthyology: standard symbolic codes for institutional resource collections in herpetology and ichthy- ology. Supplement No. 1: additions and corrections. Copeia 1988(1):280-282. Leviton, A. E., R. H. Gress, Jr., E. HEAL, AND C. E. DAwson. 1985. Standards in herpetology and ichthyology: part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia 1985(3):802— 832. MaAkusHok, V. M. 1967. Whiptails (family Macrouridae or Coryphaenoididae Auct.). Chap. IV, pp. 221-244 in The Pacific Ocean, Biology of the Pacific Ocean, Book III, Fishes of the open waters. [In Russian.] V. G. Kort, chief ed. Mos- cow. 273 pp. [English transl. by U.S. Naval Oceanographic Office, Washington, D.C. Transl. 528. 1971.] Marsua tl, N. B. 1973. Family Macrouridae. Pp. 496-665, in Fishes of the western North Atlantic. D. M. Cohen, ed. Mem. Sears Found. Mar. Res. (1)(pt.6). OKAMuRA, O. 1970. Fauna Japonica, Macrourina (Pisces). Academic Press of Japan, Tokyo. 11 + 216 pp., 44 pls. OKAMURA, O., K. AMAOKA, AND F. MITANI, EDS. 1982. Fishes of the Kyushu-Palau Ridge and Tosa Bay. Japan Fish. Res. Cons. Assoc., Tokyo. 435 pp. OKAMURA, O. AND T. KiTAsIMA, EDS. 1984. Fishes of the Okinawa Trough and the adjacent waters—I. Japan Fish. Res. Cons. Assoc., Tokyo. 414 pp. Parin, N. V. 1990. Preliminary review of fish fauna of the Nazca and Sala y Gomez Submarine Ridges (southern, East Pacific Ocean). [In Russian with English summary.] Tr. Inst. Okeanol. Akad. Nauka SSSR 125:6-36. Parin, N., G. A. GOLovAN’, N. P. PAKHoRUKOYV, Y. I. SAZONOV, AND Y. N. SHCHERBACHEV. 1981. Fishes from the Nazca and Sala-y-Gomez underwater ridges collected in cruise of R/V “Ikthiandr.” [In Russian with English summary.] Pp. 5-18 in Fishes of the open ocean. N. V. Parin, ed. P. P. Shirshov Inst. Okeanol. Akad. Nauk SSSR, Moscow. 120 pp. RapcuiFFE, L. 1912. Descriptions of a new family, two new genera, and twenty-nine new species of anacanthine fishes from the Philippine Islands and contiguous waters. Proc. U.S. Natl. Mus. 43:105-140, pls. 22-31. Sazonov, Y. I. AND T. IwAMoTo. 1992. Grenadiers (Pisces, Gadiformes, Macrouridae) of the Nazca and Sala y Gomez ridges. Proc. Calif. Acad. Sci. 48(2):27-95. 108 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 3 SHCHERBACHEV, Y. N. 1987. Preliminary list of thalasso- bathyal fishes of the tropical and subtropical waters of the Smitu, M. M. AND P. C. HEEMsTRA. 1986. Smiths’ sea fishes Macmillan South Africa (Publishers), Ltd. 1,047 pp. Indian Ocean. [In Russian.] Voprosy Ikhtiologii (1):3-11. [English version in J. Ichthyol. 27(2):37-46.] UvyeEno, T., K. MATsuuRA, AND E. Fusi. 1983. Fishes trawled Smitu, J. L. B. 1949. The sea fishes of southern Africa. Cen- off Suriname and French Guiana. Japan Mar. Fish. Resource Res. Center, Tokyo. 519 pp. tral News Agency, Ltd., Cape Town. 550 pp. © CALIFORNIA ACADEMY OF SCIENCES, 1992 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES... : LIBRARY Vol. 48, No. 4, pp. 109-118, 6 figs., 1 table. f 2002 FEB 4 | a Pres re) THREE NEW SPECIES OF STOLIDOBRANCH ASCIDIANS (CHORDATA: ASCIDIACEA) FROM THE CALIFORNIA CONTINENTAL SHELF By Gretchen Lambert Department of Biological Science, California State University, Fullerton, California 92634 Apstracr: Three new species of solitary stolidobranch ascidians are described from the California conti- nental shelf region from Point Conception to Point San Luis. Molgula napiformis was collected from soft sediment at 90-92 m. It is distinguished from other northeast Pacific species of Molgula by a constellation of characters that includes a very long, branching posterior stolon (presumably used for anchoring in the substrate); the placement of the siphons; gonad shape; and the tubular opening of its dorsal tubercle. Boltenia polyplacoderma was collected from rock surfaces at 91-117 m. It is unlike any other species of Boltenia in that the anterior and lateral tunic surfaces are divided into annulated flat plates that increase in number as the animal grows. Styela tesseris, collected from rocks at 61-88 m, is easily distinguished from other northeast Pacific species of Styela by the tessellated tunic pattern, especially when this trait is considered along with all the other body characters. A list of all the ascidians collected is given with their depths and localities. In addition, the ascidian assemblages that were present at 11 of the 18 hard-substrate stations sampled are enumerated. Received April 22, 1992. Accepted August 31, 1992. January 21, 1993 INTRODUCTION Although the intertidal and shallow subtidal ascidian fauna is well known for the Pacific coast from Alaska to Mexico and even for much of South America (Van Name 1945), the deeper water species have rarely been sampled. Of the 14 species collected by dredging off the California coast in 1904, nine were new species, indicating that this area was not well known (Ritter 1907). Between 1983 and 1988 the U.S. Department of the Interior Minerals Management Service (MMS) sponsored a major collecting effort of the benthic invertebrate fauna of the Santa Maria Basin and western Santa Barbara Channel (Blake and Lissner, in press). During this project, some of the same localities that had been sampled by Ritter in 1904 were examined by box corer and manned submersible. Ascidians were found at 11 of the 18 hard-substrate stations and at four of the 136 soft-substrate stations. Twenty species of ascidians were collected; six of these are the same species collected by Ritter, 1 1 are described species not previously recorded from this area, and three are undescribed species (Lambert, in press). In addition to the following descriptions of these three new stolidobranch species, a list of all the ascidians collected during the MMS pro- ject, with their depth and locality, is given in Table 1. The ascidian assemblages at the 1 1 hard- substrate stations that included ascidians are list- ed in Appendix A. [109] 110 TABLE 1. Order Aplousobranchia Family Didemnidae 1. Didemnum carnulentum var. lacteolum 2. Trididemnum opacum Family Polycitoridae 3. Distaplia occidentalis 4. Eudistoma carolinense (?) Family Synoicidae 5. Aplidium californicum 6. Aplidium spauldingi 7. Aplidium sp. 8. Euherdmania claviformis Order Phlebobranchia Family Ascidiidae 9. Ascidia sp. Family Rhodosomatidae 10. Chelyosoma columbianum 11. Corella sp. Order Stolidobranchia Family Molgulidae 12. Molgula napiformis 13. Molgula regularis Family Pyuridae 14. Boltenia polyplacoderma 15. Halocynthia igaboja 16. Pyura haustor 17. Pyrua sp. Family Styelidae 18. Styela gibbsii 19. Styela milleri 20. Styela tesseris 21. Styela sp. MATERIALS AND METHODS The three new species described in this paper were collected from the Santa Maria Basin and western Santa Barbara Channel’s outer conti- nental shelf during the 1983-86 Phase I assess- ment of long-term changes in biological com- munities in these areas by the U.S. Department of the Interior Minerals Management Service and the 1987-89 Phase II monitoring program by the PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 4 Ascidians collected during the MMS Phase I and Phase II Monitoring Program in the Santa Maria Basin. Depth (m) Location 61.5-73 Goleta Pt. to Pt. Arguello 61.5 Pt. Arguello 87 Purisima Pt. 86-98 Purisima Pt. 61.5 Pt. Arguello 97 Purisima Pt. 97 Purisima Pt. 117 Pt. Buchon 98.5 Purisima Pt. 54-231 Goleta Pt. to Pt. Buchon 231 Pt. Conception 90-92 Purisima Pt. to Pt. San Luis 105 Purisima Pt. 91.5-117 Pt. Conception to Purisima Pt. 88-231 Pt. Conception to Pt. Buchon 97-231 Pt. Conception to Pt. Buchon 61.5-97 Pt. Arguello 61.5 Pt. Arguello 69-117 Goleta Pt. to Pt. San Luis 61.5-88 Pt. Arguello to Purisima Pt. 69-73.5 Goleta Pt. same agency (Blake and Lissner, in press). Mol- gula napiformis n. sp. was collected from soft sediments by box corer, with the sediment sieved through a 0.1 mm mesh sieve. The rocks from which Boltenia polyplacoderma n. sp. and Styela tesseris n. sp. were taken were collected by manned submersible. After collection, the animals were preserved in 10% sea water formalin and later transferred to 70% isopropyl alcohol. Some of the animals LAMBERT: THREE NEW ASCIDIANS FROM CALIFORNIA 111 D Figure 1. Molgula napiformis n. sp. A, B, whole animals, left side, x 13 (a—atrial siphon, b—branchial siphon); C, dorsal tubercle between the peripharyngeal bands; D, three branchial tentacles, x 65; E, detail of branchial sac. C-E greatly enlarged. were placed in glycerin to clear them for micro- scopic examination and to prevent movement during the execution of the drawings; they were then returned to 70% isopropyl alcohol. Draw- ings were made with a camera lucida mounted on a stereozoom-dissecting microscope. DESCRIPTION OF SPECIES Molgula napiformis n. sp. (Figs. 1, 2) Ho.totyre.—MMS Phase II, Santa Maria Basin, soft sub- Strate, sta. R-4, cruise 2-5 (34°43.01’N, 120°47.39'W) M/V 112 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 4 FIGURE 2. Aloha at 92 m, coll. May 1988, U.S. National Museum of Natural History, catalog #20008. PARATYPES.— MMS Phase II, Santa Maria Basin, soft sub- strate, coll. by M/V Aloha: sta. R-1, cruise 2-4 (35°05.83'N, 120°49.16’W) at 91 m, 3 specimens coll. Jan. 1988, Santa Barbara Museum of Natural History #35608; sta. R-4, cruise 2-5 (34°43.01'N, 120°47.39'W) at 92 m, 5 specimens coll. May 1988, U.S. National Museum of Natural History, #20009; sta. R-8, cruise 1-2 (34°55.30'N, 120°45.87’'W) at 90 m, | speci- men, coll. Jan. 1987, California Academy of Sciences, #086764. DESCRIPTION. — Animals are almost spherical, 2-3 mm in diameter. The clear transparent tunic is covered with short hairs to which numerous sand grains firmly adhere. An extremely long, branched, root-like posterior process with long hairs projects from the posterior end of the tunic; additional long hairs may or may not be present around the main process. On one specimen the process, or stolon, is about 4 times longer than the animal and greatly branched; it gradually ta- pers to a very fine hair at its terminus. The si- phons are very short and widely separated; the branchial siphon is more noticeable on the tunic surface than the atrial siphon and produced into 6 pointed processes, while the atrial siphon is square. In the 7 specimens examined the location of the siphons is identical; the atrial siphon is anteriormost and at the opposite end of the an- imal from the stolon, while the branchial siphon is located back about one-third of the body length. The dorsal tubercle contains a narrow oval open- ing. In the largest animal examined (3 mm in diameter) there are 7 branchial folds on the left Molgula napiformis n. sp. A, right side of animal removed from tunic; B, left side (a—atrial siphon, b— branchial siphon, d—dorsal ganglion, |—liver, o—ovary, r—rectum, t—testis). x 30.7. side of the branchial sac and 6 on the right side; some of the folds are not distinct in the smaller, 2 mm individuals. There are three longitudinal vessels per fold. The infundibula are large, with only one row of them between each of the ru- dimentary branchial folds. All the animals ex- amined are more or less immature, so the adult number of infundibular spirals was not deter- mined. The dorsal lamina is a smooth mem- brane. The largest specimen has 8 large branchial tentacles, long and slender and variably branched; additional small tentacles of various sizes occur between the largest ones and were difficult to count. The mantle contains widely spaced lon- gitudinal muscles radiating from the siphons over the body and overlapping slightly; circular mus- cles are well developed only on the siphons. The intestine and the stomach with its attached lobed liver are on the left side; the kidney is not de- veloped. The left gonad lies in the secondary intestinal loop. The gonads are oval in outline, with branched or unbranched testes surrounding each ovary. EtyMoLoGy.—The specific epithet napiformis is derived from the Latin term napus describing the shape of a turnip root, which the body and long branched stolon of this ascidian resemble. REMARKS.—This species differs from Molgula pugetiensis (1) in possessing a very long, branched tunic stolon and secondary processes in addition to the numerous short tunic hairs; (2) in the si- phons being farther apart and shorter; and (3) in LAMBERT: THREE NEW ASCIDIANS FROM CALIFORNIA ‘ \. FIGuRE 3. A NCA i ATE 113 E WwW Boltenia polyplacoderma n. sp. A, anterior surface of tunic of holotype, x 7.2; B, anterior tunic surface of juvenile paratype, x10; C, branchial tentacle, x66; D, dorsal tubercle (side view) between the peripharyngeal bands, greatly enlarged; E, detail of branchial sac, x 56 (a—atrial siphon, b—branchial siphon, lv—internal longitudinal vessels, s—stigmata). the body being attached ventrally which results in the endostyle appearing to be posterior. Boltenia polyplacoderma n. sp. (Figs. 3, 4) Ho.totype.—002-BRA-01-TX, Santa Maria Basin (34°11.377'N, 120°29.318'W to 34°11.289'N, 120°28.774'W) at 116 m, 1 specimen coll. Jul—Aug. 1984. Santa Barbara Museum of Natural History, #35607. PARATYPES. —016-BRA-01-TX, 2 specimens (34°46.544'N, 120°50.197'W to 34°45.912’N, 120°49.726’W) at 98 m. U.S. National Museum of Natural History, #20010. 013-BRA-04- TX (34°42.570'N, 120°47.899'W to 34°42.107'N, 120°48.253’W) at 98 m, 2 specimens; California Academy of Sciences, #086765. 014-BRA-02-TX (34°43.589'N, 120°49.093'W to 34°42.826'N, 120°48.37’W) at 104 m, 1 spec- imen; California Academy of Sciences, #086766. All paratypes were collected during Jul—Aug. 1984. DESCRIPTION. — The body is very flattened an- teroposteriorly. The tunic is rather thin but com- posed of thickened, irregular, polygonal flat- topped plates, each plate with numerous, closely spaced annulations. The plates are straight-sided with 3-8 sides of unequal lengths. Each polygon is set off from the others by a narrow groove in the tunic, but the plates abut closely, especially in contracted regions. The inner surface of the tunic is covered by a continuous, thin sheet of white, fibrous material. The apertures are 4-sided 114 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 4 t/ Oe, {F i? ip iA fe Ficure 4. Boltenia polyplacoderma n. sp., body removed from tunic, x 8.7. A, anterior surface of mantle showing the siphons and musculature; B, posterior view of body, showing digestive and reproductive organs (a—atrial siphon, b—branchial siphon, e—endostyle, |—liver, o—ovary, od—oviduct, st—stomach, r—rectum, t—testis). and are formed by 4 main triangular plates with smaller plates forming between the bases of the larger ones in the larger animals (compare Fig. 3A, B). The mantle muscles are very regular and robust, the longitudinally radiating ones deeper and the circular ones around each aperture more superficial, creating a beautifully regular mesh- work. The dorsal tubercle has a narrow, oval opening at the end of a long tube, opening to the right between the peripharyngeal bands. There are about 18 branched oral tentacles, mostly of 2 alternating sizes. The larger tentacles are al- ternately branched, the branches containing tiny protruberances. In addition, very small projec- tions occur on the trunk of the main tentacles between the branches. Probably the tentacles would be more branched in larger animals. The branchial sac contains 6 branchial folds on each side, of unequal size; the second (counting from the dorsal lamina) is rudimentary, visible mainly as a region of more closely spaced internal lon- gitudinal vessels. The internal longitudinal ves- sels are numerous, with about 5 between bran- chial folds and about 8-12 on each fold. There are 10-12 rows of stigmata between the trans- verse vessels. Characteristic of this genus, the stigmata are arranged with their long axis run- ning dorso-ventrally in the branchial sac. The stigmata are fairly short and broad, each one crossed at its midpoint by a very thin parastig- matic internal longitudinal vessel. The dorsal lamina is cleft into a series of closely spaced lan- guets. An unidentified notodelphyid copepod was found in the branchial sac. The stomach is long and tubular, with no external grooves or plica- tions. Small lobes of the hepatic gland are visible at the anterior end of the stomach. The intestine is long, tubular, and recurved; the anus margin is slightly lobed. There is one gonad per side, the left one completely within the gut loop. Each ovary 1s long and slender, with the numerous irregularly lobular testes arranged alongside most of the length of the left ovary and along the entire length of the right ovary. The sizes of the specimens in this collection range from 3 mm in width to 10 mm as measured across the siphons. All are very flattened, no more than 4 mm measured anteroposteriorly. The go- nad is fairly well developed in the larger animals, though it is not possible to say if they are mature. ETYyMOLoGy.—The specific epithet polyplaco- derma (Gr. plakos, a flat tablet; Gr. derma, skin) aptly describes the numerous characteristic tunic plates that distinguish this species. REMARKS.—The most striking difference be- tween Boltenia polyplacoderma and the other species of Boltenia is the large number of plates on the anterior and lateral regions of the tunic. This species could even be confused with species of Chelyosoma were it not for the square 4-sided apertures, branchial folds, branched oral tenta- cles, and dorso-ventrally arranged stigmata. The tunic pattern resembles that of the European Pyura tessellata, so a reexamination was made of several specimens kindly provided by Dr. Ib Svane of the Kristineberg Marine Station, Swe- LAMBERT: THREE NEW ASCIDIANS FROM CALIFORNIA 115 Ficure 5. Styela tesseris n. sp. A, whole animal, right side, x 6.5; B, detail of tunic, x 22: C, dorsal tubercle. greatly enlarged; D, small piece of branchial sac, x 40; E, body removed from tunic, right side, x 6.8 (a—atrial siphon, b— branchial siphon, lv— internal longitudinal vessel, s—stigmata). 116 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 4 FIGURE 6. Styela tesseris n. sp. A, animal opened mid-ventrally to show internal anatomy, 9; B, enlarged detail of gonad; C, detail of stomach and anterior intestine showing the delicate connection between them (a—atrial siphon, at—atrial tentacles, b—branchial siphon, bt— branchial tentacles, e—endocarps, es— esophagus, i— intestine, o—ovary, pd—pyloric duct, s—stomach, sd—sperm duct, t—testis). den. The stigmata in P. tessellata are oriented longitudinally as in the other Pyura spp. Pyura tessellata has only four branchial folds per side in the branchial sac, a much thicker tunic, and the body is compressed laterally, while Boltenia polyplacoderma has six branchial folds per side and is flattened antero-posteriorly. Styela tesseris n. sp. (Figs. 5, 6) Hototyre.—021-BRA-02-TX, Santa Maria Basin (34°47.335'N, 120°45.903'W to 34°47.548'N, 120°46.123'W) at 87 m, on rock, Jul—Aug. 1984. U.S. National Museum of Natural History, #20011. PARATYPE.—006-BRA-02-TX, Santa Maria Basin (34°30.246'N, 120°35.555’W) at 61 m, on rock, one specimen LAMBERT: THREE NEW ASCIDIANS FROM CALIFORNIA 117 coll. Jul.—_Aug. 1984. U.S. National Museum of Natural His- tory, #20012. DESCRIPTION. — Both specimens are immature. The holotype is 15 mm long and 8 mm wide as measured across the siphons, and the paratype is 2 mm long. The animals are flattened laterally. The tunic has a very distinctive, irregularly tes- sellate pattern, with flat-topped plates raised and separated from one another by furrows over the entire tunic except adjacent to the siphonal open- ings, where the tunic is smooth. The pattern re- sembles a mosaic tilework, with a narrow groove between each tile. The tunic between the tessel- lations is flexible and changes shape as the tunic is bent or moved. The plate-like tunic thicken- ings may overlap at their edges in contracted areas of the tunic. The furrow between each raised area presumably allows for expansion and con- traction of the body. The tunic is thick, with a thick and uniformly smooth inner white layer; the tessellations are superficial and do not pen- etrate to this white layer. At the posterior end of both animals the tunic forms a wide, short, lat- erally flattened stalk; no body parts extend into this basal region of tunic. Small scattered sand grains are embedded superficially in the tunic. The tessellations are largest and most pro- nounced anteriorly; posteriorly they become pro- gressively smaller, flatter, and less distinct. They appear to be oriented in rows running more or less across the tunic dorsoventrally, especially in the more posterior region; however, there are none on the stalk. Both siphons are square, the branchial one terminal and the atrial one situated slightly posterior. On the mantle, numerous very fine longitudinal muscles extend the full length of the body. Only a few longitudinal muscles occur on the siphons, where the deeper, circular muscles predominate. The opening of the dorsal tubercle is a simple oval slit. There are about 40 long, slender, unbranched oral tentacles, mostly of 2 alternating sizes, and a large number of very tiny, simple atrial tentacles. The dorsal lamina is a wide, flat membrane. The branchial sac has 4 folds per side, with the following numbers of internal longitudinal vessels; numbers in paren- theses refer to vessels on the folds. Right side, counting from the dorsal lamina to the ventral endostyle: 6 (6) 3 (6) 3 (6) 5 (5) 3; left side, starting from the dorsal lamina: 2 (6) 3 (6) 3 (6) 3 (6) 3. On the flat areas of the sac there are 4 stigmata between the internal longitudinal vessels; each stigma is crossed at its mid-point by a very thin parastigmatic transverse vessel. The stomach is elongate, with about 23 lon- gitudinal folds or ridges. The proximal part of intestine folds back alongside the stomach, while its distal portion curves anteriorly toward the atrial siphon. There is no hepatic gland. In the larger animal there are two thin-walled gastro- intestinal connections (pyloric ducts) located within the first intestinal curve; one is much wid- er than the other. There is a single pyloric duct in the smaller animal. The larger specimen has two long, tubular ovaries on each side, with a few irregularly lobed testes grouped around the posterior end of the left anterior ovary, which is the most fully developed. The individual sperm ducts appear to coalesce to form a common sperm duct that runs along the surface of the ovary to end near the mouth of the oviduct, close to the base of the atrial siphon. The right anterior ovary is the least developed and has no eggs. No gonads are apparent in the smaller specimen. Numerous endocarps of various sizes are attached to the atrial body wall around the posterior regions of the gonads and stomach. EtTyMoLoGy.—The specific epithet tesseris de- scribes the tunic pattern. The term is derived from the Greek fessera, a small tablet usually with four sides but sometimes with five or six, used in making mosaics; a decorative pattern made with small pieces of such shape (Onions 1966). REMARKS.—The most obvious way in which this new species differs from all the described northeast Pacific species of Styela is by the tes- sellated pattern of its tunic. This species most closely resembles Styela atlantica, which is also a deeper water form, though never recorded from the Pacific. Styela tesseris differs from S. atlan- tica in the tessellated tunic, the fewer internal longitudinal vessels on and between the folds in the branchial sac, and the rather wide dorsal lam- ina even in these immature specimens. In ad- dition, the dorsal tubercle in this species is an oval slit without S. atlantica’s inrolled horns. DISCUSSION This project is the only sampling of continental shelf areas off the California coast since 1904 in which the ascidians have been identified. Table 1 gives a complete listing of the ascidians col- lected, with their depth and locality. Aplidium sp. and Styela sp. are different from the other listed species of those genera, but because of their condition or immature stage it was not possible 118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 4 to identify them to species. Pyura sp. is ajuvenile and is probably P. haustor; it was not included in the total species count as a separate species. Of the 20 species collected, therefore, three are new. Six of the 16 taxa identified to species are the same as those collected by Ritter in 1904 (Ritter 1907). Of the 14 species collected in 1904, four came only from abyssal depths (1,828-4,207 m) and were not found during the present MMS project. These four may or may not occur at shallower levels, leaving only four species from Ritter’s (1907) 14 as ones that he found but we did not at similar depths during this project. Rock samples were collected by manned sub- mersible at 12 of the 18 hard-substrate stations; ascidians were found on rocks at 11 of these stations. Appendix A lists the combined assem- blages on all the rocks analyzed at each station. This collection of ascidians from the western Santa Barbara Channel and Santa Maria Basin demonstrates that the rich diversity of ascidians that occurs in shallow California waters (Abbott 1975; Fay and Vallee 1979; Abbott and New- berry 1980) also occurs in deeper continental shelf areas. Some of the species listed here occur in both shallow and deep water; others appear to be limited to deeper habitats. Future collections in this region will likely disclose additional un- described species of ascidians. ACKNOWLEDGMENTS I thank Andrew Lissner and James Blake of Science Applications International Corp. for the opportunity to examine this ascidian collection, Leslie Harris and Gordon Hendler for the loan of specimens from the Los Angeles County Mu- seum of Natural History, Ib Svane for the spec- imens of Pyura tessellata from Sweden, and my husband Charles for his help and encouragement during all aspects of this work. I am grateful for the many significant improvements made to the manuscript by Todd Newberry and two other anonymous reviewers. Publication costs were funded by the Southern California Association of Marine Invertebrate Taxonomists. LITERATURE CITED Assott, D. P. 1975. Phylum Chordata: introduction and Urochordata. Pp. 638-655 in Light’s manual: intertidal in- vertebrates of the central California coast, 3rd edition, R. I. Smith and T. J. Carlton, eds. Univ. of Calif. Press, Berkeley, California. Assott, D. P. AND A. T. NEwperry. 1980. Urochordata: the tunicates. Pp. 177—226 in Intertidal invertebrates of Cal- ifornia, R. H. Morris, D. P. Abbott, and E. C. Haderlie, eds. Stanford Univ. Press, Stanford, California. BLAKE, J. A. AND A. LissNER, EDS. In press. Taxonomic atlas of the macroinvertebrate fauna of the Santa Maria Basin and the western Santa Barbara Channel. Science Applications Intl. Corp., Woods Hole, Massachusetts. Fay, R. C. AND J. A. VALLEE. 1979. A survey of the littoral and sublittoral ascidians of southern California, including the Channel Islands. Bull. Southern Calif. Acad. Sci. 78:122- i135), LaMBeERT, G. In press. Phylum Chordata, Subphylum Uro- chordata, Class Ascidiacea. Chapter 14 in Taxonomic atlas of the macroinvertebrate fauna of the Santa Maria Basin and the western Santa Barbara Channel, J. A. Blake and A. Lis- sner, eds. Science Applications Intl. Corp., Woods Hole, Massachusetts. Onions, C. T., ED. 1966. The Oxford dictionary of English etymology. Clarendon Press, Oxford, England. Ritter, W. E. 1907. The ascidians collected by the United States Fisheries Bureau steamer Albatross on the coast of California during the summer of 1904. Univ. Calif. Publ. Zool. 4:1-52. VAN Name, W. G. 1945. The North and South American ascidians. Bull. Amer. Mus. Nat. Hist. 84:1-476. APPENDIX A Ascidian hard substrate assemblages by station number. Number in parentheses refers to number of rocks sampled. 01 (3) Chelyosoma columbianum, Didemnum carnulentum var. lacteolum, Styela milleri, Styela sp. 02 (1) Boltenia polyplacoderma, Chelyosoma columbianum 04 (2) Chelyosoma columbianum, Corella sp., Halocynthia igaboja, Pyura haustor 06 (2) Aplidium californicum, Chelyosoma columbianum, Di- demnum carnulentum var. lacteolum, Halocynthia iga- boja, Pyura sp., Styela gibbsii, Styela tesseris, Trididemnum opacum 13 (7) Aplidium sp., Ascidia sp., Boltenia polyplacoderma, Chelyosoma columbianum, Eudistoma carolinense (?), Halocynthia igaboja, Pyura haustor, Pyura sp., Styela mil- leri 14 (4) Boltenia polyplacoderma, Chelyosoma columbianum, Molgula regularis, Pyura haustor, Styela milleri 16 (5) Aplidium spauldingi, Ascidia sp., Boltenia polyplaco- derma, Chelyosoma columbianum, Halocynthia igaboja, Pyura haustor 20 (4) Chelyosoma columbianum, Halocynthia igaboja, Pyura haustor, Styela milleri 21 (4) Chelyosoma columbianum, Distaplia occidentalis, Eu- distoma carolinense (?), Halocynthia igaboja, Styela tes- seris (1) Chelyosoma columbianum, Corella sp., Styela milleri (2) Aplidium californicum, Chelyosoma columbianum, Euherdmania claviformis, Halocynthia igaboja, Pyura haustor NN awn © CALIFORNIA ACADEMY OF SCIENCES, 1993 Golden Gate Park San Francisco, California 94118 a to tdantify ther? apes, Myuraso. ae juvenile afd is prctabay faze out wee hu inure, itr thee Novak Ageces cowl! eo nee rairetc spemign. 13 tho oglinnaly their furs, thies.ace ow. See iS aie i) min [a Ape ent sniags we "ey Um es : => rie.e 7 ive miges a) 1 al Ch : Ny) i cles Ph a] j A I 4 ' ' U i a4 = y i Wak TA, er Er ofthe cwbrete weronede Ga (Ver treme Serta Gefteire » tole Cia Wor Mote) ak C. 1 Wee Chava ie ~ <) od ee | od sgh 4 eatlige om : j deer th Sai s+ (Ra keaen ~~Arop Ole rails Oe Cee “ate ‘rte _— ¢ oe m/ NG 7 as Ka bd >i ey x nl hb let] o¢ a > ? - 7 val 7a * nis ti . nie vidios @ aaa ne ee by } ‘ Fe - as wi ‘Ta, : y. (* & NGF, - sre Vlei fie (fal i 1 : ’ j ‘ PROCEEDINGS OF THE CALIFORNIA ACADEMY OF PRR Lasatay = Vol. 48, No. 5, pp. 119-130, 4 figs. | LIBRA : J a Zl. 1993 FEB D 1 ag 1993 Ste TAXONOMIC AND GEOGRAPHIC NOTES ON |, CENTRAL AMERICAN ACANTHACEAE By Thomas F. Daniel Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 Asstract: Justicia almedae is described as a new species from a lowland rain forest in southeastern Costa Rica. Persistence of the rare Belizian endemic, Louteridium chartaceum, is noted. A new name, Streblacanthus cordifolius, is provided for the species previously known as Streblacanthus cordatus and Pseuderanthemum cordatum. Ecbolium chamaeranthemodes is transferred to Justicia and recognized as a species distinct from J. chamaephyton. Streblacanthus chirripensis is included as a taxonomic synonym of J. chamaeranthemodes. Lectotypes are designated for Streblacanthus cordatus and S. chirripensis. The generic position of Justicia albobractea is resolved, the species is adequately described for the first time, and its known distribution is extended from Belize to Guatemala. Range extensions are reported for the following species: Carlowrightia myriantha (Belize), Mendoncia guatemalensis (Honduras), Ruellia standleyi (Costa Rica), Spathacanthus hahnianus (Honduras), Staurogyne agrestis (Costa Rica), Stenandrium pedunculatum (E\ Salvador), and Streblacanthus cordifolius (Costa Rica). Received June 24, 1992. Accepted September 16, 1992. INTRODUCTION The comprehensive treatments of Acantha- ceae for Panama and Costa Rica by Durkee (1978, 1986) have stimulated considerable interest in this family in southern Central America. Recent systematic studies by Daniel and Wasshausen (1990), Gomez-Laurito (1990), Gomez-Laurito and Grayum (1991), and Daniel and McDade (in prep.) have all built on the foundations estab- lished by Durkee. In this report, a new species of Justicia is described from Costa Rica, signif- icant range extensions of Acanthaceae are re- ported from throughout Central America, per- sistence of the rare species Louteridium chartaceum in Belize is noted, and miscellaneous taxonomic notes on Central American Acantha- ceae are provided. The taxonomic notes include nomenclatural renovations in Streblacanthus and Justicia, and lectotypifications of Streblacanthus chirripensis and S. cordatus. A New SPECIES Justicia almedae T. F. Daniel, sp. nov. (Fig. 1) Typre.—COSTA RICA. Limon: Reserva Bioldgica Hitoy Cerere, trail from Rio Cerere to Cerro Bobocara, 26 February 1991, T. Daniel, F. Almeda, & B. Bartholomew 6228 (holotype: CAS!; isotypes: CR!, MO!, US!). Herba usque ad 3.5 dm alta. Folia petiolata petiolis usque ad 22 mm longis; laminae ovatae vel ovato-ellipticae, 43-134 mm longae, 20-68 mm latae, 1.7—3.6-plo longiores quam lat- iores. Spicae axillares vel terminales, pedunculares, usque ad 45 mm longae; bracteae lanceolato-lineares vel anguste ellip- ticae, 7-9 mm longae, 1.7—3 mm latae; bracteolae lanceolato- lineares, 5-7 mm longae, 1—1.3 mm latae. Flores oppositi; calyx 10-12 mm longus, inaequaliter quinquelobus lobis lanceolato- subulatis; corolla cremea, 14.5—15 mm longa; stamina 9.5 mm [119] 120 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 Ficure 1. Justicia almedae (Daniel et al. 6228). (a) Habit, x 0.5; (b) Bract, x 4.5; (c) Bractlet, x 4.5; (d) Calyx, x 6; (e) Corolla, x 3.75; (f) Distal portion of stamen, x 17.5; (g) Capsule, <3; (h) Seed, x 8. Drawn by Tina Cash. longa thecis superpositis, theca inferior calcarata. Capsula stip- itata, 11.5-12 mm longa, puberula; semina 4, papillosa. Herb to 3.5 dm tall; young stems quadrate, pubescent with antrorse to antrorse-appressed eglandular trichomes 0.1—0.2 mm long, the tri- chomes + evenly disposed. Leaves petiolate; pet- ioles up to 22 mm long; blades ovate to ovate- elliptic, 43-134 mm long, 20-68 mm wide, 1|.7- 3.6 times longer than wide, acuminate at apex, acute to attenuate at base, adaxial surface gla- brous, abaxial surface pubescent along major veins with cauline type trichomes. Inflorescence of axillary (from distalmost leaf axils) or termi- nal, pedunculate spikes; peduncles to 6 mm long; spikes to 45 mm long; rachis + evenly pubescent with straight to antrorse eglandular trichomes DANIEL: CENTRAL AMERICAN ACANTHACEAE 0.05—0.2 mm long; bracts lance-linear to narrow- ly elliptic, 7-9 mm long, 1.7-3 mm wide, abaxial surface inconspicuously sparsely pubescent with antrorse eglandular trichomes up to 0.05 mm long; bractlets lance-linear, 5-7 mm long, 1-1.3 mm wide, pubescent like bracts. Flowers oppo- site at spike nodes, sessile or borne on short (less than 1 mm long) pedicels; calyx 10-12 mm long, externally pubescent like bracts, lobes 5, lance- subulate, unequal, posterior lobe 5—7.5 mm long, 0.6 mm wide, conspicuously smaller than other lobes, the other lobes subequal, 7.5-11.5 mm long, 1-1.4 mm wide; corolla cream-white with maroon bands on lower lip, 14.5-15 mm long, externally pubescent with straight to flexuose to retrorse eglandular trichomes 0.1-0.2 mm long, limb internally pubescent with glandular tri- chomes 0.05-0.1 mm long, tube 8-8.5 mm long, upper lip 6.5—7 mm long, entire at apex, inter- nally rugulate, lower lip 7-7.5 mm long with lobes subelliptic, 2.3-2.8 mm long, 1.5—3.5 mm wide; stamens 9.5 mm long, filaments glabrous, thecae superposed, + parallel, 1-1.2 mm long, lower theca with a spur to 0.3 mm long, pollen 2-colporate, colpi flanked by unbroken bands of exine and faint pseudocolpi; style 13 mm long, pubescent with antrorse eglandular trichomes, stigma equally bilobed with lobes 0.2 mm long. Capsule 11.5—12 mm long, stipe 4—4.5 mm long, head 7-8 mm long, inconspicuously and sparsely puberulent with antrorse eglandular trichomes 0.05 mm long; seeds 4, green turning brown, subcircular, 3—3.2 mm long, 2.5 mm wide, sur- face roughened with low rounded papillae. PHENOLOGY.— Flowering and fruiting: Febru- ary—March. DISTRIBUTION AND HABITAT. — Known only from southeastern Costa Rica where plants occur in lowland rain forest (with Hura, Ceiba, Pen- taclethra, and Terminalia in the overstory and Socratea, Zamia, and Dieffenbachia common in the understory) at an elevation of about 200 me- ters. The lowland rain forests of Costa Rica’s Ca- ribbean slope are especially rich in species of Acanthaceae and continue to yield new species. Two were recently described from the Reserva Biolégica Hitoy Cerere: Herpetacanthus steno- phyllus Gomez-Laurito & Grayum and Justicia bitarkarae GOmez-Laurito. Justicia almedae cannot be readily affiliated with any of the infrageneric taxa recognized by Graham (1988). Its pollen (Fig. 2) is most similar 121 to Graham’s type 4. The only New World section of Justicia, as treated by Graham (1988), with type 4 pollen is sect. Leucoloma. Justicia al- medae differs from this section by its spikes with 2-flowered nodes, bracts without a scarious mar- gin, larger thecal appendages, and glabrous seeds. Justicia almedae differs from a similar species recently described from Cerro Bitarkara, also in the Reserva Biolégica Hitoy Cerere, by the fol- lowing couplet: Petioles 4-8 cm long; flowers solitary and al- ternate at inflorescence nodes; bracts di- morphic at a node (fertile bract considerably larger than sterile bract), petiolate, ovate to elliptic to obovate to oblanceolate, and glan- dular pubescent; calyx lobes equal, 5.5—6 mm long; corolla 10-11 mm long; style 6-7 mm long, glabrous; seeds (based on protologue) verrucose with waxy-yellow projections PULA eM Lien Pe ee See J. bitarkarae Petioles up to 22 mm long; flowers opposite at inflorescence nodes; bracts homomorphic at a node, sessile, lance-linear to narrowly elliptic, and pubescent with eglandular tri- chomes; calyx lobes unequal, 7.5-11.5 mm long; corolla 14.5—15 mm long; style 13 mm long, pubescent; seeds roughened with low, rounded papillae, lacking waxy-yellow pro- JECUONS: 2. een eS ree J. almedae Pollen of these two species (Fig. 2) resembles one another except that of J. bitakarae is con- siderably smaller and has the bands in the trema region at least partially broken up into insulae. A RARE SPECIES Louteridium chartaceum Leonard, Publ. Carne- gie Inst. Wash. 461:197. 1936. TyPe.—BE- LIZE. Belize: Gracie Rock, Sibun River, 24 March 1935, P. Gentle 1526 (holotype: US!; isotypes A, ARIZ!, K!, LL!, MICH, NY). Multi-trunked, soft-wooded tree to 3 m tall; young stems subterete, glabrous. Leaves petio- late (or distal leaves subsessile); petioles to 65 mm long, glabrous; blades ovate to ovate-ellip- tic, 70-285 mm long, 33-95 mm wide, 1.9-3.4 times longer than wide, rounded to acute (distal leaves) to attenuate (proximal leaves) at base, acuminate to subfalcate at apex, margin entire, surfaces glabrous. Inflorescence a terminal pan- icle to 400 mm long; rachis glabrous; lateral branches consisting of many-flowered short 122) PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 FIGuRE 2. Scanning electron micrographs of pollen. (a) Justicia almedae (Daniel et al. 6228), colpal view; (b) J. almedae (Daniel et al. 6228), intercolpal view; (c) Justicia bitakarnae (Herrera & Soltis 2469), colpal view; (d) Justicia albobractea (Proctor 36627), colpal view. shoots to 17 mm long; bracts subtending short shoots caducous, not seen; floral bracts caducous, triangular, 2.5-3.5 mm long, 1.5-—2.5 mm wide, glabrous; bractlets, if present, not evident. Flow- ers pedicellate; pedicels 32-43 mm long, gla- brous; calyx 3-lobed, 18-—25(—30) mm long, lobes subequal, linear, divided nearly to base, 3.5-5 mm wide, glabrous, posterior lobe conduplicate, anterior lobes planar; corolla green with maroon at distal tips of lobes, 35-40 mm long, glabrous, tube (to base of stamens) 5-8 mm long, throat (from base of stamens to corolla lobes) 14-17 mm long, lobes subequal, 14—14.5 mm long, 9.5- 10.5 mm wide, bifid with rounded segments 2- DANIEL: CENTRAL AMERICAN ACANTHACEAE FiGure 3. 5aum Scanning electron micrographs of pollen. (a) Justicia chamaephyton (Allen 5480), colpal view; (b) Justicia cha- maeranthemodes (Daniel & Almeda 6352), colpal view; (c) Ruellia standleyi (Almeda et al. 6858); (d) Louteridium chartaceum (Daniel & Butterwick 5905). 3.5 mm long, 3.5—4 mm wide; fertile stamens 4, subdidynamous, joined in pairs for 3-4 mm at base, 55-60 mm long, filaments distally glabrous, pubescent near base with eglandular trichomes, thecae 6.5-9.5 mm long, pollen (Fig. 3) spheric, pantoporate, evenly gemmate, staminode pos- terior, consisting of a subconical projection to 1.5 mm long; style 60 mm long, glabrous, stigma to 1.5 mm long with unequal lobes. Capsule el- lipsoid, 23-26 mm long, 5-7 mm across non- splitting face, glabrous, retinacula 2—2.5 mm long; seeds 16 per capsule, whitish, 4.2-4.5 mm long, 124 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 3.5-4 mm wide, thickened around periphery, margin fringed with dense, trichome-like projec- tions, surface with low rounded protrusions. PHENOLOGY.—Flowering January-March; fruiting February-March, June. DISTRIBUTION AND HaAsBitaT.—Known only from the type locality in east-central Belize where plants occur in tropical, moist forest on an iso- lated limestone hill in the coastal plain at ele- vations from 100 to 120 meters. ADDITIONAL SPECIMENS EXAMINED.— BELIZE. Belize: Gra- cie Rock Hill near Rockville Quarry between Western Hwy. and Sibun River, ca. 30 km (air) SW of Belize City, T. Daniel & M. Butterwick 5905 (C, CAS, K, MICH, MO, NY), culti- vated plants propagated from this collection, 7. Daniel s.n. (CAS); Gracie Rock, 1.5—4 mi. S of Mile 22 on Western Hwy., J. Dwyer 10959 (LL), R. Liesner & J. Dwyer 1485 (BM, DUKE, TEX). The above description substantially amplifies that of Leonard’s protologue. Louteridium char- taceum is one of only four species of Acanthaceae currently recognized as endemic to Belize (the other three are Justicia bartletii (Leonard) D. Gibson, Odontonema amicorum V. Baum, and Ruellia obtusata Blake). It is known from a single locality. Until recently, it was known only from the type collection made in 1935. The species is still extant at the type locality (as of February 1989), but its habitat is threatened with destruc- tion by quarrying activities there. Richardson (1972) included this species in sec- tion Tetrandrium, which is characterized by four diadelphous stamens. It differs from the other three species of this section, all restricted to west- central Mexico, by its linear (vs. lanceolate to ovate to triangular) calyx lobes that vary from 3.8 to 6 (vs. less than 3) times longer than wide and its glabrous (vs. pubescent) capsule. Louter- idium donnell-smithii S. Wats., the only other species of the genus known to occur in Belize, differs conspicuously by its two fertile stamens and densely pubescent shoots. MISCELLANEOUS TAXONOMIC NOTES Streblacanthus cordifolius T. Daniel, nom. nov. Eranthemum cordatum Nees in Benth. Bot. Voy. H.M.S. Sul- phur 147. 1846. —See discussion for locality, 4. Sinclair (holotype: K!). Pseuderanthemum cordatum (Nees) Radlk. Sitzungber. Math.-Phys. Kl. Bayer. Akad. Wiss., Miinchen 13:286. 1883. Streblacanthus cordatus Lindau, Fedde Rep. Sp. Nov. 11:123. 1912. Type.— PANAMA. Colon: “forests around Porto Bello,” 5-200 m, 6-8 April 1911, W. Maxon 5793 (lectotype: US!, see discussion). This species was first described by Nees based on a collection made by Sinclair at Atacames, Ecuador during the voyage of HMS Sulphur. (The species has subsequently been collected in Ec- uador, e.g., Eggers 14862 at K, M, and PR.) Al- though Nees (1846) noted that staminodes were lacking, he placed the species in Eranthemum, a genus that he characterized as having two stam- inodes (Nees 1847). Lindau (1912) correctly treated a collection of Maxon from Panama, which agrees in all characteristics with E. cor- datum, in Streblacanthus. Because transfer of the earliest known epithet of this species to Strebla- canthus would result in a later homonym, a new name has to be designated for the species ac- cording to Article 54.1 of the International Code of Botanical Nomenclature (Greuter et al. 1988). I have chosen one that, like the original, attempts to highlight a very distinctive feature of the leaves. The long, slender corolla tubes of Streblacan- thus superficially resemble those of Pseuderan- themum Radlk. Indeed, S. cordifolia has usually been treated under the latter genus (Leonard 1953; Durkee 1978). Two characters readily distin- guish these genera. Species of Streblacanthus have an androecium of two stamens and no stami- nodes, and capsules with subcircular heads. In contrast, species of Pseuderanthemum have an androecium of two stamens and two staminodes, and capsules with hourglass-shaped heads. In the protologue of S. cordatus, Lindau (1912) provided two syntypes. These consisted of col- lections of H. Pittier and W. Maxon made near Portobelo in 1911 as part of a Smithsonian Bi- ological Survey of the Panama Canal Zone. Lin- dau did not indicate where the collections he utilized were deposited. If specimens were sent to him at B and were retained there, they were destroyed in 1943. Fortunately, specimens of both collections are extant at US and identified as “Streblacanthus cordatus Lindau n. sp.” in Lin- dau’s handwriting. Maxon’s collection is here se- lected as the lectotype of S. cordatus. It is the more complete of the two collections, containing both flowers and fruits. Justicia chamaeranthemodes (Kuntze) T. F. Daniel, comb. nov. Ecbolium chamaeranthemodes Kuntze, Rev. Gen. Pl. 2:487. 1891. Type.—COSTA RICA. Limon: “Port Lemon” (Li- mon), 14 June 1874, O. Kuntze 1977 (lectotype, designated by Leonard, 1938: NY!). Streblacanthus chirripensis Lindau, Bull. Herb. Boissier, ser. 2 4:404. 1904. Type.—COSTA RICA. Cartago: prope Haci- DANIEL: CENTRAL AMERICAN ACANTHACEAE 125 Ficure 4. Map of southern Central America showing the distributions of Justicia chamaeranthemodes (closed circles) and J. chamaephyton (open squares). The precise Nicaraguan locality of J. chamaeranthemodes is not known. enda de Chirripé [ca. 9°48'N, 83°23'W, “cerca de quebrada Platanilla en un sitio conocido como Grano de Oro” fide L. Poveda, pers. comm.], 200 m, February 1900, H. Pittier 16046 (holotype: B, destroyed, photos at GH and US; lec- totype, designated here, see discussion: US!; isolectotype: CR!). Perennial herb to 3 dm tall; young stems subquadrate to quadrate, pubescent with retrorse to retrorse-appressed eglandular trichomes to 0.2 mm long. Leaves petiolate; petioles to 20 mm long; blades ovate to elliptic to obovate, 38-157 mm long, 16-73 mm wide, 1|.4—2.8 times longer than wide, rounded to acute (to subattenuate) at base, rounded to acute at apex, dark purplish abaxially. Inflorescence a terminal, pedunculate spike (sometimes branched proximally) to 160 mm long (including peduncles and excluding flowers); peduncles to 85 mm long, pubescent with straight to retrorse eglandular trichomes 0.05-0.2 mm long; rachis pubescent like pedun- cles; internodes near midspike 6-11 mm long; bracts lanceolate to subulate (to subtriangular), 2-4 mm long, 1—1.2 mm wide; bractlets lanceo- late to subtriangular, 1.5—3 mm long, 0.40.6 mm wide. Flowers opposite at nodes, usually not overlapping; calyx 5-8 mm long, lobes hetero- morphic, anterior pair 4.8-7.5 mm long, lateral pair 3.8-5 mm long, posterior lobe 2.8-4 mm long, the differences in length among the types >1 mm; corolla purple, 10-13 mm long, exter- nally pubescent with eglandular trichomes to 0.2 mm long; stamens 6-6.5 mm long, filaments glandular puberulent, thecae superposed, over- lapping by 0.2 mm, upper theca 0.8-1 mm long, dorsally pubescent, lower theca 1.3-1.5 (includ- ing a basal spur to 0.5 mm long), pollen prolate, 2-colporate, colpi flanked by 1 unbroken band and | row of insulae; style 12.5 mm long, prox- imally pubescent, stigma lobes 0.2 mm long. Capsules 9.5-12.5 mm long, pubescent with straight to flexuose eglandular trichomes 0.1-0.2 mm long, stipe 4-6 mm long, head 5.5-6 mm long; seeds 4, suboval to subcordate, 2.5-3.2 mm long, 2.2-2.5 mm wide, surface smooth. PHENOLOGY. — Flowering: February—Decem- ber; fruiting: September—June. DISTRIBUTION AND HABiTAT. — Nicaragua, Costa Rica, and Panama (Fig. 4); plants occur in lowland rain forests at elevations from near sea level to 500 meters. ADDITIONAL SPECIMENS EXAMINED. —COSTA RICA. Limon: between Siquerres and Rio Pacuare, 10°5’N, 83°29'W, W. Bur- ger & R. Liesner 6994 (CR, US); Port Limon, Zent banana farm, O. Cook & C. Doyle 470 (US); Cerro Skopte just W of 126 Rio Siori, ca. 7 km SW of Coroma, 7. Daniel & F. Almeda 6352 (CAS, CR, K, MO); Talamanca, Suretka, J. G6mez-Lau- rito 9535 (CR); Canton de Talamanca, Amubri, Rio Urén, Loma Sheuab, 9°32'N, 82°54'W, G. Herrera 2988 (CR); Re- serva Biol. Hitoy Cerere, 9°40'N, 83°03'W, G. Herrera & M. Solis 2431 (CR); Reserva Biol. Hitoy Cerere, 9°38’N, 83°05'W, G. Herrera 2464 (CR); Talamanca, Foréts de Tsaki, A. Tonduz (or H. Pittier) 9553 (CR, US); Hacienda de Zent, United Fruit Company 255 (US). NICARAGUA. Chontales: without spe- cific locality, 1886-1887, R. Tate 297 (186) (K). PANAMA. Bocas del Toro: Bocas del Toro, M. Carleton 210 (US); above RR stop at Milla, 7. Croat & D. Porter 16290 (MO); Quebrada Huron, J. Kirkbride & J. Duke 427 (MO); Sibubi Falls, Sixaola Valley, W. Rowlee & H. Rowlee 380 (US); Water Valley, H. von Wedel 602 (US), 603 (MO, US), 916 (MO, US), 917 (US); vicinity of Chiriqui Lagoon, H. von Wedel 1038 (MO, US). Chiriqui: David airport, W. Lewis et al. 851 (MO, US). San Blas: El Llano-Carti Road, Km. 19.1, 9°19’N, 78°55’W, G. de Nevers & H. Herrera 7263 (CAS). Leonard (1938) noted that the description of Ecbolium chamaeranthemodes (as E. ““Chama- eranthemum’’) was suggestive of Beloperone var- iegata Lindau. He refrained from taxonomically merging these two species because he had not seen any of the type material of the former. Ec- bolium chamaeranthemodes was based on syn- types from Nicaragua and Costa Rica (Kuntze 1891), but was not treated by Lindau (1900) or Durkee (1986). Leonard’s (1938) indication of Kuntze’s collection from Limon, Costa Rica as the type of this species effectively lectotypified E. chamaeranthemodes. Examination of both syntypes (i.e., Tate 297 (186) and Kuntze 1977) reveals that E. chamaeranthemodes differs from B. variegata in several important characters (see below). Species treated by Kuntze (1891) in Ec- bolium are now placed in Justicia (Graham 1988). Durkee (1986) concluded that Streblacanthus monospermus Kuntze, S. macrophyllus Lindau, and S. longiflorus Cufodontis represented a sin- gle species. All three were described from Costa Rican specimens. However, Durkee (1986) did not deal with S. chirripensis, a species also de- scribed from Costa Rica based on a collection of Pittier. The holotype of S. chirripensis was de- stroyed at B in 1943. Isotypes, annotated by Lin- dau as S. chirripensis, were located at CR and US, and the latter is chosen as the lectotype. The isotypes resemble a photograph (GH, US) of the destroyed holotype that shows a single plant and detailed drawings of the flower and pollen. The form of the corolla (short, ampliate tube), an- droecium (two stamens with superposed thecae, the lower theca conspicuously mucronate), and pollen (2-aperturate) all reveal this species to be congeneric with Justicia rather than Streblacan- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 thus. Indeed, Leonard (1938) included S. chir- ripensis in the synonymy of Beloperone varie- gata. Beloperone is now included within Justicia and the name J. chamaephyton D. Gibson ap- plies to B. variegata when it is treated in Justicia. Gibson (1972) excluded S. chirripensis from the synonymy of J. chamaephyton, noting that the leaf base form and corolla color of the former do not conform to the latter. Durkee (1978, 1986) noted variation in J. chamaephyton encompass- ing both rounded to attenuate leaf bases and light purple to cream colored corollas. Examination ofa large series of collections un- der these names at CAS, CR, MO, and US re- veals the presence of two distinctive species, one conforming to the types of E. chamaeranthe- modes and S. chirripensis and the other conform- ing to the type of Justicia chamaephyton. For these reasons, E. chamaeranthemodes is trans- ferred to Justicia and S. chirripensis is included as a synonym of J. chamaeranthemodes. These two species can be distinguished by the following couplet: Corolla purple; calyx lobes conspicuously un- equal, the anterior pair equal, = 1 mm longer than the lateral pair, the lateral pair equal, =1 mm longer than the posterior lobe; tri- chomes of young stems, peduncle, and ra- chis 0.05—0.2 mm long; internode length near midspike 6-11 mm long, the bracts and flowers rarely overlapping Nee te eee Le J. chamaeranthemodes Corolla cream or greenish yellow; calyx lobes subequal, the anterior pair <1 mm longer than the lateral pair, the lateral pair <1 mm longer than the posterior lobe; trichomes of young stems, peduncle, and rachis 0.2—1.5 mm long; internode length near midspike 2- 3 (-8) mm long, the bracts and flowers over- lapping 225 5 sae J. chamaephyton Pollen of both species (Fig. 3) is 2-colporate with the trema region traversed by 2 unbroken bands and 2 rows of insulae (one of each on both sides of the colpi). This type of pollen is not included among the 10 types of pollen in Justicia as delimited by Graham (1988). Using Graham’s (1988) key to infrageneric taxa of Justicia, both J. chamaeranthemodes and J. chamaephyton would be included in section Rostellaria subsec- tion Rostellaria. The two Central American spe- cies differ from this Old World assemblage by their longer corollas, pollen with solid bands in DANIEL: CENTRAL AMERICAN ACANTHACEAE the trema region, pubescent capsules, and dis- tinctly flattened seeds. Although purple corollas are noted on labels or observable on most collections of J. chama- eranthemodes, von Wedel collected plants at one locale in Panama with white corollas among plants with purple corollas. These are presumed to be either albinic corollas or a much less com- mon form of the species. In all other diagnostic characteristics, these white-flowered plants re- semble purple-flowered J. chamaeranthemodes. Figure 4 shows that the known distributions of J. chamaeranthemodes and J. chamaephyton do not overlap. Furthermore, the former species occurs primarily on the Caribbean slope whereas the latter occurs primarily on the Pacific slope. Justicia albobractea Leonard, Carnegie Inst. Washington Publ. 461:230. 1936. Type.— BE- LIZE. District unknown: Camp 32, Guatemala Survey, ca. 700 m, 2 March 1934, W. Schipp 1277 (holotype: F; isotypes: CAS!, MICH!). Shrub to 1 m tall; younger stems subquadrate, bifariously pubescent with flexuose-ascendant eglandular trichomes to | mm long. Leaves pet- iolate; petioles to 25 mm long; blades narrowly elliptic to narrowly obovate-elliptic, 45-132 mm long, 13-35 mm wide, 2.5—3.8 times longer than wide, acute to attenuate at base, acute to acu- minate at apex, abaxial surface pubescent with cauline type trichomes, the trichomes becoming restricted to major veins, adaxial surface pubes- cent along midvein, soon glabrate. Inflorescence of axillary pedunculate spikes to 65 mm long (including peduncle but excluding flowers), 12- 17 mm in diameter (measured flat) near mid- spike; peduncles to 32 mm long, pubescent like stems; rachises pubescent like peduncles; bracts “sreenish white,” ovate to elliptic, 8.5-16 mm long, 4.5-9.5 mm wide, abaxially pubescent along midvein with appressed-flexuose eglandular tri- chomes, margin ciliate with similar trichomes; bractlets narrowly elliptic to narrowly ovate-el- liptic, 7-10 mm long, 1.5—3.5 mm wide, pubes- cent like bracts or the abaxial surface glabrous. Flowers solitary, sessile at nodes; calyx 5-lobed, 4.5-5.5 mm long, lobes lance-subulate, 4-5 mm long, abaxially glabrous; corolla “‘pale yellow” or “white,” 29-31 mm long, externally pubescent with flexuose eglandular trichomes, tube 15-16 mm long, distally ampliate, upper lip 14 mm long, internally rugulate, apically emarginate, 127 lower lip 15 mm long; stamens 15 mm long, thecae superposed with a gap 0.2 mm long be- tween them, parallel, 1.6—2 mm long, lower theca calcarate with a spur 0.5 mm long, pollen (Fig. 2) prolate, 3-porate, trema region with a single row of subcircular to elongate insulae on each side of pore, exine reticulate; style 24-27 mm long, pubescent, stigma flared, 0.1-0.2 mm long. Capsule stipitate, 9.5-10.5 mm long, pubescent near apex with flexuose eglandular trichomes 0. 1- 0.2 mm long, stipe 3-4 mm long, head 6-6.5 mm long; seeds 4, flattened, subcircular to subcordate in outline, 2-2.4 mm long, 1.8-2.2 mm wide, pubescent with erect to appressed eglandular tri- chomes 0.05-0.1 mm long. PHENOLOGY. — Flowering: November—Febru- ary; fruiting: February—March. DISTRIBUTION AND Hapsirats. — Belize and Guatemala; plants occur in tropical moist forests at elevations up to about 600 meters. ADDITIONAL SPECIMENS EXAMINED.—BELIzeE. Toledo: Co- lumbia Forest Reserve, vicinity of Union Camp, 575-600 m, G. Proctor 36627 (F). GUATEMALA. Peten: Los Arcos, Ca- denas Road, W of KM 143/144, E. Contreras 9356 (US); La Cumbre, on Chocalte arriba, KM 155, E. Contreras 10554 (US), 10558 (US), 10565 (US). The above cited collections from Guatemala are the first records of J. albobractea from that country. Gibson (1974) stated that material of this species, then known only from a fruiting collection from Belize, was insufficient to allow for its description or interpretation. Thus she did not include the species in her treatment (Gibson 1974) of Acanthaceae for the Flora of Guate- mala, the coverage of which also included Belize. The species was not listed among the Acantha- ceae of Belize by Dwyer and Spellman (1981). Recent flowering and fruiting collections of J. albobractea from Belize and Guatemala permit a complete description of the species and an as- sessment of its affinities. Floral characters (especially the internally ru- gulate upper lip, the calcarate lower thecae, and the pollen), place this species into Justicia. It does not conform to any of the infrageneric taxa treat- ed by Graham (1988), however. Its pollen (Gra- ham’s type 2) resembles that of the Old World section Tyloglossa. Numerous macromorphol- ogical features of J. albobractea (e.g., shrubby habit, simple inflorescences, larger corollas, and flattened seeds) preclude its placement in that section, however. 128 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 RANGE EXTENSIONS Carlowrightia myriantha (Standley) Standley Be.ize. Corozal: Cerros Maya Ruins, Lowry’s Bight, ca. 2 mi S of Cerros, 8 April 1983, C. Crane 516 (LL). Twenty-three species of Carlowrightia, all of which occur in Mexico, are presently recognized (Daniel 1988). Carlowrightia myriantha was known previously from the Yucatan Peninsula of Mexico (Quintana Roo and Yucatan) (Daniel 1988). Crane’s flowering and fruiting collection from northern Belize is the first record of this species outside of Mexico. It is the only species of Carlowrightia known from Belize and one of only three species of the genus known to occur in Central America (Daniel 1983, 1988). The three species of Carlowrightia in Central Amer- ica can be distinguished by the following key: 1. Cauline trichomes antrorse; corolla exter- nally glabrous; capsule sparsely pubescent at apex of head, otherwise glabrous; seeds 2-3 mm long, the margin entire; Belize _. att Ale ease C. myriantha (Standley) Standley 1. Cauline trichomes straight to flexuose to retrorse; corolla externally pubescent (at least on lower-central lobe); capsule entire- ly glabrous or pubescent throughout head; seeds 3.5—4.8 mm long, the margin dentic- ulate or irregularly fringed. 2. Young stems pubescent with glandular and eglandular trichomes; corolla 17- 18.5 mm long; capsule pubescent throughout head; seeds with a promi- nent, swollen marginal ring, the margin fringed with irregular, wing-like projec- tions; El Salvador .. C. hintonii T. Daniel 2. Young stems pubescent with eglandular trichomes only; corolla 9-15 mm long; capsule entirely glabrous; seeds lacking a swollen marginal ring, the margin den- ticulate; Guatemala and Costa Rica ..... SR on 2 ae Lec 1 C. arizonica A. Gray Mendoncia guatemalensis Standley & Steyer- mark HONDURAS. Cortes: entre Agua Azul y Pito Solo, Lago de Yojoa, 600 m, 28 May 1956, A. Molina R. 7329 (LL). The range of this species was recently extended northwestward from Guatemala to Mexico (Daniel 1992). This collection extends it south- eastward into the “bosque lluvioso”’ of western Honduras. Two other species of Mendoncia have been reported from Honduras, M. retusa Turrill and M. costaricana Oersted (Molina 1975). Ten species of Mendoncia are presently recognized as occurring in Central America (Durkee 1978, Gibson 1974). Mendoncia guatemalensis is sim- ilar in many characteristics to M. gracilis, which is known from Colombia, Panama, and Costa Rica. The primary differences between them, based on published descriptions and specimens available to me, appear to be the shorter brac- teoles (15-19 mm long vs. 20-29 mm long) and corollas (25-30 mm long vs. 40-42 mm long) of the latter species. Molina’s specimen from Hon- duras lacks corollas but has bracteoles 25 mm long. The relationship between these two species deserves further study. Ruellia standleyi Leonard COSTA RICA. Cartago: Refugio Nacional de Fauna Sil- vestre Tapanti, Sendero Oropendola and vicinity, 1250-1300 m, 22 March 1986, F. Almeda, B. Anderson, & N. Zamora 5697 (CAS, CR, TEX), 1 March 1991, F. Almeda, T. Daniel, & B. Bartholomew 6858 (CAS, CR, DUKE, MO), 13 February 1992, T. Daniel & F. Almeda 6336 (CAS, CR, K, MICH); pasture and forested slope N of Quebrada Casa Blanca, Ta- panti, 9°47’N, 83°48'W, 1350 m, 25 November 1984, M. Gra- yum 4591 (CAS). This species was previously known only from Guatemala (Gibson 1974) and Nicaragua (Dur- kee 1988). The Costa Rican specimens listed above were collected in a region of montane rain forest to cloud forest on the Carribean escarp- ment where they were locally common. Ruellia standley is a viscid herb to 1 m tall with greenish yellow to greenish white corollas (15-20 mm long) and stipitate-ellipsoid capsules. Gibson (1974) provided a detailed description of the species and Leonard (1941) included an illustration of it in the protologue. Pollen of this species (Fig. 3) is similar to that described for all other Mesoamerican Ruellia. Spathacanthus hahnianus Baillon HONDURAS. Yoro: ca. 16 km from Yaryuha on Quebrada de Oro to Cerro Bufalo, 900-950 m, 15 August 1982, W. Holmes 4392 (TEX). Three species are recognized in this neotrop- ical genus of tall shrubs. Spathacanthus hahni- anus (including S. simplicifolius (J. Donnell Smith) Leonard) was previously known from Guatemala (Gibson 1974) and southern Mexico (Daniel 1986). This collection extends its range into northwestern Honduras. The distinctions DANIEL: CENTRAL AMERICAN ACANTHACEAE among the three species of Spathacanthus and their distributions are provided in the following key: 1. Calyx at anthesis equally or subequally di- vided into 2 prominent segments; corolla 73-95 mm long; capsule 55-85 mm long; Mosia Rica... .. S. hoffmannii Lindau 1. Calyx at anthesis unequally divided into 2 (rarely more) prominent segments; corolla 23-70 mm long; capsule 40-62 mm long. 2. Corolla 23-29 mm long, the throat 5- 7.5 mm in diameter, the limb 5-8 mm in diameter with lobes 1-3 mm long; capsule 40-43 mm long; Mexico and Guatemala _ S. parviflorus Leonard 2. Corolla 50-70 mm long, the throat 9.5- 15 mm in diameter, the limb 25-40 mm in diameter with lobes 9-15 mm long; capsule 55-62 mm long; Mexico, Gua- temala, Honduras _ S. hahnianus Baillon Staurogyne agrestis Leonard COSTA RICA. Guanacaste: Exsiccated bog, near Finca Es- cameka, ca. 10 km S of Las Cafios, 10 March 1965, R. Godfrey 66973 (CR). This is the initial report of this genus in Costa Rica. The rarely collected species, S. agrestis, was described based on collections from Nicaragua and Panama. Daniel and Lott (in prep.) also re- port S. agrestis from Mexico and Venezuela. Staurogyne consists of about 80 species dis- tributed worldwide in tropical regions. Most of the American species occur in Brazil. Taxonom- ically, the genus is usually included in subfamily Nelsonioideae (Lindau 1895), which consists of capsular-fruited herbs with retinacula papilli- form or lacking. The other Costa Rican genera of this subfamily are Ne/sonia R. Br. and Ely- traria Michx. Staurogyne differs from these gen- era by the following combination of character states: leaves opposite, anterior lobe of calyx di- vided nearly to base into two similar segments, stamens four, and stigma divided into two lobes. Stenandrium pedunculatum (J. Donnell Smith) Leonard EL SALVADOR. Santa Ana: an der Strasse zwischen dem Lago de Guija und Metapan, 300 m, 7.5.1958, F. Weberling & F. Schwanitz 2273 (M). This collection documents Stenandrium pe- dunculatum in El Salvador for the first time. The species was previously known from western and 129 southern Mexico, Guatemala, Honduras, and Nicaragua (Daniel 1984). The three species of Stenandrium known from Central America can be distinguished by the following key: 1. Plants acaulescent; bracts subulate to lan- ceolate to elliptic. 2. Leaf blades truncate to cordate at base, not decurrent on petiole; bracts subu- lates: = lesammiwide =2.4-4 ak eee AE Src nat le rea S. subcordatum Standley 2. Leaf blade acute to attenuate at base, often decurrent on petiole; bracts lan- ceolate to elliptic, =2 mm wide .............. lee Ue Pe eae S. dulce (Cav.) Nees 1. Plants caulescent; bracts obovate a taae, F: S. pedunculatum (J. Donnell Smith) Leonard Streblacanthus cordifolius T. Daniel COSTA RICA. Puntarenas: hills above Palmar Norte, trail to Buenos Aires, 17 February 1951, P. Allen 5903 (F). Streblacanthus Kuntze comprises fewer than five species of perennial herbs in Central and South America. Durkee (1986) recognized a sin- gle species of the genus, S. monospermus, as Oc- curring in Costa Rica (see above). Allen’s col- lection from southern Puntarenas adds S. cordifolius (see above for synonyms) to the known flora of the country. These two species can be distinguished by the following couplet: Young stems evenly pubescent with trichomes up to 0.1 mm long; leaf base subcordate to cordate; rachis and bracts inconspicuously glandular, if at all; bracts and bractlets short- er than calyx; calyx 5-lobed, 10-25 mm long; capsule 13=20 mmilong 2 ee Dk Res LS eh ee S. cordifolius T. Daniel Young stems pubescent in 2 lines with tri- chomes greater than 0.3 mm long; leaf base acute to attenuate; rachis and bracts con- spicuously glandular; bracts and bractlets longer than calyx; calyx 4-lobed, 4-6 mm long; capsule 20-25 mm long srllere tg) RORY en Te S. monospermus Kuntze Streblacanthus cordifolius is now known from Costa Rica, Panama, Colombia, and Ecuador. ACKNOWLEDGMENTS Funding for my field studies in Costa Rica was provided by the La Amistad Research Fund of the California Academy of Sciences. I am grateful 130 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 5 for the logistical support provided by the Museo Nacional de Costa Rica, the Servicio de Parques Nacionales de Costa Rica, and the Ministerio de Recursos Naturales, Energia y Minas de Costa Rica (Proyecto Reserva de la Biosfera de La Amistad). For their valuable assistance in the field, I thank Frank Almeda, Bruce Bartholo- mew, Mary Butterwick, Juan Diego Alfaro, Her- man Gonzalez, Carlos Hasbun, Quireco Jimé- nez, Pablo Sanchez, and Gina Umana. Luis Poveda kindly provided locality information. The various herbaria cited in text are acknowledged for generously providing specimens for study. Lisa Borok and Darrell Ubick supervised oper- ation of the scanning electron microscope. RESUMEN Se describe Justicia almedae como especie nueva que se encuentra en una selva lluviosa de tierra baja de la parte sudeste de Costa Rica. Se hace cuenta de que Louteridium chartaceum, una especie rara, persiste en Belice. Se provee un nombre nuevo, Streblacanthus cordifolius, para la especie anteriormente conocido como S. cor- datus y Pseuderanthemum cordatum. Se trans- fiere Ecbolium chamaeranthemodes a Justicia como especie distinta de J. chamaephyton. Se incluye S. chirripensis como sindnima taxon- omica de J. chamaeranthemodes. Se designa lec- totipos para S. cordatus y S. chirripensis. Se re- porta registros distribucionales nuevos para Carlowrightia myriantha (Belice), Mendoncia guatemalensis (Honduras), Ruellia standleyi (Costa Rica), Spathacanthus hahnianus (Hon- duras), Staurogyne agrestis (Costa Rica), Sten- andrium pedunculatum (El Salvador), y Strebla- canthus cordifolius (Costa Rica). LITERATURE CITED DanteL, T. F. 1983. Carlowrightia (Acanthaceae). Fl. Neo- trop. 34:1-116. 1984. A revision of Stenandrium (Acanthaceae) in Mexico and adjacent regions. Ann. Missouri Bot. Gard. 71: 1028-1043. 1986. Acanthaceae. Pp. 27-30 in Listados floristicos de México IV. Flora de Chiapas. D. E. Breedlove, ed. In- stituto de Biologia, UNAM, México. 1988. Taxonomic, nomenclatural, and reproductive notes on Carlowrightia (Acanthaceae). Brittonia 40:245-255. . 1992. Acanthaceae: Mendoncioideae of Mexico. Acta Bot. Mex. 17:17-24. DaniEL, T. F. AND E. Lotr. In prep. Staurogyne agrestis, a genus and species new to Mexico. DaniEL, T. F. AnD L. McDape. In prep. Taxonomic notes on the Acanthaceae of Panama. DaniEL, T. F. AND D. C. WAssHAUSEN. 1990. Three new species of Justicia (Acanthaceae) from Panama. Proc. Calif. Acad. Sci. 46:289-297. Durkee, L. H. 1978. 1978. Acanthaceae. Pp. 155-284 in Flora of Panama. R. Woodson et al., eds. Ann. Missouri Bot. Gard. 65. 1986. Acanthaceae. Pp. 1-92 in Flora Costaricensis. W. Burger, ed. Fieldiana, Bot. 18. 1988. A checklist of Acanthaceae in Costa Rica, Nicaragua, and Panama. Acanthus 3:3,4. Dwyer, J. D. AND D. L. SPELLMAN. 1981. 1981. A list of the Dicotyledoneae of Belize. Rhodora 83:161-236. Grsson, D. N. 1972. Studies in American plants, III. Field- iana, Bot. 34:57-87. 1974. Acanthaceae. Pp. 328-461 in Flora of Gua- temala. P. C. Standley et al., eds. Fieldiana, Bot. 24(10). Gomez-LauritTo, J. 1990. 1990. Two new species from the Caribbean of Costa Rica. Brenesia 33:139-144. Gomez-LauriTo, J. AND M. Grayum. 1991. Herpetacanthus stenophyllus (Acanthaceae), a new species from Costa Rica. Novon 1:15-16. GraHaM, V. A. W. 1988. Delimitation and infra-generic classification of Justicia (Acanthaceae). Kew Bull. 43:551- 624. GreEuTER, W. ET AL., EDS. 1988. International code of bo- tanical nomenclature. Regnum Veg. 118:1-328. Kuntze, O. 1891. Revisio genera plantarum, Vols. 1, 2. Ar- thur Felix, Leipzig. LEONARD, E. C. 1938. Acanthaceae. Pp. 1188-1263 in Flora of Costa Rica. P. C. Standley, ed. Field Mus. Nat. Hist., Bot. Ser. 18(2). . 1941. New Acanthaceae from Guatemala. J. Wash. Acad. Sci. 31:96-105. . 1953. The Acanthaceae of Colombia, II. Contr. U.S. Natl. Herb. 31:119-322. Linpau, G. 1895. Acanthaceae. Pp. 274-354 in Die natiir- lichen Pflanzenfamilien, Vol. 4(3b). H. Engler and K. Prantl, eds. Wilhelm Engelmann, Leipzig. 1900. Ord. Acanthaceae. Pp. 299-317 in Primitiae florae Costaricensis, Vol. 2(4). H. Pittier, ed. Instituto Fisico- Geografico Nacional, San José de Costa Rica. 1912. Einige neue Acanthaceen. Fedde, Reper. 11: 122-124. Moun, R., A. 1975. Enumeracion de las plantas de Hon- duras. Ceiba 19:1-118. NEES VON EseNnBECK, C. G. 1846. Acanthaceae. Pp. 145-149 in The botany of the voyage of H. M. S. Sulphur, Pt. 6. G. Bentham, ed. Smith, Elder and Co., London. 1847. Acanthaceae. Pp. 46-519 in Prodromus sys- tematis naturalis regni vegetabilis, Vol. 11. A. de Candolle, ed. Victoris Masson, Paris. Ricuarpson, A. 1972. Revision of Louteridium (Acantha- ceae). Tulane Stud. Zool. and Bot. 17:63-76. © CALIFORNIA ACADEMY OF SCIENCES, 1993 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 6, pp. 131-140 December 21, 1993 FIFTY YEARS OF PROGRESS IN RESEARCH ON SPECIES AND SPECIATION :nc Biological Laboratory/ By Ernst Mayr Woods Hole Oceanographic Institution Library DEC 27 1993 Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138 Woods Hole, MA 02543 Adapted from a lecture delivered at the Golden Jubilee Celebration of the publication of Systematics and the Origin of Species at the California Academy of Sciences on October 16, 1992. Received December 31, 1992. Accepted February 11, 1993. Historians of science have taught us how much one can learn from studying the history ofa field of science. This is excellently illustrated by the history of evolutionary biology as a whole, and by our growing understanding of species and spe- ciation, in particular. After 1859, two of Darwin’s theories were ac- cepted almost at once. First, evolution as such, and secondly, the branching theory of common descent. Natural selection was with almost equal unanimity rejected, being accepted only by a small group of naturalists. This was not too surprising since at that time no one understood variation and its origin. Finally, an active, almost violent, controversy developed over two other major Darwinian theories, that of speciation and that of evolutionary gradualism. As a matter of fact, Darwin himself was vacillating with respect to these two theories. Let us now proceed to the year 1900 and the rediscovery of Mendel’s laws. At that time, two camps became established in evolutionary bi- ology. One consisted of the Mendelians repre- sented by Bateson, DeVries, and Johannsen. They were strict typologists who saw discontinuity ev- erywhere in nature and applied this correctly to the nature of the genetic material, resulting in the theory of particulate inheritance. However, they drew from this the wrong conclusion as far as evolution is concerned, claiming that new spe- cies were produced by new mutations in a single saltation, completely rejecting Darwin’s theory of gradualism. Their opponents were the biom- etricians, such as Pearson and Weldon, who cor- rectly insisted on the gradualness of evolution but incorrectly claimed that inheritance was equally gradual, that is, blending. As far as ge- netics is concerned, the Mendelians were right; as far as evolution is concerned, the biometri- cians were right. There was no genuine popula- tion thinking in either camp and the biometri- cians and other opponents of Mendelism adopted Lamarckian inheritance in order to account for the gradualness of evolution. In the ensuing years, the gap between the two camps narrowed appreciably as a result of the new findings of genetics and systematics. Even- tually the biometricians disappeared from the scene and were replaced by a group of evolu- tionists I shall call the naturalists. At the same time the Mendelians were replaced by the pop- ulation geneticists. [131] 132 The major difference between these two new constellations of evolutionists was their sphere of interest. To document this it is necessary to define “‘evolution.”’ The geneticists adopted a re- ductionist definition that, as we now see it, was quite misleading. They defined evolution as the change of gene frequencies in populations. This definition emphasized the wrong level of activ- ity. Evolution is the story of adaptation, of the develoment of new kinds of animals and plants, of the origin of modes of reproduction, and of all aspects of the history of organisms. Evolu- tionary biology deals not merely with genes but with two major sets of problems: (1) the acqui- sition and maintenance of adaptedness, and (2) the origin and nature of organic diversity. The geneticists dealt only with the problem of adaptedness. Their approach was reductionist, concerned with the genetic changes within a pop- ulation. It dealt only with the time dimension, with what one might call “vertical evolution.” The naturalists concentrated on the other as- pect of evolution, the nature of organic diversity. Their interest was in populations, species, and macroevolution, with particular emphasis on the process of speciation and the geographical com- ponents of evolution, what one might call the “horizontal dimension.” This difference in basic concerns was, in my opinion, more important than their differences in the genetic interpreta- tion. There were also conspicuous national differ- ences. Evolutionary genetics flourished in the Anglophone countries as indicated by the names T. H. Morgan, H. J. Muller, S. Wright, R. A. Fisher, and J. B. S. Haldane, while the study of organic diversity in an evolutionary manner flourished in Russia, Scandinavia, and Germany, but was poorly represented in English-speaking countries, in the United States by F. B. Sumner, L. R. Dice, D. S. Jordan, and J. Grinnell, and in England by E. Poulton and E. B. Ford. We thus had two rather different traditions in the 1920s and early 1930s: an Anglophone ge- netic tradition studying the vertical component of evolution, that is, adaptive genetic change, and an essentially continental European tradition in systematics studying the horizontal component, that is the geographical changes of populations leading to speciation and macroevolution. As late as the early 1930s (up to 1935 and 1936), several authors declared that the gap between the two camps was unbridgeable. The confusion about PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 6 the causes of evolutionary change and speciation is well-documented by the widely read work of Robson and Richards (1936). At that time there appeared on the scene a young beetle systematist who had grown up in the thinking of the Russian tradition with its emphasis on organic diversity but who also had had the advantage of nearly 10 years of work in (and stimulation by) an American laboratory in population genetics. I am, of course, referring to Theodosius Dobzhansky, who integrated the two great traditions I have just described and who produced what could almost be called the Bible of the evolutionary synthesis, his magnificent Genetics and the Origin of Species (1937). In spite of some omissions and even a few outright errors, this work contained the gist of the new paradigm of the evolutionary synthesis. Within a few years the synthesis was completed in zoology, as doc- umented by the publications of J. Huxley (1942), E. Mayr (1942), G. G. Simpson (1944), and B. Rensch (1947). In 1950 G. Ledyard Stebbins brought in botany, showing in his monumental Variation and Evolution in Plants that the prin- ciples developed in the synthesis were equally applicable to plants, contrary to the claims of some other botanists. This is also true even for the numerous special phenomena and processes encountered in plants (Mayr and Provine 1980). The synthesis of the 1930s—40s was the end of old arguments, in particular the final refutation of the various non-Darwinian theories of evo- lution, but it was also the beginning of a new set of controversies. My limitation of time permits me only to deal with those concerning species and speciation. Even under that limitation one has to cover scores of books and thousands of smaller publications. This forces me to present my findings as the final conclusions of long drawn- out arguments and perhaps seemingly in a rather dogmatic manner. I am often asked, what in particular had been my own contributions to the Evolutionary Syn- thesis? They can be recorded under the two head- ings: species and speciation. SPECIES As far as species are concerned, I demonstrated the weakness, if not invalidity, of the previously most popular species criteria, particularly as stat- ed in the morphological and genetical species MAYR: SPECIES AND SPECIATION definitions. Instead, I promoted acceptance of the biological species concept with its emphasis on populations and on reproductive isolation: ““A species is a group of interbreeding natural populations that is reproductively isolated from other such groups.” I was not the first to adopt the biological species concept but there is little doubt that it was the support I gave it in my 1942 book that led to its rapid subsequent adoption. In particular, I pointed out the weakness of the morphological definition because it provided no criterion by which to determine the status of highly distinct intraspecific variants. Another weakness of the morphological definition was its inability to cope with a phenomenon for which I introduced the term “‘sibling species,” that is, morphologically virtually identical populations that were nevertheless reproductively isolated. I expanded my treatment of sibling species in 1948 and 1963, and their extreme frequency is now generally acknowledged. Previously most of them, if recognized at all, had been listed as biological races. After 25 years of argument, I finally per- suaded even Tracy Sonneborn to recognize the so-called varieties of Paramecium as sibling spe- cies. I believe there are some 14 such sibling species in the Paramecium aurelia group alone (Sonneborn 1975). Recent molecular studies have shown that most sibling species are as different from each other on a genetic-molecular basis as are morphologically distinct species. The adoption of the biological species concept was perhaps even more important for the field worker. Ecologists and students of behavior usu- ally work in a local situation, and there the rec- ognition of species as non-interbreeding, coex- isting populations is of the utmost importance. I also showed that in most species that are not host-specific, there is considerable geographic variation requiring the adoption of polytypic species. Such geographic variation may be clinal when populations are continuous, or discontin- uous when populations, particularly peripheral populations, are spatially isolated from each oth- er. Throughout I continued to emphasize that species are not types but populations. The biological species concept, based as it is on population thinking, was not palatable to workers in several fields. The paleontologists, for instance, who study species in the time dimen- sion, looked for a species concept that would be particularly suitable for the delimitation of fossil species. Here, Simpson (1961:153) proposed this 133 definition: ‘an evolutionary species is a lineage (an ancestral-descendant sequence of popula- tions) evolving separately from others and with its own unitary role and tendencies.” The re- placement of the clear-cut criterion of the bio- logical species (reproductive isolation) by such undefined vague terms as “evolutionary tenden- cies” and “evolving separately” does not permit discrimination between good species and iso- lates. It is not applicable to polytypic species that contain geographical isolates. Nor does it even permit the delimitation of an “‘evolutionary spe- cies” within a phyletic lineage. (For a more de- tailed discussion see Mayr 1988a:323.) Simpson was not raised as a naturalist and, in spite of his biometric work on samples of fossils, his mate- rial did not allow a study of geographic specia- tion. Another group of opponents of the biological species concept consisted of certain museum and herbarium taxonomists. They had to assign spec- imens to species, particularly such from widely distant geographic locations, and were puzzled what criteria to use in order to infer whether or not these isolates were reproductively isolated. As a result, they returned to a typological/mor- phological species concept. But this at once bur- dened them again with the two formidable de- ficiencies of the morphological species concept, the treatment of sibling species and of polymor- phism. Both problems require a biological spe- cies concept for their resolution. Whatever de- cision one makes, it can only be inferred but not proven. The procedure of such inference is de- scribed in Mayr and Ashlock 1991:100-105. Here I must make a short aside. Those who study species most intensely are naturalists work- ing at a given locality. They study the interaction of species either as members of an ecosystem or in connection with the study of behavioral in- teractions among different species. The ranking of geograpically distant populations is usually irrelevant for their objectives. For instance, for someone who studies the song sparrows of the San Francisco Bay region from the point of view of ecology, adaptation, or behavior, it is quite irrelevant whether one calls the song sparrows of the Aleutian Islands conspecific or a full spe- cies. However one ranks the Aleutian song spar- row will have no effect whatsoever on the study of the song sparrows of the San Francisco Bay region. In a study of the over 600 species of North 134 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 6 American birds undertaken jointly with L. L. Short (1970), we found that with one single ex- ception (Pipilo) all of them were fully consistent with the biological species concept. In a recent analysis of the vascular plants of Concord town- ship in Massachusetts (Mayr 1992a), I found that the number of cases where the biological species concept led to an ambiguous decision was less than 10%. Even here, the difficulties were mostly those of insufficient scientific analysis rather than a consequence of using the biological species con- cept. Professor Stebbins has recently informed me that he found the number of species causing difficulties in two local floras to be around 4-6%. I am quoting these figures merely in order to make the point that the claim the biological spe- cies concept is inapplicable to plants is not sub- stantiated when an actual analysis ofa local flora is undertaken. In their endeavor to apply cladistic principles even to intraspecific populations, that is to the very lowest branching points, some cladists have recently proposed a so-called “‘phylogenetic spe- cies” concept. Indeed there are now at least three versions of this concept in existence (Nixon and Wheeler 1990). This concept was first suggested by Rosen (1979:277) who proposed to consider the lowest population or population aggregate showing a new character (apomorphy) as a sep- arate species: “‘a ‘species’ is merely a population or group of populations defined by one or more apomorphous features, it is also the smallest nat- ural aggregation of individuals with a specifiable geographic integrity that can be defined by any current set of analytical techniques.” His defi- nition would have required raising the popula- tion of just about every tributary of every Central American river to species rank, for nearly all of them have some special color gene or other pe- culiar characteristic. The most widely accepted definition of the phylogenetic species is: the smallest cluster of organisms which is diagnos- ably distinct from other such clusters. The real purpose of the phylogenetic species is to serve as “the smallest unit suitable for cladistic analysis” (Nixon and Wheeler 1990:212). I am not aware of any biological significance of this unit. To adopt this reductionist approach would lead to a mas- sive increase in the number of recognized species in all groups with geographical variation and iso- lation. I estimate that birds would have over 20,000 phylogenetic “‘species,”’ as defined above, instead of the about 9,500 or so now recognized. In other groups with many localized diagnosable populations the inflation might be even greater. New papers supporting or proposing unortho- dox species concepts continue to be published quite frequently, sometimes entire volumes stress such definitions (e.g., Otte and Endler 1989). As far as I am concerned, none of these proposals strikes me as particularly convincing. Coyne (1988), in a review in Nature of one of these volumes, refers to the commotion produced by the new proposals and then continues, ““When the dust has settled only Mayr is still on his feet with his original concept remaining the simplest and most useful view of species.”’ For a while H. Paterson’s (1981) “recognition concept” was quite popular but neither Paterson himself nor anyone else seems to support it any longer after it was shown that, first, it did not really differ from the biological species concept, and secondly, that Pa- terson’s attack of the latter concept was based on a number of misconceptions (Mayr 1988b, and Coyne et al. 1988). The major reason why it is impossible at this time to say the last word on species concepts is that the biological species concept is based on the study of diploid, sexually reproducing organ- isms, forming standard biological populations. Yet our knowledge of the population structure and mode of reproduction of many groups of lower invertebrates, lower plants, protists, and particularly prokaryotes is altogether insufh- cient. Many of them clearly do not fit the stan- dard biological species concept. The place where the biological species concept runs into particular difficulties is in its application to asexually (uni- parentally) reproducing organisms. These do not form populations in the sense of the populations of diploid sexually reproducing organisms, and therefore a species definition based on popula- tions is inapplicable. It will be up to the spe- cialists of such organisms to develop a species definition that is particularly suitable for them. The question is often asked what influence mo- lecular biology has had on our understanding of species and speciation. I conclude that it has made three major contributions. First, it led to the discovery or confirmation of many sibling species. Biological races and oth- er suspected sibling species were often shown by molecular methods to be just as different as mor- phologically highly distinct species. Further- more, in the case of morphologically highly sta- ble groups of species, relationships can sometimes MAYR: SPECIES AND SPECIATION be worked out very reliably by molecular meth- ods, and more than that, the branching point between various lineages can often be assigned to definite points in the geological time scale. Second, when there is doubt concerning a par- ticular variety, whether it is merely an intraspe- cific variant or a good species, molecular meth- ods can usually give a clear-cut answer. No matter how different they may appear, intraspecific var- iants usually differ from other members of their population by only very few genes. Third, in asexually reproducing organisms, molecular methods have cast a great deal of light on the amount of difference among various clones and on various cryptic methods of gene ex- change. It must be stressed that, opposing claims not- withstanding, molecular methods have not, in any way, weakened the biological species concept nor have they affected the standard interpreta- tion of speciation. SPECIATION The recognition that species are populations, not types, was particularly important in the ex- planation of speciation. I reported in 1942 that in birds, mammals, butterflies, and snails, that is, in all taxonomically well-studied groups, spe- ciation invariably turned out to have been geo- graphical. This means that a population that had been isolated by geographical or vegetational barriers had acquired genetical isolating mech- anisms during this geographical isolation and that this subsequently permitted it to coexist with the parental species without interbreeding. Here my studies of island faunas were particularly im- portant because it enabled me to show that geo- graphic speciation is a continuous process: pop- ulations on the most recently colonized islands are still almost identical with the source popu- lation while the longer an island population had been isolated, the more different it was, until finally after a sufficient time interval, complete species status had been reached. How the isolating mechanisms were acquired continued to be controversial. There were two opposing theories, that of Alfred Russel Wallace, who thought they were acquired by selection when the previously isolated populations came into secondary contact, and that of Darwin, who said that selection could never complete the process of speciation under those circumstances. Dob- 135 zhansky favored the Wallace theory, Muller and I the Darwin theory. All the recent studies of secondary hybrid zones indicate that Darwin was, on the whole, right. Recent developments indi- cate that behavioral isolating mechanisms in an- imals may well be due to a change of function of properties acquired as a result of sexual selec- tion during the previous isolation (Mayr 1988b). In 1942 I distinguished four types of specia- tion: (1) geographical (allopatric), (2) semi-geo- graphical (now called parapatric), (3) sympatric, and (4) instantaneous. I did not reject other forms of speciation outright, but I insisted that, as far as higher organisms are concerned, allopatric speciation was the most common mode. I think it is legitimate to state that this evaluation is still valid today. It is of interest to note that by far the longest chapter of my 1942 book was devoted to what I called the “biology of speciation,”’ that is, all the ecological and behavioral factors in- volved. So far as I know there was no such treat- ment in the earlier literature. Perhaps my major contribution was that I solved the old conundrum of how one could rec- oncile the sharp demarcation of species in a local fauna and flora with the Darwinian concept of gradual evolution. I demonstrated that in the local situation species are indeed sharply sepa- rated by gaps, but that if one looks at a species taxon in its total geographical representation through its entire range, one finds that most spe- cies consist of a large aggregation of local pop- ulations. Some of these, particularly those iso- lated at the periphery of the species range, are actually incipient species, that is, in transition from the status of local population to that of an independent new species. This refuted the old claim of Darwin’s opponents that the sharp de- limitation of local species, emphasized by nat- uralists from Linnaeus on, was incompatible with Darwin’s concept of gradual evolution. The puz- zle is solved by expanding the non-dimensional species of the local naturalist to the geographi- cally variable species in a multi-dimensional ap- proach. Furthermore this showed that it was un- necessary, indeed incorrect, to postulate speciation by saltation because geographic spe- ciation is a gradualistic process. To explain how new species originate had clearly been the most important objective of my Systematics and the Origin of Species. As is well known, Darwin had supported geographic spe- ciation up to the 1840s but had become uncertain 136 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 6 when encountering certain situations in plants and eventually allowed for massive sympatric speciation (Mayr 1992b). And that was still the majority opinion among biologists when I gave the Jesup lectures in 1941. The number of confusions existing at that time was considerable. Many authors, for instance, made no distinction between phyletic change (as observed by paleontologists) and an actual mul- tiplication of species. Other authors, such as Goldschmidt (1940), in the tradition of the orig- inal Mendelians, still thought of speciation as a phenomenon involving a single individual giving rise to a new species. Soon after 1942 it became clear to me that geographic speciation was not the simple unitary phenomenon I had first thought. I discovered that there are indeed two kinds of geographic speciation. In the classical type, now also called the dumbbell model or dichopatric speciation, the range of a more or less widespread species is divided by a barrier, resulting in two isolated parts of the species. These in due time may be- come sufficiently different to act toward each oth- er as good species. The other mode, later called by me peripatric speciation (Mayr 1982), takes place when some dispersing individuals of a spe- cies establish an isolated founder population be- yond the species periphery and this population becomes in due time sufficiently different and acquires the necessary isolating mechanisms to rank as a separate species. Such speciation through the founder effect is of double significance. First, the genetic variation among the few founder individuals, sometimes only a single fertilized female, is only a small sample of the total genetic variability of the pa- rental species. Inbreeding during the founder pe- riod will lead to further loss of genetic variability. This bottleneck effect may lead to a shift in ep- istatic interactions and to a considerable restruc- turing of the genotype. Such a genetic reorgani- zation takes place only in maybe | out of 50 or 1 out of 500 of the founder populations. But if it takes place it provides an opportunity for ma- jor shifts in adaptation, particularly since the se- lecting factors in the founder population are apt to be quite different from those in the parental population. I arrived at this model of peripatric speciation not on the basis of theoretical consid- erations but rather because I observed that in South Sea islands’ birds, nearly always the most isolated populations, or the populations most distant from the core area of the species, tended to be the ones most different. Indeed, some of them were so different that they had been de- scribed as different genera (Mayr 1954). And this indicates the second reason for the importance of peripatric speciation. It provides a bridge from speciation to macroevolution. I published these ideas in 1954, and Eldredge and Gould based their theory of punctuated equilib- ria on it in 1972. This theory, in its most sim- plified version, states that many seeming gaps in the fossil record are due to the fact that such highly localized events, as genetic reorganiza- tions in founder populations, are not likely ever to be found in the fossil record, and secondly that after such speciation is completed and a new well-balanced genotype has been formed, evo- lution will greatly slow down and there may fol- low a period of stasis lasting millions of years. In the long controversy about punctuated equi- libria, it would seem to me that more facts were published supporting the theory (in the simpli- fied version here presented by me) than facts opposing it. However, most likely we have plu- ralism here as in most evolutionary phenomena, and in a few cases a phyletic lineage may dras- tically change in the course of time, also under- going changes of macroevolutionary significance. When we go back to the classical writings of Fisher and Haldane of the early 1930s we find that they, by only considering additive gene ef- fects, assumed that evolution would proceed most rapidly in large, populous species. The more re- cent evidence refutes this belief. Let me add that the isolation of an incipient species has to be pretty nearly complete. Situations such as those described by Sewall Wright of temporarily iso- lated demes may contribute to the variability of a species, but not to the formation of new species. Evolutionary changes corresponding to peri- patric speciation take place also in refugia. As was first pointed out by Stresemann (1919), later by myself (1942), and more recently particularly by Haffer (1969), contracting relict populations resulting from climatic changes may also become incipient species and with a greatly accelerated rate of divergence during their period of reduced population size. UNORTHODOX MODES OF SPECIATION Even though geographic speciation was ac- cepted by most evolutionists as by far the most MAYR: SPECIES AND SPECIATION frequent mode of speciation, a minority of au- thors promoted the occurrence of alternate modes. One can assign these various proposals to three classes. Instantaneous speciation occurs by the pro- duction of an individual that is reproductively isolated from the parental species. Actually, one type of instantaneous speciation, allopolyploidy, produced by the doubling of the chromosomes of a sterile species hybrid, is quite common in plants and not controversial. A different type of such instantaneous speciation is not infrequent in certain groups of animals, that is, a shift of species hybrids to parthenogenetic reproduction. Such cases have been particularly well analyzed in lizards and fishes (White 1974). White (1978) proposed one further type of in- stantaneous chromosomal speciation, called by him stasipatric speciation. In this model, a new chromosome is produced through a chromosom- al mutation which, although somewhat inferior as heterozygote, can conquer a new superior niche when homozygous. And then it can live side by side with the parental species owing to selection against the deleterious heterozygotes. White was led to this model by the observation that allo- patric species in morabine grasshoppers invari- ably differ from each other in some chromosomal rearrangement that is inferior in the heterozy- gotes. However, as Key (1981), Mayr (1982), and others have since shown, White’s model is not supported by any facts. Ifit were valid, one should find many cases of reproductively isolated new chromosome types inside the range of the pa- rental species but one never finds this. It is far simpler to postulate, and this is supported by all known facts, that the chromosomal rearrange- ment takes place in a small founder population. The mistake made by White was to ask, is spe- ciation chromosomal or geographical? In fact, however, all chromosomal speciation is simul- taneously also geographical, and there is no dif- ficulty in developing a model which combines both geographical and chromosomal speciation. A second model not involving complete geo- graphic isolation is so-called parapatric specia- tion. Here it is postulated that isolating mecha- nisms between two continuously distributed populations arise along an ecological escarp- ment. This model of speciation, particularly fa- vored by Murray and Clarke (1980) and by En- dler (1977), has not been substaniated in any of the very numerous test cases analyzed in recent 137 years. It is clearly associated with the Wallace- Dobzhansky theory of the sympatric origin of isolating mechanisms. The genetic analysis of numerous cases of secondary hybrid belts has shown that such belts are indeed secondary and do not lead to a gradual strengthening of the isolating mechanisms. The only seeming excep- tions are cases of rapidly spreading species over- running part of the range of a closely related species, a situation in which the first colonists have a greatly reduced opportunity for mating with conspecific individuals, with the result that occasional hybridization occurs. Such hybridiza- tion ends when the colonization is completed. As far as the reinforcement of isolating mech- anisms is concerned, one must make a distinc- tion between species and populations. Where two closely related species overlap, indeed one finds sometimes a character divergence in the isolating mechanisms. On the other hand, it is now well established that gene flow and recombination prevent such an occurrence in the area of meeting of conspecific populations. Finally, there is the possibility of sympatric speciation, that is, the development of isolating mechanisms within the cruising range of indi- viduals of the parental species. As is well known, Darwin, for various reasons, finally accepted the widespread occurrence of sympatric speciation (Mayr 1992b). For 80 years it was considered the prevailing mode of speciation by most of his followers, particularly the entomologists. Even though some perceptive evolutionists, like Ed- ward Poulton and particularly Karl Jordan, had refuted numerous putative cases of such sym- patric speciation, it was still a popular, if not majority, belief in 1942. William Thorpe (1930), for instance, published several papers describing biological races believed by him to be incipient sympatric species. My own researches led me to claim in 1942, 1947, and 1963 that all these cases could be as well or better explained by geographic speciation and furthermore, that most so-called biological races were nothing but good biological species even though morphologically very sim- ilar, that is, they were sibling species. I never denied the possibility of sympatric speciation, but I denied that this form of speciation was indeed substantiated by the proposed cases. In recent years, Guy Bush has advanced a great deal of evidence believed by him to support sympatric speciation (Bush and Zwéolfer 1984). He states, for instance, that there are more than 100,000 138 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 6 species of insects where mating and egg laying takes place on the host plant. He believes a shift to a new host plant might quickly result in sym- patric speciation. He minimizes, however, the numerous difficulties as, for instance, the occur- rence of back colonization of the parental host and the development of a two-host species, pointed out by various authors. Such difficulties disappear when one assumes that the host shift takes place in a founder population. Bush’s ev- idence has been severely criticized by Futuyma and Mayer (1980), by Jaenike (1981), and by Paterson (1981). In the best studied case, the original host of the fruit fly species Rhagoletis pomonella is hawthorne (Crataegus). After the fly had colonized apple orchards around 1850, the apple population developed certain differ- ences in life cycle and morphology from the orig- inal hawthorne population. However, vast monocultures, like apple orchards, provide an unusual situation. I spend my summers in south- ern New Hampshire, where there are square miles of apple orchards, and yet in all my botanizing in that area I have not yet found a single wild Crataegus tree. This population of Rhagoletis is therefore virtually isolated on apple trees. Attempts were also made in the laboratory to produce sympatrically reproductively isolated species. One or two of these experiments, par- ticularly one by Thoday, have indeed been seem- ingly successful (Thoday and Gibson 1962). However, there are literally scores of other such endeavors that led to negative results. Disruptive selection has been suggested as an effective mechanism to produce different species. If we had within a single population different morphs specializing in different food sources, this might in due time lead to their isolation as dif- ferent species, it was believed. Such cases of tro- phic polymorphism have indeed been found in various species of fishes, particularly cichlids (Meyer 1990) and in a species of birds (Smith 1990). However, in none of the cases is there any indication of a development of isolating mech- anisms between the trophic types. Indeed, as a Darwinian, one should expect that a lineage hav- ing both trophic types would have greater inclu- sive fitness than one giving up one of the two sources of food. This would be true in any vari- able environment, and that of course means it is true for all environments. Up to now the best evidence for sympatric speciation is provided by ants, where sometimes in a flourishing colony of one species, inquilines of another species are found (Buschinger 1990). The latter are so-to-speak parasitic because they have no worker caste and benefit from the labor of the workers of the host species. In the few cases that have been well described, the inquiline species seems to be most closely related to the host species and there is no mode of geographic speciation that could produce such a result. In spite ofall these advances and clarifications, there are still areas of great uncertainty. For in- stance, how do pelagic species speciate? Each of such species is usually associated with a partic- ular water mass, but how can such a species be divided into two? There is a possible scenario. One can assume that some peripheral portion of such a water mass becomes isolated by a different water mass intervening between it and the re- mainder of the species population and that even- tually this isolated population acquires species status. However, except for one or two cases of fossil species, there is so far very little evidence indicating the validity of this scenario. As in the case of species, our conclusions about speciation are based on the situation in diploid sexual species. There is still great uncertainty about speciation in asexual organisms, particu- larly in view of the fact that it is so uncertain what a species is in these organisms. Hence, what is speciation? However, we have the fact of the bdelloid rotifers, a taxon with some 200-odd spe- cies, all of them asexual with no males ever en- countered. This is best explained by the as- sumption that there is a continuing production of asexual clones, and that subsequent extinction of the less-viable clones leads to gaps between what ultimately will be called species. GENETICS OF SPECIATION In view of the intensive work on speciation during the last 100 years, it is shocking to have to admit how little we actually know about the genetics of speciation. This has a number of rea- sons. The first is that many geneticists did not fully understand the difference between phyletic change and multiplication of species. They thought that by extrapolating from the changes within a gene pool they would be able to explain speciation. However, it is now quite clear that the solution will come from a comparison of dif- ferent populations. Everything we know, partic- ularly through the study of peripatric speciation, MAYR: SPECIES AND SPECIATION indicates that certain genetic turnovers may go on in a founder population that are not encoun- tered in a normal species population. The second reason for the slow progress is the heterogeneity of the genotype. Owing to the ease of the elec- trophoresis method, most authors have studied the variation of enzyme genes. However, there is no evidence that this class of genes is partic- ularly involved in the origin of isolating mech- anisms. On the contrary, many of the enzyme alleles are apparently in Kimura’s class of quasi- neutral genes. It would certainly be misleading to base one’s explanation of speciation on this class of genes. There is much to suggest that the genes for behavioral isolating mechanisms be- long to different classes of genes from the genes for sterility factors. But very little concrete evi- dence on these classes of genes is available. There is little hope for a valid explanation of the genetics of speciation until genes are classified into different categories, some of which do and some of which do not play a role in speciation. Furthermore, it is quite likely that different kinds of genes are involved in the speciation of differ- ent kinds of organisms. The work on the African cichlid fishes indicates that behavior genes may be very important in this taxon, becoming iso- lating mechanisms by way of sexual selection, and may permit the evolution of reproductively isolated populations within incredibly short pe- riods, periods of only a couple of hundred years (Mayr 1988a). On the other hand, the conspe- cificity of certain eastern North American plants with their representatives in eastern Asia indi- cates that in this case an isolation of 5, 8, or 10 million years was not sufficient for the origin of efficient isolating mechanisms. Wherever we have drastically different rates of speciation, one can be reasonably sure that different kinds of isolat- ing mechanisms and their genetics are involved. SUMMARY In conclusion, I think it is legitimate to say that the basic picture of species and speciation developed by taxonomists during the Evolution- ary Synthesis and presented in 1942 in my book Systematics and the Origin of Species did not have to be changed subsequently. Most of the attacks against it have been thoroughly refuted, even though there have been clarifications and the development of a more pluralistic approach. However, there are still vast uncertainties, par- 139 ticularly with respect to the lower invertebrates, lower plants, fungi, prokaryotes, and also such ecological specialists as pelagic animals. Finally, the area where there is perhaps still the greatest uncertainty is the genetic basis of the isolating mechanisms. But here again the researches are not likely to lead to any refutation of currently accepted views. To repeat, the concepts of spe- cies and speciation as developed during the evo- lutionary synthesis are likely to endure. LITERATURE CITED Buscuincer, A. 1990. Sympatric speciation and radiative evolution of socially parasitic ants— Heretic hypotheses and their factual background. Z. Zool. Syst. Evolutionsforsch. 28:241-260. Busu, G. L. AND H. ZwO6.rer. 1984. Sympatrische und para- patrische Artbildung. Z. Zool. Syst. Evolutionsforsch. 22: 211-233. Coyne, J. A. 1988. What do we know about speciation? Nature 331:485-486. Coyne, J. A., H. A. Orr, AND D. J. FuruyMA. 1988. Do we need a new species concept? Syst. Zool. 37(2):190-200. DoszHANsky, T. 1937. Genetics and the origin of species. Columbia University Press, New York. ELprREDGE, N. AND S.J. GouLp. 1972. Punctuated equilibria: an alternative to phyletic gradualism. Pp. 305-332 in Models in paleobiology, T. J. M. Schopf, ed. Freeman, Cooper, San Francisco. EnpLerR, J. A. 1977. Geographic variation, speciation and clines. Princeton University Press, Princeton, New Jersey. FutuyMA, D. J. AND G. C. Mayer. 1980. Non-allopatric speciation in animals. Syst. Zool. 29:254-271. GOLDSCHMIDT, R. B. 1940. The material basis of evolution. Yale University Press, New Haven, Connecticut. Harrer, J. 1969. Speciation in Amazonian forest birds. Sci- ence 165:131-137. Huxtey, J. 1942. Evolution: the modern synthesis. George Allen and Unwin, London. JAENIKE, J. 1981. Criteria for ascertaining the existence of host races. Am. Nat. 117:830-834. Key, K. 1981. Species, parapatry and the morabine grass- hoppers. Syst. Zool. 30:425-458. Mayr, E. 1942. Systematics and the origin of species. Co- lumbia University Press, New York. . 1947. Ecological factors in speciation. Evolution 1:263-288. 1948. The bearing of new systematics on genetical problems. The nature of species. Adv. Genet. 2:205-237. . 1954. Change of genetic environment and evolution. Pp. 157-180 in Evolution as a process, J. Huxley, A. C. Hardy, E. B. Ford, eds. Allen and Unwin, London. 1963. Animal species and evolution. Harvard Uni- versity Press, Cambridge, Massachusetts. 1982. Process of speciation in animals. Pp. 1-19 in Mechanisms of speciation, C. Barigozzi, ed. Alan R. Liss, New York. . 1988a. Toward anew philosophy of biology. Harvard University Press, Cambridge, Massachusetts. . 1988b. The why and how of species. Biol. & Philos. 3:431-441. 140 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 6 1992a. A local flora and the biological species con- cept. Am. J. Bot. 79(2): 222-238. . 1992b. Darwin’s Principle of Divergence. J. Hist. Biol. 25:343-359. Mayr, E. AND W. PRovine. (EDS) 1980. The evolutionary synthesis. Harvard University Press, Cambridge, Massa- chusetts. Mayr, E. AND L. L. SHort. 1970. Species taxa of North American birds: a contribution to comparative systematics. Nuttall Orn. Club, Publ. No. 9, Cambridge, Massachusetts. Meyer, A. 1990. Ecological and evolutionary consequences of the trophic polymorphism in Cichlasoma citrinellum (Pi- sces, Cichlidae). Biol. J. Linn. Soc. 39:279-299. Murray, J. AND B. CLARKE. 1980. The genus Partula on Moorea. Speciation in progress. Proc. R. Soc. Lond. Ser. B. Biol. Sci. 211:83-117. Nixon, K. C. AND Q. D. WHEELER. 1990. An amplification of the phylogenetic species concept. Cladistics 6:211-223. Orte, D. AND J. E. ENDLER, EDS. 1989. Speciation and its consequences. Sinauer, Sunderland, Massachusetts. Paterson, E. H. 1981. The continuing search for the un- known and unknowable: a critique of contemporary ideas on speciation. S. Afr. J. Sci. 77:113-119. Renscu, B. 1947. Die Transspezifische Evolution. Ferdinand Enke, Stuttgart. Rosson, G. C. AND O. W. RicHARDs. 1936. The variations of animals in nature. Longmans, Green, London. Rosen, D. 1979. Fishes from the upland intermontane basins of Guatemala. Bull. Am. Mus. Nat. Hist. 162:269-375. Smpson, G. G. 1944. Tempo and mode in evolution. Co- lumbia University Press, New York. 1961. Principles of animal taxonomy. Columbia University Press, New York. Smitu, T. B. 1990. Patterns of morphological and geographic variation in trophic billmorphs of the African finch Pyr- enestes. Biol. J. Linn. Soc. 41:381-414. SONNEBORN, T. M. 1975. The Paramecium Aurelia complex of fourteen sibling species. Trans. Am. Microsc. Soc. 94(2): 155-178. Setssins, G. L. 1950. Variation and evolution in plants. Columbia University Press, New York. STRESEMANN, E. 1919. Uber die europdischen Baumlaufer. Verh. Ornithol. Ges. Bayern. 14:39-74. Tuopay, J. M. AND J. B. Grsson. 1962. Isolation by disrup- tive selection. Nature 193:1164-1166. Tuorre, W. H. 1930. Biological races in insects and allied groups. Biol. Rev. Camb. Philos. Soc. 5:177-212. Wuite, M. J.D. 1974. Animal cytology and evolution. Cam- bridge University Press, Cambridge, Massachusetts. 1978. Modes of speciation. W. H. Freeman, San Francisco. © CALIFORNIA ACADEMY OF SCIENCES, 1993 Golden Gate Park San Francisco, California 94118 Re be ey" < *—-2 ¥ Os SF ba ee nem Gh,” ana uaa pias enalipnigee . | ny ready der art phen” 4 ie, Bee 3279-199 Sree, FB ~ oy i Geis, ie Te ene Phe oe CRiipabed Siesetday Piel, + P a Se -apige han A Shee See Ett Cia Gia "cis a? is Capp TNe, eels aes ee 2. f —* ep Oe (egeeessiin acet, | th; dno f, Bllipoan Cee eee cum tyweert. wks FT Sip aabectiod Manure 198 ea] P< ou . f= , =" eC ie oan. FITS Heyes. a emer Tce oui eA ‘lear iors dure Wal Gets Cate Fa ys = ie 1% Fie hie Ona > - Weve Ot PT 19th Aninags Die. Gnd ee! a ea eee ts ee Priduer Luntieuy On, Cunt merits t ‘ Shee ‘ Wicdkey oF igre ay | " Ce Le eee wu } bs ype Ae . Pa huts 2 miei PLS ane te | $ = far spout, 1 m8 i i Re * aa erue?! Ja Fel tt SCRANCSS, 7603 tates Cate Path alifivnda Oa bth PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCI Vol. 48, No. 7, pp. 141-152, 2 figs. lO;GGical Laboratory, 7 ogical Laboratory/ ods Hole Oceanographic Institution Library December 21, 1993 DEC 27 1993 AN EVALUATION OF THE MESOAMERIGAN. SPREGIES OF MERIANIA (MELASTOMATACEAE: MERIANIEAE) Frank Almeda Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 Asstract: Meriania, one of 17 genera in the neotropical tribe Merianieae, ranges widely from southern Mexico, Central America, and the Greater Antilles south to the tropical Andes, the Guayana Highland region, and southeastern Brazil. In this study, Meriania is reported from Mexico for the first time; a new species, M. odorata is described from Costa Rica and Panama; and two species of Centronia, C. grandiflora, and C. phlomoides, are transferred to Meriania. This summary of the five Mesoamerican species includes a key, descriptions, geographical and phenological notes, diagnostic illustrations, and citation of representative specimens. Received 24 May 1993. Accepted 23 June 1993. INTRODUCTION The genus Meriania is a unifying basal group among the 17 genera comprising the Merianieae, a neotropical tribe characterized by prevailingly terminal inflorescences, capsular fruits, and ob- long-pyramidate, angulate or winged seeds with a smooth or poorly defined surface patterning. With over 50 described species, Meriania is the largest genus in the tribe and one of two wide- spread genera ranging from southern Mexico, Central America, and the Greater Antilles, south through the tropical Andes of South America and east to the Guayana highlands and southeastern Brazil. Preparation of a treatment of the Melas- tomataceae for Flora Mesoamericana has neces- sitated an evaluation of Meriania and the five other genera of Merianieae reported for the re- gion (Almeda 1981, 1989, 1990; Eves 1936; Gleason 1958; Standley and Williams 1963). Pri- or to this study, only two species of Meriania [141] were known from Central America based on few collections (Gleason 1958; Standley and Wil- liams 1963). In this study, the range of Meriania is extended to Mexico, a new species is described from Costa Rica and Panama, and the two Central American species previously included in Centronia are transferred to Meriania. To facilitate recognition and comparison, a regional revision is presented with a key, descriptions, diagnostic illustrations, and citation of representative specimens. All spe- cies treated here represent outliers ofa genus with major diversity in South America. The five Me- soamerican species can be sorted into three groups based on androecial morphology and details of the indument. Each group appears to have its closest affinities with a different assemblage of South American taxa. A better understanding of species relationships, therefore, must await a comprehensive revision of Meriania and its sat- ellite genera. 142 SYSTEMATIC TREATMENT Meriania Swartz, Fl. Ind. Occ. 2:823. (Nom. cons.) 1798 Type.—M. leucantha (Swartz) Swartz, type cons. [= Rhexia leucantha Swartz]. Trees or shrubs with glabrous to variously pu- bescent distal branchlets. Leaves opposite, de- cussate, and often long-petiolate, those of a pair equal to somewhat unequal, glabrous to vari- ously pubescent. Flowers typically borne in ter- minal multiflowered panicles but occasionally solitary, ternate, or verticillate. Hypanthium mostly free from the ovary, campanulate, hemi- spheric, or urceolate, terete or costate in fruit. Calyx tube usually well-developed, truncate and flangelike, calyx lobes and external teeth obsolete or the calyx deeply lobed with well-developed exterior teeth, or calyx sometimes fused in bud and rupturing irregularly at anthesis. Petals 5 (in our species), free and typically glabrous, oblique- ly to broadly rounded or truncate at the apex. Stamens 10 (in our species); nearly isomorphic to dimorphic, glabrous, strongly geniculate at the filament insertion; filaments usually declined in one direction transversely across the flower; the- cae subulate with a dorsally or ventrally inclined apical pore; connective usually thickened, pro- longed, or variously modified dorsally into a dor- so-basal spur and an ascending blunt or pro- longed dorsal appendage. Ovary superior (3-)5— 6-celled, glabrous, the apex prolonged and toothed, truncate, or umbilicate. Style straight, somewhat sigmoid or declinate with a truncate or punctiform stigma. Fruit a many-seeded lo- culicidal capsule. Seeds narrowly oblong-pyr- amidate to cuneiform, angulate, truncate to somewhat uncinate at the wider distal end or with tails at both ends. Among the 17 genera of Merianieae, Adelo- botrys, Axinaea, Centronia, and Graffenrieda are close relatives of Meriania. There is a particu- larly close relationship with the Centronia/Graf- fenrieda alliance on one hand and a cluster of Adelobotrys species on the other. It has even been suggested that the circumscription of Meriania could logically be expanded to include these four closely related genera (Wurdack 1973). As is true of other tribes in the Melastoma- taceae, the Merianieae exhibit an impressive range of form in calyx structure and anther mor- phology. It is the spectacular diversity and mod- ifications in these features that pose problems in PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 fixing generic limits because delimitation of Mer- iania and its satellite genera relies heavily on these characters. A reconsideration of generic limits in the tribe is beyond the scope of this study. Hopefully, new sources of comparative data such as pollen morphology and chromo- some numbers will provide information needed to reevaluate relationships among these genera. In the absence ofan optimal generic classification of the tribe, Adelobotrys, Axinaea, Graffenrieda, Meriania, and the more distantly related Tess- mannianthus will be recognized in the melas- tome treatment for Flora Mesoamericana. The two Mesoamerican species of Centronia have androecial morphology and dorsally in- clined anther pores like Meriania and are here transferred to that genus. They are clearly dis- cordant in Centronia, because they do not have arcuate essentially isomorphic anthers with ven- trally inclined pores and a connective modified into a solitary acute dorso-basal spur. Among Mesoamerican genera of Merianieae the only species that could be confused with Mer- iania is Adelobotrys panamensis Almeda. This species is similar to some species of Meriania in being shrubby or arborescent, lacking malpigh- iaceous hairs, and in having an anther connective modified dorso-basally into an erect spur and an acute appendage directed more or less parallel to the thecae (Almeda 1981). Adelobotrys pana- mensis is readily distinguished from all species of Meriania, however, by its compressed, 2-edged uppermost branchlets and umbelliform pseu- dolateral inflorescences borne on short stubby branchlets. KEY TO THE MESOAMERICAN SPECIES OF MERIANIA 1. Leaf blades variously pubescent on the ab- axial (lower) surface. 2. Abaxial foliar surface covered with shaggy-plumulose hairs and a sparser stellate-hirsute indument between the elevated veins; pedicels and hypanthia copiously pubescent; calyx fused into a bud cap that ruptures irregularly and falls away after anthesis; petals white or greenish-white; stamens nearly isomor- phic. 3. Inflorescence a few-flowered panicle with ultimate units consisting of simple cymes or paired or solitary ALMEDA: MESOAMERICAN SPECIES OF MERIANIA flowers; floral buds 15-25 x 10-13 mm; petals 2-2.7 x 1.5-3 cm; an- ther thecae 10-11 mm long; fruiting hypanthia 10-14 x 15-20 mm __. (a. een M. grandiflora 3. Inflorescence a multiflowered verti- cillate panicle with ultimate units consisting of fascicled or umbelli- form clusters; floral buds 9-13 x 4— 7 mm; petals 1.1-1.6 x 0.8-1.3 cm; anther thecae 6.5 mm long; fruiting hypanthia 6-7 x 5-7 mm ied sees M. phlomoides 2. Abaxial foliar surface uniformly cov- ered with a pulverulent or floccose in- dument; pedicels and hypanthia gla- brous; calyx truncate and flangelike with persistent lobes that are ill-defined de- pressed-ovate undulations; petals ma- genta; stamens dimorphic te, ares M. macrophylla 1. Leaf blades glabrous on the abaxial surface. 4. Leaves entire; calyx on fruiting hypan- thia truncate and flangelike, the lobes obsolete or consisting of ill-defined un- dulations; petals 1.7-2.5 x 1-1.7 cm; fruiting hypanthia 6-7 x 7mm Sy ot EN Ss 0 M. odorata 4. Leaves coarsely dentate along the apical half or two-thirds of the blade; calyx on fruiting hypanthia consisting of bluntly rounded or truncate lobes 2-4 mm long; petals 2.6-3.9 x 1.6-3.7 cm; fruiting hypanthia 9-11 x 12-13 mm _. Ne ree Re ae M. panamensis Meriania grandiflora (Standl.) Almeda, comb. nov. (Figs. 1C, D, E) Centronia grandiflora Standl., Field Mus. Nat. Hist., Bot. Ser. 22:95. 1940. Type. — PANAMA. Chiriqui: Bajo Chorro, 2,100 m, 2 Mar 1938 (fl), Davidson 360 (Holotype: F!; isotypes: A!, MO!, US!). Tree 3-15 m tall. Older cauline internodes + terete and glabrate; the bluntly quadrate, upper- most branchlets, inflorescence axes, floral buds, and hypanthia moderately to densely covered with brown, basally enlarged hairs (0.5-2 mm long) that are shaggy-plumulose basally but spar- ingly barbellate to glabrous toward the apex. Leaves of a pair equal to somewhat unequal in size; petioles 2-7.5 cm long and 2-3 mm wide; blades membranaceous when dry, 11.2-21.3 x 143 5.5-14 cm, elliptic to elliptic-ovate, (5—)7—9-pli- nerved with a well-defined elevated network of secondary and tertiary veins, base obtuse to rounded or varying to obscurely cordate, apex acuminate to attenuate, the margin entire to den- ticulate, the adaxial surface glabrous and retic- ulate-bullate, the abaxial surface covered with basally plumulose hairs on the primary and sec- ondary elevated veins and a stellate-hirsute in- dument between the veins. Inflorescence a ter- minal few-flowered, openly branched, pedunculate panicle 10-21 cm long, the ultimate units consisting of simple cymes or reduced to paired or solitary flowers; bracteoles 15-30 x 2.5 mm, linear-lanceolate, early deciduous and typically seen only on young inflorescences. Ped- icels 9-15 mm long, lengthening to as much as 27 mm in fruit. Hypanthium (at anthesis) hemi- spheric, 8—12 mm long to the torus. Flower buds rounded-ellipsoid, 15-25 x 10-13 mm, shortly (2-3 mm) blunt-rostrate. Calyx fused in bud but rupturing into irregular rounded-deltoid lobes of varying size that fall away after anthesis. Petals 2-2.7 cm long, 1.5—3 cm wide, white or greenish- white, translucent when fresh, sometimes with darkened conspicuous venation when dry, erect, obovate, + concave and connivent, rounded to shallowly emarginate apically, the margin entire. Stamens nearly isomorphic, very similar in size but differing slightly in connective modification. Antepetalous stamens: filaments 10-11 mm long, anther thecae 10-11 mm long, white, subulate, horizontal and essentially straight with a dorsally inclined pore; connective thickened, + flat ven- trally but somewhat ridged dorsally along the 2 mm segment prolonged below the thecae, dilated dorso-basally into a divergent longitudinally ridged and adaxially furrowed appendage (4 x 2 mm) and a smaller adjacent spur elevated 1-1.5 mm. Antesepalous stamens: filaments 10-11 mm long, anther thecae 10-11 mm long, otherwise like antepetalous stamens in color, posture, and pore inclination; connective flattened ventrally but vaguely ridged dorsally along the 0.5 mm segment prolonged below the thecae, dilated dor- so-basally into an erect compressed, longitudi- nally ridged appendage (4 x 3 mm) anda smaller adjacent tooth elevated 1 mm. Ovary (at anthe- sis) ovoid but truncate apically, glabrous, 5—6(- 7)-celled, becoming rounded-lobulate to round- ed-truncate apically at maturity, superior or ad- nate to the hypanthium basally for about 3 mm. Style 21-25 mm long, glabrous, erect or slightly 144 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 can Meriania. Meriania macrophylla: A, antepetalous 15. M. grandiflora: C, antepetalous stamen (left) and plumulose hair (left) of branchlets, inflorescences, and dary foliar veins, and stellate-hirsute hair (right) of abaxial foliar x ca. 2; G, seeds, x ca. 9. M. Ficure 1. Representative stamens, seeds, and hairs of Mesoameri stamen (left) and antesepalous stamen (right), x ca. 3; B, seeds, x antesepalous stamen (right), x 1 1/2; D, seeds, x ca. 13; E, shaggy- hypanthia, shaggy hair (middle) of abaxial primary and secon surface, x 15. M. panamensis: F, antepetalous stamen (left) and antesepalous stamen (right), ALMEDA: MESOAMERICAN SPECIES OF MERIANIA declinate and gently curved apically below the truncate stigma. Fruiting hypanthium campan- ulate, terete or irregularly rugose or bullate, 10- 14 x 15-20 mm. Seeds brown to beige, cunei- form or oblong-pyramidate and angulate, 0.5-1 mm long. DIsTRIBUTION AND PHENOLOGY.— Endemic to Panama where it is local in cloud forests at (880)1,350—2,500 m. Flowering collections have been made from January through August; fruit- ing specimens have been collected in all months except March, April, June, and December. REPRESENTATIVE SPECIMENS EXAMINED.— PANAMA. Bocas del Toro: Robalo trail, N slopes of Cerro Horqueta, 5-7 Aug 1947 (fl, fr), Allen 4992 (MO, NY); E slope of La Zorra to divide on Chiriqui trail, 20 Apr 1968 (fl), Kirkbride & Duke 833 (NY). Chiriqui: vicinity of Cerro Punta, 24 May 1946 (fl), Allen 3492 (BM, BR, G, NY); 1-2 km N of Las Nubes and 5 km NW of Cerro Punta, 28 May 1972 (fl), Almeda & Wilbur 1537 (DUKE, MO); trail to Cerro Pate Macho, 8°49’N, 82°24’W, 13 Mar 1988 (fl), Almeda et al. 6140 (CAS, MO, PMA, TEX); Monte Azul, 1.4 mi N of Entre Rios on E slopes of Cerro Punta, 22 Nov 1979 (fr), Antonio 2702 (CAS); vicinity of Las Nubes, 2.7 mi NW of Rio Chiriqui Viejo, 27 Feb 1973 (fl), Croat 22397 (C, F, MO, NY); Bajo Chorro, Boquete district, 25 Jan 1938 (fr), Davidson 212 (A, F, MO); Guadalupe Arriba, above Cerro Punta, 8°52’N, 82°33’ W, 23 Jul 1985 (fl), de Nevers & Charnley 6055 (CAS); E of Guadalupe along Rio Chiriqui Viejo, 2 mi NE of Cerro Punta, 13 Jan 1971 (fl), Wilbur et al. 13093 (CAS, DUKE); slopes above Rio Caldera beyond Bajo Mono, 17 Jan 1970 (fl), Wilbur et al. 11082 (DUKE, F, GH, MO); vicinity of Bajo Chorro, 20-22 Jul 1940 (fl, fr), Woodson & Schery 6574 (NY). Coclé: Cerro Pilon, spring 1968 (fr), Lallathin 9 (NY). Meriania grandiflora and its close relative, M. phlomoides, differ from other Mesoamerican congeners in androecial morphology and com- plexity of the indument. The upper branchlets, inflorescences, floral buds, and hypanthia of these species are covered with hairs that are here de- scribed as shaggy-plumulose. These hairs have an enlarged roughened or shaggy base (Figs. 1E, 1J) and a smooth tapering shaft. The elevated primary and secondary veins on abaxial foliar surfaces of these species are also covered with similar hairs but the base of each is not as prom- inently expanded and the roughening often ex- tends for a greater distance up the hair shaft. The 145 hair covering between the veins on abaxial leaf surfaces is stellate-hirsute. I follow Gleason (1940: 339) in using this term for straight or curved essentially smooth, simple hairs with a base that is more or less stellate with numerous radiating arms. (Figs. 1E, 1J). For scanning electron mi- crographs of comparable hairs see Wurdack (1986:53). The stamens of M. grandiflora and M. phlo- moides are also noteworthy in having a pseudo- inversion of the connective that was first noted for other species of Meriania by Triana (1871: 165). In each stamen the two anther cells are completely separated by the thickened connec- tive. The anther cells are not only separated but strongly positioned upward toward the dorsal ap- pendage-bearing side (Figs. 1C, 1H). This in- verted posture of the anther cells superficially makes the dorsal side appear as though it 1s ven- tral. Gleason (1958) erroneously applied the name Centronia tomentosa Cogn. to this species and relegated C. grandiflora (= M. grandiflora) to its synonymy. Wurdack (1976) subsequently trans- ferred C. tomentosa to Meriania and correctly noted that these two species are not conspecific (Wurdack 1980). Among South American spe- cies of Meriania, M. tomentosa resembles M. grandiflora in having an irregularly rupturing de- hiscent calyx, and anther thecae that show the pseudo-inversion described above. It consis- tently differs from M. grandiflora in having or- ange-red petals, yellow anther thecae, larger seeds (1.5 mm) with an elongate lateral raphe, and an appressed-setose hypanthial indument of fine hairs with an expanded roughened base and an attenuate cauducous apex. Meriania macrophylla (Benth.) Triana, Trans. Linn. Soc. London 28:66. 1871. (Figs. 1A, B) Davya macrophylla Benth., Pl. Hartw. 75. 1841. TYPE.— GUATEMALA. Quezaltenango: Rancho Palo Hueco, Har- tweg s.n. (Holotype: K!). Conostegia excelsa Pittier, Jour. Wash. Acad. Sci. 14:450. 1924. _ phlomoides: H, antepetalous stamen (left) and antesepalous stamen (right), x ca. 11; I, seeds, x ca. 11; J, shaggy basally enlarged hair (left) of branchlets, inflorescences, and hypanthia, shaggy hair (middle) of abaxial primary and secondary foliar veins, and stellate-hirsute hair (right) of abaxial foliar surface, x ca. 2 1/4. (A from Breedlove 47700; B from Breedlove & Almeda 47727, C from Almeda & Wilbur 1537; D from Antonio 2702; E from Almeda et al. 6140; F from Antonio 1040; G from Lao 399; H from Almeda et al. 2728; 1 from Almeda & Daniel 7076; J from Almeda & Nakai 4674.) 146 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 TYPE.—PANAMA. Chiriqui: southern slope of Cerro Hor- queta, Pittier 3196 (Holotype: US!; isotype: F!). Tree 6-21 m tall with terete to bluntly quad- rate cauline internodes. Distal branchlets, pe- duncles and pedicels of the inflorescence, young hypanthia, and abaxial foliar surfaces caducously pulverulent or floccose. Leaves ofa pair typically somewhat unequal in size; petioles (2.3—)4—9 cm long and 3.5—6 mm wide; blades coriaceous when dry, 9.8-34 x (5—)10-28 cm, elliptic to elliptic- ovate, sometimes varying to subrotund, 5(—7)- nerved with secondaries also prominent and elevated abaxially, base broadly rounded to cor- date and typically subpeltate, apex obtuse to rounded and commonly mucronate, the margin entire, glabrous adaxially at maturity. Inflores- cence a terminal, erect, multiflowered panicle mostly 8-30 cm long, the ultimate units con- sisting of umbelliform or subverticillate clusters; bracts and bracteoles 0.5-2.5 x 0.25-0.5 mm, subulate, sessile, caducous and typically absent at anthesis. Pedicels 4-12 mm long. Hypanthium (at anthesis) campanulate, 4-6 mm long to the torus. Calyx tube (0.5—)1—2 mm long, ascending or spreading and flangelike; calyx lobes evident as ill-defined undulations or depressed-ovate to semicircular, 1-1.5 mm long and 4-5 mm wide basally between sinuses. Petals 1.5—1.8 cm long, 0.9-1.2 cm wide, magenta, glabrous, oblong- ovate, obliquely rounded apically, the margin entire. Stamens dimorphic and geniculate at the point of filament insertion; filaments (6-)7.5-10 mm long, magenta, glabrous and complanate; anther thecae purple, glabrous, strongly subulate, 7 mm long and 1.5—2.5 mm wide with a dorsally inclined apical pore. Connective of the antepe- talous thecae prominently thickened dorsally and modified at the filament insertion into an erect, acutely bifid appendage 3 mm long with a for- wardly projecting arm 4 mm long and an out- wardly projecting arm mostly 2 mm long. Con- nective of the antesepalous thecae also thickened dorsally and modified into an incurved bifid ap- pendage 4 mm long with each divergent arm typically 1.5-2 mm long. Ovary (at anthesis) subglobose, glabrous, 5-celled, becoming round- ed-lobulate and umbilicate apically at maturity, superior or adnate to the hypanthium only ba- sally. Style 16 mm long, glabrous, somewhat de- clinate and slightly incurved apically just below the truncate stigma. Fruiting hypanthium essen- tially glabrous, shallowly campanulate to hemi- spheric, terete to bluntly costate when dry, 6-8 mm long to the torus and 6-7 mm wide. Seeds brown, vernicose, straight to somewhat curved, cuneiform and angulate, 0.5-1.5 mm long. DISTRIBUTION AND PHENOLOGY.—Local and uncommon from Chiapas, Mexico, and western Guatemala, disjunct to central Costa Rica and western Panama, in montane forests at 1,400- 2,600 m. Flowering from September through January and July; fruiting from January through April. REPRESENTATIVE SAMPLES EXAMINED. — MEXICO. Chiapas: SW slope of Volcan Tacana above Talquian, 13 Dec 1974 (fl), Breedlove 42550 (CAS, DS); above Talquian on slopes of Vol- can Tacana, 18 Nov 1988 (fl), Breedlove 71535 (CAS); SE side of Volcan Tacana, 23 Nov 1980 (fl), Breedlove & Almeda 47700 (CAS, TEX); SE side of Volcan Tacana, municipio of Union Juarez, 23 Nov 1980 (fr), Breedlove & Almeda 47727 (CAS, TEX). GUATEMALA. San Marcos: Volcan Tajumulco, 8 Apr 1965 (sterile), Andrle 521 (US); near Aldea Fraternidad, be- tween San Rafael Pie de la Cuesta and Palo Gordo, 10-18 Dec 1963 (fl), Williams et al. 26200 (F, G, US). COSTA RICA. Cartago: Volcan Turrialba, between Hacienda Volcan and Pueblo Santa Cruz, 14 Oct 1967 (fl), Maguire & Maguire 61415 (NY, US). Heredia: NW slopes of Volcan Barva near Rio San Rafael, 17 Sep 1967 (fl), Lent 129] (CR, F, MO). PANAMA. Chiriqui: Monte Rey above Boquete, 21 Jul 1971 (fl), Croat 15713 (F, MO); Boquete, 20 Mar 1977 (fr), Folsom 2207 (CAS); N side of Rio Caldera between Horqueta and Bajo Mono, 3 Jan 1975 (fl), Luteyn & Wilbur 4583 (CAS, DUKE, F, MO); valley of Rio Chiriqui Viejo, 14 Apr 1938 (fr), White 79 (MO, NY); slopes above Rio Caldera beyond Bajo Mono, 17 Jan 1970 (fl), Wilbur et al. 11086 (DUKE, GH, LL, MO, NY). All Mesoamerican populations of this species are assignable to the nominate subspecies that does not occur in South America. The Colom- bian population of M. macrophylla, which is known only from fruiting material, appears to represent an undescribed subspecies whereas the populations known from the northern and south- western mountains of Venezuela are attributable to subspecies costanensis and subspecies meri- densis respectively (Wurdack, 1978). Despite its broad but disjunct distribution, subspecies macrophylla is rare throughout its known range. Ongoing destruction of its cloud forest habitat presents a serious threat to its sur- vival. Unfortunately, it is not yet known from any protected park or reserve in the Mesoamer- ican region. The large, handsome foliage, straight bole with sparingly flaky bark, and showy floriferous pan- icle make M. macrophylla one of the most spec- tacular arborescent melastomes in the neotrop- ics. ALMEDA: MESOAMERICAN SPECIES OF MERIANIA The staminal morphology of this species is ex- traordinary. The dorsally thickened connective on the antepetalous stamens (Fig. 1A) is es- pecially thick and fleshy but nothing is known about its functional significance, if any, in pol- lination. Meriania odorata Almeda, sp. nov. (Fig. 2) Type.— PANAMA. Bocas del Toro: vicinity of Fortuna Dam, along road to Chiriqui Grande, 08°50'N, 82°15'W, ca. 470 m, 6 Sep 1987 (fl), McPherson 11651 (Holotype: CAS!; isotypes: MO!, PMA!, US!). Arbor 8-15 m; ramuli rotundato-quadrangulati demum ter- etes sicut folia inflorescentia hypanthiaque ubique glabri. Pe- tioli 2-5 cm longi; lamina 6.5-25.6 x 4.2-10.8 cm elliptica vel elliptico-ovata apice acuminata basi acuta vel obtusa, 5-nervata, coriacea et integra. Panicula 6-20 cm longa mul- tiflora, flores 5-meri, pedicellis (ad anthesim) 6-9 mm longis. Hypanthium (ad torum) 4-5 mm longum; calyx 1.5 mm longus truncatus, dentibus exterioribus non evolutis. Petala 1.7—2.5 x 1-1.7 cm asymmetrice obovata. Stamina dimorpha glabra; antheris obongo-subulatis poro 0.75-1 mm diam. Stamina maiora; filamenta 9-10 mm longa; antherarum thecae 9-12 x 1.5 mm, poro dorsaliter inclinato; connectivum 0.5 mm pro- longatum dente basali 0.5 mm acuto, appendice ascendenti 3 mm longa. Stamina minora: filamenta 13-14 mm longa; an- therarum thecae 5 x 1.5 mm, poro ventraliter inclinato; dente basali 0.25 mm longo acuto, appendice ascendenti 1.5-2 mm longa. Ovarium 5-loculare glabrum; semina 2-3 mm longa. Tree 8-15 m tall, essentially glabrous through- out with rounded-quadrate internodes. Leaves ofa pair equal to slightly unequal in size; petioles 2-5 cm long and 2 mm wide; blades coriaceous when dry, 6.5-25.6 x 4.2-10.8 cm, elliptic to elliptic-obovate, S-nerved with a prominulous network of secondary and higher order veins ev- ident abaxially, base acute to obtuse, apex blunt- ly acuminate, the margin entire. Inflorescence a terminal multiflowered panicle 6—20 cm long, the ultimate units consisting of umbelliform or ver- ticillate clusters; bracts and bracteoles evidently early caducous and not seen at anthesis. Pedicels 6-9 mm long. Hypanthium (at anthesis) cam- panulate, 4-5 mm long to the torus. Calyx tube 1.5 mm long, spreading and flangelike; calyx lobes obsolete or consisting of ill-defined depressed un- dulations on flowering and fruiting hypanthia; calyx teeth obsolete. Petals 1.7-2.5 cm long, 1- 1.7 cm wide, pale pink, glabrous, obovate, rounded to obliquely emarginate or irregularly lobulate apically, the margin otherwise entire. Stamens dimorphic and unequal in length with the larger (antepetalous) stamens inserted on the torus opposite the petals and the smaller (ante- 147 sepalous) stamens inserted opposite the calyx lobes. Antepetalous stamens: filaments 9-10 mm long, anther thecae 9-12 mm long, yellow, linear- subulate and arcuate with a dorsally inclined api- cal pore, connective prolonged 1 mm below the thecae and dilated dorso-basally into an acute red spur 0.5 mm long and an oblong red ap- pendage 3 mm long directed + parallel to the thecae. Antesepalous stamens: filaments 13-14 mm long, anther thecae 5 mm long, yellow, su- bulate, and somewhat rostrate with a ventrally inclined apical pore; connective dilated dorso- basally into an acute red spur 0.25 mm long and a forwardly projecting oblong appendage 1.5-2 mm long. Ovary (at anthesis) subcylindric to nar- rowly ellipsoid, glabrous, 5-celled, becoming rounded-lobulate and umbilicate apically at ma- turity, superior. Style 12-14 mm long, glabrous, declinate and hooked apically below the truncate stigma. Fruiting hypanthium glabrous, subur- ceolate, terete, 6-7 x 7 mm. Seeds dark brown, vernicose, straight or somewhat curved, linear- cuneate with an inflated winglike appendage at the wider end but tapered and + caudate at the opposite end, 2-3 mm long. DISTRIBUTION AND PHENOLOGY.—Local and uncommon in montane rainforest from southern Costa Rica (Lim6n) to western and central Pan- ama at elevations of 470-1,450 m. Flowering in May, July, and September; fruiting in February, May, and December. PaRATYPES.—CosTA Rica. Limon: Parque Internacional La Amistad, Fila Tsiurabeta, entre Rios Urén y Lani, 9°27'30’N, 83°00'00”W, 25 Jul 1989 (fl), Chacén 288 (CAS, CR, MO). PANAMA. Chiriqui: vicinity of Fortuna Dam, 8°45'N, 82°5’W, 15 Dec 1987 (fr), McPherson 11794 (CAS). Cocle: Cerro Gaital, E slope and ridges leading to the summit, 8°40'N, 80°07’W, 24 Feb 1988 (fr), Almeda et al. 5900 (CAS, CR, MO, PMA, US); hills N of El Valle, E slope and ridges leading to Cerro Gaital, 8°40'N, 80°07'W, 30 May 1982 (fl), Knapp 5328 (CAS). Veraguas: above Santa Fe and above Alto de Piedra, on ridge trail to top of Cerro Tute-Arizona, 8°30'N, 81°10'W, 6 Feb 1988 (fr), McPherson 12080 (MO). All but one of the six known collections of this species were made in the last six years. It seems surprising that an arborescent melastome with showy, fragrant flowers should have gone un- noticed for so long. Rainy season flowering and the apparent restriction of M. odorata to the wet- ter, less accessible slopes of the Caribbean drain- age in southern Central America account, in part, for the paucity of collections. The distinctive features of M. odorata include 148 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 Se PAE ES Ficure 2. Meriania odorata. A, habit, x ca. 1/3; B, representative leaf (abaxial surface), x 1/4 with enlargement (left) of area between two transverse secondary veins; C, fruiting hypanthium, x 2; D, seeds, x 7; E, petals, x 1; F, antesepalous stamen (left) and antepetalous stamen (right), x ca. 2. (A, B, E, F from the holotype; C from Knapp 5328; D from Almeda et al. 5900.) ALMEDA: MESOAMERICAN SPECIES OF MERIANIA its overall glabrosity, entire leaves, pale pink pet- als, flangelike calyx with obsolete or ill-defined lobes (Fig. 2C), absence of calyx teeth, and com- paratively large, appendaged, linear-cuneate seeds (Fig. 2D). Among its congeners, M. odorata appears to be most closely allied to M. panamensis, which is known only from higher elevations (1,850- 3,000 m) on the northern slopes of Volcan Bara (= Volcan Chiriqui) in western Panama. Both species are glabrous throughout and share a di- morphic androecium with each set of stamens differing in apical pore orientation (Figs. 1F and 2F). In addition to the diagnostic characters pre- sented in the key, M. panamensis also differs in having an ill-defined dorso-basal appendage on the antesepalous stamens (Fig. 1 F) and markedly different seeds (compare Figs. 1G and 2D). Label information on Chac6n 288, Knapp 5328, and McPherson 11651 note that the an- thers of this species are yellow. The latter two collectors describe the anthers as yellow and the anther tails as red. I interpret this to mean that the anther thecae are yellow and the appendages are red. If corroborated by additional field ob- servations, coloration of the androecium will provide another difference between M. odorata and M. panamensis; the latter has magenta an- ther thecae and yellow appendages. The epithet for this species is derived from odorus, Latin for fragrant or having a smell, to emphasize the rose-scented fragrance of the flow- ers. Meriania panamensis Gleason in Woodson and Schery, Ann. Missouri Bot. Gard. 28:436. 1941. (Figs. 1F, G) TyPe.— PANAMA. Chiriqui: Casita Alta to Cerro Copete, Vol- can de Chiriqui, 2,300-—3,300 m, 10 Jul 1940 (fl, fr), Woodson & Schery 361 (Holotype: NY!; isotype: MO!). Tree 4-10 m tall, essentially glabrous through- out with terete internodes. Leaves ofa pair equal or only slightly unequal in size; petioles 1.8—4.5 cm long and 1.5—2 mm wide; blades coriaceous when dry, (3.2)7-18.5 x 3.2-8 cm, broadly el- liptic varying to elliptic-ovate, 5-nerved with a prominulous network of secondary and higher order veins especially evident abaxially, base acute, apex cuspidate, the margin coarsely and sometimes remotely dentate along the apical half or two-thirds of the blade. Inflorescence a ter- 149 minal, openly branched multiflowered panicle mostly 9-20 cm long, the ultimate units con- sisting of simple cymes; bracts and bracteoles 0.3-2.5 cm long and 0.5—5 mm wide, linear-ob- long with a cuneate base and rounded apex, the petioles 1-1.5 cm long, early caducous and rarely seen at anthesis. Pedicels 2-5 mm long. Hypan- thium (at anthesis) campanulate, 4-7 mm long to the torus. Calyx truncate and flangelike in bud but rupturing (on fruiting hypanthia) into bluntly rounded or truncate lobes 2-4 mm long and 5-— 7 mm wide basally between sinuses, each calyx lobe inconspicuously beset with a blunt subter- minal abaxial tooth or callosity. Petals 2.6-3.9 cm long, 1.6-3.7 cm wide, magenta or violet, glabrous, obovate, rounded to emarginate api- cally, the margin entire. Stamens dimorphic and unequal in length with the larger (antepetalous) stamens inserted on the torus opposite the petals and the smaller (antesepalous) ones inserted op- posite the calyx lobes. Antepetalous stamens: fil- aments | 1—12 mm long, anther thecae 10-12 mm long, magenta dorsally and white ventrally, lin- ear-subulate and somewhat arcuate with a dor- sally inclined apical pore; connective prolonged below the thecae for 1 mm and dilated dorso- basally into a short acute + erect yellow spur and a blunt linear-oblong yellow appendage 3-3.5 mm long directed + parallel to the thecae. Antese- palous stamens: filaments 15-16 mm long, an- ther thecae 6.5-7 mm long, brownish dorsally and white ventrally, subulate with a ventrally inclined apical pore; connective dilated dorso- basally into a yellow erect spur and a forwardly projecting rounded appendage 0.5 mm long. Ovary (at anthesis) ellipsoid, glabrous, (4-)5- celled, becoming rounded-lobulate and umbili- cate apically at maturity, superior or adnate to the hypanthium basally for about 1 mm. Style 15-17 mm long, glabrous, somewhat declinate and strongly hooked apically. Fruiting hypan- thium glabrous, campanulate, essentially terete or bluntly costate when dry, 9-11 x 12-13 mm. Seeds brown, vernicose, straight or curved, nar- rowly cuneiform and angulate, tapered to a beak at the wider end when dry, 1-2 mm long. DISTRIBUTION AND PHENOLOGY.— Known only from western Panama where it is local and un- common in cloud forests at 1,860-3,000 m. Flowering collections have been made from Jan- uary through March and in June and September; fruiting specimens have been gathered in Janu- ary, June, and September. 150 REPRESENTATIVE SPECIMENS EXAMINED.—PANAMA. Chi- riqui: 8 km W of Cerro Punta in vicinity of Las Nubes, 11 Feb 1978 (fl), Almeda & Nakai 3526 (CAS); Bajo Chorro, E of Cerro Punta, 8°53'N, 82°32'W, 5 Jun 1979 (fl, fr), Antonio 1040 (CAS, MO); camino de Los Pozos a Cotito y Nueva Zelandia, 13 Mar 1990 (fl), Aranda et al. 1074 (CAS, PMA); 2.7 mi NW of Rio Chiriqui, W of Cerro Punta, 27 Feb 1973 (fl), Croat 22351 (CAS, MO, NY); Cerro Punta, 14 Sep 1971 (fr), Lao 399 (MO, US), along road from Cerro Punta towards Boquete near Bajo Grande, 8°50'N, 82°35’W, 5 Jun 1986 (fl), McPherson 9334 (CAS, MO); above Cerro Punta, 8°50’N, 82°35'W, 9 Feb 1986 (fl), McPherson & Morello 8380 (CAS, MO); Paseo de Respinga on way to Boquete at top of divide, 14 Jan 1971 (fl), Wilbur et al. 13196 (DUKE, F, GH, MO, NY). Meriania panamensis is readily recognized even in sterile condition by its glabrous leathery leaves that are coarsely dentate along the apical half or two-thirds of the blade. In addition to the line drawings of stamens and seeds presented here (Figs. 1F, 1G) see Gleason (1958:208) for other illustrations of some diagnostic characters. See M. odorata for a discussion of the probable relationships of M. panamensis. Meriania phlomoides (Triana) Almeda, comb. nov. (Figs. 1H, I, J) Centronia phlomoides Triana, Trans. Linn. Soc. London 28: Vs Ws7 Me Type.—Costa Rica, without exact locality, Oersted 55 (Syn- type: B, destroyed; photographs: CAS!, F, US!). Another syntype, Hoffmann s.n., was cited in the protologue but, to date, I have been unable to locate original or duplicate ma- terial of either collection. Oersted’s collections from Costa Rica are preserved at Copenhagen (C) but no type material of this species appears to be among them (B. Hansen, in litt., 25 Aug 1988). According to Standley (1937), Carl Hoff- mann’s collections were at Berlin (B), all of which were ev- idently destroyed during World War II. I refrain from neo- typifying this species pending a more extensive search of European herbaria. Shrub or tree 2-12 m tall. Older cauline in- ternodes + terete to rounded-quadrate and gla- brate, the rounded-quadrate uppermost bran- chlets, inflorescence axes, pedicels, floral buds, and hypanthia covered with brownish basally en- larged hairs (0.5—1.5 mm long) that are shaggy- plumulose basally but sparingly barbellate to glabrous toward the apex. Leaves of a pair es- sentially equal to slightly unequal in size; petioles 1.5-1.75 cm long and 1.5—-4 mm wide; blades membranaceous when dry, 10.5-26 x 6-15.5 cm, broadly elliptic to elliptic-lanceolate, 5-7- plinerved with a well-defined network of second- ary and tertiary veins, base obtuse to rounded or PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 obscurely cordate, apex acute to short-acumi- nate, the margin denticulate to subentire, the ad- axial surface glabrous but not bullate, the abaxial surface covered with shaggy-plumulose hairs on the primary and secondary elevated veins and a stellate-hirsute indument between the veins. In- florescence a terminal multiflowered, long-pe- dunculate (9-17 cm) verticillate panicle (9-)13- 21 cm long, the ultimate units appearing fasci- cled or umbelliform; bracteoles 20-30 x 3-4 mm, linear-lanceolate, early deciduous and usually seen only on young inflorescences. Pedicels 6— 12 mm long, lengthening to as much as 23 mm in fruit. Hypanthium (at anthesis) campanulate, 5-6 mm long to the torus. Flower buds rounded- ellipsoid, 9-13 x 4-7 mm, rounded to bluntly acute or bluntly rostrate apically. Calyx fused in bud but rupturing into two irregular rounded- deltoid lobes that fall away after anthesis. Petals 1.1-1.6 cm long, 0.8-1.3 cm wide, greenish-white, somewhat translucent when fresh, the venation darkened and conspicuous when dry, erect, ob- ovate, + concave and connivent, rounded api- cally, the margin entire. Stamens nearly isomor- phic, very similar in size, differing slightly in details of connective modification. Antepetalous stamens: filaments 8-9 mm long, anther thecae 6.5 mm long, white, subulate, horizontal and es- sentially straight with a dorsally inclined apical pore; connective thickened, + flat ventrally but rounded and somewhat ridged dorsally along the 1 mm segment prolonged below the thecae, di- lated dorso-basally into a spreading ridged and adaxially furrowed appendage (2 x 1.5 mm) and a blunt adjacent spur elevated about 0.5 mm. Antesepalous stamens: filaments 7-8 mm long, anther thecae 6—6.5 mm long, otherwise like an- tepetalous stamens in color, posture, and pore inclination; connective flattened ventrally but somewhat elevated and rounded along the 1 mm segment prolonged below the thecae, dilated dor- so-basally into a spreading, flattened, longitu- dinally ridged appendage (1.5-1.75 mm) and a smaller adjacent tooth elevated 0.5 mm. Ovary (at anthesis) ovoid, glabrous, 5-celled, rounded- lobulate apically, superior. Style 16-17 mm long, glabrous, erect or only slightly declinate and barely curved apically below the truncate stigma. Fruiting hypanthium cupulate, costulate, 6-7 x 5-7 mm. Seeds beige, cuneiform or oblong-pyr- amidate and angulate, 0.5—1 mm long. DISTRIBUTION AND PHENOLOGY.—Local and uncommon from northwestern Costa Rica (Cor- ALMEDA: MESOAMERICAN SPECIES OF MERIANIA dillera de Tilaran) south through the Cordillera Central to Tapanti and El Empalme on the north- ern slopes of the Cordillera de Talamanca, dis- junct to western Panama and southwestern Co- lombia (departments of Huila, Narifio, and Valle) at elevations of 900-2,400 m. Flowering and fruiting probably occur sporadically all year but no flowering collections have been seen for Jan- uary and April, and no fruiting collections have been seen for January, April, and September. REPRESENTATIVE SPECIMENS EXAMINED.—CosTA Rica. Ala- juela: ca. 10 km S of the summit of Volcan Poas, | Mar 1981 (fr), Almeda & Nakai 4674 (CAS, CR); Cerro Pata de Gallo, San Ramon, 6 Sep 1980 (fl), G6mez-Laurito 5834 (CR); Fila Volcan Viejo, San Carlos, 11-14 Feb 1986 (fr), Gé6mez-Laurito 11106 (CR); Volcan Poas, 5 Jun 1966 (fr), Schnell 759 (CR, F, NY, US); Reserva Forestal de Grecia, orillas del Rio Achiote, Umaria & Campos 46 (CR). Alajuela/Puntarenas Border: Cor- dillera de Tilaran, Monteverde Cloud Forest Reserve, Sendero Brillante along continental divide, 25 Feb 1992 (fr), Almeda & Daniel 7076 (CAS, CR). Alajuela/Puntarenas/Guanacaste Border: Monteverde Reserve, Vert. Pacifico cerca de Ventana, 10 Jun 1976 (fl, fr), Dryer 810 (CR, F). Cartago: 4-6 km N of Trinidad on S facing slopes of Volcan Turrialba, 4 Jul 1977 (fr), Almeda et al. 2872 (CAS, CR); Reserva de Tapanti, Nov 1982 (fl), G6mez 18751 (CAS); ca. 6 km SE of Tapanti, near and along crest of ridge S of Alto Patillos, 9°43’N, 83°46.5’W, 9 Oct 1986 (fl), Grayum & Herrera 7730 (CAS, CR); Cerro Carpintera, Feb 1924 (fr), Standley 34373 (US); Sendero Pal- mito, Tapanti, 31 Aug 1989 (fl), Umajia et al. 379 (CR). He- redia: Varablanca, 20 Oct 1971 (fl), Holdridge 6566 (CR); Monte de la Cruz, 16 Feb 1972 (fl, fr), Poveda 380 (CR, MO); Cerro de las Caricias, N of San Isidro, 11 Mar 1926 (fl), Standley & Valerio 52252 (US). Heredia/San José Border: ridge and slopes along Rio Para Blancas (Pacific drainage), Cerros de Zurqui, 10°3’N, 84°1’W, 13 Sep 1978 (fl), Burger & Antonio 11029 (CR). San José: Cordillera de Talamanca, ca. 4 km beyond jet. of Interamerican Hwy. and C.R. #222, 19 Dec 1975 (fl, fr), Almeda 2776 (CAS, CR); 3-6 km beyond Las Nubes near Cascajal on the NW flanks of Volcan Irazi, 10 Dec 1975 (fl, fr), Almeda et al. 2635 (CAS); La Palma, pass between Volcan Barba and Volcan Irazu, 4 Oct 1968 (fl), Davidse & Pohl 1283 (CR, F, US); entre La Sierra y El Empalme, 26 Nov 1964 (fl, fr), Jiménez 2635 (BM, F, G); La Hondura de San José, 15 Aug 1933 (fi, fr), Valerio 694 (CR, F). Puntarenas: Reserva Biol6gica Monteverde, swamp on continental divide (Sendero Pantanoso) and Sendero Chomogo, 10°18'N, 84°47’W, 16 Jul 1990 (fl), Haber & Zuchowski 10019 (CAS, CR, MO). PAN- AMA. Chiriqui: el camino a Jurutungo, hasta la cima del cerro que esta enfrente, luego que se pasa el rio, 15 Mar 1990 (fr), Aranda & Aratiz 1180 (CAS, PMA); camino de Los Pozos a Cotito y Nueva Zelandia, 13 Mar 1990 (fr), Aranda et al. 1103 (CAS, PMA). Although local and uncommon like the other four taxa of Meriania in Mesoamerica, M. phlo- Moides is the most frequently collected species and the only one with a range extending to south- western Colombia. Meriania phlomoides and M. grandiflora are 151 often confused because they are so similar in indument details, calyx morphology, petal color, and androecial characters. A collection of M. grandiflora, Woodson & Schery 657, for exam- ple, was the basis for an erroneous report of M. phlomoides (as C. phlomoides) from Panama (Woodson and Schery 1941). The only authentic Panamanian collections of M. phlomoides are the two cited above, which were gathered in 1990. Meriania phlomoides \acks the reticulate-bul- late leaves that are so characteristic of M. gran- diflora and has a verticillate multiflowered pan- icle instead of a few-flowered cymose panicle. One of the most striking differences between these two species, however, is simply size. The anther thecae and fruiting hypanthia of M. grandiflora are about twice as large as those of M. phlo- moides. Field experience with these two species indicates that this combination of quantitative and qualitative characters provides an infallible guide to their recognition. ACKNOWLEDGMENTS Fieldwork and other aspects of this study were supported by U. S. National Science Foundation Grants DEB 76-83040, DEB 78-25620, and BSR 8614880 (Flora Mesoamericana), the G. Lindsay Field Research Fund and the Nathan Jay and Virginia Friedman Fund of the California Acad- emy of Sciences. For field assistance during trips that yielded material pertinent to this study, I thank T. F. Daniel, K. Nakai, G. McPherson, J. F. Utley, K. Burt-Utley, and R. L. Wilbur. For logistical support in the field I am grateful to the Museo Nacional de Costa Rica, the Missouri Bo- tanical Garden, the Organization for Tropical Studies, the Tropical Science Center (Costa Rica), and the Smithsonian Tropical Research Insti- tute. My special thanks go to J. J. Wurdack for information on Meriania and to Ellen del Valle for the line drawings. At one time or another, I have studied in or received material as a gift or on loan from several herbaria. I thank the cu- rators and staffs of the following herbaria (ac- ronyms fide Holmgren et al. 1990) for their help and professional courtesies: A, BM, BR, C, CAS, CR, DS, DUKE, F, G, GH, K, LL, MA, MEXU, MICH, MO, NY, P, PMA, TEX, US, USJ, and WIS. RESUMEN Meriania con mas de 50 especies descritas es el género mas grande en la tribu neotropical Me- 152 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 7 rianieae. Meriania se distribuye desde la parte sur de México continuado a través de Centro América y Las Antillas hasta los Andes, la parte sudeste de Venezuela, y la parte sudeste de Brasil. Los géneros Meriania, Adelobotrys, Axinaea, Centronia, y Graffenrieda estan intimamente re- lacionados pero es Util reconocerlos hasta que dispongamos de una monografia de la tribu Me- rianieae. Se provee una revision para las especies Mesoamericanas que incluye una clave, descrip- ciones, discusiones, ilustraciones y notas sobre distribucién y fenologia para todas las especies. Se reconocen cinco especies. Se describe uno nuevo (M. odorata de Costa Rica y Panama) y se transfieren dos especies de Centronia a Me- riania. Estos cambios se basen en caracteristicas de las anteras. LITERATURE CITED AtMeDA, F. 1981. The Mexican and Central American spe- cies of Adelobotrys (Melastomataceae). Ann. Missouri Bot. Gard. 68:204-212. . 1989. Tessmannianthus, an arborescent genus of Me- lastomataceae new to Panama. Ann. Missouri Bot. Gard. 76:1-6. 1990. A third species of Tessmannianthus (Melas- tomataceae: Merianieae) from Panama. Brittonia 42:7-11. Eves, D. S. 1936. A revision of the genus Axinaea (Melas- tomaceae). Bull. Torrey Bot. Club 63:211-226. Geason, H. A. 1940. The Melastomataceae of the Yucatan Peninsula. Carnegie Inst. Wash. Publ. 522:325-373. 1958. Melastomataceae. Pp. 203-304 in Flora of Panama. Ann. Missouri Bot. Gard., Vol. 45. R. E. Woodson, Jr. and R. W. Schery, eds. HoLmaren, P. K., N. H. HOLMGREN, AND L. C. BARNETT, eds. 1990. Index Herbariorum. Part I, 8th ed. The herbaria of the world. Regnum Veg. 120:1-693. STANDLEY, P. C. 1937. Botanical Exploration in Costa Rica. Pp. 45-57 in Flora of Costa Rica. Field Mus. Nat. Hist., Bot. Ser. 18(1). STANDLEY, P. C. AND L. O. WiLtiAMs. 1963. Melastomaceae. Flora of Guatemala. Fieldiana Bot. 24:407-525. TrIANA, J. 1871. Les Mélastomacées. Trans. Linn. Soc. Lon- don 28:1-188. Woobpson, R.E., JR. AND R. W. ScHery. 1941. Contributions toward a Flora of Panama. Ann. Missouri Bot. Gard. 28: 409-492. Wurback, J. J. 1973. Melastomataceae. Pp. 1-819 in Flora de Venezuela, 8. T. Lasser, ed. . 1976. Certamen Melastomataceis. XXV. Phytologia 35:1-13. . 1978. Suplemento a las Melastomaceas de Venezue- la. Acta Bot. Venez. 13:125-172. . 1980. Melastomataceae. Pp. 1406 in Flora of Ecua- dor, 13. G. Harling and B. Sparre, eds. 1986. Atlas of hairs for neotropical Melastomata- ceae. Smithsonian Contr. Bot. 63:1-80. © CALIFORNIA ACADEMY OF SCIENCES, 1993 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 8, pp. 153-169, 10 figs., 3 tables Marine Biolo Woods Hole Oce gical Laboratory/ Library DEC 27 1993 December 21, 1993 Woods Hole, MA 02543 A REVISION OF THE SNAKE EEL GENUS MYRICHTHYS (ANGUILLIFORMES: OPHICHTHIDAE) WITH THE DESCRIPTION OF A NEW EASTERN PACIFIC SPECIES By John E. McCosker California Academy of Sciences, San Francisco, California 94118 and Richard H. Rosenblatt Scripps Institution of Oceanography, La Jolla, California 92093 Asstract: Species of the tropical snake-eel genus Myrichthys, family Ophichthidae, subfamily Ophi- chthinae, are reviewed. An identification key, synonymies, and diagnoses are provided for two subgenera, Chlevastes and Myrichthys, and each of the valid species: breviceps, colubrinus, maculosus, magnificus, ocel- latus, pantostigmius, pardalis, and tigrinus. Myrichthys aspetocheiros n. sp. is described from deepwater eastern Pacific specimens from Mexico to Panama, and differs from its congeners by its large pectoral fin, coloration, and vertebral condition. Eastern Pacific species previously synonymized with the Indo-Pacific M. maculosus are resurrected. The following new synonymies are proposed: Myrichthys bleekeri = M. colubrinus; Ophichthus miyamotonis and Myrichthys aki = M. maculosus; Ophichthus quincunciatus = M. tigrinus. The status of all nominal species of Myrichthys is described. We propose that M. colubrinus is a Batesian mimic of venomous hydrophiid sea snakes and that limited variability in eel coloration at some locations is concordant with snake/ eel sympatry. Myrichthys colubrinus coloration variability may also be explained by a “sheep in wolf's clothing” mimicry, a phenomenon we report for the first time in a vertebrate. Received 19 March 1993. Accepted 30 June 1993. INTRODUCTION The spotted and banded snake-eels of the ge- nus Myrichthys are commonly found along most tropical shores. Their striking patterning, shal- low habitus, and tendency to leave the substrate occasionally during daytime makes them per- haps the most recognized of ophichthids. It has long been suggested that their appearance and behavior, which ranges from the general simi- larity of spotted species (Caldwell and Rubinoff 1983) to the precise mimicry of the banded Indo- Pacific species M. colubrinus (Cott 1940), are mistaken by potential predators as that of a ven- omous sea snake (family Hydrophiidae). In our examination of M. colubrinus we have discov- ered what appears to be two mimetic associations between it and banded hydrophiids, either or both of which may limit the variability in eel coloration. And, as an aside, we note that despite the absence of sea snakes in the tropical Atlantic, [153] anographic Institution 154 the sight of a vividly colored (albeit spotted, not banded) eel swimming sinusoidally at or near the surface is often cause for the misreporting of hy- drophiids. Although little is known of the ecology of snake eels, we presume from their abundance in rote- none collections that they are not uncommon. We know that they do not create or inhabit per- manent burrows, and they are adept at burrowing rapidly into sand either head or tail first. Their diet, based upon gut content examination (Rand- all 1967; and our data), consists primarily of crustaceans, as one might deduce from their den- tition. Myrichthys can be seen searching for prey more commonly at night than during the day (Starck and Davis 1966; Hobson 1968; and our observations), and we have observed M. tigrinus to be attracted to carrion associated with fishing boats at the Galapagos and Revillagigedo is- lands. The species of Myrichthys share common fea- tures that make them easily recognizable among ophichthines. These include the anterior dorsal fin origin, granular dentition, broad-based and short pectoral fins, and vivid coloration. The ge- nus can be divided into several lineages, includ- ing: the banded Indo-Pacific species colubrinus, subgenus Chlevastes; the four closely related spe- cies of the Indo-Pacific “‘maculosus complex”’; the two Atlantic morphotypes; and a large-finned, deepwater species from the eastern Pacific that we describe herein. MATERIALS AND METHODS Measurements are straight-line, made either with a 300 mm ruler with 0.5 mm gradations (for total length, trunk length, and tail length) and recorded to the nearest 0.5 mm, or with dial calipers (all other measurements) and recorded to the nearest 0.1 mm. Body length comprises head and trunk lengths. Head length is measured from the snout tip to the posterodorsal margin of the gill opening; trunk length is taken from the end of the head to mid-anus; maximum body depth does not include the median fins. Mean vertebral formula (MVF) represents the average vertebral number at the location of the dorsal fin origin, mid-anus, and the last vertebral element. Vertebral counts (which include the hypural) are taken from radiographs. A minimum of 10 spec- imens (except for M. pardalis, for which we lack adequate material) of varying sizes of each spe- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 cies were measured for morphometric analysis. Material examined is deposited in the following institutions: The Academy of Natural Sciences of Philadelphia (ANSP); the British Museum (Natural History) (BMNH); the Bernice P. Bish- op Museum (BPBM); the California Academy of Sciences (CAS and SU); Grice Marine Biological Laboratory (GMBL); Instituto del Mar del Peru (IMARPE); Harvard Museum of Comparative Zoology (MCZ); the Musée Nationale d’Histoire Naturelle, Paris (MNHN); Princeton University (PU); Queensland Museum, Brisbane, Australia (QM); Scripps Institution of Oceanography (SIO); Natur-Museum und Forschungs-Institut Senck- enberg (SMF); Tokyo Imperial University, Sci- ence College Museum (TIU); Universidad de Costa Rica, San José (UCR); Rosenstiel School of Marine and Atmospheric Science (UMML); the National Museum of Natural History (USNM); University of Washington, Seattle (UW); Universitat Hamburg Zoologisches Insti- tut und Museum (ZMH); and the Zoological Mu- seum of the University of Tokyo (ZUMT). Genus Myrichthys Girard, 1859 Myrichthys Girard 1859:58 (type species: Myrichthys tigrinus Girard 1859, by monotypy). Chlevastes Jordan and Snyder 1901:867 (type species: Mu- raena colubrina Boddaert 1781, by original designation). DraGcnosis. —Ophichthid eels, subfamily Ophichthinae, tribe Ophichthini (sensu Mc- Cosker 1977) with body moderately to very elon- gate, its depth 25-70 times in total length, com- pressed posteriorly, shorter than tail. Dorsal fin origin behind nape, well before gill openings. Pectoral fins short and broad-based, less than or slightly longer than their bases. Snout short. Gill openings lateral, inclined anterodorsally, and crescentic. Anterior nostrils tubular and elon- gate; posterior nostrils open into mouth. Teeth molariform, multiserial. Third preopercular pore absent. Coloration spotted and/or banded. EtTyMOLoGy.— From the Greek myrus (eel) and ichthys (fish), masculine. REMARKS.—In creating the genus Chlevastes for the species Muraena colubrinus, Jordan and Snyder (1901:867) noted its relationship to Myr- ichthys but considered the foreshortened anal fin ending to be generically distinct. McCosker (1977) examined the osteology of M. maculosus, M. col- ubrinus, M. tigrinus, and the new species, and found them to fall within the generic limits of McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 155 Myrichthys. The synapomorphies of M. colubri- nus, including its coloration, foreshortened anal fin, slender body, and head length are deserving of subgeneric rank, for which the name Chle- vastes is available. KEY TO THE SPECIES OF MyRICHTHYS la. Body coloration pale, overlain with 25- 35 black or brown rings or saddles, sep- arated by white or pale interspaces, en- circling or partially encircling the body; anal fin ends well in advance of end of dorsal fin; body very elongate, its depth 50-70 times in total length 0. foe subgenus Chlevastes _....M. colubrinus (Indo-Pacific) 1b. Body coloration pale or dark, overlain with round spots that are either dark, pale, or dark diffuse spots with bright centers; anal and dorsal fin end about equally relative to tail tip; body moderately elongate, its depth 25-45 times in total length OR Uae subgenus Myrichthys 2 2a. Body coloration of round pale spots on a dark background of brown or green ........... pe 2. Le M. breviceps (western Atlantic) 2b. Body coloration pale, overlain with dark spots or diffuse dark spots with bright cen- “305 oa i 2 Bee 3 3a. Spots on body diffuse with bright centers E02: TEES) a a 4 3b. Spots on body dark and distinct, without palcror brighticentérs:sus O00 5 aa.) Fotal vertebrae 15 1=159 2) M. pardalis (eastern Atlantic) 4b. lotal vertebrae 164-173... meee ieee) it M. ocellatus (western Atlantic) 5a. Pectoral fin reduced, its length less than the width at its base; dorsal fin origin above Ist, 2nd or 3rd vertebra, 1.4—2.0 in head length; total vertebrae 149-197. 6 5b. Pectoral fin developed, longer than the width at its base; dorsal fin origin above 6th, 7th or 8th vertebra, 1.1-1.4 in head length; total vertebrae 159-167 CO eS EE La M. aspetocheiros (eastern Pacific) Gaye otal vertebrae’ 177-197 2 es 7 6bs Total'vertebrae 149-168 8 7a. Total vertebrae 177-183; spotting on chin and throat of adults smaller than or equal to eye, spotting on flanks round noe M. magnificus (Hawaii, Leeward, and Johnston islands) 7b. Total vertebrae 180-197; spotting on chin and throat of adults larger than eye, spot- ting on flanks generally ovoid chit, deta POS VO M. maculosus (Indo-Pacific) Sa. otal vertebrae 149-156. 2. Sean eS M. tigrinus (eastern Pacific) Sbi otal’ vertebraew/58=163i es Bund ref M. pantostigmius (Revillagigedo and Cipperton islands) Myrichthys aspetocheiros, new species (Figs. 1, 2; Tables 1, 2) Myrichthys sp. McCosker 1977:16, 78. DraGcnosis. —An elongate species of Myr- ichthys with depth 26.9-36.8 times and tail 1.66— 1.79 in TL; pectoral fin elongate, longer than snout; numerous large brown spots on head and body; and total vertebrae 159-167, mean 163.2 (n = 12). CoUNTS AND MEASUREMENTS OF HOLOTYPE (in mm).—Total length 405; head 44.1; trunk 128.9; tail 232; predrosal distance 31.0; pectoral fin length 8.1; pectoral fin base 5.6; body depth 14 at gill openings; body width 11 at gill openings; snout 7.2; tip of snout to rictus of jaw 12.4; eye diameter 3.4; interorbital distance 5.2; gill open- ing height 5.7; isthmus width 6.5. Vertebral for- mula 5-60-162. Ten lateral line pores in left bran- chial region; the remainder small and difficult to discern. DESCRIPTION. — Body elongate, depth at gill openings 26.9-36.8 in TL. Head and trunk 2.27- 2.51 and head 8.5-11.1 in TL. Snout rounded, conical when viewed from above. Lower jaw in- cluded, its tip reaches base of anterior nostrils. Eye large, 2.9-4.2 in upper jaw, its center well behind midpoint of upper jaw. Anterior nostrils tubular, less than twice in eye, with a small lappet extending from each side. Posterior nostrils in upper lip, not visible externally, beginning before eye and ending beneath middle of pupil. Upper lip papillate, particularly between anterior and posterior nostrils. A broad fleshy chevron di- vides snout between anterior nostrils (Fig. 2). Dorsal fin origin on head, well in advance of gill opening, 1.11-1.36 in HL. Pectoral fin well-de- veloped for a Myrichthys, longer than its base, equal to or longer than snout. Head pores barely visible, but typically Myr- ichthys-like (cf. McCosker, 1977:78). Preoper- 156 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 Ficure 1. cular, temporal, suborbital, postorbital, and oth- er series present. Single median interorbital and temporal pores. Two preopercular pores. Four mandibular pores. Lateral line pores present, but difficult to ascertain. Teeth granular, small and fixed (Fig. 2), irreg- ularly biserial in jaws and on vomer. A small intermaxillary chevron anteriorly, followed by a gap. Color in isopropyl alcohol tan dorsally, paler ventrally, overlain with two irregular rows of large brown spots, one along lateral line and the other flanking the dorsal fin. Median fins mostly un- pigmented, with occasional smudges or spots. Smaller spots on chin and snout. Throat, pectoral fin, and tail tip unpigmented. Eyes blue. S1zE.—The largest specimen we examined is 513 mm long. EryMoLocy.—From the Greek aspetos (un- speakably great) and cheiros (hand), treated as a noun in apposition. RANGE.— Mexico to the Gulf of Panama, trawled over sand bottoms between 8 and 64 m depth. REMARKS.— The new species differs from its congeners on the basis of its large pectoral fin and posterior dorsal fin origin. It is most closely related to the dark-spotted Indo-Pacific species, and differs from them in its spot size and distri- bution and in its vertebral number. All speci- mens were captured by benthic trawls; most oth- Holotype of Myrichthys aspetocheiros n. sp., SIO 65-166, 405 mm. er Myrichthys are captured using ichthyocides or are netted at the surface. We presume that unlike its congeners, the new species inhabits uniform, rock-free deeper-water habitats of sand and mud. MATERIAL EXAMINED.— Ho .otyPe: SIO 65-166, 405 mm, Mexico—Gulf of Tehuantepec: 16°03.6’N, 95°19’W, otter trawled over a sand bottom in 24 fms by T. Matsui and C. W. Jerde on 7 June 1965. PARATYPES. —Mexico—SIO 65-166, 311.6 mm, collected with the holotype. SIO 73-229, 210 mm, Islas Tres Marias: NE of Maria Madre, 45 fms, 30 Mar. 1973. Costa Rica.— SIO 73-298, 304 mm, SSW of Pta. Judas, 09°27.3'N, 84°30. 1'W, 17 fms. CAS 33299 (originally GMBL 73-162), 513 mm, 09°49.5'N, 84°48.8'W, 30 fms. SIO 73-296, 304.5 mm, Gulf of Nicoya: 09°37'N, 84°50'W, 47 fms, 22 Apr. 1973. UCR 4-8, 226.2 mm, Gulf of Papagayo: N of Nicoya Peninsula, 64 m. PANAMA.—SIO 64-357, 483.9 mm, off Viejo Panama, 4 fms. SIO 64-371, 386 mm (cleared and stained), Viejo Panama: 28 Jan. 1960. UMML 26901, 3 (319-325 mm), 08°54.5’N, 79°11'W. UMML 26710, 295.5 mm, 08°40.5’N, 79°30.7'W. USNM 323889 (formerly UMML 29145), 279 mm, 08°26'N, 79°43'W. ANSP 169606 (formerly UMML 29146), 335 mm, 08°18'06”N, 80°00'30'W, Bay of Panama: otter trawled in 18.3 m on | May 1967. Myrichthys breviceps (Richardson) (Fig. 3, Table 2) Ophisurus breviceps Richardson 1844:99 (locality unknown, holotype BMNH uncatalogued). Muraena acuminata Gronow in Gray 1854: 21 (“in insula Divi Eustachii in America,” holotype unknown). Pisodonophis guttulatus Kaup 1856a:49 (Martinique, syntypes MNHN B.2693, B.2694). Pisoodonophis coronata Kaup 1860:14 (St. Thomas, holotype ZMH 391). McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 157 FicureE 2. Dentition of the holotype of Myrichthys aspe- tocheiros n. sp., SIO 65-166, 405 mm. Ophisurus longus Poey 1867:254 (Cuba, holotype MCZ 9155). Ophichthys pisavarius Poey 1880:253 (Cuba, holotype un- known). Leptocephalus undulatus Str6mman 1896:27 (29°N, 72°W, a leptocephalus, holotype unknown). DiaGcnosis.—Body moderately elongate, its depth 33-40 in TL. Head and trunk 2.4—2.6 and head 11.6—13.3 in TL. Snout rounded, 4.5-6.2 in HL. Dorsal fin origin 1.4—1.6 in HL. Pectoral fin length less than its base, 12.6—20 in HL, about 3 in snout. Body and trunk with two to four rows of pale round spots overlaying brown or green background; numerous smaller spots on snout, cheek, and nape; a pair of spots in advance of dorsal fin origin. Total vertebrae 165-175, MVF 3-62-170 (n = 25). DisTRIBUTION.—An insular species, known from Bermuda, the Bahamas, the Florida Keys, throughout the West Indies and offshore islands south to Brazil. Collected over sand and turtle TaBLE 1. Proportions (in thousandths) and vertebral counts of the holotype and 10 paratypes of Myrichthys aspetocheiros. (TL = total length; HL = head length.) Mean Range Total length 226-513 Total vertebrae 163.4 159-167 Head length/TL 107 89-125 Trunk length/TL 312 297-333 Tail length/TL 580 559-601 Body depth/TL 33 30-37 Dorsal fin origin/HL 798 736-898 Pectoral fin length/HL 202 160-261 Snout/HL 180 160-202 Upper jaw/HL 278 250-328 Eye/HL 79 70-96 Interorbital width/HL 130 116-151 grass beds at depths to 9 m, usually less than 3 m (McCosker et al. 1989:376). SizE.—The largest specimen we examined is 783 mm long. ETYMOLOGY.—From the Latin brey (short) and ceps (head). A noun in the genitive case. REMARKS.— The above diagnosis and synon- ymy is based on that of McCosker et al. (1989: 374-377). Randall (1967) reported that this spe- cies (as M. acuminatus) fed mainly on crabs and stomatopods. Thresher (1984) reported large ag- gregations of what he tentatively identified as Mpyrichthys (which, if correct, would likely be this species) near the surface well off the Atlantic coast of Panama, and suggested that they migrate to deeper waters to spawn. Leiby (1989:827-829) described and illustrated the leptocephalus and glass eel stages of M. breviceps. MATERIAL EXAMINED.—A complete listing of the material that we examined is in McCosker et al. (1989:377). Myrichthys colubrinus (Boddaert) (Figs. 4-5, Table 2) Muraena colubrina Boddaert 1781:56 (Amboina, type un- known). Muraena annulata Ahl 1789:8 (East Indies, type unknown). Muraena fasciata Ahl 1789:9 (East Indies, type unknown). Ophisurus alternans Quoy and Gaimard 1824:243 (Guam, type unknown). Ophisurus fasciatus var. latifasciata Bleeker 1864:64 (Indo- nesia, type not designated). Ophisurus fasciatus var. oculata Bleeker 1864:64 (Indonesia, type not designated). Ophisurus fasciatus var. semicincta Bleeker (non O. semicinc- tus Lay and Bennett 1839:66) non O. semicinctus Richardson (1844:99), 1864:64 (Indonesia, type not designated). 158 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 Ficure 3. Myrichthys breviceps, ANSP 98359, 714 mm (from McCosker et al. 1989, Fig. 383). Ophichthys elapsoides Castelnau 1875:47 (Cape York, Austra- lia, type unknown). Ophichthys naja de Vis 1884:455 (South Sea Islands, syntype QM I. 864). Chlevastes elaps Fowler 1912:13 (Philippine Islands, holotype ANSP 1001). Myrichthys bleekeri Gosline 1951:314 (substitute name for Ophisurus fasciatus var. semicincta Bleeker 1864, preoccu- pied). D1aGnosis. — Body very elongate, its depth 51- 68 in TL. Head and trunk 2.0-2.3 and head 17- 20 in TL. Snout rounded, 5.5—-6.0 in HL. Dorsal fin origin 1.7—2.3 in HL. Anal fin ends well in advance of tail and dorsal fin ending. Pectoral fin length much reduced, about 3-5 in snout. Head, body, and tail coloration pale to cream, encircled (completely or as a saddle) with 25-35 brown or black bands that are equal to or nar- rower than their pale interspaces; dark bands ex- tend onto dorsal fin; some larger specimens with large spots between bands on trunk and tail (see TABLE 2. Total vertebrae of species of Myrichthys. Confi- dence limits are the limits of the mean. n Mean Range 95% conf. M. aspetocheiros 12 163.2 159-167 161.9-164.5 M. breviceps 25 170.2 165-175 170.0-170.4 M. colubrinus 30 197.0 193-202 196.5-197.5 M. maculosus 30 192.1 185-199 191.9-192.4 M. magnificus 23 179.0 177-183 178.9-179.2 M. ocellatus 25 168.1 164-173 167.9-168.3 M. pantostigmius 18 162.2 158-168 161.2-163.2 M. pardalis 5 154.8 151-159 153.1-156.5 M. tigrinus 53 153.3 148-157 152.6-153.9 remarks for additional color information). Total vertebrae 193-202, MVF 1-85-197 (n = 30). SizE.—The largest specimen we examined is 970 mm long. ETyMOLOGY.—From the Latin colubrinus (snakelike). An adjective. DISTRIBUTION. — Widespread in the Indo-Pa- cific, from east Africa and the Red Sea to the Central Pacific, including Johnston Island but not Hawaii. REMARKS. — The variety of colorations that this harlequin eel displays has resulted in the descrip- tion of 10 nominal species and the vexation of all who have studied it. Bleeker (1864) was among the first to recognize the range of banding and spotting patterns and named three varieties to include them. Subsequent taxa were created, syn- onymized, and often resurrected, resulting in the general acceptance of but one (J. Smith 1962) or as many as three species (Schultz 1943; Schultz et al. 1953), Myrichthys colubrinus, M. elaps, and M. bleekeri. We have examined a large series of specimens that display a broad range of size, geo- graphic origin, and coloration, and have con- cluded that there is but a single species whose livery is related to size, maturity, and the pres- ence or absence of hydrophiid sea snakes. In brief, we propose that Myrichthys colubri- nus is a Batesian mimic of strikingly banded, venomous sea snakes that occupy shallow coral reefs over much, but not all, of the eel’s range. Although sea snakes are eaten by some birds and fishes, there is evidence that they are avoided by many predators (Heatwole 1975, and references therein). We therefore attribute the near unifor- McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 159 Ficure 4. Myrichthys colubrinus, 555 mm (from Herre 1923, Pl. 4). mity in many collections of the “colubrinus” (Figs. pletely encircles the eel is typical of Japanese, 4-5) phase to the overlap in its distribution with Philippine, Palauan, Guamanian, and Indone- sea snakes of the genera Hydrophis and Laticau- sian specimens; hydrophiins are similarly dis- da (M. Smith 1926). Wide banding that com- _ tributed (McDowell 1972). Banded Myrichthys Figure 5. Myrichthys colubrinus, “‘elaps’” coloration, based on the holotype of Chlevastes elaps, ANSP 1001, 705 mm (from Fowler 1912, Fig. 3). 160 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 Ficure 6. Adult banded sea snake, Hydrophis melanocephalus, length unstated (from Halstead 1970, Pl. XXI, Fig. 3). extralimital to the distribution of banded sea snakes (e.g., Polynesia and the Red Sea) are more variable in coloration and often demonstrate the “‘bleekeri’ condition wherein the dark bands are much narrower and do not meet ventrally. Con- cordance in color pattern by sympatric mimics and models has been demonstrated for such Batesian mimics as New World coral snakes (H. Greene and McDiarmid 1981) and African egg- eating snakes (Gans 1961). We are unaware of other ichthyological examples. We further propose that an additional mimicry phenomenon limits the variability of coloration of Myrichthys colubrinus in sympatry with band- ed sea snakes. This form of sign stimulus mim- icry, whereby prey resemble predators in order to avoid predation, is akin to ‘“‘a sheep in wolf’s clothing.” It has been recently described for tephritid flies that mimic jumping spiders that prey upon them (Mather and Roitberg 1987; E. Greene et al. 1987), and may also serve to explain the Myrichthys situation. Eels and other elongate fishes, and ophichthid eels in particular, are the primary food of banded sea snakes (McCosker 1975). There is no evidence that sea snakes feed upon conspecifics or other snakes, therefore such a mimicry whereby predators are unable to dis- criminate between conspecifics and mimics would benefit M. colubrinus and other possible sea snake mimics (most notably, Leiuranus semicinctus, a common ophichthid that is similar in appear- ance to M. colubrinus and many hydrophiids). In order to test the hypothesis that coloration could be related to sex or sexual condition, we dissected and sexed several M. colubrinus. Of nine females examined (497-815 mm), the smallest ripe individual was 600 mm. Of three males examined, (520-970 mm), only the largest individual appeared to be ripe. Spotting was present in both sexes, and we are therefore unable to relate coloration to sex or sexual condition. The holotype of Chlevastes elaps (Fig. 5) has 2-86-198 vertebrae, well within the range of M. colubrinus (Table 2). The syntype of Ophichthys naja is an “‘elaps”-like specimen of M. colubri- nus. (Most references to this taxon have incor- rectly cited the volume of the reference.) Oph- ichthys elapsoides Castelnau (1875) has escaped notice by most authors. It was summarily syn- onymized with colubrinus by Whitley (1951) but subsequently included as Chlevastes elapsoides in a list of Australian fishes (Whitley, 1964). It is apparent from Castelnau’s description that his species was colubrinus. MaTERIAL EXAMINED.—ANSP 1001, 705 mm, Philippine Islands, the holotype of Chlevastes elaps. QM I. 864, an adult, South Sea Islands, the syntype of Ophichthys naja. New Gutn- EA—CAS 63529, 2 (425-495 mm), Madang. PHitipprnes—CAS McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS FIGURE 7. 79417, 405 mm; CAS 79381, 460 mm; CAS 79384, 525 mm; CAS 79379, 510 mm; CAS 79529, 630 mm; CAS 79383, 300 mm; and CAS 79398, 3 (200-370 mm), Batanes Province. CAS 79394, 3 (665-815 mm); CAS 79365, 800 mm; CAS 79366, 732 mm; CAS 79384, 530 mm; CAS 79368, 642 mm; and CAS 53688 (formerly UW 6525), 575 mm, Dumaguete. PaLtau—CAS 79382, 3 (335-365 mm), off Chol. CAS 79380, 480 mm, Koror. CAS 79375, 415 mm; and CAS 79388, 525 mm, Peleliu. CAS 79389, 355 mm, Auluptagel. CAS 79386, 540 mm, Babelthaup. CAS 79377, 2 (240-312 mm), Nardueis. CAS 79364, 4 (497-970 mm), Urukthapel. CAS 79373, 351 mm, Iwayama Bay. CAROLINE ISLANDs —Ifaluk: CAS 79392, 12 (195-635). Kapingamarangi: CAS 79371, 2 (242-292 mm); CAS 79369, 575 mm; CAS 79385, 425 mm; CAS 79370, 2 (315-600 mm); and CAS 79390, 290 mm. Yap: CAS 79378, 455 mm; CAS 79372, 325 mm; and CAS 79376, 390 mm. Marianas IsLANDs — Guam: CAS 79387, 3 (235-515 mm); and CAS 79367, 2 (435-600 mm). Ponape: CAS 79374, 315 mm. Tahiti: CAS 79395, 510 mm, Moorea. CAS 79391, 3 (520- 690 mm), Taravao. Palmyra Atoll: CAS 79313, 488 mm; CAS 79530, 2 (504-640 mm). Johnston Island: CAS 33498 (for- merly UH 1421), 358 mm; CAS 79397, 2 (335-360 mm); and 161 Myrichthys maculosus, BPBM 7006, 226 mm (photo by J. E. Randall). CAS 79396, “Sandwich Islands,” locality questionable, 625 mm. Myrichthys maculosus (Cuvier) (Fig. 7, Table 2) Muraena maculosa Cuvier 1816: 232 (“European seas,” ho- lotype MNHN B.2730). Muraena tigrina Riippell 1830: 118 (Mohili, Comoro Islands, holotype SMF 3525). Ophichthys dromicus Giinther 1870: 80 (‘West Africa,” ho- lotype in the British Museum, unnumbered). Myrichthys rupestris Snyder 1911: 490 (Okinawa, Ryukyu Is- lands, holotype USNM 74048). Ophichthus miyamotonis Tanaka 1913: 195 (Oshima, south of Kagoshima, Japan, holotype TIU 2923). Mpyrichthys aki Tanaka 1917: 458 (Japan, holotype ZUMT 7612). D1aGnosis.—Body moderately elongate, its depth 33-46 in TL. Head and trunk 2.3-2.45 162 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 and head 12.7-15.2 in TL. Snout rounded, 4.7— 5.7 in HL. Dorsal fin origin 1.5—2.0 in HL. Pec- toral fin length less than its base, 2—3 in snout. Head and body color yellow to tan, overlain with brown to brownish-black spots, variable in size but mostly larger than eye, the spots becoming larger and more numerous in larger specimens (see Remarks below). Total vertebrae 185-199, MVF 1-75-191 (n = 30). SizE.—The largest specimen we examined is 990 mm long. ETYMOLOGY.—From the Latin maculosus (spotted). An adjective. DISTRIBUTION. — Widespread in the Indo-Pa- cific, from east Africa and the Red Sea to the central Pacific, not including Hawaii and the Lee- ward Islands. Collected over sand, usually in shallow waters, to depths of 20-30 m. REMARKS.—We refer all dark-spotted Indo- west Pacific Myrichthys to M. maculosus. As in other species of Myrichthys, there is considerable variation in the size and number of spots with growth, and we feel that the differences ascribed to certain nominal species can be explained in that way. In particular, M. aki, said to be dis- tributed from the Izu Peninsula to Kochi Pre- fecture, Japan (Masuda et al. 1984), was differ- entiated from M. maculosus in having more spots. Our comparison of large individuals from broad geographic ranges (e.g., CAS 35204, 845 mm TL, from Grande Comore with ZUMT 54,141, 840 mm TL, from Hochijo Island, Japan) demon- strated that spotting differences are ascribable to growth. The spotting of M. maculosus may be summarized in the following manner: (1) juve- niles (< 25 cm) have a single row of nearly sub- equal spots above the lateral midline to the base of the dorsal fin, and lack chin, throat or ventral spotting; (2) intermediate individuals (30—50 cm) have larger, equally spaced spots (about equal in size to the snout) along the flanks, centered along the lateral midline, alternating with pairs of spots that meet along the dorsal midline; and (3) large adults (> 50 cm) have 2-3 rows of larger, more ovoid dark spots along the dorsal midline and along the flanks, with 2-3 rows of smaller spots along the ventral surface, the spots extending onto the dorsal fin but not the anal fin, and numerous spots (variable in size) on chin and throat. The smallest specimen we examined (CAS 48460) was 82 mm long and lacked pigmentation entirely. Blache (1975) reexamined the type of Oph- ichthys dromicus Giinther (1870) and on the ba- sis of its coloration and vertebral count (192 to- tal, 81 “abdominal’’) synonymized it with M. maculosus. Blache questioned the purported provenance of dromicus and suggested that it was erroneously identified as from “West Africa” rather than “East Africa.”” We concur with the long-recognized synonymy of Muraena tigrina Rippell (1830); the type (SMF 3525) has 194 total vertebrae, 2 predorsal, and 75 preanal. The type specimen of M. maculosus has 197 total vertebrae, 3 predorsal (E. B. Bohlke, in litt. 29 March 1979). We propose the following new synonymies of this species: With advice from Yoshiaki Tom- inaga (in litt. 6 June 1978), we synonymize Oph- ichthus miyamotonis Tanaka (1913) from Japan with M. maculosus. As mentioned above, we consider Myrichthys aki Tanaka (1917) to also be a junior synonym; the type of M. aki (ZUMT 7612) has 196 total vertebrae, 1-2 predorsal. MATERIAL EXAMINED.—MNHN B. 2730, 260 mm, “Euro- pean Seas,” the holotype of Muraena maculosa. BMNH un- numbered, 730 mm, “West Africa,” the holotype of Oph- ichthys dromicus. ZUMT 7612, 990 mm, radiograph of the holotype of Myrichthys aki. USNM 74048, 415 mm, Okinawa, holotype of Myrichthys rupestris. JAPAN—ZUMT 19.148, 930 mm; and ZUMT 54.141, 840 mm, Hochijo Id. PHupprnes—CAS 30491, 167 mm, Luzon. CAS 79302, 483 mm, S. of Dumaguete. CAS 79335, 365 mm, Batanes Province. PALAU—CAS 79311, 249 mm; and CAS 79309, 435 mm, Urukthapel Id. CAS 79331, 331 mm, Au- luptagel. CAS 79313, 225 mm; and CAS 79306, 435 mm, Babelthaup Id. CAS 79303, 396 mm; and CAS 79305, 410 mm, Angaur Id. CAS 79329, 109 mm, Iwayama Bay. CAS 79328, 165 mm, Nardueis Id. CAS 79330, 310 mm, Peleliu. CAS 79319, 345 mm, Kayangel Id. CAS 79318, 370 mm, Urukthapel Id. CAROLINE ISLANDs— Yap: CAS 79308, 349 mm, Gagel Id. Ifaluk: CAS 79301, 3(182-334 mm); GVF 53-32, 247 mm; CAS 79315, 252 mm; CAS 79320, 2(222-233 mm); CAS 79324, 130 mm; CAS 79326, 200 mm; and CAS 79317, 365 mm. Kapingamarangi: CAS 79323, 3 (145-285 mm); and CAS 79321, 360 mm. MARIANAS IsLANDs—Guam: CAS 79300, 6 (180-460 mm); CAS 79312, 191 mm; CAS 79310, 325 mm; and CAS 79316, 280 mm. Saipan: CAS 79307, 205 mm. Amer- ican Samoa: CAS 38564, 185 mm, Tutuila. Marcus Island: BPBM 7006, 226 mm. Pitcairn IsLANDs—Oeno Atoll: BPBM 16456, 500 mm. Palmyra Island: CAS 79322, 4 (110-285 mm). Wake: CAS 79332, 140 mm. Kirrati—Christmas Is- land: CAS 79327, 222 mm. SouTH CHINA SEA — Vietnam: CAS 79304, 290 mm, 10°29'15’N, 108°57'30’E. Santa Cruz Is- lands: SU 66887, 310 mm, Vanikolo. Hong Kong: SU 66884, 268 mm. New Caledonia: SU 19047, 485 mm, Noumea. Gran- de Comore Island: CAS 35204, 845 mm, N. of Moroni. CAS 48460, 82 mm, Itsandra. RED SEA: CAS 52099, 105 mm, Dahab. Vertebral data only.—ANSP 119340, 1, Kenya. ANSP 119342, 3, Kenya. ANSP 109987, 1, Kenya. ANSP 119341, McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 163 FIGURE 8. 1, Aldabra. ANSP 119338, 2, Sri Lanka. ANSP 119339, 2, Sri Lanka. ANSP 90780, 1, Guam. ANSP 119592, 1, Saipan. ANSP 119346, 3, Howland Id. ANSP 109993, 1, Solomon Is. ANSP 119335, 1, Solomon Is. ANSP 119337, 1, Solomon Is. Myrichthys magnificus (Abbott) (Table 2) Pisoodonophis magnifica Abbott 1861:476 (Hawaiian Islands, holotype ANSP 1013). Ophichthus stypurus Smith and Swain 1882:120 (Johnston’s Island, holotype USNM 26817). D1AGnosis.—Body moderately elongate, its depth 32-45 in TL. Head and trunk 2.22-2.30 and head 13.1—14.1 in TL. Snout rounded, 4.4— 5.2 in HL. Dorsal fin origin 1.5—2.0 in HL. Pec- toral fin length less than its base, about 2 in snout. Head and body color yellow to tan, overlain with brown to brownish-black spots, variable in size but about equal to or smaller than eye, the spots becoming more numerous in larger specimens (see Remarks below). Total vertebrae 177-183, MVF 2-76-180 (n = 23). S1zE.—The largest specimen we examined is 780 mm long. ETYMOLOGY.—From the Latin magnificus (splendid). An adjective. DISTRIBUTION. — A common eel, found in shal- low water and occasionally at the surface over sand and coral in the Hawaiian Islands, Johnston Island, the Leeward Islands, and Midway Atoll. REMARKS. — The Hawaiian form of Myrichthys has been most recently placed in the synonymy of M. maculosus (cf. Gosline 1951; McCosker 1979). We herein resurrect M. magnificus on the basis of differences in coloration and vertebral number. Although patterned much like its wide- Myrichthys ocellatus, ANSP 98410, 494 mm (from McCosker et al. 1989, Fig. 387). spread Indo-Pacific congener, the size and num- ber of spots, particularly as seen on large indi- viduals, is different enough to be recognized. For example, our comparison of a 78 cm specimen from Oahu (CAS 34961) with an 84.5 cm spec- imen from Grande Comore (CAS 35204) clearly demonstrates that the Hawaiian form has small- er (diameter less than snout length) and rounder (not ovoid) spots on its head and flanks, and fewer and smaller (not larger than the eye) spots on the chin and throat. The holotype and the paratype of magnifica each have 2-76-179 ver- tebrae. The holotype of Ophichthus stypurus has two predorsal vertebrae, but was apparently damaged and had healed, thereby disallowing a total vertebral count. MATERIAL EXAMINED. —ANSP 1013, 685 mm, the holotype, from the Sandwich Islands (= Hawaiian Islands), and ANSP 1014, 483 mm, a paratype, collected with the holotype. USNM 26817, 619 mm (tail damaged and healed), the holotype of Ophichthus stypurus, from ‘“‘Johnston’s Island.” SU 8584, 3 (305-400 mm), Honolulu, Oahu. CAS 34961, 780 mm, Oahu. CAS 79333, 435 mm (tail damaged and regrown), between French Frigate Shoals and Kauai. CAS 79336, 2 (345-440 mm), NW of Laysan Id., Hawaiian Leeward Is. CAS 79334, 480 mm, NW of Maro Reef, Hawaiian Leeward Is. SIO 68- 497, 3 (375-390 mm), Midway Id. Vertebral data only: ANSP 87609, 2, Hawaii. ANSP 119347, 3, Hawaii. Myrichthys ocellatus (Lesueur) (Fig. 8, Table 2) Muraenophis ocellatus Lesueur 1825:108 (Barbados, neotype ANSP 107304). Pisodonophis oculatus Kaup 1856a:49 (Curacao, holotype un- known). Ophisurus latemaculatus Poey 1867:252 (Cuba, holotype MCZ 27223). Mpyrichthys keckii Silvester 1916:214 (Puerto Rico, holotype PU 3082, not extant). 164 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 9) 7 WR aT PUES ‘Sa. ee a ——— « ail ed .. 2 3 e oe 2. es zs BYERS Danesh Lge Sy Oe Nee Ficure 9. Holotype of Myrichthys pantostigmius, SU 5710, 494 mm (from Jordan and McGregor 1899, Pl. 4). D1AGnosis.—Body moderately elongate, its depth 33-40 in TL. Head and trunk 2.5-2.7 and head 10.6-13.2 in TL. Snout rounded, 5.0-5.9 in HL. Dorsal fin origin 1.5—1.7 in HL. Pectoral fin length less than its base, about 3 1/2 in snout. Head and body pale, overlain with diffuse dark spots with small bright centers. Total vertebrae 164-173, MVF 2-59-168 (n = 24). S1zE.—The largest specimen we examined is 1080 mm long. ETYMOLOGyY.—From the Latin ocellatus (spot- ted as with little eyes). An adjective. DIsTRIBUTION.—An insular species, known from Bermuda, the Florida Keys, throughout the West Indies, and south to Brazil. A common species, taken more often than Myrichthys brev- iceps (except in the Florida Keys where it is rare). Found in shallow grass beds and coral reefs, to depths of 7 m. REMARKS. — Data for the diagnosis of this spe- cies is based primarily on McCosker et al. (1989: 377-379). Randall (1967) reported that this spe- cies (as M. oculatus) fed mainly on crabs, sto- matopods, and other crustaceans. Leiby (1989: 829-830) described and illustrated the lepto- cephalus of M. ocellatus. MATERIAL EXAMINED.—A complete listing of the material that we examined is in McCosker et al. (1989:378-379). Myrichthys pantostigmius Jordan and Mc- Gregor (Fig. 9, Table 2) Myrichthys pantostigmius Jordan and McGregor 1899:2802 (type locality Clarion Island, Mexico, holotype SU 5710). DiaGnosis.—Body moderately elongate, its depth 32-39 in TL. Head and trunk 2.22-2.45 and head 11.2-13.6 in TL. Snout rounded, 4.7- 5.4 in HL. Dorsal fin origin 1.5-1.62 in HL. Pectoral fin length much less than its base, short- er than snout. Head and body tan, overlain with numerous large brown to brownish-black spots along sides and at base of dorsal fin (see Remarks below); chin and throat with smaller spots; dorsal fin margin white, anal fin plain. Total vertebrae 158-168, MVF 2.5-61-162 (n = 18). SizE.—The largest specimen we examined is 494 mm long. ETYMOLOGY.—From the Greek pan (whole, entire) and stigma (spot). An adjective. DIsTRIBUTION. — Islas Revillagigedos (Mexico) and Clipperton Island. Found in shallow water (to 20 m) among rocks and sand. REMARKS. — The Clipperton and Revillagigedo populations of Myrichthys are separable from eastern Pacific Myrichthys on the basis of their higher vertebral numbers. They do not appear to differ in morphometry or coloration (largely because of the variability in mainland M. figri- nus), however, a comparison of similar-sized i1n- dividuals suggests that a difference in spot size and number may exist. For example, we com- pared 48-cm individuals from Clarion (CAS 33509) with those from Sonora, Mexico (CAS 18490) and observed that M. tigrinus has more and larger body and tail spots (= snout length) than does M. pantostigmius (spots =< snout length). We have also observed that the basal body color of Myrichthys from Clipperton and the Revillagigedos often has a greenish tinge in formalin and in alcohol; we are unable to ascribe significance to that fact. MATERIAL EXAMINED.—SU 5710, 494 mm, Clarion Island, Revillagigedo Islands, Mexico, the holotype. CAS 33509, 5 (434-489 mm), Clarion Id. SIO 73-64, 412 mm, Socorro Id. W 58-297, 20 (62-274), Clipperton Id. Myrichthys pardalis (Valenciennes) (Table 2) Ophisurus pardalis Valenciennes 1835: 90 (type locality, Ca- nary Islands, holotype MNHN 2130). D1aGnosis.—Body moderately elongate, its depth 33-43.5 in TL. Head and trunk 2.4-2.6 and head 1.1-1.4 in TL. Snout rounded, 4.5-6.0 in HL. Dorsal fin origin 1.5-1.8 in HL. Pectoral McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 165 SS ee By Le = a . = a SS Soa Saks Se See ‘ aN 2 : ae ilar ae > z : SF =e = ee ae a LEZ i = : Pr Stig eo sé ® = ey LEZ oss Leyes TIO Ficure 10. Myrichthys tigrinus, 617 mm (from Jordan and Evermann 1900, Fig. 166). fin length shorter than base, about 2 in snout. Head and body yellow-orange, overlain above lateral line with a series of brown spots (23 on the holotype) with bright centers, interspaced be- neath lateral line (posterior to the pectoral fin) with smaller, paler spots, also with bright centers. A single brown spot with a pale center above nape. Total vertebrae 151-159, mean = 154.5 (n = 5). S1zE.— The largest specimen described by Blache and Cadenat (1971) is the holotype, 648 mm long. EtyMoLocy.— From the Greek pardalis (leop- ard). A noun in apposition. DISTRIBUTION. — Known from the oceanic is- lands of the eastern Atlantic, including Sao Tomé, the Cape Verdes, and the Gulf of Guinea. REMARKS.—The information and data con- cerning this species are based upon Blache and Cadenat (1971) and McCosker et al. (1989). Blache and Cadenat erred in including Oph- ichthys (Sphagebranchus) guineensis Osorio (1894) in the synonymy of pardalis. It is now recognized as a trans-Atlantic species of Cal- lechelys (Blache et al. 1979; McCosker et al. 1989). MATERIAL ExAMINED.—MNHN 2130, 648 mm, the holo- type, from the Canary Islands. BMNH 1935.7.10.1 and BMNH 1937.2.10.2, Cape Verde Islands. Mpyrichthys tigrinus Girard (Fig. 10, Table 2) Mpyrichthys tigrinus Girard 1859:58 (“Adair Bay, Oregon,” most probably from Bahia Adair, Sonora, Mexico, holotype USNM 8810). Ophichthys quincunciatus Giinther 1870:83 (“Habitat —-?”, holotype BMNH 1864.12.12.32). Ophichthys xysturus Jordan and Gilbert 1882:346 (Mazatlan, Mexico, syntype USNM 28247). Callechelys peninsulae Gilbert 1892:548 (La Paz Bay, Gulf of California, holotype USNM 44297). D1AGNosis.—Body moderately elongate, its depth 26-37 in TL. Head and trunk 2.2-—2.5 and head 10.7-12.4 in TL. Snout rounded, 4.9-5.5 in HL. Dorsal fin origin 1.44-1.58 in HL. Pec- toral fin length much less than its base, shorter than snout. Head and body tan, overlain with numerous large brown spots along sides and at base of dorsal fin. Total vertebrae 148-157, MVF 3-59-153 (n = 53). SizeE.—The largest specimen we examined is 729 mm long. ETYMOLOoGY.—From the Latin ftigrinus (like a tiger). An adjective. DISTRIBUTION. — Baja California to Peru, in- cluding the Galapagos Islands. Generally found over sand and rock bottoms from tidepool depths to 60 m. REMARKS.—The abundance of specimens of this species has convinced us of the variability in its coloration. Like its dark-spotted congeners, larger specimens typically have more and pro- portionately larger spots. The spots of larger specimens (> 50 cm) reach and may extend onto the dorsal fin base. The dorsal fin margin is white; the anal fin is pale. There are approximately three rows of spots on the flanks, with smaller spots on the chin, throat, and belly. Small specimens (< 20 cm) have dorsal and lateral spotting, but few on the chin and ventral surface. 166 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 TaBLeE 3. Status of the nominal species of Myrichthys. Sources are: 1, This study; 2, Blache 1975; 3, Blache and Cad- enat 1971; 4, McCosker 1979; 5, McCosker et al. 1989; 6, Tominaga in litt. 1978; 7, Weber and deBeaufort 1916. As described Current status Source Muraena acuminata Gray 1854 breviceps Myrichthys aki Tanaka 1917 maculosus 1 Ophisurus alternans Quoy & Gai- mard 1824 colubrinus 7 Muraena annulata Ahi 1789 colubrinus 7 Mpyrichthys aspetocheiros McCosker & Rosenblatt, n. sp. aspetocheiros 1 Myrichthys bleekeri Gosline 1951 colubrinus 1 Ophisurus breviceps Richardson 1844 breviceps 5 Muraena colubrina Boddaert 1781 colubrinus Pisoodonophis coronata Kaup 1860 _ breviceps 5 Ophichthys dromicus Giinther 1870 maculosus 2 Chlevastes elaps Fowler 1912 colubrinus 7 Ophichthys elapsoides Castelnau 1875 colubrinus l Muraena fasciata Ahl 1789 colubrinus i Ophisurus fasciatus var. latifasciata Bleeker 1864 colubrinus 1 Ophisurus fasciatus var. oculata Bleeker 1864 colubrinus 1 Ophisurus fasciatus var. semicincta Bleeker 1864 colubrinus Pisodonophis guttulatus Kaup 1856a_breviceps 5 Pisoodonophis guttulatus Kaup 1856b breviceps 5 Myrichthys keckii Silvester 1916 ocellatus 5 Ophisurus latemaculatus Poey 1867 ocellatus 5 Ophisurus longus Poey 1876 breviceps 5 Muraena maculosa Cuvier 1816 maculosus Pisoodonophis magnifica Abbott 1861 magnificus 1 Ophichthus miyamotonis Tanaka 1913 maculosus 6 Ophichthys naja de Vis 1884 colubrinus 7 Muraenophis ocellatus Lesueur 1825 ocellatus 5 Pisodonophis oculatus Kaup 1856a _ocellatus 5 Pisoodonophis oculatus Kaup 1856b__ ocellatus 5) Myrichthys pantostigmius Jordan & McGregor 1899 Ophisurus pardalis Valenciennes pantostigmius 1 1835 pardalis 3 Callechelys peninsulae Gilbert 1892 tigrinus 1 Ophichthys pisavarius Poey 1880 breviceps 5 Ophichthys quincunciatus Ginther 1870 tigrinus 1 Myrichthys rupestris Snyder 1911 maculosus 1 Ophichthus stypurus Smith & Swain 1882 magnificus 1,4 Muraena tigrina Riippell 1830 maculosus 1 Leptocephalus undulatus Str6mman 1896 breviceps 5 Ophichthys xysturus tigrinus 1 A large series (CAS 16626) of this species col- lected on 4 June 1950 at the entrance of San Carlos Bay, Sonora, Mexico, included numerous sexually mature individuals. There is broad overlap in the size of males and females and no apparent difference in coloration associated with sex. The ripe females ranged between 385-695 mm and the males 515-655 mm. The type of Myrichthys tigrinus was described as being from “‘Adair Bay, Oregon.”’ Written on the original orange label accompanying the type is “Ore. ? C. P. Stone.”’ An additional and newer label in the same jar states ““Ophichthys tigrinus (TYPE?) Oregon ? Capt. C. P. Stone.” In that no subsequent specimens have been discovered north of the lower Baja Peninsula, and there is no Adair Bay in Oregon, we suggest that the spec- imen more likely came from Bahia Adair, So- nora, Mexico. Harry (1948) cited a specimen (CAS 20236, 508 mm) that “bore no other data than ‘San Francisco Bay, Calif. Presented by a fisherman, June 1931’ ’’. We discount that record and suggest that it probably refers to Bahia San Francisco, Baja California. Further evidence that it is from a more tropical locale is provided by its reproductive state; it is a female with well- developed ova. The existence of Muraena tigrina Rippell (1830), a junior synonym of Myrichthys macu- losus (Cuvier 1817), creates a potential hom- onymy with Myrichthys tigrinus Girard. On that basis, recent authors (McCosker 1977, 1979; Thomson et al. 1979) have adopted the name Mpyrichthys xysturus (Jordan and Gilbert 1882) for Girard’s species. Subsequently, the late W. I. Follett advised us that such action is unnecessary if the binomen Myrichthys tigrinus (Rtippell) had not been used and that Myrichthys tigrinus Gi- rard had not been replaced at any time before 1961. We presume that to be the case. Follett therefore suggested that Girard’s binomen is but a potential junior homonym and that it has not been permanently rejected in accordance with Article 59 of the Rules of Zoological Nomencla- ture. Storey (1939) provided an extensive treatment of Myrichthys tigrinus and its synonymy. As well, she expanded upon Gilbert and Starks’ (1904) observation that larger specimens have more nu- merous and more granular dentition, and ob- served, as did Gilbert and Starks (1904) and Meek and Hildebrand (1923) and as we have, that there McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 167 is considerable variation in spotting pattern with size. Like Myrichthys maculosus, smaller M. ti- grinus have fewer and smaller (proportionately) spots than do larger individuals. We include Ophichthys quincunciatus in the synonymy of Myrichthys tigrinus. The type spec- imen (BMNH 1864.12.12.32, total length 655 mm) is in adequate condition, has 2-61-148 ver- tebrae, and is recognizable as a specimen of M. tigrinus. Jordan and Davis (1891:618) included ““2?Ophisurus breviceps Richardson . . . (locality unknown)” in their treatment of Myrichthys ti- grinus. McCosker et al. (1989) examined the ho- lotype (BMNH uncat., total length 880 mm) and recognized it to be the senior synonym of the Atlantic species formerly known as Myrichthys acuminatus. We have examined numerous examples of spotted eastern Pacific Myrichthys and have dif- ficulty in separating them, other than by verte- bral counts, from their Pacific congener, M. mac- ulosus. McCosker (1979:63-64) anguished over this apodal anomaly and was “... resigned to recognize these populations at a subspecific level.” After subsequent examination of more than 11,000 vertebral centra, we have concluded that the eastern Pacific populations of Myrichthys dif- fer from those of the central and western Pacific and Indian oceans, and may be recognized as M. tigrinus and M. pantostigmius. MATERIAL EXAMINED. — USNM 8810, 640 mm, a female with eggs, “Adair Bay, Oregon,” the holotype of Myrichthys tigri- nus. USNM 28247, 291 mm, Mazatlan, a syntype of Ophi- chthys xysturus. USNM 44297, 264 mm, La Paz, Mexico, the holotype of Callechelys peninsulae. Mexico—CAS 26030, 601 mm; and SU 33922, 575 mm, San Felipe. CAS 79429, 399 mm, Bahia de Los Angeles. CAS 79411, 2 (201-268 mm), Bahia Concepcion. CAS 79406, 2(176— 187 mm), Bahia Santa Inez. CAS 79405, 13 (142-265 mm), Pta. Pordeones. CAS 79407, 3 (201-230 mm); and CAS 79404, 75 mm, Pta. de las Cuevas. SU 39792, 325 mm; and CAS 79414, 604.5 mm, Isla Partida. SIO 61-277, 2 (153-161 mm), Isla Espiritu Santo. CAS 20236, 508 mm, “San Francisco Bay,” presumably Bahia San Francisco, Baja California. CAS 79412, 215 mm, Tetas de Cabra. SU 17944, 108 mm, Isla Ceralbo. CAS 79409, 520 mm, La Paz. SIO 65-182, 8 (581-729), Al- mejas Bay. CAS 20240, 246 mm, Bocachibompo. CAS 79408, 616 mm, Sonora. CAS 794332, 190 mm; CAS 79431, 20 (110- 140 mm); and CAS 16626, 29 (185-695 mm), San Carlos. CAS 18490, 2 (240-305 mm); and CAS 79413, 635 mm, Guaymas. SU 3238, 2 (600-682 mm), Mazatlan. SU 19045, 75 mm, Islas Tres Marias, Cleopha Id. NicaracuA—SU 46974, 435 mm, Cornito. Costa Rica—SU 46971, 4 (111-147 mm), Uvita Bay. CAS 79401, 146 mm; SU 46402, 123 mm; SU 57312, 69 mm; and SU 46975, 2 (232-312 mm), Port Parker. PANAMA— SU 46970, 20 (63-406); SU 57314, 86 mm; and CAS 79402, 71 mm, Bahia Honda. CAS 79403, 97 mm, Ft. Amador. GALA- PAGOS IsLANDs: CAS 23736, 474 mm; and CAS 3962, 2 (295- 380 mm), Isla Santa Cruz. SU 37372, 298 mm, Albemarle. CoLomBIA—SU 46972, 241 mm; SU 46973, 3 (61-86 mm); and SU 68868, 71 mm, Gorgona Island. PERRU—IMARPE S54al: 282, 683 mm, Isla Lobos de Tierra. Also examined were 518 specimens in 81 lots from the Gulf of California, retained in the SIO Fish Collection. ACKNOWLEDGMENTS Many individuals have generously assisted with this study, however we especially thank Eugenia Bohlke (ANSP) for sharing her vertebral and oth- er data with us. As well, we thank the following: David Catania (CAS) for the preparation of ra- diographs; Elizabeth Parker for the preparation of Figures 1-2; the late W. I. Follett (CAS) for nomenclatural advice; the Janss Foundation for our participation in the SEARCHER cruises; Ira Rubinoff for aiding us while studying eels in Pan- ama; Susan Jewett for assistance with USNM illustrations; Bruce Halstead for permission to use Figure 6; and Carl J. Ferraris, Jr., for reading an early draft of this manuscript. Numerous cu- rators have allowed us to examine specimens in their care, including: E. B6hlke (ANSP); J. Pax- ton and D. Hoese (Australian Museum, Sydney); O. Crimmin (BMNH); J. E. Randall (BPBM); W. Eschmeyer and T. Iwamoto (CAS); W. D. An- derson (GMBL); L. Fishelson (Hebrew Univer- sity); K. Hartel (MCZ); M. Bauchot and J. Blache (MNHN); R. McKay (QM); the late M. M. Smith (Rhodes University); H. J. Walker (SIO); B. W. Walker (UCLA); W. Bussing (UCR); W. A. Gos- line (then of UH); C. R. Robins (UMML); the staff of the National Museum (USNM); T. Pietsch, (UW); and Y. Tominaga (ZUMT). LITERATURE CITED Assott, C. C. 1861. Description of new species of apodal fishes in the museum. . . . Proc. Acad. Nat. Sci. Phil. 12(1860): 475-479. Aut, J. N. 1789. Dissertatio de Muraena et Ophichtho. Dis- sert. Acad. Upsaliae habitae sub praesidio C. P. Thunberg 3(1):1-12. Biacu_, J. 1975. Description et révision de la position sys- tématique d’Ophichthys dromicus Gthr., 1870 (Pisces, An- guilliformi, Ophichthidae). Bull. Mus. Nat. Hist. Natur., 3° sér. no. 307, Zoologie 217:799-803. BLAcHE, J. AND J. CADENAT. 1971. Contribution 4a la con- naissance des poissons anguiliformes de la céte occidentale 168 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 8 d’Afrique. Dixiéme note: les genres Myrichthys, Bascani- chthys et Callechelys (Fam. des Ophichthidae). Bull. del’Inst. Fond. d’Afrique Noire, sér. A 33(1):158-201. BLAcHE, J., M. L. BAUCHOT, AND L. SALDANHA. 1979. Dés- ignation d’un neotype pour Callechelys guineensis (Osorio, 1894), et validation de l’espéce (Pisces, Anguilliformes, Ophichthidae). Cybium, sér. 3 3(7):97-98. BLeeker, P. 1864. Atlas ichthyologique des Indies Oriental Neerlandaises, publié sous les auspices du gouvernement colonial. Tome 4, Murenes, Synbranchies, Leptocéphales. Leiden: 132 pp. Boppaert, P. 1781. Beschreibung zweier merkwiirdiger Fische (Sparus palpebratus und Muraena colubrina). Neue Nord. Beitr. 1781(2):55-57. CALDWELL, G. S. AND R. W. RusinorF. 1983. Avoidance of venomous sea snakes by naive herons and egrets. Auk 100: 195-198. CASTELNAU, F. De. 1875. Researches on the fishes of Aus- tralia. Phil. Centen. Exhib., 1876 (Melbourne, 1875): Official Record. Intercolonial Exhibition Essays No. 2:1-52. Cort, H. B. 1940. Adaptive Coloration in Animals. Methu- en, London, 508 pp. Cuvier, G. 1816. Le Régne Animal distribué d’aprés son organization . . . ’anatomie comparée. Les reptiles, les pois- sons, les molusques et les annélides. Edit. 1. 532 pp. De Vis, C. W. 1884. Fishes from South Sea Islands. Proc. Linn. Soc. N. S. W. 8(4):444-457. Fow.Ler,H.w. 1912. Descriptions of nine new eels, with notes on other species. Proc. Acad. Nat. Sci. Phil. 64:8-33. Gans, C. 1961. Mimicry in procryptically colored snakes of the genus Dasypeltis. Evol. 15(1):72-91. GILBERT, C. H. 1892. Descriptions of thirty-four new species of fishes collected in 1888 and 1889, principally among the Santa Barbara Islands and in the Gulf of California. Proc. U. S. Natl. Mus. 14(880):539-566. GILBERT, C.H. ANDE.C. Starks. 1904. The fishes of Panama Bay. Mem. Calif. Acad. Sci., Vol. 4:1-304. GrrRARD, C. 1859. Ichthyological notices. Proc. Acad. Nat. Sci. Phil. 11:56-68. Gos.ine, W. A. 1951. The osteology and classification of the ophichthid eels of the Hawaiian Islands. Pacific Sci. 5(4): 298-320. Gray, J. E. 1854. Catalogue of fish collected and described by Laurence Theodore Gronow, now in the British Museum. British Museum, London: vii + 196 pp. GREENE, E., L. J. ORSAK, AND D. W. WHITMAN. 1987. A tephritid fly mimics the territorial displays of its jumping spider predators. Science 236:310-312. GREENE, H. W. AND R. W. McDiarmip. 1981. Coral snake mimicry: does it occur? Science 213(4513):1207-1212. GUnTHER, A. 1870. Catalogue of the Fishes in the British Museum. Vol. 8. Catalogue of the Physostomi, containing the families Gymnotidae . . . in the British Museum. British Museum, London: xxv + 549 pp. HAtsTEAD, B. W. 1970. Poisonous and venomous marine animals of the world. Vol. III, Vertebrates. U. S. Govt. Print- ing Office. Washington, D.C. 1006 pp. Harry, R. R. 1948. New records for the fish, Myrichthys tigrinus, a snake eel of the eastern tropical Pacific, with a relocation of the type locality. Copeia 1948(2):145-146. HEATWOLE, H. 1975. Predation on Sea Snakes. Pp. 233-249 in W. A. Dunson, ed., The Biology of Sea Snakes. Univ. Park Press, Baltimore. Herre, A. W.C. T. 1923. A review of the eels of the Phil- ippine Archipelago. Philippine J. Sci., 23(2):123-236. Hopson, E. S. 1968. Predatory behavior of some shore fishes in the Gulf of California. U. S. Dept. Int., Fish and Wildl. Serv. Res. Report 73. 92 pp. JORDAN, D. S. AND B. M. Davis. 1891 (1892). A preliminary review of the apodal fishes or eels inhabiting the waters of America and Europe. Report of the Commissioner, United States Commission of Fish and Fisheries for 1988, 16(9): 581-677. JorDAN, D. S. AND B. W. EvERMANN. 1900. The fishes of North and Middle America: a descriptive catalogue of the species of fish-like vertebrates found in the waters of North America, north of the Isthmus of Panama. Part IV. Bull. U. S. Natl. Mus. No. 47: 3137-3313, Pls. 1-392. Jorpan, D. S. AND C. H. GitpertT. 1882. Descriptions of thirty-three new species of fishes from Mazatlan, Mexico. Proc. U.S. Natl. Mus. 4(237):338-365. Jorpan, D. S. AND R. C. McGreGcor. 1899. List of fishes collected at the Revillagigedo Archipelago and neighboring islands. Rep. U. S. Fish Comm. v. 24 (1898):271-284. JorpDan, D. S. AND J. O. Snyper. 1901. A review of the apodal fishes or eels of Japan, with descriptions of nineteen new species. Proc. U. S. Natl. Mus. 23 (1239):837-890. Kaup, J. J. 1856a. Uebersicht der Aale. Archiv fiir Nautr. 22(1):41-77. 1856b. Catalogue of apodal fish, in the collection of the British Museum. British Museum, London: viii + 163 pp. . 1860. Neue aalaehnliche Fische des Hamburger Mu- seums. Abhandlungen aus dem Gebiete der Naturwissen- schaften Herausgegeben von dem Naturwissenschaftlichen Verein in Hamburg 4(2):1-29 + 4 pp. Lay, G. T. AND E. T. BENNETT. 1839. Fishes. Pp. 41-75 in The zoology of Captain Beechey’s voyage ... in 1825-28. London. Lerpy, M. M. 1989. Family Ophichthidae. Pp. 764-897, vol. 2, in E. B. Bohlke, ed., Fishes of the Western North Atlantic, part Nine. Yale Univ., New Haven. Lesugur, C. A. 1825. Descriptions of four new species of Muraenophis. J. Acad. Nat. Sci. Phil. 5(1):107-109. Masupa, H., K. AMAOKA, C. ARAGA, T. UYENO, AND T. YOSH- INO. 1984. The fishes of the Japanese Archipelago. Tokai Univ. Press. 437 pp. Matuer, M. H. AND B. D. RorrBerG. 1987. A sheep in wolfs clothing: tephritid flies mimic spider predators. Science 236: 308-310. McCosker, J. E. 1975. Feeding behavior of Indo-Australian Hydrophiidae. Pp. 217-232 in W. A. Dunson, ed., The Bi- ology of Sea Snakes. Univ. Park Press, Baltimore. . 1977. The osteology, classification, and relationships of the eel family Ophichthidae. Proc. Calif. Acad. Sci. 41(1): 1-123. 1979. The snake eels (Pisces, Ophichthidae) of the Hawaiian Islands, with the description of two new species. Proc. Calif. Acad. Sci. 42(4):57-67. McCosker, J. E., E. B. BOHLKE, AND J. E. BOHLKE. 1989. Family Ophichthidae. Pp. 254-412, vol. 1, in E. B. Bohlke, ed., Fishes of the Western North Atlantic, part Nine. Yale Univ., New Haven. McDoweE t, S. B. 1972. The genera of sea-snakes of the Hydrophis group (Serpentes: Elapidae). Trans. Zool. Soc. London 32:189-247. McCOSKER AND ROSENBLATT: REVISION OF SNAKE EEL GENUS MYRICHTHYS 169 MEEK, S. E. AND S. F. HILDEBRAND. 1923. The marine fishes of Panama. Field Mus. Nat. Hist. Pub. 215, Zool. Ser. 15(1): x1 + 330 pp. Osorio, B. 1894. Estudos ichthyologicos acerca da fauna dos dominios portuguezes da Africa. 3° nota: Peixes maritimos das ilhas de S. Thomé, do Principe e ilheo das Rolas. J. Sci. Math., Phisicas e Nat. da Acad. Real das Sci. de Lisboa, ser. 2, 3(9):173-182. Poey, F. 1867. Peces Cubanos. Monografia de las morenas Cubanas. Repert. Fisico-Nat. de la Isla de Cuba, Habana, 2(10-12):245-264. . 1876. Enumeratio Piscium Cubensium. Parte secun- do. Anal. de la Soc. Espafiola de Hist. Nat. 5:131-218. 1880. Revisio piscium cubensium. Anal. de la Soc. Espanola de Hist. Nat. 9:243-260. Quoy, J. R. C. AND P. GAIMARD. 1824. Remarques sur qu- elques poissons de mer. Pp. 183-401 in L. de Freycinet. Voyage autour du monde... et “La Physicienne,” pendant les années. . . 1820. Vol. 3. Zoologie. Pillet Ainé, Paris. 712 pp. RANDALL, J. E. 1967. Food habits of reef fishes of the West Indies. Stud. Trop. Oceanog. 5:665-847. RICHARDSON, J. 1844. Ichthyology. Pp. 51-150 in R. B. Hinds, ed., The zoology of the voyage of H. M.S. Sulphur. . . 1836- 42. Smith, elder and Col, London 1: iv + 150 pp. RUppeLL, E. 1830. Atlas zu der Reise im nérdlichen Africa. Fische des Rothen Meeres. Frankfurt-am-Main. Part 3:95- 141. ScHuttTz, L. P. 1943. Fishes of the Phoenix and Samoan Islands collected in 1939 during the expedition of the U. S. S. “Bushnell’’. Bull. U. S. Natl. Mus. no. 180. 316 pp. ScHuttz, L. P. AND COLLABORATORS. 1953. Fishes of the Marshall and Marianas Islands. Families from Asymme- trontidae through Siganidae. Bull. U. S. Natl. Mus. no. 202 (1). 685 pp. SILvEsTER, C. F. 1916. Fishes new to the fauna of Porto Rico. Carnegie Inst. Wash., Year Book No. 14 (1915):214-217. SmitH, J.L. B. 1962. Sand-dwelling eels of the western Indian Ocean and the Red Sea. Rhodes Univ. Ichthyol. Bull. 24: 47-466. SmitH, M. 1926. Monograph of the Sea-Snakes (Hydrophi- idae). British Mus. Nat. Hist., London. 130 pp. Situ, R. AND J. Swain. 1882. Notes ona collection of fishes from Johnston’s Island including descriptions of five new species. Proc. U. S. Natl. Mus. 5:119-143. SnypDer, J.O. 1911. Descriptions of new genera and species of fishes from Japan and the Riu Kiu Islands. Proc. U. S. Natl. Mus. 40(1836):525-549. Starck, W. A. AND W. P. Davis. 1966. Night habits of fishes of Alligator reef, Florida. Ichthyologica, The Aquarium Journal 38(4):313-356. Storey, M.H. 1939. Contributions toward a revision of the ophichthyid eels. 1. The genera Callechelys and Bascani- chthys, with descriptions of new species and notes on Myr- ichthys. Stanford Ichthyol. Bull. 1(3):61-84. StTROMMAN, P. H. 1896. Leptocephalids in the University Zoological Museum at Upsala. Almquist and Wicksell, Up- sala: vi + 53 pp. TANAKA, S. 1913. Figures and descriptions of the fishes of Japan, including Riukiu Islands, Bonin Islands, Formosa, Kurile Islands, Korea and southern Sakhalin. Vol. 11:187- 198. 1917. Ibid. Vol. 26:455-474. TuHomson, D. A., L. T. FINDLEY AND A. N. KerstitcH. 1979. Reef fishes of the Sea of Cortez. J. Wiley & Sons, New York. 302 pp. THRESHER, R. E. 1984. Reproduction in reef fishes. T. F. H. Pub., Inc., Neptune City, New Jersey. 399 pp. VALENCIENNES, A. 1835. Ichthyologie des Isles Canaries. Pp. 1-109 in P. B. Webb and S. Berthelot. Histoire Naturelle des Iles Canaries 2(2). WeseR, M. AND L. F. DE BEAUFORT. 1916. Fishes of the Indo- Australian Archipelago. Vol. 3. Ostarophysi: II Cyprinoidea, Apodes, Synbranchi. E. J. Brill, Leiden. 455 pp. Wuit.ey, G. P. 1951. New fish names and records. Proc. Royal Zool. Soc. NSW 1949-50:61-86. 1964. Presidential address. A survey of Australian ichthyology. Proc. Linn. Soc. NSW 89(1):11-127. © CALIFORNIA ACADEMY OF SCIENCES, 1993 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 9, pp. 171-183, 10 figs. August 16, 1994 NEW RECORDS OF FLABELLINIDAE (OPISTHOBRANCHIA: AEOLIDACEA) FROM THE TROPICAL AMERICAS, WITH DESCRIPTIONS OF TWO NEW SPECIES... 32°01 Leborstony By s Hole Ocean : ~2NOGtaphic Institut Libre aPhie Institution > {\ Terrence M. Gosliner © | IG 22 1994 California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 Woscde . Voods Hole, MA Ucoay ABsTRACT: Two new species of Flabellina are described from the tropical Americas. Flabellina vansyoci sp. nov. is widely distributed from Bahia Magdalena on the Pacific coast of Baja California to the Gulf of Chiriqui. Panama. It is compared to its close relative, F. pedata (Montagu, 1815), which is first recorded here from the Azores Islands. Flabellina hamanni sp. nov. has been found from the Bahamas to Venezuela in the Caribbean and is most closely related to F. marcusorum Gosliner and Kuzirian, 1990. Flabellina bertschi Gosliner and Kuzirian, 1990, previously recorded only from the Gulf of California, is also found along the Pacific coast of Panama. Flabellina marcusorum, was known previously from the Gulf of California and the Caribbean and is here reported for the first time from Panama. Flabellina stohleri Bertsch and Ferreira, 1974, is considered a junior synonym of F. telja Marcus and Marcus, 1967. The range of F. telja is extended southward to the Islas Revillagigedos, Panama, and the Galapagos. Received October 13, 1993. Accepted December 3, 1993. INTRODUCTION Recently, Gosliner and Kuzirian (1990) re- viewed the Flabellinidae from the tropical coasts of the Americas. Both this work and that of Gos- liner and Willan (1991) examined the phyloge- netic relationships within Flabellina. Since then, additional collections of material from both the eastern Pacific and Caribbean have yielded spec- imens of two undescribed species, as well as new records for some previously named taxa. This paper describes the morphology of these taxa and adds new biogeographical data regarding tropical American Flabellinidae. DESCRIPTIONS Flabellina bertschi Gosliner and Kuzirian, 1990 MATERIAL. —CASIZ 088179, on Eudendrium sp., 22 m depth, Hill Rocks (7°17'20” N, 81°40'65” W), SW of Punta Anegada, Isla Coiba, Gulf of Chiriqui, Pacific coast of Panama, 18 April, 1993. DIsTRIBUTION. —Flabellina bertschi is known within the Gulf of California from Puerto Pen- asco south to Isla San Diego (Gosliner and Ku- zirian, 1990) and from the Pacific coast of Pan- ama. EXTERNAL MorPHOLOGY. — Living animal thin, elongate and 4 mm in length. Body color trans- lucent white with overlying opaque white pig- ment on distal two thirds of the rhinophores and oral tentacles. Each translucent ceras with central red digestive diverticulum, terminating in a large, opaque white cnidosac. Rhinophores smooth. Oral tentacles approximately as long as rhino- [171] Wii phores. Cerata of uniform diameter throughout most of their length. Many cerata damaged and partially regenerating. Cerata arranged in dis- crete clusters. Anteriormost cluster on either side containing four rows of cerata with 1-3 cerata per row. Posterior digestive branch with four clusters of cerata, each with 1-3 rows of cerata. Each row with 1-3 cerata. Genital apertures lo- cated on right side of body, ventral to second and third ceratal rows. Pleuroproctic anus situ- ated within interhepatic space, below notum, im- mediately anterior to anteriormost cerata of right posterior digestive branch. Anterior border of foot with elongate corners. Discussion. — There 1s no doubt that the pres- ent specimen is Flabellina bertschi. It agrees with the original description of this species, but is slightly smaller than any of the type specimens. Nevertheless, it is sexually mature with well-de- veloped gonads and reproductive organs visible through the transparent body wall. Flabellina marcusorum Gosliner and Kuzirian, 1990 MATERIAL. —CASIZ 088212, two specimens, 15 m depth, Islas Secas, Gulf of Chiriqui, Pacific coast of Panama, 21 April 1993, T. M. Gosliner. CASIZ 088180, one specimen, 22 m depth, Spanish Waters, Curacao, April 1990, Jeff Hamann. DISTRIBUTION. —Flabellina marcusorum has been collected from the Atlantic coast of Curacao (present study), Venezuela (Jeff Hamann, pers. comm.), and Brazil (Marcus and Marcus 1961). Flabellina marcusorum is known from the Pa- cific coast of Mexico, from Isla Cedros south to Sayulita, Nayarit, and from San Agustin, Sonora, south to Los Islotes, north of La Paz, within the Gulf of California (Gosliner and Kuzirian 1990). The present specimens extend the range south to Panama in the Pacific and north to Venezuela and Curacao in the Caribbean. EXTERNAL MorpHoLoGy.—Living animals 4 and 10 mm in length. General body color trans- lucent rose pink. Oral tentacles, foot corners, rhinophores rose basally with purple pigment in their middle and opaque white present on apical portions. Oral tentacles with wide band of pur- ple. Cerata with only narrow purple ring. Rhino- phores elongate, bearing numerous elongate pa- pillae on their posterior face. Oral tentacles thin and elongate, longer than rhinophores. Cerata arranged in distinct groups. Anterior, precardiac cluster with 2-4 rows of cerata on either side of animal with 2-4 cerata per row. Postcardiac cer- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 ata arranged in five paired clusters. Anterior 2— 4 postcardiac groups containing complete arch of cerata while posterior 1-3 groups with only single row of cerata. Gonopore situated on right side of body, ventral to third and fourth ceratal rows. Pleuroproctic anus situated at posterior end of interhepatic space. Foot with elongate, ten- tacular corners. Discussion. — The external morphology of the present material agrees entirely with the original description. Flabellina telja Marcus and Marcus, 1967 (Figs. 1, 2) Flabellina telja Marcus and Marcus, 1967:223 (in part), Figs. 76-82. Flabellina stohleri Bertsch and Ferreira, 1974:347, Figs. 3, 16- 21. syn. nov. MaTERIAL EXAMINED.—CASIZ 019047, five paratypes of Flabellina stohleri, 2-3 m depth, lower end of bay, Bahia de San Francisquito, S of Punta San Francisquito, Baja California, Mexico, 24 August 1971, H. Bertsch, E. Coan, R. Holiday. CASIZ 088160, one specimen, 16 m depth, N side Isla Mon- tuosa, Gulf of Chiriqui, Pacific coast of Panama, 15 April 1993, T. Gosliner. CASIZ 088154, one specimen, 13 m depth, N side of Isla Montuosa, Gulf of Chiriqui, Pacific coast of Pan- ama, 15 April 1993, T. Gosliner. CASIZ 079164, 3 m depth, W end Gardner Bay, Isla Espafiola, Islas Galapagos, 7 Septem- ber 1991, T. Gosliner. CASIZ 071303, one specimen, 5 m depth, S point, Academy Bay, Isla Socorro, Islas Revillagige- dos, Mexico. DISTRIBUTION. — Flabellina telja was originally described from Bahia Cholla, Puerto Pefiasco, in the northern extreme of the Gulf of California and subsequently recorded from Bahia San Fran- cisquito (Bertsch and Ferreira 1974, as F. stoh- leri) on the western side of the Gulf of California. Specimens examined in this study extend the range of this species to the Islas Revillagigedos, Gulf of Chiriqui, and the Galapagos Islands. EXTERNAL MorpuHococy.— Living animals 3- 9 mm in length. General body color translucent white, sometimes with bluish purple cast. Rhino- phores translucent whitish with opaque white apex. Basal portion of oral tentacles same color as body. Distal third of oral tentacles opaque white. Entire surface of body, including basal portion of cerata, frosted with dense opaque white spots. Through translucent surface of cerata, salmon to red pigment of digestive diverticula visible. Apex of cerata opaque white. Rhino- phores bulbous and densely perfoliate in their distal halves. Oral tentacles thin and elongate. Cerata are arranged in distinct pedunculate clus- ters with 3-4 ceratal rows in the anteriormost GOSLINER: TROPICAL AMERICAN FLABELLINIDAE 173 Ficure 1. Flabellina stohleri. Scanning electron micrographs of paratype. (A) Half row of radular teeth, scale = 60 um. (B) Ventral view of rachidian tooth, scale = 30 um. cluster. Posterior peduncles with only a single row of cerata. Gonopore situated on right side of body, ventral to second and third ceratal rows of precardiac cluster. Pleuroproctic anus situated within interhepatic space. Anterior corners of foot tentacular. INTERNAL MoORPHOLOGY.— Buccal mass short and muscular. Jaws thin and coriaceous. Several rows of denticles present along masticatory bor- der. Radular formula 17-31 x 1.1.1., in two specimens examined (CASIZ 079164, and the paratype of F. stohleri, CASIZ 019047, respec- tively). Rachidian teeth (Fig. 1A, B) simply arched with 6-11 triangular denticles on either side of elongate central cusp. Central cusp slightly wider than adjacent denticles and depressed ventrally from their level. Lateral teeth (Fig. 1 A) triangular in shape with broad base extending towards outer edge. Laterals with single prominent, acutely pointed apex. Inner margin of tooth with series of 0-12 irregular denticles. Arrangement of reproductive organs (Fig. 2) essentially the same as described by Marcus and Marcus (1967). Narrow preampullary duct wid- ens into ampulla consisting of two convolutions. Postampullary duct again narrowing and divid- ing into oviduct and vas deferens. After short distance, oviduct joining short stalked recepta- culum seminis. Oviduct entering female gland mass in region of albumen gland and continuing towards genital apertures. Large bursa copulatrix FicureE 2. Flabellina telja. Reproductive system, al = al- bumen gland; am = ampulla; be = bursa copulatrix; me = membrane gland; mu = mucous gland; pe = penis; pr = pros- tate; rs = receptaculum seminis, scale = 1.0 mm. 174 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 Ficure 3. Living animals. (A) Flabellina vansyoci, photograph of holotype. (B) Flabellina hamanni, photograph of specimen from Turks and Caicos Islands. adjacent to female gonopore. Bursa exiting via uniform diameter. Vas deferens entering bulbous moderately long duct. Female gland mass con- _ penis. Penis ornamented with many small, glan- sisting of membrane, albumen, and mucous dular papillae. glands. Vas deferens moderately short and of Discussion.—There has been considerable GOSLINER: TROPICAL AMERICAN FLABELLINIDAE C 175 Ficure 4. Flabellina vansyoci. (A) Ceras, scale = 1.0 mm. (B) Lateral view of body showing ceratal arrangement, a = anus; ga = genital aperture; r = rhinophore, scale = 2.0 mm. (C) Ventral view of foot, scale = 2.0 mm. (D) Reproductive system, am = ampulla; fgm = female gland mass, pe = penis; rs = receptaculum seminis; v = vagina, scale = 0.75 mm. confusion surrounding the systematics of several species of Flabellina occurring in the eastern Pa- cific. Gosliner and Kuzirian (1990) noted that Marcus and Marcus’s (1967) description of Fla- bellina telja also included specimens that were subsequently described as F. bertschi Gosliner and Kuzirian, 1990. Gosliner and Kuzirian also indicated that specimens identified as F. telja by Ferreira and Bertsch (1972) and Kerstitch (1989) were F. marcusorum Gosliner and Kuzirian, 1990. Gosliner and Willan (1991), in comparing Flabellina rubropurpurata Gosliner and Willan, 1991 to other taxa, indicated that Flabellina telja and Flabellina stohleri are likely synonymous with each other. Bertsch and Ferreira (1974) stated that F. stohleri differed from F. telja in its color, number of radular rows, and shape and dentic- ulation of the radular teeth. Both taxa were de- scribed as having a translucent body densely cov- ered by opaque white spots. The orange body color described by Bertsch and Ferreira is due to the color of the viscera rather than the epidermis. The bluish purple color described by Marcus and Marcus is evident in some specimens and absent in others, even from the same locality. The mor- phology of specimens of Flabellina telja studied 176 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 Ficure 5. Flabellina vansyoci. Scanning electron micro- graph of masticatory border of paratype, scale = 43 um. here from Islas Revillagigedos, Panama, and the Galapagos is entirely consistent with morphol- ogy described by Marcus and Marcus (1967) and the holotype of F. telia (USNM 678418) exam- ined at the Smithsonian. The Galapagos speci- men examined here had fewer radular rows than the type material, but was of smaller body size. The morphology of a paratype of F. stohleri was examined in this study (Fig. 1A, B). Its number of radular rows (31) is more similar to the 30 rows described for F. telja than the 14 rows de- scribed for F. stohleri. Also, the number of den- ticles and shape of teeth are intermediate to the material examined by Marcus and Marcus (1967) and Bertsch and Ferreira (1974). The reproduc- tive system of F. stohleri was not originally de- scribed by Bertsch and Ferreira. Examination of two paratypes of F. stohleri (present study) in- dicates that these specimens have a papillate pen- ial papilla as described for F. te/ja. There remains no basis for separation of the taxa, and F. stohleri is here regarded as a junior synonym of F. telja. Flabellina vansyoci sp. nov. (Figs. 3A, 4-6) Flabellina sp. 2 Behrens, 1991:86, Fig. 173. Type MaTERIAL.— Holotype, California Academy of Sci- ences, San Francisco, CASIZ 088274, S end of Isla Magdalena, Magdalena Bay, Pacific coast Baja California, 16 January, 1984, R. Van Syoc. Paratypes, CASIZ 088081, two specimens, one dissected, 10 m depth, anchorage Islas Ladrones, Gulf of Chi- riqui, Pacific coast of Panama, 13 April 1993, T. Gosliner. DISTRIBUTION. — Flabellina vansyoci has been collected from Magdalena Bay on the Pacific coast of Baja California and from Islas Ladrones along the Pacific coast of Panama. EtyMoLocy.—Flabellina vansyoci (pro- nounced “‘van psyche’’) is named for my friend and colleague, Robert Van Syoc, of the California Academy of Sciences, who collected the first specimen of this species. EXTERNAL MORPHOLOGY. — Living animals (Fig. 3A) 15-30 mm in length. General body color translucent rose pink to purple. Cerata translucent purplish with red digestive gland vis- ible through surface of cerata. Distal half of each ceras bearing congested spots of opaque white. Whitish cnidosac visible at apex of ceras. Rhin- ophores and oral tentacles same color as body. Rhinophores elongate (approximately 4 mm in length) and conical. Rhinophores each bearing indistinct rugose markings along their entire length. Oral tentacles thin and elongate, tapering to a rounded apex. Oral tentacles slightly longer than rhinophores. Cerata (Fig. 4A) cylindrical and elongate with rounded apex and elongate cnidosac. Thick core of digestive gland filling most of diameter of ceras. Cerata arranged in distinct clusters of rows (Fig. 4B). Only anterior cluster slightly elevated from notum. Anterior, precardiac cluster with five rows of cerata on either side of animal with 1-5 cerata per row. Postcardiac cerata arranged in 6-8 clusters per side of body. Generally, each group containing 1-3 tightly packed rows with 1-5 cerata per row. Only posteriormost 1-2 clusters consist of single ceratal row. Gonopore situated on right side of body, ventral to second and third ceratal rows of precardiac cluster. Pleuroproctic anus situated near middle of interhepatic space, anterior to first postcardiac ceratal row. Nephroproct immedi- ately anterodorsal to anus. Foot (Fig. 4C) grooved anteriorly and possessing elongate, tentacular foot corners. Foot tapering gradually to narrow tail posteriorly. GOSLINER: TROPICAL AMERICAN FLABELLINIDAE Ficure 6. Flabellina vansyoci. Scanning electron micrographs of paratype. (A) Dorsal view of radula, scale = 60 um. (B) Rachidian teeth, scale = 20 um. (C) Lateral view of radula, scale = 25 um. (D) Lateral tooth, scale = 15 um. INTERNAL MorPHOLOGY.— Buccal mass short and muscular. Large digitate oral gland extend- ing from anteroventral portion of either side of buccal mass and continuing into widened portion of notum in region of precardiac cerata. Large pair of chitinous jaws situated within buccal mass. Jaws ovoid with elongate masticatory margin (Fig. 5). Margin bearing five or six rows of denticles with about 34 denticles on outer row. Radular formula 36 x 1.1.1., in one paratype examined. Rachidian teeth (Figs. 6A—C), simply arched with 5-6 triangular denticles on either side of more 178 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 FicureE 7. Flabellina pedata. Scanning electron micrographs. (A) Dorsal view of rachidian teeth, scale = 20 um. (B) Lateral view of radula, scale = 20 um. elongate central cusp. Central cusp approxi- mately same width as adjacent denticles, but de- pressed ventrally from their level (Fig. 6C). Lat- eral teeth (Fig. 6A, D) triangular with a broad base extending towards outer edge. Single prom- inent, acutely pointed apex present on lateral. Inner margin of tooth with series of 15-20 tri- angular denticles. Laterals on right side of the radula of paratype entirely lack denticles (likely aberration rather than characteristic feature of species). Pair of elongate salivary glands that ex- tend posteriorly on dorsolateral surface of stom- ach present at posterior limit of buccal mass, near its junction with esophagus. Arrangement of reproductive organs essen- tially triaulic (Fig. 4D). Narrow preampullary duct widening into simply saccate ampulla. Ampulla narrowing into elongate hermaphroditic duct and dividing into vas deferens and oviduct. Oviduct curving and joining pair of equally sized recep- tacula seminorum and vaginal duct. More dis- tally, vaginal duct entering female gland mass in region of albumen gland. Vaginal duct continu- ing towards genital apertures. Nidamental glands comprising bulk of reproductive system. Mucous gland largest portion by far. Albumen and mem- brane glands smaller. Nidamental glands emp- tying via their own gonopore, ventral to vaginal and penial apertures. Vas deferens narrow, but appearing to contain prostatic cells throughout most of its length. Vas deferens entering short penial sac, wider than the vas deferens. Penial papilla simple and unarmed. Discussion.—Flabellina vansyoci differs markedly from other eastern Pacific species of Flabellina. It is the only species with rugose rhin- ophores. The other eastern Pacific species all have annulate, perfoliate, or smooth rhinophores. Fla- bellina vansyoci has a reddish-purple body sim- ilar to that of F. marcusorum. It can be readily distinguished by the opaque white spots and a lack of purple bands on the cerata. Flabellina telja may also have a purplish body color and opaque white spots. However, its spots are scat- tered over the body surface, not restricted to the cerata. This species also has perfoliate rather than rugose rhinophores. Flabellina vansyoci is related to other flabel- linids with a bilobed receptaculum seminis. It is most similar to F. pedata (Montagu, 1815) and the sympatric F. bertschi, both of which have smooth rather than rugose rhinophores. All other species with a bilobed receptaculum have papil- late rhinophores. In order to make comparisons with F. vansyoci, six specimens of F. pedata (CASIZ 072597, one dissected, intertidal pools, GOSLINER: TROPICAL AMERICAN FLABELLINIDAE C 179 Ficure 8. Flabellina hamanni. (A) Ceras, scale = 0.5 mm. (B) Lateral view showing arrangement of cerata, a = anus; ga = genital aperture; r = rhinophore, scale = 2.0 mm. (C) Ventral view of anterior end of foot, scale = 2.0 mm. (D) Reproductive system, al = albumen gland; am = ampulla; be = bursa copulatrix; me = membrane gland; mu = mucous gland; pe = penis; pr = prostate; rs = receptaculum seminis; v = vagina, scale = 1.0 mm. Ponta Delgada Harbor, Ilha Sao Miguel, Acores, Portugal, 17 July 1988, T. Gosliner) were ex- amined. The specimens of F. pedata studied here represent the first record of this species from the Acores. The ground color of F. pedata is a deep purple violet while F. vansyoci is reddish purple. Flabellina pedata lacks the opaque spots on the basal portion of the cerata that are present in F. pedata. As noted above, F. pedata has smooth rhinophores, while in F. vansyoci they are rugose. In F. pedata, the postcardiac ceratal groups are arranged in arches while in F. vansyoci the groups contain 1-3 rows of cerata. While the rachidan teeth of F. pedata (Fig. 7A, B) are similar to those of F. vansyoci, the lateral teeth are markedly dif- ferent. In F. pedata, the lateral teeth bear 6-9 coarse denticles, while F. vansyoci has 15-20 fine denticles. The reproductive systems are very similar, but it appears that the prostate is thicker in F. pedata than in F. vansyoci (Schmekel and Portmann, 1982; present study). Flabellina hamanni sp. nov. (Figs. 3B, 8-10) Type MATERIAL.—Holotype, California Academy of Sci- ences, San Francisco, CASIZ 088275, 3 m depth, Freeport, Grand Bahama Island, Caribbean Sea, June 1986, J. Hamann. Paratypes, CASIZ 088276, one specimen dissected, same lo- 180 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 Ficure 9. Flabellina hamanni. Scanning electron micrographs. (A) Jaw, scale = 430 um. (B) Masticatory margin, scale = 43 um. cality and date as holotype. Paratype, CASIZ 072269, one specimen, living on hydroids, 20 m depth, Thea’s Wreck, Ba- hamas, Caribbean Sea, April 1986, B. Rose. Paratypes, CASIZ 088882, three specimens and egg mass, reef S of West Caicos, Turks and Caicos Islands, 20 m depth, December 1983, Jeff Hamann. DISTRIBUTION. — Flabellina hamanni has been collected from the Turks and Caicos Islands and Grand Bahama Island (Jeff Hamann, pers. comm.) EtyMoLoGcy.—This species is named for my friend and colleague, Jeff Hamann, who first col- lected this species. EXTERNAL MorPHOLOGY.— Living animals (Fig. 3B) 25-32 mm in length. General body col- or translucent rose pink. Oral tentacles, foot cor- ners, and rhinophores rose basally with narrow band of purple pigment and opaque yellowish- white present on apical portion. Oral tentacles with wide area of opaque yellow covering two- thirds of their length. Cerata and rhinophores bearing only proportionately narrower band of yellow. Surface of cerata covered with dusting of opaque white. Dusting absent from inner basal portion of cerata. Posterior end of foot also pur- ple with opaque yellowish line or spot on its posterodorsal end. Digestive gland within cerata charcoal gray and visible through inner trans- parent portion of ceratal bases. Rhinophores elongate (approximately 4 mm in length) and conical. Each rhinophore bearing approximately 100 elongate papillae on its posterior face. Pa- pillae approximately 0.5 mm in length and ar- ranged in indistinct rows, covering middle two- thirds of rhinophores. Apex of rhinophores conical and devoid of papillae. Oral tentacles thin and elongate, tapering to acutely pointed apex. Cerata narrow at base and widening to just below their apices. Cerata terminating in acute apex (Fig. 8A). Thin core of digestive gland filling much of diameter of each ceras. Cerata arranged in distinct groups (Fig. 8B). Only anterior cluster slightly elevated from notum. Anterior, precar- diac cluster containing four rows of cerata on either side of animal, with 3-4 cerata per row. Postcardiac cerata arranged in six groups of arch- es per side of body. Generally, each arch con- GOSLINER: TROPICAL AMERICAN FLABELLINIDAE Ficure 10. Flabellina hamanni. Scanning electron micrographs. (A) Dorsal view of entire width of radula, scale = 60 um. (B) Rachidian teeth, scale = 20 um. (C) Lateral view of radula, scale = 25 um. (D) Lateral tooth, scale = 15 um. taining 2-7 cerata. Only posteriormost arch pres- ent aS row containing two cerata. Gonopore situated on right side of body, ventral to second and third ceratal rows of precardiac cluster. Pleu- roproctic anus situated at posterior end of inter- hepatic space, just anterior to first postcardiac ceratal arch. Nephroproct immediately antero- dorsal to the anus. Foot (Fig. 8C) grooved an- teriorly and possessing elongate, tentacular foot corners. Foot tapering gradually to narrow tail posteriorly. INTERNAL MorpHOLOGY.— Buccal mass short and muscular. Large digitate oral gland extend- ing from anteroventral portion of either side of buccal mass and continuing into widened portion of notum in region of precardiac cerata. Pair of ovoid chitinous jaws (Fig. 9A) within buccal mass. Elongate masticatory margin (Fig. 9B) bearing several rows of triangular denticles. Radular for- mula 40 x 1.1.1., in one paratype examined. Rachidian teeth (Fig. 10A—C) simply arched with 8-11 triangular denticles on either side of elon- 182 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 9 gate central cusp. Central cusp slightly wider than adjacent denticles and depressed ventrally from their level (Fig. 10C). Lateral teeth (Fig. 10A, C, D) triangular in shape with broad base extending towards outer edge. Lateral tooth with single prominent, acutely pointed apex. Series of 8-9 triangular denticles present on inner margin of lateral tooth. Pair of elongate salivary glands present at posterior limit of buccal mass, near its junction with the esophagus. Glands extend- ing posteriorly on dorsolateral surface of stom- ach. Arrangement of reproductive organs triaulic (Fig. 8D). Narrow preampullary duct widening into ampulla consisting of two convolutions. Postampullary duct narrowing slightly and pass- ing between lobes of albumen and membrane glands. Preampullary duct dividing into oviduct and vas deferens. Oviduct joining pair of recep- tacula seminorum of equal size after short dis- tance. Oviduct again narrowing and entering nidamental glands near the middle of albumen gland. Distinct, narrow vaginal duct continuing towards the genital apertures from this junction. Vagina joining with large saccate bursa copu- latrix immediately before exiting at its own gen- ital pore. Bursa thin walled and short stalked. Nidamental glands comprising the bulk of re- productive system. Mucous gland by far largest portion, while albumen and membrane glands smaller. Nidamental glands empty via their own gonopore, ventral to vaginal and penial aper- tures. Vas deferens expanding slightly into smooth, slightly convoluted prostatic portion. Vas deferens wide throughout its length and expand- ing into slightly wider apex of penial papilla. No armature associated with the penial papilla. Discussion. — The systematics and phylogeny of flabellinids have recently been reviewed (Gos- liner and Kuzirian 1990; Gosliner and Willan 1991). Five previously described species of Fla- bellina have a triaulic reproductive system and papillate rhinophores (Gosliner and Willan 1991), which are also present in F. hamanni. The only other species with a well-developed bursa cop- ulatrix (a plesiomorphic feature) is Flabellina marcusorum. Both of these species are found in the western Atlantic, though F. marcusorum is also present in the tropical eastern Pacific. Externally, the two species are similar in ap- pearance. F/abellina hamanni has narrower bands of purple on the cerata than does F. marcusorum. Specimens of F. hamanni also have opaque white pigment covering the basal portion of the cerata, which is absent in F. marcusorum. The cerata of F. hamanni are blunter than those of F. mar- cusorum and terminate in an abruptly pointed rather than gradually tapering apex. The radular morphology differs significantly between F. hamanni and F. marcusorum. The single specimen of F. hamanni had 40 rows of radular teeth. No specimen of F. marcusorum has been described with more than 34 rows of radular teeth. More significantly, the rachidian teeth of F. hamanni have 8-11 denticles on either side of the central cusp, while F. marcusorum has only 5-6 denticles per side. The rachidian teeth of F. hamanni are more similar to those of F. exoptata Gosliner and Willan, 1991, and F. delicata Gosliner and Willan, 1991. The lateral teeth of F. hamanni are similar to those of F. marcusorum F. rubrolineata (O’Donoghue, 1929) and F. poenicia (Burn, 1957), with relatively few denticles along their masticatory margin. The reproductive system is most similar to that of F. marcusorum, in that the bursa copu- latrix remains large and well developed. The re- productive systems of the two species differ in one important regard. The vas deferens of F. hamanni is uniform throughout its length, while in F. marcusorum it 1s constricted immediately prior to its entrance into a more distinct, ex- panded penial sac. ACKNOWLEDGMENTS Several individuals assisted in the successful completion of this project. Jeff Hamann and Robert Van Syoc kindly provided specimens of the two species described in this paper. Field work for this study was generously supported by the Lindsay Field Research Fund of the Califor- nia Academy of Sciences and funds from Ka- tharine Stewart. Roy Eisenhardt kindly printed the photographs used in the color plate. Hans Bertsch assisted with the translation of the ab- stract to Spanish. RESUMEN Dos especies nuevas de Flabellina son iden- tificadas de las aguas tropicales de América. La primera, Flabellina vansyoci sp. nov., ocupa un area que se extiende desde la Bahia Magdalena, en la costa pacifica de Baja California, hasta el Gulfo de Chiriqui, Panama. Esta especie se com- para a la especie mas cercana en la cadena evo- GOSLINER: TROPICAL AMERICAN FLABELLINIDAE lucionaria, F. pedata (Montagu, 1815), que ha sido encontrada por primera vez en las Islas Azores. La secunda, Flabellina hamanni sp. nov., ha sido encontrada en el Mar Caribe, de las Islas Bahamas hasta Venezuela. Evolucionariamente, esta nueva especie se relaciona mas cerca a la especie F. marcusorum Gosliner y Kuzirian, 1990. Flabellina bertschi, concocida previamen- te solamente en el Mar de Cortés, ha sido en- contrada en la costa pacifica de Panama. Fla- bellina marcusorum fue inicialmente encontrada en el Mar de Cortés y en el Mar Caribe y ahora ha sido encontrada por primera vez en Panama. Voy a designar F. stohleri Bertsch y Ferreira, 1974, como sinonimo de F. telja Marcus y Mar- cus, 1967. El rango de distribucion de F. telja se extiende hacia al sur, a las Islas Revillagigedos, Panama y las Islas Galapagos. LITERATURE CITED BEHRENS, D. 1991. Pacific coast nudibranchs. Sea Challeng- ers, Monterey. 107 pp. 183 BERTSCH, H. AND A. J. FERREIRA. 1974. Four new species of nudibranchs from tropical West America. Veliger 16(4):343- 353. FERREIRA, A. AND H. Bertscu. 1972. Additional data on Flabellina telja (Gastropoda: Opisthobranchia). Veliger 14(4): 414-415. GosLiner, T. M. AND A. M. KuziriAn. 1990. Two new spe- cies of Flabelliniae (Opisthobranchia: Aeolidacea) from Baja California Proc. Calif. Acad. Sci. 47(1):1-15. GosLiner, T. M. AND R. C. WILLAN. 1991. Review of the Flabellinidae (Nudibranchia: Aeolidacea) from the tropical Indo-Pacific, with the descriptions of five new species. Ve- liger 34(2):97-133. KerstitcH, A. 1989. Sea of Cortez marine invertebrates. A guide for the Pacific coast, Mexico to Ecuador. Sea Chal- lengers, Monterey. 112 pp. Marcus, Er. AND Ev. Marcus. 1967. American opistho- branch mollusks. Stud. Trop. Oceanogr. 6:1-256. Marcus, Ev. AND Er. Marcus. 1961. On Coryphellina rub- rolineata O’Donoghue, 1929 (Opisthobranchia, Eolidacea). Proc. Malac. Soc. Lond. 34(4):224—227. SCHMEKEL, L. AND A. PoRTMANN. Opisthobranchia des Mit- telmeeres. Springer Verlag, Berlin. 410 pp. © CALIFORNIA ACADEMY OF SCIENCES, 1994 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 10, pp. 185-220, 6 figs., 5 tables August 16, 1994 PHILIPPINE FROGS OF THE FAMILY RHACOPHORIDAE By Walter C. Brown Department of Herpetology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 and Angel C. Alcala Department of Environment and Natural Resources, Republic of the Philippines, Quezon City, Philippines Marine Biolo Woods Ho gical Laboratory/ le Oceanographic Institution any UG 22 1994 Wo. wie 3 Role, MA uz543 Asstractr: Philippine rhacophorid frogs include 18 species assigned to four genera. None of the genera but 12 of the species are endemic. Two species, Philautus poecilus and P. surrufus, are described as new. The distribution and ecology of the Philippine rhacophorids are discussed. Received June 25, 1993. Accepted December 15, 1993. INTRODUCTION The most recent systematic review of the Phil- ippine rhacophorid frogs is that of Inger (1954). He referred the then known species to two gen- era, Rhacophorus (nine species) and Philautus (eight species). He did note, however, that the diagnoses of the genera were arbitrary and in- adequate. Liem (1970) addressed that problem, providing a clearer understanding of the evolu- tionary trends within the Rhacophoridae and much more satisfactory diagnoses of the genera based on adult morphology. He assigned the Philippine species to four genera and redistrib- uted some species between Rhacophorus and Philautus. Inger (1985b) reported that differences in the morphology and development of larvae further supported Liem’s recognition of the gen- era Hazelia (= Nyctixalus) and Philautus, but that larval similarities only reaffirmed the close relationship of Polypedates and Rhacophorus. Dring (1987) reexamined the Bornean Philautus, describing several new species and establishing species groups within the genus. Dubois (1987) recognized three subgenera of Philautus and two subgenera of Rhacophorus. As to Philautus, we have followed Dring’s Group classification; and as to Rhacophorus, we have not placed the R. bimaculatus Group in a separate subgenus. Since 1954, extensive field surveys by the authors and others have greatly expanded the sample sizes for some of the Philippine species and have also provided samples from previously unknown populations on a number of islands. In view of these, a review of the Philippine rhacophorids is appropriate at this time. MATERIALS AND METHODS Type material, if available, has been exam- ined. The largely unreported, extensive collec- tions made by the authors over the past four [185] 186 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 WD G d Ficure 1. Male reproductive organ, ventral view (modified from Liem 1970); T = testis; K = kidney; WD = Wolffian duct: (a) Philautus acutirostris; (b) Polypedates dugritei. Distal pha- langes of third finger, dorsal view (after Liem 1970); TP = terminal phalanx; IC = intercalary cartilage: (c) Philau- tus acutirostris; (d) Polypedates colletti. decades and other recent collections, particularly those at the Field Museum and the United States National Museum, have also been studied. Repositories for specimens are referred to in the text by the following standard acronyms: ANSP, BMNH, CAS, CAS-SU, CM, FMHN, MNHN, MCZ, MSNG, MTKD, NMW, SMF, USNM, ZMB (Leviton et al. 1985). Except for the new species, specimens examined are cited in Appendix B rather than in species accounts. Characters used to differentiate species in- clude: color features, cutaneous ornamentation, extent of webbing of fingers and toes, size-range of adults, digital disks, presence or absence of vomerine teeth and vocal sacs, nuptial pads, presence or absence of a cutaneous pectoris mus- cle, insertion of the lateralis externis muscle, type of eggs, and measurements such as snout-vent length (SVL), head length (HL), head breadth (HB), snout length (SnL), eye diameter (ED), tympanum diameter (TD), tibia length (TL), hindlimb length, third finger length (measured to proximal edge of basal tubercle), diameter of third finger disk, and diameter of third toe disk. These measurements were made to the nearest 0.1 mm, using a Helios dial-caliper. Ratios given in the descriptions are based on these measure- ments. Brief notes on color in life are given for species for which they are available. Only general habitat is noted in the species accounts. More information on habitats is provided in the sec- tion on ecology and zoogeography. SYSTEMATIC SECTION Rhacophoridae Hoffman Polypedatidae Giinther, 1859:9. Rhacophoridae Hoffman, 1932 (1859):562 Under the provisions of Article 40 of the In- ternational Code of Zoological Nomenclature (Ride 1985), the name Rhacophoridae has pri- ority over the name Polypedatidae Giinther 1859, but it takes the date of Giinther’s publication. Whether the status of this taxon should be that of a subfamily of Ranidae, to which it is closely related, or a family equal in rank to Ranidae and Hyperolidae has been a matter of debate for the past century and a half. We follow Liem (1970), Dubois (1981), and Inger (1985) in recognizing the Rhacophoridae as a family. KEY TO PHILIPPINE GENERA OF RHACOPHORIDAE la. Skin of dorsum and upper limb surfaces with prominent, pointed to spiny, scat- tered to numerous, whitish tubercles or pointed spicules; eggs pigmented, grayish to, brownish... =. = 4 eee Nyctixalus lb. Skin of dorsum and upper limb surfaces lacking prominent whitish tubercles or spicules; eggs whitish to creamy, without dark pigment,«.<)\s0) 0% = 2 2a. Vomerine teeth greatly reduced or absent; Wolffian ducts not convoluted (Fig. 1A); PHILIPPINE RHACOPHORIDAE terminal phalanx of digits notched or bluntly rounded (Fig. 1C) _.. Philautus 2b. Vomerine teeth prominent; Wolffian ducts convoluted (Fig. 1B); terminal phalanx of Gisits ¥-shaped \(Fig:(1D) 22 Sew 3 3a. Skin on head partly co-ossified with skull; head longer than broad; dorsal color pat- tern usually dull grayish with various dark markings; vertebral column diplasiocoe- lous (sacral vertebra convex anteriorly with double condyle posteriorly; eighth verte- bra biconcave, and first seven procoelous) 1 2) 1) a ee Polypedates 3b. Skin of head free from skull; head as broad as or broader than long; dorsal color pat- tern bright green or brownish; vertebral column procoelous (vertebrae concave at anterior face and convex at posterior face) prameers BRIG!) Wi). Janus). xr Rhacophorus Nyctixalus Boulenger Nyctixalus Boulenger, 1882b:35. Type species: Nyctixalus margartifer, by monotypy. Hazelia Taylor, 1920:292 (not Hazelia Walcott, 1920:281). Type species: Hazelia spinosa, by original designation. Edwardtayloria Marx, 1975:2. Nomen substitutum for Hazelia Taylor, 1920. Dubois (1981) resurrected the genus Nyctix- alus from the synonomy of Philautus and re- ferred four, possibly five, species to it, four that Liem (1970) had placed in the genus Hazelia and one from the genus Theloderma as possibly be- longing in Nyctixalus. Inger (1985a) recognized only three species as belonging in Nyctixalus (margartifer, pictus, and spinosus). D1aAGnosis.—Size moderate (SVL 30-40 mm at maturity for Philippine species); skin of dorsal surfaces with blunt or sharply pointed spicules; belly coarsely granular; disks of fingers large, completely circumscribed by groove; fingers without or only with basal webs; tips of digits with prominent disks, those of toes smaller than those of fingers; hindlimbs much longer than body; toes one-third to one-half webbed; outer metatarsals not fused; males have nuptial pads; Wolffian ducts not convoluted; no vomerine teeth; males without vocal sacs; terminal pha- langes Y-shaped; omosternum forked; humero- dorsalis muscle two slips; eggs large, heavily pig- mented; clutch size 10-30 (for Philippine species), eggs deposited on walls of tree holes; feeding- larval stages are flatter in lateral view than larvae of other rhacophorid genera (Inger 1985b). 187 KEY TO PHILIPPINE SPECIES OF NYCTIXALUS la. Cranialand supratympanic crests present; numerous, pointed, whitish spicules cov- ering dorsal surfaces; posterior surface of thighs with coarse granules ..... N. spinosus 1b. Cranial and supratympanic crests absent; scattered, low, whitish tubercles covering dorsal surfaces; posterior surface of thighs MNearly:smOOth t= 22s ee N. pictus Nyctixalus pictus pictus (Peters) Ixalus pictus Peters, 1871:580 (type loc.: Sarawak, Borneo; holotype in MSNG). Philautus anodon: van Kampen 1923:271. DESCRIPTION. —SVL 30.4—34.5 for two males; HB 96-99% of HL and 34-37% of SVL; snout nearly truncate, jutting sharply at level of nos- trils, SnL 41% of HL and 42% of HB; eye mod- erate, ED 67-74% of SnL and 25-32% of HB; tympanum fully exposed, TD 66-85% of ED; canthus sharply rounded; lores nearly vertical, shallowly concave; fingers without webs; diam- eter of third finger disk 25-27% of third finger length and 54-61% of TD; subarticular tubercles low, slightly rounded; supernumerary tubercles present; inner metacarpal tubercle poorly de- fined; SVL 61-62% of hindlimb length; TL 56- 58% of SVL; toes webbed: to distal edge of tu- bercle on first, between tubercle and disk on in- side of second, to tubercle on third, and nearly to basal tubercle on fourth; diameter of third toe disk 71% of third finger disk; subarticular tu- bercles low, slightly pointed; metatarsal tubercles present; prominent inner but no outer metatarsal tubercle; skin of dorsum and limbs with scattered low, whitish tubercles; underside of head and throat nearly smooth; belly with flat granules; prominent tubercle on both sides dorsolateral to vent, and several below vent. CoLor.—In preservative, tan on upper sur- faces with scattered, brown-ringed, whitish spots concentrated along upper lip, upper eyelid, can- thus, and tip of snout, a large one dorso-posterior to tympanum; venter creamy white. In life, the dorsal color is cinnamon-brown to chocolate (In- ger 1966). COMPARISONS. — The Palawan population, based on two examples, is judged conspecific with samples of Nyctixalus p. pictus from northern Borneo. This species differs from N. spinosus from the southeastern Philippines in the less numer- 188 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 ous and less prominent dorsal tubercles, the lack of prominent cranial and supratympanic crests, less numerous whitish spots, and slightly broader head relative to SnL or HL. REPRODUCTION. — This species attaches its eggs to walls of tree holes about 5-10 cm above the water level in lowland forest in north Borneo (Inger 1966). Larvae at pre-limb bud (pre-stage 25, Gosner 1960) stages also occur in similar tree-hole sites (Inger 1985b). One of the speci- mens, laboratory-reared to early metamorpho- sis, confirms the identification of the larvae. Gravid females from the Bornean population ex- hibit a low number (8-10) of large grayish eggs. Neither egg-deposition site nor larvae have been observed for the Philippine population. HaBITAT.—The two known Philippine speci- mens are from lowland forest at 150-300 m el- evation. RANGE.— Malay, Singapore, Borneo, and Pa- lawan Island, Philippines. Nyctixalus spinosus (Taylor) Hazelia spinosa Taylor, 1920:292 (type loc.: Bunawan, Agusan Prov., Mindanao Island; holotype in CM); 1921:93; Liem 1970:95. DESCRIPTION. —SVL 30.4—37.2 mm for seven males and 36.8-39.0 mm for four females; HB 88-94% of HL and 34-39% of SVL; snout nearly truncate to obtusely pointed, jutting out at level of nostrils; SnL 39-44% of HL and 43-50% of HB; ED 68-86% of SnL and 31-38% of HB; tympanum bordered by supratympanic crest; TD 60-80% of ED; canthus sharply rounded; lores nearly vertical, slightly concave; fingers without webs; diameter of third finger disk 23-29% of third finger length and 53-70% of TD; subarti- cular tubercles round-pointed; supernumerary tubercles prominent; large inner metacarpal tu- bercle, SVL 56-61% of hindlimb length; TL 56- 61% of SVL; toes webbed: to tubercle on first toe, beyond tubercle on inside of second, be- tween basal and distal tubercle on third, nearly to proximal tubercle on fourth, and to base of tubercle on fifth; diameter of third toe disk 57- 75% of third finger disk; subarticular tubercles round-pointed; supernumerary tubercles numer- ous; prominent inner but no outer metatarsal tubercle; skin of dorsum and limbs with prom- inent, pointed, whitish spicules; cranial and su- pratympanic crests present; underside of head and throat with small granules; belly and pos- terior thighs coarsely granular; whitish tubercle dorso-lateral to vent on both sides and several tubercles below vent. CoLor.—In preservative, pinkish gray to tan or light brown on upper surfaces with a pair of prominent, whitish, brown-rimmed spots dorso- posterior to the tympanum and usually a few scattered, similar but smaller spots on dorsum, hindlimbs, canthus, snout, and sometimes the edge of upper eyelid; venter creamy white. In life, dorsum reddish to tannish brown with a pair of prominent white spots dorsolaterally at shoulder level and a few smaller, scattered white spots posteriorly; several short, white, transverse bars on hind limbs. REPRODUCTION. — Three adults and about 120 tadpoles in a water-containing hole near the base of the trunk of an Agathis alba tree located in the dipterocarp forest on Dapitan Peak, Min- danao Island, are identified with this species (Al- cala and Brown 1982). The three adults clung to the charred wood above the level of the water. The larval stages varied from 1-44 (Gosner 1960), indicating that multiple clutches were laid at dif- ferent times. Color markings, dorsal spicules and webbing of toes, evident in late stages, positively identify the larvae with this species. Because gravid females have 30-40 eggs per clutch, the number of egg-depositions was probably four or five. The mature eggs from gravid females are brownish on one hemisphere and black on the other. Although not observed, the egg-deposition site is probably on walls of tree holes above the water level as has been reported for the closely related Nyctixalus pictus (Inger 1966). CoMPARISONS. —See Nyctixalus pictus. HABITAT.—Specimens are from dipterocarp forest areas at elevations of 500-1100 m. RANGE.—Basilan, Mindanao, and Bohol is- lands. Philautus Gistel Orchestes Tschudi, 1838:35 (not Illiger, 1798:498). Type spe- cies: Hyla aurifasciata Schlegel, 1837, by monotypy. Ixalus Duméril and Bibron, 1841:523 (not Ogilby, 1837:119). Nomen substitutum for Orchestes Tschudi, 1838. Philautus Gistel, 1848:10. Nomen substitutum for Orchestes Tschudi, 1838. Liem (1970) redefined the genus Philautus and assigned to it three species that had been pre- viously placed in Rhacophorus (emembranatus, lissobrachius, and surdus). Dring (1987) further PHILIPPINE RHACOPHORIDAE TABLE 1. 189 Snout—vent length and pertinent body proportions for Philippine species of the aurifasciatus group of Philautus. (R = range, M = mean, n = number in sample, 3fd = third finger disk, 3fl = third finger length, 3td = third toe disk.) Adult SVL ED TD 3fd 3d Species Male Female HB ED 3fl 3fd P. acutirostris R 16.1-22.7 22.0—27.8 0.31-0.38 0.35-0.50 0.36-0.46 0.80-0.97 M 0.342 0.422 0.423 0.897 n 15 17 16 16 1 P. leitensis R 19.2-21.9 22.6-26.4 0.23-0.35 0.29-0.50 0.26-0.44 0.68-0.92 M 0.283 0.381 0.312 0.787 n 10 12 12 12 12 P. longicrus R 18.0-21.9 21.2-25.5 0.27-0.33 0.15-0.29 0.19-0.29 0.40-0.71 M 0.301 0.214 0.246 0.587 n 12 24 15 15 15 P. schmackeri R 19.0-21.5 22.6-28.0 0.25-0.34 0.16-0.35 0.15-0.28 0.50-0.71 M 0.291 0.21 0.225 0.624 n 14 15 18 18 18 divided the Bornean and Philippine Philautus into species-groups (Hhosei, vermiculatus, surdus, tectus, and aurifasciatus). Further discussion of these species groups was provided by Inger (1989). Both Dring and Inger referred the then known Philippine species to the surdus and aurifasciatus groups. DraGcnosis. — Size small to moderate (SVL 16- 44 mm at maturity for known Philippine spe- cies); body tapered from widest point of head to groin, skin of dorsum shagreened with or without tubercles, belly with flattish granules; digital disks moderate to large, completely circumscribed by a groove, disks of toes somewhat smaller than those of fingers; fingers without or only with basal webs; nuptial pads present or absent; hindlimb much longer than body; toes moderately to al- most fully webbed; vomerine teeth present or absent; usually males with vocal sacs; terminal phlanges notched or bluntly rounded; omoster- num usually not forked at base; Wolffian ducts not convoluted; petrohyoideus posterior muscle usually two slips; eggs large, unpigmented; clutch size 6-30 (Philippine species); eggs, where known, deposited in leaf axils, epiphytic ferns, and Ne- penthes pitchers; development direct, within egg capsule for all species for which development is known. This montane genus, with nine species, exhib- its greater diversity than other rhacophorid gen- era occurring in the Philippines. KEY TO THE PHILIPPINE SPECIES OF PHILAUTUS SVL for mature males usually less than 23 mm, for females less than 28 mm; vo- merine teeth usually absent 2 . SVL for mature males usually greater than 23 mm, for females greater than 28 mm; vomerine teeth usually present 0. 5 . Diameter of third toe disk rarely less than 70% of third finger disk; TD usually great- er than 35% of ED, (Table 1); except oc- casionally for P. /eitensis, upper eyelids and dorsum without tubercles (or very rarely a few); dorsal color pattern not ex- hibiting a dark cruciform blotch 3 . Diameter of third toe disk rarely greater than 70% of third finger disk; TD very rarely as great as 35% of ED (Table 1); dorsum with some tubercles at least on eyelids and a pair at shoulder level; dorsal color pattern usually with some modifi- cation of dark cruciform blotch 4 . SnL usually greater than 40% of HB; snout sharply pointed, often with pale-colored projection at tip; webbing of fourth toe usually does not reach tubercle (Fig. 3A); (known from Mindanao and Basilan is- lands) ) passin) tgs ere P. acutirostris . SnL usually less than 40% of HB; snout round pointed, without pale-colored pro- 190 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 TasLe 2. Snout—vent length and pertinent body proportions for species of the surdus group of Philautus and P. poecilus. (R = range, M = mean, n = number in sample, 3fd = third finger disk, 3fl = third finger length, 3td = third toe disk.) Adult SVL ED TD 3fa 31d Species Male Female HB ED 3fl 3fd P. emembranatus R 25.1-32.4 30.9-43.9 0.25-0.31 0.18-0.34 0.22-0.36 0.65-0.86 M 0.29 0.255 0.281 0.771 n 7 8 15 15 15 15 P. surdus R 22, 5-31.0 28.5-38.4 0.30-0.37 0.32-0.49 0.36-0.45 0.83-1.00 M 0.34 0.379 0.403 0.912 n 25 20 25 25 25 25 P. surrufus R 20.0-27.6 26.6-35.4 0.31-0.39 0.34-0.50 0.34-0.44 0.81-1.00 M 0.35 0.437 0.379 0.909 n 16 12 20 20 20 20 P. poecilus R isos 0.28-0.32 0.23-0.38 0.21-0.38 0.64-0.80 M 0.296 0.267 0.333 0.723 n 7 7 7 7 7 jection at tip; webbing of fourth toe usu- for adult males 21-28 mm (Table 2); di- ally to proximal edge of tubercle (Fig. 3B); ameter of third finger disk usually less than (known from Leyte and Bohol islands) ... 35% of third finger length; (known from ahd een ALAA eNO Yi router P. leitensis northeastern Mindanao Island) 4a: Dorsuny heavily shaereened: with prom- § © 2... 5 Soo 2 ae P. poecilus inent tubercles at shoulder level on both 6b. Dorsal color pattern reddish brown or 4b. Sa. Sb. 6a. sides and on eyelids, usually other scat- tered tubercles; (known from Mindoro Is- land) -tpenetan wPloraci «fier Dorsum usually with tubercles in blackish blotch at shoulder level on both sides and on eyelid, rarely with other scattered tu- bercles; (known from Palawan and Bor- TCO) as einai ao ties cheney P. longicrus Dorsum with pair of distinct tubercles in darkly pigmented spots at shoulder level; other scattered tubercles on eyelids and dorsum; diameter of third finger disk usu- ally greater than 35% of third finger length; all toes but fourth usually webbed to be- tween tubercle and disk (Fig. 4B); (known from Greater Mindanao and Greater Lu- TOT ARCAS) pam heel snes i » Wy Mer P. surdus Dorsum relatively smooth, usually with- out or with very few tubercles; webs of toes usually not distal to tubercles except on second and fifth for some specimens offemembranatus eae fee Ae Dorsal color pattern usually mottled black (or dark brown) and ivory or tannish (Fig. 2A), or with dark cruciform blotch; SVL P. schmackeri mottled reddish brown and tannish (Fig. 2B); ED greater than 30% of HB; diameter of third finger disk usually greater than 35% of third finger length; SVL for adult males 20-28 mm (Table 2); (known from northwestern Mindanao Island) RRC oe P. surrufus 6c. Dorsal color pattern golden-brown to gray, usually not heavily blotched; ED usually less than 30% of HB; diameter of third finger disk usually less than 35% of third finger length; SVL for adult males 25- 33 mm (Table 2); (known from Mindanao Island) ek: taners a Warnes ee P. emembranatus Inger (1954) regarded four Philippine species (acutirostris, leitensis, longicrus, and schmackeri) as probably closely related. Later (1966) he placed acutirostris, longicrus, and the Bornean species petersi and mjobergi in the synonymy of auri- fasciatus, but he did not comment on the status of leitensis or schmackeri. Dring (1987) resur- rected all of these species and restricted aurifas- ciatus to Java. He also extended the range of longicrus to include some populations in north- PHILIPPINE RHACOPHORIDAE 191 FIGURE 2. west Borneo as well as Palawan. He further noted the general similarities between /ongicrus and mjobergi. AURIFASCIATUS GROUP Dring (1987) defined the aurifasciatus Group as: small to moderate in size, vomerine teeth absent, cutaneous pectoris muscle present, gen- iohyoideus lateralis externis insertion includes membrana thyroideus, nuptial pad small and smooth, and ova unpigmented. He assigned four Philippine species (acutirostris, leitensis, longi- crus, and schmackeri) to this group. Liem (1970) stated that these species lack a nuptial pad, because he did not find it for spec- imens that he examined. We found small to mod- erate pads for males of P. acutirostris and P. longicrus, but large for P. schmackeri. The nup- tial pad is evident for only some of the males of any of these species. This may be seasonal or caused by some other regulating mechanism. We found vomerine teeth present for most examples of P. acutirostris but lacking for other Philippine species of this group. Philautus acutirostris (Peters) Ixalus acutirostris Peters, 1867:32 (type loc.: Eastern Minda- nao, Philippines; holotype in ZMB). Philautus woodi Stejneger, 1905:346 (type loc.: Mt. Apo, Min- danao I., Philippines; holotype in USNM). (a) Philautus poecilus (23 mm SVL); (b) Philautus surrufus (28 mm SVL). Rhacophorus (Philautus) woodi: Ah\ 1931:66. Philautus basilanensis Taylor, 1922a:169 (type loc.: Basilan Island; holotype in CAS). We examined the holotypes of Ixalus acuti- rostris, Philautus basilanensis, and Philautus woodi and found them to be, for characters used, within the range of variation exhibited by a large sample of the P. acutirostris populations in northeastern and northwestern Mindanao. Our findings are in agreement with those of Inger (1954), and we treat both names as synonyms of acutirostris. DESCRIPTION. — Body tapering posteriorly; SVL 16.1-22.7 mm for 15 males and 22.0-27.8 for 17 females (The holotype at 17 mm SVL and lacking vocal slits is assumed to be an immature female.); HB 99-108% of HL and 39-44% of SVL; obtusely to acutely pointed, frequently with narrowly rounded, pale-colored projection at tip of snout; SnL 41-47% of HL and 39-46% of HB; eye moderate, ED 76-97% of SnL and 31-38% of HB; tympanum distinct, bordered dorsally by shallow fold extending from corner of eye to ax- illa, TD 35-50% of ED; canthus moderately to sharply rounded; lores somewhat oblique, con- cave; vomerine teeth present; males with vocal sacs; fingers without webs; diameter of third fin- ger disk 39-46% of third finger length and 118- 156% of TD; subarticular tubercles low, round- ed, nearly as broad as fingers except for basal one 192 on third toe; palmer area coarsely granular; an elongate inner metacarpal tubercle and usually a less distinct outer one; nuptial pad of males, small rounded, smooth; SVL 49-67% of length of hind- limb; tibia length 59-69% of SVL; toes webbed: at most to distal edge of subarticular tubercle on first and second, to middle or outer edge of distal subarticular tubercle on outside of third and in- side of fifth, and to penultimate tubercle on fourth (Fig. 3A); diameter of third toe disk 80-97% of diameter of third finger disk; plantar nearly smooth or with some low tubercles; elongate in- ner but no outer metatarsal tubercle; outer edge of fifth toe smooth or with fringe of low tubercles; skin of dorsum shagreened, usually without tu- bercles, or with a few low tubercles, most prom- inent for males; belly, undersurface of thighs, and to lesser extent posterior thighs coarsely granu- lar; under surface of head and throat finely gran- ular to nearly smooth. CoLor.—In preservative, dorsal ground color variable, from grayish tan to reddish brown, or sometimes sooty gray-brown; nearly uniform, or with a darker interorbital area and sometimes small, dark blotches associated with tubercles; lores and tympanic area usually somewhat darker than dorsum; posterior surface of thighs and low- er leg, anterior surface of thighs and groin rusty- brown; upper surface of hindlimbs with or with- out vague dark, transverse barring; sometimes tip of snout, canthus, border of upper eyelid, and/ or supratympanic fold pale relative to rest of dorsum; venter creamy to ivory with some brownish blotches, usually most prominent un- der head and throat. In life, usually dorsum and limbs grayish brown, lilac-brown, or reddish brown nearly uni- form (rarely with darker hourglass pattern); darker on lateral surface of head. REPRODUCTION. —Gravid females of P. acuti- rostris possess 5—10 large non-pigmented eggs in each ovary, thus a clutch size of 10-20 eggs. Sev- eral clutches of very similar eggs were in leaf axils of wild banana and aerial ferns on Dapitan Peak in western Mindanao in 1959 and on Mt. Hil- onghilong in eastern Mindanao in 1971 (Alcala and Brown 1982) and may belong to this species. Lack of late developmental stages precluded pos- itive identification although the series of early stages were similar to those of other members of the genus Philautus. CompaArIsons. — Philautus acutirostris is very similar to and doubtless related to P. Jeitensis. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 The color patterns are similar, and SVL for both sexes exhibit essentially the same range. Philau- tus acutirostris differs in: the less extensive web- bing of the toes (Figs. 3A, B); the more pointed, longer snout relative to HB, usually exhibiting a pale-colored projection at the tip; and the some- what smaller eye relative to the HB (Table 1). Philautus acutirostris differs from longicrus in: the lack of tubercles on the dorsum, the larger tympanum relative to the eye, and the more ex- panded finger disks relative to finger length. HasitTAT.— Based on 90 specimens from Dap- itan Peak on Zamboanga Peninsula and Mt. Hil- onghilong in northeastern Mindanao, P. acuti- rostris ranges from about 400 m in dipterocarp forest to more than 2000 m in montane forest. RANGE. — Basilan and Mindanao islands, Phil- ippines. Philautus leitensis (Boulenger) Ixalus leitensis Boulenger, 1897:107 (type loc.: Leyte Island, Philippines; holotype in BMNH). Boulenger (1897) described this species from Leyte Island, based on a single specimen. Stejne- ger (1905), Taylor (1920) and Inger (1954) con- tinued to recognize the species but with few or no additional specimens on which to determine variability or assess relationships to other spe- cies. Our series from Bohol and one specimen from Leyte are in general agreement with the holotype and serve to establish the differences that we use to distinguish it from P. acutirostris. These species are closely related and may simply represent geographic races (subspecies). How- ever, pending the availability of more material from Leyte (type locality) as well as Samar and possibly small islands between them and Min- danao, we continue to treat /eitensis as a distinct species. DESCRIPTION.—SVL 19.2-21.9 mm for 10 males and 22.6-26.4 mm for 12 females; HB 103- 111% of HL and 38-43% of SVL; snout rounded to round pointed, SnL 36-45% of HL and 34- 41% of HB; eye large, ED 68-93% of SnL and 23-35% of HB; tympanum distinct, bordered dorsally by shallow fold, TD 29-50% of ED; can- thus sharply rounded; lores slightly oblique, usu- ally somewhat concave; vomerine teeth absent; fingers without webs; diameter of third finger disk 26-44% of third finger length and 96-140% of TD (only two of 15 specimens below 109%); su- barticular tubercles prominent, outer tubercle PHILIPPINE RHACOPHORIDAE low, oval to elongate; nuptial pad small, round, smooth; SVL 49-62% of hindlimb length; TL 55-66% of SVL; toes webbed: to between su- barticular tubercle and disk on inside of first, second, third and fifth, and to distal edge of pen- ultimate tubercle on fourth toe (Fig. 3B); diam- eter of third toe disk 68-92% of third finger disk; subarticular tubercles moderate, low; few small, supernumerary tubercles; inner metatarsal tu- bercle present; no outer tubercle; skin of dorsum shagreened, very rarely with few tubercles; un- dersurface of head and throat relatively smooth, except for low, whitish tubercles along border of lower jaws for most specimens; belly and pos- terior surface of thigh with flattish granules. CoLor.—In preservative, dorsum and upper lateral surfaces grayish-tan or tan to dark brown, nearly uniform, sometimes with a few small dark spots or rarely lighter blotches; venter pale ivory to creamy usually with brownish flecks or heavily mottled with brown especially under head and throat; posterior and anterior surface of thighs rusty-brown, extending onto the lateral surface at the groin. In life, dorsal ground color orange-brown, red- dish brown, or lilac-brown, nearly uniform or with some darker spots; hindlimbs usually with dark crossbars. REPRODUCTION.— Three gravid females with greatly enlarged, non-pigmented eggs in the ova- ries provide a clutch size of 6-14. ComMPARISONS.—See Philautus acutirostris. Philautus leitensis also differs from /ongicrus and schmackeri in the larger tympanum relative to the eye (Table 1) and the less tuberculate dorsum. HABITAT.—AIl Bohol specimens came from dipterocarp or submontane forest at elevations between 150 and 700 m. RANGE.—Leyte and Bohol islands, Philip- pines. Philautus longicrus (Boulenger) Exalus longicrus Boulenger, 1894:88 (type loc.: Palawan Is- land; syntypes in BMNH). DESCRIPTION.—SVL 18.0-—21.9 mm for 12 males and 21.2—25.5 for 24 females (one female, 19.9 mm, is probably mature, based on appear- ance of oviduct); HB 106-117% of HL and 39- 44% of SVL; snout rounded to round—pointed, protruding but without pale, conical tip, SnL 30- 36% of HB and 33-41% of HL; eye large; ED 193 83-100% of SnL and 27-33% of HB; tympanum covered by skin, TD 15-29% of ED; canthus sharply to moderately rounded; lores slightly oblique and concave; vomerine teeth absent; males with vocal sacs; fingers without webs; di- ameter of third finger disk 19-29% of third finger length and 140-233% of TD; subarticular tuber- cles round to round-pointed; a few, low super- numerary tubercles; elongate inner metacarpal tubercle; low fringe on outside of fourth finger; lower arm smooth; nuptial pad small, nearly round, smooth; SVL 52-59% of hindlimb length; TL 57-64% of SVL; toes webbed: at most nearly to disk on inside of first and second, to tubercle or its midpoint on inside of third and fifth, short of penultimate tubercle on fourth (Fig. 3C); di- ameter of third toe disk 40-71% of diameter of third finger disk; subarticular tubercles rounded, small, low; supernumerary tubercles few and small or absent; inner metatarsal tubercle elon- gate, outer small or absent; skin of dorsum sha- greened, nearly uniform or with a few scattered, small tubercles; under surface of head and throat nearly smooth to faintly granular (especially for males); usually some tubercles on edge of lower jaw; belly and proximal under surface of thighs coarsely granular. CoLor.—In preservative, dorsal ground color grayish or grayish tan to brown or dark brown, nearly uniform or with scattered dark or pale spots (24 of 60), with pale, median band (4 of 60), or with dark cruciform or hourglass pattern (32 of 60); hindlimbs usually with dark cross bars; posterior and anterior surface of thighs with reddish-brown bands and sometimes similarly colored blotches on posterior lateral surfaces; venter cream to creamy tan usually with small, dark flecks under head and throat and sometimes brown spots or blotches on belly. In life, dorsum nearly uniformly dark brown (sometimes almost black) or ochre-brown with a darker hourglass spot or blotches; snout area usually more pale; hind limbs usually with black- ish transverse bars; lateral surfaces lighter brown; hindlimbs with dark cross bars. REPRODUCTION. — Two gravid females have 9 and 13 large, non-pigmented eggs in the ovaries. CompPARISONS. — Philautus longicrus differs from schmackeri in: more extensive webbing of toes (Fig. 3C, D); slightly less prominent tuber- cles on dorsum; and the slightly larger eye rela- tive to SnL. See CoMPARISONS under P. acuti- rostris and P. leitensis. 194 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 5mm 5mm Ficure 3. Ventral surface of foot, showing extent of webbing and tubercles: (a) Philautus acutirostris; (b) Philautus leitensis; (c) Philautus longicrus; (d) Philautus schmackeri. PHILIPPINE RHACOPHORIDAE HasiraAT.—This species occurs from near sea level in dipterocarp forest to nearly 1300 m in montane forest on Thumb Peak. Philautus lon- gicrus occurs in moss forest at 3,000 m in north- west Borneo (Inger and Stuebing 1992). RANGE.— Borneo and Palawan Island, Phil- ippines. Philautus schmackeri Boettger Philautus schmackeri Boettger, 1892:17 (type Loc.: Mount Halcon, Mindoro Island; holotype in SMF). Ixalus mindorensis Boulenger, 1897:107 (type loc.: Mount Du- langan, Mindoro Island; syntypes in BMNH). DESCRIPTION. —SVL 19.0-21.5 mm for 14 males and 22.6-—28.0 mm for 15 females; HB 100- 111% of HL and 38-44% of SVL; snout round pointed without conical tip, SnL 38-44% of HL and 31-41% of HB; eye moderate, ED 68-93% of SnL and 25-34% of HB; tympanum covered by skin, TD 16-35% of ED; canthus sharply rounded; lores slightly oblique and concave; vo- merine teeth absent; males with vocal sacs; fin- gers without distinct webs; diameter of third fin- ger disk 15-28% of third finger length and 100- 225% of TD; subarticular tubercles rounded; su- pernumerary tubercles few and small; elongate inner and outer metacarpal tubercles; low flange of skin on outer edge of hand; row of pale tu- bercles along lower arm; nuptial pad small, round, smooth; SVL 54-63% of hindlimb length; TL 53-60% SVL; toes webbed: at most to distal edge of tubercle or beyond on inside of first, second, and third, usually to middle or distal edge of tubercle on third and fifth, and to penultimate tubercle on fourth (Fig. 3D); diameter of third toe disk 50-7 1% of third finger disk; subarticular tubercles rounded, small, low; scattered small tubercles on plantar surface; an outer but no in- ner metatarsal tubercle; skin of dorsum sha- greened, uniform or with few scattered tubercles; undersurface of head and throat nearly smooth or with fine granules; belly and posterior thighs with longer, flattish granules. Cotor.—In preservative, highly variable; dor- sal ground color grayish-tan, tan, brown, dark brown and reddish brown; relatively uniform or with scattered dark or pale spots (19 of 60), with pale, median dorsal band (15 of 60), or with dark cruciform or hourglass pattern (26 of 60); limbs with vague to distinct, dark transverse bands or rarely nearly uniformly dark; a rusty-red, lon- gitudinal band on posterior and anterior surface of thighs; venter pale cream to creamy tan, usu- 195 ally with dark flecks (especially anteriorly), or with scattered, more dense brownish spots. In life, dorsal ground color light brown to dark brown with variable, scattered, dark or pale markings; lateral surfaces to dorsolateral line tannish anteriorly, usually darker posteriorly; dark transverse bars on hindlimbs and interor- bital region. REPRODUCTION. —Clutch size varies from 6- 15 for this species, based on the number of en- larged ovarian eggs in several gravid females and two clutches of 12 and 15 embryos, respectively, from leaf axils of bird’s nest ferns (Alcala and Brown 1982). The oldest larva in these clutches was about stage 31 (Gosner 1960). The examples are typical of that stage in other species of the genus Philautus. CoMPARISONS.—See Philautus longicrus and acutirostris. In terms of the tuberculation of the dorsum and the extent of the webbing of the toes, Philautus schmackeri is more similar to P. surdus than it is to P. /ongicrus, aurifasciatus, or acu- tirostris. Philautus schmackeri, however, was re- ported by Liem (1970) as being more closely re- lated to P. aurifasciatus (and presumably longicrus), being in agreement for all 36 character states; whereas P. schmackeri differs from surdus (= lissobrachius of Liem) in the character state for the geniohyoideus muscle, the vomerine teeth (not absolute), and the smaller size. We have followed Liem in including schmackeri in the aurifasciatus group. HasitaAT.—Specimens of Philautus schmack- eri occur from lower dipterocarp to montane for- ests at elevations from 30-1450 m. RANGE.— Mindoro Island, Philippines. VERMICULATUS GROUP Dring (1987) defined the vermiculatus Group as: comprising species of moderate-sized indi- viduals with vomerine teeth absent, nuptial pad absent, cutaneus pectoris absent, geniohyoideus lateralis externis inserts on hyoid plate and pos- terolateral process, and ova half pigmented. Vo- merine teeth absent (present but very reduced in Philautus poecilus). Seven specimens collected on south side of Mt. Hilonghilong in northeastern Mindanao Island represent an undescribed species of Philautus. They are from montane forest (1600-1700 m). Based on the general slender appearance and ab- sence of the cutaneous pectoral muscle, we have tentatively placed this species in the vermiculatus 196 Group of Dring (1987). A nuptial pad 1s present for some males. Philautus poecilus new species (Figure 2A) Hototyre.—CAS 133526, south side of Mt. Hilonghilong, Agusan del Norte Prov., Mindanao I., collected by L. Alcala et al., May 1971. PARATYPES. —CAS 133524-25, 133532, 133543-44; FMNH 250821 (same general locality as holotype). Diacnosis.—Differs from other Philippine species of Philautus in the following combination of characters: an intermediate-sized species, males 21.4-28.1 mm SVL; cutaneous pectoris muscle absent; usually more mottled, ivory-tan and blackish-brown, dorsal pattern (two of seven with dark cruciform mark); small web at base of fin- gers; SnL 28-34% of HB; TD 23-38% of ED; and lack of dorsal tubercles other than one on eyelid. DESCRIPTION. — Body moderately tapered from widest part of head to groin; SVL 21.4-28.1 mm for seven males; head broadest at angle of jaws, HB 105-118% of HL and 38-41% of SVL; snout rounded to round-pointed, SnL 28-34% of HB and 33-40% of HL; eye relatively small, ED 73- 89% of SnL and 28-32% of HB; tympanum dis- tinct but covered by skin, small, TD 23-38% of ED; snout moderately projecting; nostril much closer to tip of snout than eye; vomerine teeth in very small, low patches for three specimens, absent for one (CAS 133530); vocal sacs present in males; canthus rounded; lores slightly oblique and slightly concave; cutaneous pectoris muscle absent; fingers without webs or with small web at base; disk of third finger largest, its diameter 21-36% of third finger length and 111-—200% of TD; subarticular tubercles large (except for basal one on third finger), round, low; palmar area wrinkled; inner metacarpal tubercle elongate; low fringe of skin distally on outer edge of hand; tubercles on outer edge of lower arm variable; nuptial pads not evident for any of the males; hindlimbs long, SVL 58-61% of hindlimb length, TL 50-57% of SVL; toes webbed: to center or distal edge of tubercle for first, to distal edge or slightly beyond on inside of second, third, and fifth, to basal tubercle for fourth; disks of toes smaller than those of fingers; diameter of third toe disk 64-80% of diameter of third finger disk; plantar surface with few, small tubercles; inner metatarsal tubercle elongate, outer lacking; skin of dorsum shagreened without tubercles, except PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 for very small one on eyelid; belly, under surface of head, throat and proximal part of thighs coarsely granular. CoLor.—In preservative, dorsum and lateral surfaces mottled blackish brown and ivory tan, occasionally nearly uniform grayish tan with a few dark blotches or a dark cruciform pattern; belly ivory to dusky ivory; under surface of head and throat similar but variously flecked or spot- ted with brown; hindlimbs with irregular dark and pale transverse bands. In life, dorsal ground color brownish slate, lightly to heavily mottled with ivory to creamy blotches; lateral surfaces and limbs predomi- nantly ivory colored but with some brownish slate markings. REPRODUCTION. — No information is available for this species. ETYyMOLoGY.—The name poecilus is based on the Greek word poikilas meaning mottled. CompPARISONS. — Philautus poecilus differs from other Philippine Phi/autus in the absence of the cutaneous pectoris muscle and nuptial pad, and for this reason it was assigned to the vermiculatus Group. It also differs in the more mottled color pattern frequently evident. Based on SVL of adult males (Table 2), it is somewhat larger than the four Philippine species of the aurifasciatus Group and somewhat smaller than P. emembranatus and P. surdus in the surdus Group. The minute web at the base of the fingers further distinguishes it from the species of the aurifasciatus Group. The tubercle-free dorsum also distinguishes it from P. longicrus, P. schmackeri, and P. surdus. HABITAT.—The seven known specimens are from montane (moss) forest on Mt. Hilonghilong in northeastern Mindanao Island at an elevation of 1600-1900 meters. Six are from leaf axils of tree Pandanus and one from a bird’s nest fern. The heights of these sites above the ground ranged from 4-8 m. RANGE.— Known only from the type locality, Mindanao Island. SURDUS GROUP Dring (1987) defined the surdus Group as: comprising species with small-sized individuals with vomerine teeth present; nuptial pad large, smooth (absent for species of this group other than surdus, Liem 1970); cutaneus pectoris pres- ent; geniohyoideus lateralis externis inserts on hyoid plate and posterolateral process; and ova unpigmented. The species of this group are ac- PHILIPPINE RHACOPHORIDAE tually moderate-sized, based on Dring’s inter- pretation of moderate for the vermiculatus Group (maximum SVL for males of the four species 27.1-33.6 mm). Maximum SVL for males of the three species referred to the surdus Group is 27.6— 32.4 mm. Philautus emembranatus (Inger) Rhacophorus emembranatus Inger 1954, 33:392 (type loc., Mount McKinley, Davao Prov., Mindanao Island; holotype in FMNH). Inger (1954) described emembranatus on the basis of a unique, female specimen from the mountains of south-central Mindanao Island. To the best of our knowledge no other examples of this species have been recorded from that area. However, a small series from Mt. Hilonghilong in northeastern Mindanao and from Dapitan Peak, Zamboanga Peninsula, Mindanao, are in close agreement with the type and provide us with some information on the variability for this species and a better understanding of how it dif- fers from other species of the surdus Group. DESCRIPTION.—SVL 30.9-43.9 mm for eight females (one measuring 29 mm is not fully ma- ture); 25.1-32.4mm for seven males; head broadest at angle of jaws; HB 103-122% of HL and 37-46% of SVL; snout round-pointed or ending in a knob-like protrusion, most promi- nent for several large females, (similar to that exhibited by Rhacophorus appendiculatus); SnL 30-35% of HB and 34-41% of HL; eye large; ED 80-97% of SnL and 25-31% of HB; tympanum covered by skin, TD 18-33% of ED; nostril much closer to tip of snout than eye; males with vocal sacs; vomerine teeth reduced to small, low, wide- ly separated patches (absent in one example); canthus rounded; lores moderately oblique, slightly concave; fingers with small web at base; disk of third finger largest, its diameter 22-36% of third finger length, and 133-244% of TD; su- barticular tubercles, large low, rounded; no su- pernumerary tubercles but palm wrinkled; inner metacarpal tubercle elongate, outer indistinct; narrow ridge of flesh on outer edge of fourth finger; outer edge of lower arm smooth or nearly so; nuptial pad large, elongate, smooth; SVL 53- 63% of hindlimb length; TL 50-64% of SVL; toes webbed: to approximately distal edge of tubercle or somewhat beyond on inside of first, second, third, and fifth, between penultimate and distal tubercles on fourth (Fig. 4A); diameter of third 197 toe disk 65-86% of diameter of third finger disk; subarticular tubercles low, rounded; inner meta- tarsal tubercle elongate; outer indistinct; skin of dorsum lightly shagreened; tubercles usually ab- sent; venter and lower surface of thighs with moderate to large flattish granules. Cotor.—In preservative, dorsum and upper lateral surfaces gray to dusky tan or golden brown, nearly uniform or occasionally with vague, darker blotches; venter creamy white, uniform except for dark flecks along edge of lower jaw and for some individuals a few on posterior abdomen and under-surface of thighs; limbs without dis- tinct transverse bands. In life, dorsal ground color tannish brown, slate- gray, or dark brown, usually relatively uniform or with a few scattered, darker markings. REPRODUCTION. — The ovaries of the largest fe- male (CAS 183415) held 22 large (2.25 mm), uniform creamy-white eggs. Deposition site and developmental stages have not been observed. CoMPARISONS. — This is one of three Philippine species placed in the surdus Group (Dring, 1987). It is the largest species of the group. It differs from the other species in: the more uniform col- oring of the dorsum; the absence of tubercles on the dorsal surfaces and limbs; the small eye, as a ratio of ED/HB; the small tympanum, as a ratio of TD/ED; and the less dilated toe disks, as a ratio of diameter of third toe disk over diameter of third finger disk (Table 2). Three examples from a population on Mt. Kitanglad, Mindanao are tentatively assigned to this species. The skin is smoother than is typical of the species. HasitAT.—Specimens are from submontane and montane forest, between 800-2100 m ele- vation. RANGE.—Recorded from mountain ranges in Agusan del Norte, Davao, Misamis Occidental, and Zamboanga del Norte Provinces, Mindanao Island. Philautus surdus (Peters) Polypedates surdus Peters, 1863:459 (type loc.: Luzon Island; holotype in ZMB). Philautus williamsi Taylor, 1922:167 (type loc.: Polillo Island; holotype in Philippine Bureau of Science, destroyed during World War II). Rhacophorus lissobrachius Inger, 1954:390 (type loc.: Mount McKinley, Davao Prov., Mindanao Island; holotype in FMNBR). Peters (1863) described Polypedates surdus on the basis of a unique specimen from Luzon Is- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 198 “Snfnadns SnINDpIY_ (9) ‘snpans sninvjIYd (Q) ‘snjouDAquiaa snInDpIYd (B) :S2[d19QN} puke BUIQGaM JO 1U91X9 BUIMOYS ‘1OOJ JO JdRJINS [eIIUDA “p TUNOL] WUWG PHILIPPINE RHACOPHORIDAE land. Taylor (1920) continued to recognize the species as a polypedatid. In 1922 he collected a series of specimens from Mt. Maquiling, Luzon that he identified as Rhacophorus surdus. He did not publish this information and presumably did not compare the specimens with those that he described as P. williamsi. Inger (1954) described Rhacophorus lisso- brachius on the basis of a single female specimen from Davao Province, Mindanao Island. He not- ed that the /issobrachius resembled surdus, fol- lowing comparison with Taylor’s series of that species from Mt. Maquiling. He stated that /is- sobrachius differed in the following characters: snout less projecting, more extensive webbing of fingers and toes, and larger areas of brown on the posterior surfaces of the thighs. A series of about 50 specimens from the moun- tains of northeastern Mindanao as well as north- western Mindanao and Bohol Island are exam- ples of the species represented by the unique type of lissobrachius. The variation exhibited by our series for the characters noted above as well as other characters overlap the variation exhibited by the series of surdus from Luzon, although the mean extent of the webbing of the toes appears to be greater for populations on the southern islands than on Luzon. Because we cannot dem- onstrate any other clear differences in the char- acter states for any of the characters we have used, we regard these populations from Luzon, Bohol, and Mindanao as conspecific. We also predict that exploration of suitable microhabitats in the mountain forests in Leyte and Samar will reveal populations of surdus on those islands. Taylor (1920) described Philautus williamsi, based on four examples from southern Polillo Island and one from Mauban, Tayabas on the east coast of Luzon. Part of the series, including the holotype, deposited in the Philippine Bureau of Science, were destroyed during World War II. We have compared the remaining paratypes (MCZ 14473 and CAS 62254) with the holotype and a series of Philautus surdus from Luzon. Al- though the examples of wi/liamsi are apparently immature, they are in close agreement with the series of surdus in terms of webbing of the toes, the pair of tubercles on the dorsum at the shoul- der level, the proportional characters used, and the color pattern. We regard these populations as conspecific. DESCRIPTION.—SVL 22.5-31.0 for 25 males and 28.5-38.4 for 20 females (The holotype at 199 25.2 mm SVL and lacking vocal slits is assumed to be an immature female); head broadest at an- gle of jaws; HB 107-120% of HL and 39-45% of SVL; snout rounded to round-pointed, occa- sionally strongly projected (particularly for some larger individuals), SnL 30-40% of HB and 33- 43% HL; eye large, ED 86-103% of SnL and 30- 37% of HB; tympanum distinct but covered by skin, strong supratympanic fold, TD 33-49% of ED; upper jaw projecting slightly beyond lower jaw (strongly in a few large individuals); nostril closer to tip of snout than eye; vomerine teeth in small, low patches of only a few teeth (occa- sionally absent or prominent, CAS-SU 20356); vocal sacs present in males; canthus rounded; lores moderately oblique, slightly concave; cu- taneous pectoris muscle present but thin; fingers not conspicuously webbed, no more than result of joining of marginal flanges of skin of fingers; diameter of third finger disk 36-45% of third finger length and 111-192% of TD; subarticular tubercles (except for basal one on third finger) large, round, low; palmar area with folds; nuptial pad large, smooth, elongate; SVL 55-69% of hindlimb length and TL 48-60% of SVL; toes webbed: to distal edge of tubercle or somewhat beyond on inside of first and second, nearly to disk for third and fifth toes, and to penultimate tubercle or beyond on fourth toe (Fig. 4B); di- ameter of third toe disk 83—100% of diameter of third finger disk; subarticular tubercles rounded; supernumerary tubercles elongate inner meta- tarsal tubercle; outer indistinct; skin of dorsum shagreened, nearly always with a very distinct dorsolateral tubercle at shoulder level on either side (22 in a sample of 30), and usually other scattered tubercles including one or more on up- per eyelid, especially for juveniles and males; under surface of head and throat granular for males, less so for females; belly and proximal portion of thighs coarsely granular. Cotor.—In preservative, dorsal ground color from tan to dark, rusty brown (rarely grayish- brown), nearly uniform or with a few darker spots (one pair usually prominently enclosing the pair of shoulder tubercles) or with prominent dark interorbital bar and/or large, dark blotch on body; venter and lower lateral surfaces creamy, lightly to heavily spotted with brown; hindlimbs usually with two or three dark cross-bands; anterior and posterior surfaces of thighs with vague to con- spicuous, rusty colored areas; upper lip usually with a vague to prominent pale bar (of varying 200 width) slanting posteriorly from eye toward or to margin of upper lip. In life, dorsal ground color grayish or slate- brown to dark brown, usually with some darker spots; hindlimbs usually with dark crossbars. REPRODUCTION. — Several batches of frog em- bryos and larvae are from the leaf axils of Pan- danus and the base of arboreal bird’s-nest ferns in original forest at elevations of 1200-1500 m on Mt. Hilonghilong (Alcala and Brown 1982). These batches range in number from 5 to 19. Based on the narrow range of developmental stages exhibited by each batch at the time of collection, each batch probably consists of a sin- gle clutch. Several batches were allowed to con- tinue development by placing them in plastic bags surrounded by moist root-mass material from aerial ferns. In two instances the fully meta- morphosed froglets resulting from this direct de- velopment were identified as Philautus lisso- brachius (now P. surdus). This identification is based on color pattern, rugosity of dorsum, toe structure, and extent of webbing of the toes. The estimated developmental time is 38 to 41 days. For details on developmental stages see Alcala and Brown (1982). ComparRIsons.—See Philautus surrufus and P. emembranatus. HaBITAT.—Specimens from Bohol and Min- danao Islands are primarily from dipterocarp to montane forest at elevations from 500-2000 m (rarely below about 650 m). RANGE.— Bohol, Mindanao, Luzon, and Po- lillo islands. Philautus surrufus new species (Figure 2B) Rhacophorus surdus (part): Rabor and Alcala 1959:340. Brown and Alcala (1970b) referred the speci- mens ina large collection of Philautus from Dap- itan Peak, northwestern Mindanao, to four spe- cies: P. acutirostris, Rhacophorus (= Philautus) emembranatus, lissobrachius, and surdus. Those referred to surdus actually represent a previously undescribed species. Ho totyre.—CAS-SU 21013, a female, col- lected in submontane forest (1800-1900 m) on the west side of Dapitan Peak, about 10 km SE of Masawan, Misamis Occidental Prov., Min- danao Island, by Angel Alcala et a/., April 14, 1959. PARATYPES.— Mindanao Island, Zamboanga PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 del Norte Prov., west side of Dapitan Peak, six to eight km SE Buena Suerte; CAS-SU 20920, 21189-91; Mt. Malindang: FMNH 96043-44; Misamis Occidental Prov., west side of Dapitan Peak, about 12 km E of Masawan: CAS-SU 20955-60; about six to ten km SE of Masawan: CAS-SU 20952, 20962-74, 20976-89, 20995-99, 21023-60, 21062-68, 21070-89, 21092-126, 21145-53, 21156-74, 21181-82, 21194-95, 21658-63, 21667-68, 21670-76; 14 to 15 km SE of Buena Suerte: CAS-SU 20336, 20367, 20375, 20942, 20946-48, 20950-51, 20975, 21183-84, 21654. D1aGcnosis.— Differs from other Philippine species of the surdus Group of Philautus in the following combination of character states: the somewhat smaller size, a reddish dorsal color pattern, the reduced webbing between the toes, and the relatively smooth skin. DESCRIPTION. —SVL 20.0-27.6 mm for 16 males and 26.6-35.4 mm for 14 females; head broadest at angle of jaws, HB 104-119% of HL and 40-47% of SVL; snout rounded to round pointed, SnL 32-39% of HB and 33-43% of HL; eye large, ED 92-110% of SnL and 31-39% of HB; tympanum distinct but covered by skin, strong supratympanic fold; TD 34-50% of ED; snout projecting moderately; nostril closer to tip of snout than eye; vomerine teeth in small, low, widely separated patches of only a few teeth; vo- cal sacs present in males; canthus rounded; lores slightly oblique and slightly concave; cutaneous pectoris muscle present but small; fingers not or barely webbed at base; nuptial pad not discern- able for sample of 12 males; diameter of third finger disk 34-44% of third finger length and 106- 150% of TD; subarticular tubercles large (except for basal one on third finger), rounded; palmar area with some supernumerary tubercles; inner and outer metacarpal tubercles elongate; SVL 56-— 64% of hindlimb length and TL 52-61% of SVL; toes webbed: usually to distal edge of tubercle on inside of first, second, third, and fifth; to or nearly to penultimate tubercle on fourth (Fig. 4C); disks of toes usually smaller than those of fingers; di- ameter of third toe disk 81-96% of that of third finger disk; plantar area with small, scattered tu- bercles; inner metatarsal tubercle elongate; outer lacking; skin of dorsum finely shagreened, tu- bercles on dorsum absent even on eyelids or very rarely a few; belly and basal part of posterior surface of thighs covered with large, flattish gran- ules; those under head and throat usually smaller PHILIPPINE RHACOPHORIDAE or area simply more wrinkled; outer edge of hand with low flange of skin or tubercles; outer edge of lower arm usually with few, low tubercles; prominent whitish tubercles are usually present on the outer edge of the foot. CoLor.—In preservative, dorsum and upper lateral surfaces pale to dark reddish-brown near- ly uniform for some specimens, with scattered reddish-brown flecks or vague spots, or mottled reddish and tan; occasionally specimens are more pale, grayish to tannish with a few large, reddish- brown blotches. Dark crossbands on hindlimbs are most clearly evident for some of the paler specimens. A rusty-red band occurs on the pos- terior surface of the thighs, and sometimes there is a less conspicuous band on the anterior surface. The venter is dirty cream or ivory with few to numerous darker (usually brown) flecks or blotches, most concentrated under the head and throat. Some juveniles lack the dark flecks or spots on the venter. In life, dorsum usually light brown, sometimes limited to the middle region, being darker (black- ish) dorsolaterally, and a reddish blush over the light brown region for many examples; hind limbs with dark transverse bars; and sometimes a dark interorbital bar. REPRODUCTION. — Several gravid females have 12-22 large, unpigmented eggs in the ovaries. Deposition site and development are unknown. Comparisons. — Philautus surrufus differs from P. emembranatus in: its smaller size, larger eye relative to HB, larger tympanum relative to eye, larger third finger disk relative to third finger length (Table 2), and some features of the color pattern. It differs from P. surdus in: the some- what smaller size, (most evident for males, Table 2); the smoother, less tuberculate dorsum; and some features of the color pattern. It differs from both species in the less extensive webbing of the toes (Figs. 4A, B, and C). HaBITAT.—This species was found at eleva- tions ranging from 800 m in upper dipterocarp forest to about 2300 m in montane (moss) forest. It was most abundant at elevations above 1600 m. RANGE.— Recorded thus far from Mt. Malin- dang, Dapitan Peak area in northwestern Min- danao Island. Philautus alticola Ahl Philautus montanus Taylor, 1920:305 (type loc.: Bongao Is- land, Sulu Archipelago; holotype in Philippine Bureau of Science destroyed during World War II). 201 Philautus alticola can only be evaluated on the basis of Taylor’s original description (Taylor 1920) and those of later authors (Ahl 1931; Inger 1954), because the unique holotype in the Phil- ippine Bureau of Science was destroyed during World War II. DESCRIPTION. — Based on those of Taylor and Inger, SVL39 mm for the unique male; HB 93% of HL and 36% of SVL; snout round-pointed, SnL 50% of HB and 47% of HL; ED 84% of SnL and 42% of HB; a weak supratympanic fold, TD 54% of ED; canthus rounded; lores oblique, slightly concave; fingers webbed at base; diam- eter of third finger disk 78% of TD; indistinct fold of skin on outer edge of fourth finger and on outside of lower arm; large nuptial pad on first finger; SVL 60% of hindlimb length; TL 56% of SVL; toes webbed: to disk on outside of second toe, otherwise about two thirds webbed; subar- ticular tubercles well developed; small, conical inner metatarsal tubercle; no fold on outer edge of fifth toe or foot. Cotor.—In life, dorsum uniform lavender- gray, with numerous, small, poorly outlined, yel- low spots; lores and upper lip with minute brown dots; creamy yellow spot anterior to and below tympanum; dark spots along sides; venter dusky white to cream, with small dusky spots on chin and larger spots in belly area; anal region dark with light yellow border; hindlimbs transversely barred. COMPARISONS. — Taylor (1920) suggested that the closest relative might be P. vittiger from Java; Inger (1954) noted that Taylor’s description could just as easily justify a relationship with P. acu- tirostris, leitensis, or petersi. The size, however, SVL 39 mm for a male (Taylor mentions a large nuptial pad on the first finger) is much greater than the maximum (24 mm) for males of acu- tirostris, leitensis, or petersi. In fact, it is greater than that recorded for the otherwise largest Phil- ippine species, Philautus emembranatus, with a maximum SVL of about 33 mm for males. Ac- tually, the size (SVL 39 mm for the unique male), the proportions SnL/HB, ED/HB, TD/ED, and 3FD/TD (based on measurements in Taylor’s original description), head less broad than long, color pattern, general appearance (See Plate 3, Fig. 5, Taylor 1920), and habitat suggest to us that Taylor may have erroneously based his de- scription of Philautus montanus on a specimen of Polypedates macrotis. We hesitate to refer this species to the synonymy of macrotis, however, 202 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 until the presence of macrotis on Bongao Island is confirmed. REPRODUCTION. — Data are not available for this species. HasitAT.— Taylor commented that the spec- imen was found in a small pool near the summit of Bongao Peak (700 m), Bongao Island. RANGE. —Bongao Island, Sulu Archipelago. SPECIMENS EXAMINED. — None. Polypedates Tschudi Polypedates Tschudi, 1838:75. Type species Hyla leucomystax Gravenhorst, 1829:26. Elophila Duméril and Bibron, 1841:517. Trachykyas Fitzinger, 1843:31. Type species by original des- ignation: Polypedates rugosus Duméril and Bibron, 1841. D1AGnosis.—Size moderate to large (SVL at maturity 40-90 mm for Philippine species); skin of dorsal surfaces smooth or shagreened, usually with some ossification with skull; disks of fingers and toes dilated, completely circumscribed by groove; fingers webbed only at base (P. eques an exception); toe disks smaller than those of fin- gers; toes webbed almost to disks except for fourth toe (for Philippine species); hindlimbs long; anal ornamentation prominent but other dermal or- namentation usually absent; vomerine teeth present; vocal sacs present in some species, ab- sent in others; neuropophysis of eighth vertebra attached dorsolaterally on centrum; terminal phalanges Y-shaped; omosternum slightly forked at base; HB less than HL for Philippine species; Wolffian ducts convoluted; petrohyoideus pos- terior muscle three slips; eggs small, without pig- ment (clutch size 150-400 for Philippine spe- cies), and deposited in foam nests on vegetation near quiet water pools; feeding larvae are ovoid and robust (Inger 1985b). KEY TO PHILIPPINE SPECIES OF POLYPEDATES la. Males without vocal sacs; dorsum uni- form or with two vague, broad, dark, lon- Pitudinalssteipesieeee wr. ee) Te 2 1b. Males with vocal sacs; dorsum usually with four prominent, dark longitudinal stripes (carelysuniform)) 222 = & P. leucomystax 2a. Usually dark brown stripe from eye through tympanum to axilla or beyond; skin of head usually involved in ossifi- cation in frontoparietal area; two promi- nent whitish tubercles below vent Weis MERI antd es Ts P. macrotis 2b. No dark brown stripe on side of head; skin of head free of ossification; two prominent whitish tubercles above vent ... P. hecticus Polypedates hecticus (Peters) Polypedates hecticus Peters, 1863:457 (type loc.: Loquilocum, Samar Island; holotype lost). This species was based on a single example from Samar Island. Giinther (in litt., October, 1991) states that the holotype cannot be found in the collections at the ZMB. It is further un- likely that the holotype of P. hecticus was ever transferred to any other institution, because the holotype (ZMB 4920) of P. surdus from Luzon, which Peters described in the same short paper in 1863, is there. The specimen, furthermore, was examined by Wolf prior to publication of his revision of the genus Rhacophorus (1936). Wolf retained the species, indicating that he re- garded it as being closely related to P. leuco- mystax. Inger (1954) also retained hecticus as a valid species of Rhacophorus. In view of the fact that the herpetofauna of Samar and its small satellite islands has not been surveyed exten- sively and is poorly represented in collections, we follow Wolf and Inger in retaining hecticus as a possible valid species. However, based on Wolf's opinion as to relationships and the narrow head relative to length, we tentatively assign hec- tus to the genus Polypedates to which leucomys- tax and macrotis (includes /inki) are already as- signed. DESCRIPTION. — Based on the descriptions of Peters, Ahl, Wolf, and Inger: SVL 51 mm for one male (Ahl 1931); HB 82% of HL and 33% of SVL; snout truncated, projecting; ED about equal to SnL; tympanum distinct; TD about two thirds of ED; canthus acute; lores vertical, con- cave; vomerine teeth in two small, oblique patch- es; males without vocal sacs; fingers webbed at base; diameter of first finger disk about half of TD; subarticular tubercles weakly developed; SVL 61% of hindlimb length; TL more than one half SVL: toes, except the fourth, webbed almost completely; toe disks similar to finger disks; su- barticular tubercles large; inner metatarsal tu- bercle elongate; skin of dorsum granular; a dis- tinct dorsolateral fold beginning at eye; belly and undersurface of thighs granular; two enlarged, whitish tubercles above vent. PHILIPPINE RHACOPHORIDAE CoLor.—In preservative, dorsal ground color grayish blue; dorsolateral fold whitish, bordered by black; white line from tip of snout along upper lip, widening in front of tympanum; dark flecks in the temporal region, anterior and posterior to tympanum; light blue with blackish flecks on sides; limbs brownish with more or less distinct flecks, particularly on posterior surface of legs; venter whitish. CoMPARISONS. — Wolf (1936) indicated a close relationship to P. Jeucomystax linki. Absence of vocal sacs, if verified, is in agreement with P. macrotis as opposed to P. /. leucomystax for this character state. Based on the unique example and Peter’s description, it would appear to differ from P. macrotis in some features of the color pattern, the slightly narrower head, and shorter hin- dlimbs. REPRODUCTION. — No information is available for this species. HasitaT. — Unknown. RANGE.— Known only from the type locality, Samar Island. MATERIAL EXAMINED. — None. Polypedates leucomystax leucomystax (Graven- horst) Hyla leucomystax Gravenhorst, 1829:26 (type loc.: Java; syn- types in Breslau Mus., according to Stejneger 1907). Hyla sexvirgata Gravenhorst, 1829:28 (type loc.: Java; holo- type in Breslau Mus., according to Stejneger 1907). Hyla quadrilineata Weigman, 1835:260 (type loc.: Manila, Lu- zon Island, Philippines; holotype disposition unknown). Polypedates rugosus Duméril and Bibron, 1841:520 (type loc.: Java and Manila, Luzon Island; syntypes in MNHN). Polypedates megacephalus Hallowell, 1860:507 (type loc.: Hong Kong; holotype in ANSP). Hylarana longipes Fischer, 1885:47 (type loc.: Pagat, Borneo; holotype in BMNH). Rhacophorus braueri Vogt, 1911:180 (type loc.: Formosa; ho- lotype in ZMB). Bartlett (1894) was the first to note that both the striped and unstriped color patterns occurred in some Bornean populations. He, therefore, placed Rhacophorus quadrilineata in the syn- onymy of Rhacophorus leucomystax. DESCRIPTION.—SVL 49.7-56.6 mm for males and 57.5-75.2 mm for 15 females (one at 59 mm was, however, immature); HB 89-100% of HL and 31-38% of SVL; snout round pointed, SnL 41-48% of HB and 40-45% of HL; eye moderate, ED 61-80% of SnL and 27-38% of HB; tym- panum large, TD 64-88% of ED; canthus sharply rounded; lores moderately oblique to nearly ver- 203 tical, not or shallowly concave; fingers with basal webs between first and second and second and third; diameter of third finger disk 27-33% of third finger length and 52-82% of TD; subarti- cular tubercles large, elongate; supernumerary tubercles present; inner metacarpal tubercle elongate; outer, small, rounded; hind limbs long, SVL 57-67% of hind limb length and TL 49- 55% of SVL; toes webbed: between tubercle and disk for first, second, third, and fifth toes, and to distal tubercle on fourth; diameter of third toe disk 57-78% of diameter of third finger disk; subarticular tubercles large, pointed distally; in- ner metatarsal tubercle large, oval; outer absent or very small; skin of dorsum smooth or nearly so; skin of head partially ossified with skull for adults; venter coarsely granular except beneath head and throat which is smooth to finely gran- ular. CoLor.—In preservative, dorsal ground color grayish to grayish-tan or brown, with four vague to conspicuous, narrow, dark, longitudinal stripes (sometimes broken or incomplete), occasionally only with dark spots; usually a narrow brownish band on the upper half of the lores and between the eye and forelimb; limbs with dark, transverse bars; venter ivory-white, with or without dark spots on chin and throat. In life, dorsal ground color grayish or yellowish green to green or brownish, usually with four dark, longitudinal stripes or rows of spots (oc- casionally uniform); hindlimbs with dark trans- verse bars. REPRODUCTION. —On Negros Polypedates leu- comystax breeds from April to December, the rainy months (Alcala 1982). The foamy egg masses of this species occur on leaves of trees, shrubs, or grasses overhanging isolated pools and ponds or quiet pools of temporary streams. Oc- casionally the masses occur on rocks on the banks of such pools. The egg masses have 150-225 eggs or embryos. Nests with as many as 800-900 eggs have been reported by Taylor (1921) and Vil- ladolid and del Rosario (1930). Larvae break free from the foam-mass and fall or wriggle into the water two to three days after deposition (Alcala 1962). For a discussion of the habits of the free swimming larvae and a description of develop- mental stages, see Alcala (1962). COMPARISONS.—See comparisons under Po- lypedates macrotis. As to geographic variation among populations of P. /. leucomystax, Inger (1966) notes that the spotted pattern on the dor- 204 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 5mm 5mm 5mm 5mm Ficure 5. Ventral surface of foot, showing extent of webbing and tubercles: (a) Polypedates macrotis; (b) Rhacophorus appendiculatus; (c) Rhacophorus bimaculatus; (d) Rhacophorus everetti. PHILIPPINE RHACOPHORIDAE sum is more frequent in Javan and Sumatran populations than in Bornean and Philippine pop- ulations, and that samples of males from Phil- ippine, Bornean, Javan, and Sumatran popula- tions exhibit some significant differences in body proportions when measurements such as head breadth, third-finger disk diameter, or tympa- num diameter are expressed as ratios of SVL. HABITAT. — This species occupies a wide range of habitats including lowland forest and most man-controlled lowland cultivated areas (Brown and Alcala 1964, Inger 1966). RANGE.—India; Taiwan; northern Indochina to Borneo, Sulawesi, and the Philippines where it is found on many of the small as well as large islands. Polypedates macrotis (Boulenger) Rhacophorus macrotis Boulenger, 1891:282 (type loc.: Baram district, Sarawak, Borneo; holotype in BMNH). Polypedates linki Taylor, 1922b:276 (type loc.: Jolo Island, Philippines; holotype in CAS). DESCRIPTION. —SVL 34.8-46.7 mm for 16 ma- ture males and 59.9-71.3 mm for six mature fe- males; HB 89-103% of HL and 30-34% of SVL; snout rounded as viewed ventrally; SnL 38-46% of HL and 42-48% of HB; eye large, Ed 71-92% of SnL and 32-41% of HB; supratympanic fold encroaching on tympanum dorsally; TD 50-75% of ED (one of eight greater than 66%); canthus sharply rounded; lores slightly concave, oblique; vomerine teeth in oblique rows from anterior edge of choanae; males lack vocal sacs; fingers webbed at base; diameter of third finger disk 25- 34% of third finger length and 56-93% of TD; subarticular tubercles rounded; supernumerary tubercles present; inner metacarpal tubercle elongate; SVL 58-68% of hindlimb length; TL 51-58% of SVL; toes webbed: almost to disk on inside of first, second, third, and fifth, and be- yond penultimate tubercle for fourth (Fig. SA); diameter of third toe disk 64-85% of third finger disk; subarticular tubercles rounded; no inner metatarsal tubercle present, outer usually absent; skin of dorsum nearly smooth; low whitish ridge on outer forearm; underside of head and poste- rior to axillary region smooth; belly and posterior surface of thighs coarsely granular; dark colored patch around vent usually bordered by two prominent white tubercles below and occasion- ally by conspicuous whitish granules above. Cotor.—In preservative, dorsal ground color grayish, tan or medium brown, usually with few 205 to numerous small, darker spots; or two broad, dark, longitudinal stripes, usually a faint to dis- tinct brown band narrowing posteriorly extend- ing from eye to axilla or beyond and covering or partially covering tympanum; limbs with dark cross bars; venter cream to light tan, frequently with brown spots laterally and under head and throat. REPRODUCTION. —Eggs, clutch size 300-400, are deposited in foam nests attached to leaves of shrubs or trees and occasionally grasses above or bordering pools in small streams (Inger 1966, Brown and Alcala 1983). See Inger (1985b) for a description of larva and larval development. CoMPARISONS. — Polypedates macrotis is sim- ilar to P. leuacomystax and may be closely related to that species. It differs from /eucomystax in the absence of vocal sacs, the typical color pattern, the longer hindlimbs relative to snout-vent length, the broader head relative to HL, and size of the tympanum relative to the eye. As noted by Inger (1966), the available sample from the Palawan population indicates a smaller size (SVL) than for members of the Bornean population. The limited sample from the Jolo Island (Sulu Ar- chipelago) population does not permit a similar comparison. HABITAT.—Specimens from Palawan were from banks of streams or ditches at the edge of the forest or in coconut groves at low elevations. RANGE. — Philippines: Palawan I., Busuanga I., Calauit I., Jolo I., Dumaran I., Borneo. Rhacophorus Kuhl and van Hasselt Rhacophorus Kuhl and van Hasselt, 1822:104. Type species by monotypy: Rhacophorus moschlatus Kuhl and van Has- selt, 1822. Racophorus Schlegel, 1826:239. Emendation of Rhacophorus Kuhl and van Hasselt, 1822. Leptomantis Peters, 1867:32. Type species by monotypy: Lep- tomantis bimaculata Peters, 1867. Liem (1970) states that the group of species referred to Rhacophorus share a combination of character states that separate them from other genera by definite gaps. He also notes that three species, R. appendiculatus, R. everetti, and R. microtympanun represent a subgroup that differs in several respects from other species assigned to the genus. Two of these three are among the four species occurring in the Philippines. D1AGnosis. — Size moderate to large (SVL 30- 70 mm at maturity for Philippine species); HB equal to or greater than HL; dermal ornamen- 206 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 tation on limbs of some species and around vent; disks of fingers and toes moderate to large, com- pletely circumscribed by groove; finger webbing varying from basal to almost fully webbed; hin- dlimbs long; vomerine teeth usually present (present in all Philippine species except R. bi- maculatus), vocal sacs present in males; terminal phlanges Y-shaped; omosternum forked at base; Wolfhan ducts convoluted; petrohyoideus pos- terior muscle three slips; eggs small to moderate, unpigmented; clutch size 50-400 for those spe- cies known for this character; eggs deposited in foam nests attached to leaves of trees or shrubs overhanging quiet pools or streams (site un- known for R. bimaculatus and R. gauni); larvae adapted to quiet pools or, in the case of R. bi- maculatus and R. gauni, to streams with mod- erate to strong currents (Inger, 1985b). KEY TO THE PHILIPPINE SPECIES OF RHACOPHORUS la. Web between outer fingers reaching distal subarticular tubercle or beyond 2 lb. Web between outer fingers not reaching distal subarticular tubercle 3 2a. First finger webbed to distal tubercle, oth- er fingers to disk; a flap of skin dorsal to anus; prominent whitish anal tubercles AC Kini gyi tit maemieme Ba. write axe R. pardalis 2b. Inner fingers webbed only at base; outer fingers to distal tubercle; a small flap of skin may be present dorsal to anus; usu- ally several prominent white tubercles mostly ventral to vent ........ R. bimaculatus 3a. Fingers webbed beyond basal tubercle, web reaching the distal subarticular tubercle of fourth finger; a dermal flap on either side below the vent _. R. appendiculatus 3b. Fingers webbed only at base, web not reaching the distal subarticular tubercle of fourth finger; no dermal flap below vent RUG iit A 4 2 Borneo l 12 11 4 13 6 Sulawesi 1 3 3) Philippines 2 1 9 8 3 it 4 * Mainland Southeast Asia includes: Burma, Thailand, Laos, Vietnam, Cambodia, and the Malay Peninsula. > Polypedates hecticus is probably not a valid species. mantellid genera and Boophis in the Rhaco- phorinae were isolated (or persisted) in Mada- gascar. The rhacophorine stock remaining in Af- rica is represented by the genus Chiromantis. The rhacophorid stocks, ancestral to the seven genera occuring in Asia, were present in India when the Indian plate collided with the Eurasian plate in the Oligocene, about 35 million years BP. They propose over-water dispersal into the Philip- pines. Channing (1989), like Duellman and Trueb (1985), attributed the present distribution of gen- era to vicariance, but dates it from the breakup of Pangaea about 200 million years ago. Chan- ning believes that the Asiatic rhacophorids are polyphyletic in origin, and that the stocks an- cestral to the present genera were in Pangaea. Phylogenetic relationships of the genera of the Rhacophoridae were reported by Liem (1970) and Channing (1989). Liem’s analysis placed the Asiatic Buergeria as the sister group to the other of the Asiatic genera plus the African Chiro- mantis, and Philautus as the sister groups to the remaining Asiatic genera. He placed Nyctixalus as the sister group to Chiromantis and Polype- dates as the sister group to Rhacophorus. Chan- ning, in his analysis of Liem’s data, placed Phi- lautus as the sister group of the African Chiromantis, Rhacophorus as the sister group of Chirixalus, and Nyctixalus as the sister group of Theloderma. It is proposed that the rhacophorid genera ex- panded their ranges by dispersing eastward and later some of them southeastward through main- land Asia during the Oligocene. At least five of the seven generic stocks reached the Sunda Shelf area, possibly as early as the late Oligocene. At least four genera subsequently dispersed to Bor- neo and the Philippines. The nature and number of species associated with the generic stocks at the time of the early dispersal events are unknown. During the course of dispersal numerous speciation events oc- curred, resulting in current species richness and endemism for the subareas recognized in this study. The genus Chirixalus (seven species) is known from India, China, southeast Asia, Tai- wan, and Japan. The genus Theloderma (eight species) is recorded from China, southeast Asia, and Sumatra. Buegeria (seven species) is record- ed from Taiwan and the Ryukyu Islands. Generic ranges and subarea species richness and endem- ism patterns for the four genera that reach the Philippines are summarized in Table 4. PHILIPPINE GEOLOGICAL HISTORY The Philippines are a group of islands sepa- rated by the China Sea from the Asiatic mainland and separated by relatively narrow marine chan- nels from Borneo to the southwest and the San- gihe chain and Sulawesi (Celebes) to the south. The latter, like the Philippine Islands, are part of Wallacea. The Philippine Islands in their present form rose from the ocean floor since the Eocene. The older rock formations are capped by thick marine sediments (see Hamilton 1979, Heaney 1991). The oldest sub-aerial islands of the present Phil- ippines are believed to be parts of Mindanao and PHILIPPINE RHACOPHORIDAE Luzon (late Oligocene to beginning Miocene, 25- 30 million years BP). Palawan (as a group of small islands) probably became sub-aerial by the beginning of the Pliocene (6 million years BP); Leyte and Mindoro by the late Miocene to early Pliocene (8-10 million years BP); and Negros, Panay, and the Sulus by the middle to late Plio- cene (2-4 million years BP). Pleistocene activi- ties (uplift, volcanic, and tectonic) greatly altered the configuration of the islands. Current hypotheses concerning Pleistocene sea- level changes propose a lower level of 160 m or possibly 200 m about 160,000 years BP and a lower level of 120 m about 16,000-18,000 years BP (Ollier 1985, Heaney 1985). These changes resulted in greatly enlarged aggregate islands and narrowing of persisting channels (Fig. 6). Min- doro and some of the smaller volcanic islands such as Camiguin, Tablas, Romblon, Sibuyan, as well as Siquijor, Camotes, Lubang, and Ma- rinduque were not joined with any of the large island-aggregates in the late Pleistocene. Frag- mentation of the larger islands, including Pala- wan and Mindanao on the entryways, also oc- curred during Pleistocene periods of reduced glaciation but are not yet adequately documented (see McLean 1980). PHILIPPINE RHACOPHORIDS AND THEIR DISTRIBUTION PATTERNS The living Philippine rhacophorids are a clas- sic example of a depauperate island fauna. The fauna, as presently known, is an assemblage of 18 species compared to 30 in Borneo (the im- mediate source area) and 88 in the Asiatic main- land (the general source area). The present as- semblage includes populations of species conspecific with populations in Borneo or else- where and endemic species. The endemics may be the result of speciation events that have taken place within the Philippines, or may be the only known populations of species that evolved else- where. At least five generic stocks dispersed into the Sunda Island subarea. One (Theloderma) is pres- ently known only from Sumatra in this subarea. The other four (Nyctixalus, Philautus, Polype- dates, and Rhacophorus) are recorded through- out the Sunda Shelf Islands and have dispersed into the Philippines. Nyctixalus.—This genus fits the category of a typical relict, based on its present peripheral range relative to probable area of origin, the limited 213 ranges (peripheral island areas for two of the spe- cies), and the small number of living species, three recognized (Inger 1985). It is represented in the Philippines by two, dipterocarp forest spe- cies, one on Palawan that also has populations on Borneo, Sumatra, and Malaya, and an en- demic species in the greater Mindanao island group (Table 5). The Palawan species, N. pictus, is probably a relatively recent entry, perhaps as recently as the low sea-level period in the late Pleistocene. The ancestral stock to N. spinosus probably invaded the greater Mindanao group at an earlier point in time. Neither species has suc- cessfully colonized more peripheral Philippine islands. Polypedates and Rhacophorus.—These two genera are represented in the Philippines by three and four species respectively. They have no en- demic representatives. Both genera are highly successful in the sense that their current ranges extend from India (the ultimate source area) into the Philippines as compared to Nyctixalus, whose range is presently in peripheral subareas. One species of Polypedates (leucomystax) 1s well adapted to cultivated fields and other man- controlled habitats. Dispersal opportunities re- sulting from these adaptations may account for its wide range from India to the Philippines and throughout the latter except for the greater Pa- lawan group. Three additional species are present in Borneo, but only one of these (P. macrotis) is in the Philippines, and it is limited to greater Palawan and the Sulus (Tables 4 and 5). This shallow dispersal into the Philippines suggests a recent arrival, perhaps the time of most recent lower sea levels. This is supported by the fact that P. macrotis occurs along forest margins and some cultivated fields, a habitat that may en- hance dispersal opportunities. Polypedates hec- ticus is a doubtful species. There are no endemic Philippine species of Rhacophorus, but four of the 13 Bornean species are now known to occur in the Philippines. All are primarily in the dipterocarp forest habitat, although R. pardalis is also recorded from the submontane forest zone. Rhacophorus everetti 1s limited to the greater Palawan group of islands at the western entryway, but the other three spe- cies have dispersed from Mindanao into greater Negros and greater Luzon islands but not Min- doro (see Fig. 6 and Table 4). This expanded dispersal within the Philippines as compared to Polypedates macrotis suggests an earlier arrival 214 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 G122° Batan Islands Se ge KEY Larger Pleistocene islands. © Knowledge of the faunas of these islands is very limited. 4 Range on both Bohol and Leyte. in the Philippines. This is further supported by the fact that these Rhacophorus species are pri- marily forest species and thus have more re- stricted dispersal opportunities than does P. ma- crotis. Philautus.—Like Rhacophorus, Philautus is highly successful in terms of both its extensive range and a large number of living species (Table 4). It differs in that speciation events in the su- bareas rather than dispersal events account pri- marily for species richness in each region. This genus has nine species in the Philippines, a great- er diversity than occurs in any other rhacophorid genus. Eight of these are endemic there. Simi- larly, Borneo has 12 species with 11 endemic of —_— them. The one non-endemic species in each in- stance, P. /ongicrus, is the only species recorded from greater Palawan, the western entryway. The other eight species, all endemic, are in islands along the eastern entryway. Six species are lim- ited to the greater Mindanao group, one has dis- persed into the greater Luzon group, and one is endemic to Mindoro. The Philippine species are assigned to three of the five generic subgroups recognized by Dring (1987). The species of Phi- lautus from Borneo and the Philippines also oc- cur primarily in the submontane and montane forest habitats. All of these factors support the hypothesis that the Philippine Philautus (except for P. ongicrus) Ficure 6. Map of the Philippines (modified from Heaney 1990), showing present islands (pale stippled areas) and enlarged, late Pleistocene islands (dark stippled areas) based on the presumed extent of lower sea levels at that time. 216 are probably derived from at least two stocks that dispersed into greater Mindanao from Bor- neo somewhat earlier than did Polypedates or Rhacophorus. The limited dispersal within the Philippines of species of Philautus as compared to Rhacophorus is probably the result of reduced opportunities due to the more montane habitats occupied by the species of Philautus. The paucity of rhacophorids on greater Pala- wan (four nonendemic species) compared to the rest of the Philippine islands is not readily ac- counted for based upon our present knowledge. Another peculiar feature of the rhacophorid fau- na of greater Palawan is the absence of the wide- spread, highly adaptable Polypedates leucomys- tax that is widely dispersed through the rest of the Philippines. Information on rhacophorid populations in the forests of Leyte, Samar, and some of the moun- tain areas of Luzon is needed to better under- stand some distribution patterns and to deter- mine whether or not additional areas of endemism exist. This has some urgency, because the rain forests, especially the dipterocarp for- ests, are being cut at an accelerated rate on most of the large islands. The microhabitat preferences of the Philippine rhacophorids as well as their modes of reproduction, with the exception of P. leucomystax and P. macrotis, suggest that they will be endangered to the extent that the rain forests are depleted. SUMMARY Four of the seven recognized genera and 18 of the 159 recognized species of Asiatic rhaco- phorids occur in the Philippines. Ten of the spe- cies (including the doubtful Philalutus aticola and Polypedates hecticus) are endemic, and eight have ranges that include Borneo (the immediate source area), and in some instances, the Sunda Islands and even areas of mainland Asia. Nyctixalus has the characteristics of a relict genus. Philautus is a highly successful genus in both number of species and its extensive range, from India through southern Asia to the Phil- ippines. The genus is highly endemic in all su- bareas. For the Philippines, eight of nine species are endemic. Polypedates and Rhacophorus have even larger ranges than Philautus extending into southern temperate regions in eastern Asia. Ex- cluding the very doubtful Polypedates hecticus, only two nonendemic species of Polypedates and PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 four nonendemic species of Rhacophorus occur in the Philippines. These differences in species richness, endem- ism, and distribution patterns among the genera of Philippine rhacophorids are attributed to: probable differences in time of dispersal into the archipelago, differences in their ecology and its possible effects on dispersal within the Philip- pines, and geological events that also served to regulate dispersal opportunities. RESUME Quatre des septs genres reconnus et 18 des 159 éspéces de rhacophorides asiatiques reconnues surviennent dans les Philippines. Dix des éspéces (y compris le douteux Philalutus aticola et Po- lypedates hecticus) sont endémiques, et huit ont des rangées qui comprennent Borneo (l’immé- diate source), et quelques fois, les Iles Sunda et méme certains régions du continent asiatique. Nyctixalus a les charactéristiques d’un genre veuf. Philautus est un genre plain de succés dans les deux éspéces et sa vaste rangé, qui s’étend de l’Inde aux Philippines passant par le Sud de l’Asie. Le genre est fortement endémique dans toutes les sous-régions. Aux Philippines, huit sur neuf des éspéces sont endémiques. Polypedates et Rhacophorus ont méme des rangées plus éten- dues que Philautus, s’*étendant en régions plus tempérées a l’Est de I’Asie. Sans compter le dou- teux Polypedates hecticus, seulement deux és- péces non-endémiques de Rhacophorus survien- nent dans les Philippines. Ces différences dans la richesse des éspéces, endémisme, et la distribution des modéles par- mis les genres des rhacophorides Philippins sont attribuées a: de probables differences en temps dans la dispersion dans |’archipel, des differences dans leur écologie et ses effets possible dans la dispersion dans les Philippines et, des événe- ments géologiques qui ont aussi servi a ajuster des opportunités dispersées. ACKNOWLEDGEMENTS Weare deeply indebted to the American Philo- sophical Society for funds that enabled the senior author to complete a field trip to the Philippines in February 1992. We are indebted to E. N. Ar- nold (BMNH), A. Dubois (NMHN), K. Klem- mer (SMF), R. Giinther (ZMB), R.R. Inger (FMNH), E. Williams (MCZ), G. Zug and R. Crombie (USNM), C. J. McCoy (CM), and A. E. PHILIPPINE RHACOPHORIDAE Leviton (CAS) for permission to examine col- lections in their institutions. We also wish to thank our colleagues, A. Leviton, R. Drewes, and J. Vindum, for many helpful suggestions and for verification of some anatomical observations. We especially thank R. F. Inger and G. Zug for their thorough reviews and constructive comments during preparation of the final manuscript. We thank R. Heaney for permission to use the map that was modified for this paper and D. S. S. Liem for permission to reproduce Figure 1. Drawings and maps were prepared by P. Kaczmarek and C. Sudekum, California Academy of Sciences, and Panfilo Legaspi, Silliman University, Phil- ippines. APPENDIX A DISTRIBUTION OF RHACOPHORID FROGS IN THE PHILIPPINE ISLANDS Western Region. — Palawan Island: Nyctixalus picta, Philautus longicrus, Polypedates macrotis, Rhacophorus everetti. Balabac Island: Philautus longicrus, Polypedates macrotis. Dumaran Is- land: Polypedates macrotis. Busuanga Is- land. Polypedates macrotis. Calauit Island: Po- lypedates macrotis. Southern and Eastern Region. — Bongao Island (Sulus): * Philautus alticola. Jolo Island (Sulus): Polypedates macrotis, P. leuacomystax. Minda- nao Island: Nyctixalus spinosus, Philautus acu- tirostris, P. emembranatus, P. poecilus, P. surdus, P. surrufus, Polypedates leucomystax, Rhaco- phorus appendiculatus, R. bimaculatus, R. par- dalis. Basilan Island: Nyctixalus spinosus, Phi- lautus acutirostris, Rhacophorus appendiculatus, R. bimaculatus. Camiguin Island: Polypedates leucomystax, Rhacophorus pardalis. Dinagat Is- land: Rhacophorus appendiculatus, R. pardalis. Bohol Island: Nyctixalus spinosus, Philautus ley- tensis, P. surdus, Polypedates leucomystax, Rha- cophorus appendiculatus, R. bimaculatus, R. par- dalis. Leyte Island: Philautus leytensis, Polypedates leucomystax. Samar Island: * Poly- pedates hecticus, P. leucomystax. Poro Island: Polypedates leucomystax. Pacijan Island: Poly- pedates leucomystax. Central Region.—Cebu Island: Polypedates leucomystax. Masbate Island: Polypedates leu- comystax. Cagayan Island: Polypedates leuco- mystax. Siquijor Island: Polypedates leucomys- tax, Rhacophorus pardalis. Calagaan Island: Polypedates leucomystax. Carabao Island: Po- 217 lypedates leucomystax. Gigante South Island: Polypedates leuacomystax. Guimaras Island: Po- lypedates leucomystax. Inanipulagan Island: Po- lypedates leucomystax. Pamionlon Island: Po- lypedates leucomystax. Pan de Aczucar: Polypedates leucomystax. Sicogon Island: Poly- pedates leucomystax. Negros Island: Polypedates leucomystax, Rhacophorus appendiculatus, R. pardalis. Panay Island: Polypedates leucomys- tax. Northern Region.—Luzon Island: Philautus surdus, Polypedates leucomystax, Rhacophorus bimaculatus, R. pardalis. Polillo Island: Philau- tus surdus, Rhacophorus pardalis. Mindoro Is- land: Philautus schmackeri. Tablas Island: Po- lypedates leucomystax. Caluya Island: Polypedates leucomystax. Semirara Island: Po- lypedates leucomystax. Sibay Island: Polypedates leucomystax. * doubtful species APPENDIX B SPECIMENS EXAMINED For brevity, only island names and museum numbers are provided. Precise locality data are available from the museum catalogs. Nyctixalus spinosus.—Basilan I.: CAS 60168, 60379; Mindanao I., Agusan del Norte Province, west slopes of Mt. Hilong-hilong: CAS 139319, 133561, 133629-32, Bunawan: CM 3420 (ho- lotype of Hazelia spinosa); Bohol I., about 10 km SE of Sierra Bullones: CAS-SU 22788. Philautus acutirostris.—Mindanao I., eastern region: ZMB 5690 (holotype of Ixalus acutiros- tris), Mt. Hilonghilong, Agusan del Norte Prov.: CAS 133140, 133164-7, 133211-2, 133259, 133262, 133289-90, 133298-300, 133305-6, 133308-9, 133311-2, 133334-6, 133394, 133649- 50; west side of Dapitan Peak, Zamboanga del Norte Prov.: CAS-SU 20883-4, 20886-7, 20890- 7, 20900, 20903, 20912-3; west side of Dapitan Peak, Misamis Occ. Prov.: CAS-SU 20644, 20646-58, 20660, 20686-90, 20692-4, 20695- 706, 20728-30, 20780-92, 20798-801, 20803-6, 20851, 20861-7. Saub, Cotobato Coast: MCZ 23194-97: Davao Prov., Mt. Apo: USNM 34781 (holotype of Philautus woodi), USNM 34782 (paratype); Abungabung, Basilan I.: CAS 60145 (holotype of Philautus basilanensis); MCZ 14467- 68 (paratypes of Philautus basilanensis). Philautus leitensis.— Bohol I., Sierra Bullones (Dusita) area: CAS-SU 22775, 22819-53; Can- 218 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 taub area CAS 139011-15; Leyte I., Calabian: MCZ 23192; BMNH 96.12.11.32 (holotype). Philautus longicrus.— Palawan I.: BMNH 1947.2.6.29 (cotype); CAS 64093; Malabo (near Central Pk.): CAS-SU 24174; 2 km W of Iwahig: CAS-SU 24246; 3-5 km SW of Iwahig: CAS- SU 24211-12, 24218, 24236, 24272; south slope of Thumb Peak: CAS-SU 24204-210, 24220-25, 24227-35, 24237-39, 24241, 24243-44, 24247- 60, 24262-71, 24273-79, 24286, FMNH 105501- 02; Brooke’s Point: FMNH 51351. Philautus schmackeri.— Mindoro I., Mindoro Occidental Prov., Mangarin: CAS-SU 9840; Mindoro Oriental Prov., west of Tarogin Barrio about 30 km south of Calapan: CAS-SU 22314, 22345-48, 22363-64, 22379, 22394, 22507, 22611, 22623-24, 22645-51; southeast slope of Barawanan Peak: CAS-SU 16859-86, 22315-24, 22349-56, 22365-78, 22381-93, 22395-415, 22459-63, 22470-80, 22495-506, 22508-75, 22606-10, 22613-19, 22625-28, 22644, 22652- 76, 22735-39, 22750-73, 24054; Mt. Halcon: SMF 7035 (holotype); south slope of Mt. Halcon: CAS-SU 22467-69, 22490-94: east slope of Mt. Halcon: CAS-SU 22605, 22629-41, 22642-43; Mt. Dulangan: BMNH (syntypes of [xalus min- dorensis). Philautus emembranatus.— Mindanao I., Da- vao Prov., Mt. McKinley: FMNH 50684 (ho- lotype); Agusan del Norte Prov., south and southwest slopes of Mt. Hilonghilong: CAS 133237, 133252, 133305-06, 183415; Misamis Occidental Prov., W slope of Dapitan Peak: CAS-SU 20338, 20341, 20349, 20356, 20362- 63, 20366, 20368, 20371-73, 20949, 20961, 21175, 21641-42; Zamboanga Prov., slopes of Dapitan Peak: CAS-SU 20365, 21192, MCZ 39054; Bukidnon Prov., Mt. Kitanglad: FMHN 250626-28. Philautus surdus.—Luzon I.: ZMB 4920 (ho- lotype of Polypedates surdus); Mt. Maquiling: MCZ 23280-83, 23285-87; 87400-17; Polillo L.: MCZ 14473 and CAS 62254 (paratypes of Phi- lautus williamsi), Mindanao I., Davao Prov., Mt. McKinley: FMNH 50683 (holotype of Jisso- brachius), Agusan del Norte Prov., Mt. Hilon- ghilong: CAS 133263, 133343, 133523, 133534, 133646-47, 133791, 133793, 182565, 182568; Zamboanga del Norte Prov., 6-9 km southeast of Buena Suerte on Dapitan Peak: CAS-SU 20353-55, 20643; MCZ 39055; Misamis Occi- dental Prov., 8-22 km southeast of Buena Suerte on Dapitan Peak: CAS-SU 20337, 20339, 20342- 46, 20348-50, 20356, 20359-60, 20370, 20940- 41, 20945, 21061, 21069, 21144, 21176-80, 21185, 21651-53, 21655, 21665-66; 21669) 21677-80; Bohol I., Sierra Bullones area, Teach- ers Park: CAS-SU 23351-52; Dusita (10-13 km SE of Sierra Bullones): CAS-SU 23336, 23349- 50, 23437, CAS 145700-701; Cantaub (9-16 km SE of Sierra Bullones): CAS-SU 17143-45, 22811-12, 22814, 23341-49, 23353-54, CAS 136845, 136870, 136875, 136887, 139016- 139018. Polypedates leucomystax leucomystax.—Phil- ippines, Alabat I.: CAS-SU 3795; Bohol I.: CAS 139029: Cagayan I.: CAS 157189-208; Caluya I.: CAS 127512-14, 127540-58, 127592-95, 127608-10, 127647-49, 127657, 127670-76, 127680-97; Calagaan I.: CAS 124158, 124192, 124204; Camiguin (south) I.: CAS-SU 23045- 63, 24055, 24111-13; Carabao I.: CAS 128114; Cebu I.: CAS-SU 16501, 23863-64, 24287, CAS 129130, 131915-18, 131969, 136833-36; Gi- gante South I.: CAS 125034-37; Guimaras L: CAS 125342-43; Inampulugan I.: CAS-SU 23950-51, 23959-60, 23964-65; Jolo I.: CAS 60632-51; Leyte I.: CAS-SU 9388, 14817; Luzon I.: CAS-SU 9825-30, 14057-60, 14075, 14760- 63, 14791-97, CAS 15575-76, 15667-78, 15699- 706, 15713, 15730, 42012-20, 61044-47, 61368- 74, 61388-460, 61655-70, 73876-78, 140070; Masbate I.: CAS 144275-79, 144281, 144320; Mindanao I.: CAS-SU 13970-75, 13991, 14770- 79, 14799-800, 16239-42, 16493, CAS 15722- 24, 133626-28, 133643, 133667-71, 133712-25, 137520-24; Mindoro I.: CAS-SU_ 11117-19, 13976-90, 14063-74, 14076, 16237-38, 16246, 22113-24, 22244-45, 22261-64, 22313, 22589, CAS 73879; Negros I.: CAS-SU 9833-39, 10795- 97, 14765-69, 14780-90, 16210-35, 17719-20, 18108, 18141-43, 23529-37, CAS 92960-62, 92981, 131835, 139274, 154185; Pacijan I.: CAS 124025-26, 124339-42, 124352-56; Pan de Azu- car I.: CAS 124176, 125031-32; Panay I.: CAS- SU 11113-16, 14764, CAS 137599; Poro I.: CAS 124521; Semirara I.: CAS 127841; Sibay I.: CAS 127721-26; Sicogon I.: CAS 124110-13, 124420, 124422-25, 124581; Tablas I.: CAS 137274-80, 137297, 139209. Polypedates macrotis.—Philippines, Jolo I.: CAS 60684 (holotype of Polypedates linki), 60804; Bud Dako Mountain: CAS 60630-31; Pa- lawan I.: CAS 64074, 64089-92; Iwahig area: MCZ 14436-38; 8km S of Iwahig: CAS-SU 21640; Malabosog Creek, 95 km N of Puerta PHILIPPINE RHACOPHORIDAE Princesa: CAS 157330-42; Brooke’s Point: MCZ 14439-50; Calauit I.: CAS 180833; Dumaran L.:: MCZ 14435; Busuanga I.: CAS-SU 13969. Rhacophorus appendiculatus.— Philippine Is.: BMNH 1947.2.9.17 (cotype); Negros I., Bagtic River Valley: CAS-SU 18114-23; Bohol I., Sierra Bullones area: CAS 10216, CAS-SU 22818; Ley- te I.: CAS-SU 14750-54; Basilan I.: CAS 60169- 74; Polillo I.: CAS 62261; Mindanao, Cotabato: CAS 64078-86; Philippines: CAS-SU 14755-59. LITERATURE CITED Aut, E. 1927. Zur systematic der asiatischen arten der frosch- gattung Rhacophorus. Sitzber. Gesell. Naturf. Berlin 1927: 35-47. 1-478. AtcaLa, A.C. 1962. Breeding behavior and early develop- ment of frogs of Negros, Philippine Islands. Copeia 1962: 679-726. ALCALA, A. AND W.C. Brown. 1956. Early life history of two Philippine frogs with notes on deposition of eggs. Her- petologica 12:241-246. . AND . 1982. Reproductive biology of some spe- cies of Philautus (Rhacophoridae) and other Philippine anu- rans. Kalikasan, Philipp. J. Biol. 11:203-226. BarRTLETT, E. 1894. Notes on the Batrachians or frogs and toads of Borneo and the adjacent islands. Pp. 7-14 (reprint) in The Ethnological, Zoological, and Botanical Note Book of Sarawak, 1896, E. Bartlett, ed., Domingo J. J. Rodriques Co., Kucking, Sarawak. Berry, P. Y. 1975. The Amphibian Fauna of Peninsular Malaysia. Tropical Press, Kuala Lumpur. 1-127. BoettTGer, O. 1886. Augzdlung der von den Philippinen be- kannten Reptilien und Batrachier. Ber. Senck. Naturf. Ges. 1886:9 1-134. . 1892. Katalog der Batrachier Sammlung im Museum der Senckenbergisher naturforschenden Gesellshaft in Frankfurt am Main. Gebruder Knauer, Frankfurt a. M., l- 73}. 1931. Anura III, Polypedatidae. Das Tierreich 55: . 1899. Rhacophorus rizali ein neuer Baumfrosch von Mindanao, nebst Fundortnolizen von den Philippinen tiber- haupt. Abh. Ber. Mus. Dresden, 1899, no. 1:1-3. BouLenGER, G. A. 1882a. Catalogue of the Batrachia Sal- ientia s. Ecaudata in the Collection of the British Museum. Taylor & Francis, London, ed. 2:1-503. 1882b. Description of a new genus and species of frogs of the family Ranidae. Ann. Mag. Nat. Hist.(5) 10:35. . 1891. Descriptions of new reptiles and batrachians. Ann. Mag. Nat. Hist. (6) 7:279-283. . 1894. On the herpetological fauna of Palawan and Balabak. Ann. Mag. Nat. Hist. (6) 14:81-90. . 1897. Descriptions of New Malay Frogs. Ann. Mag. Nat. Hist. (6) 19:106-110. Brown, W. C. AND A.C. AtcaLa. 1961. Populations of am- phibians and reptiles in the submontane and montane forests of Cuernos de Negros, Philippine Islands. Ecology 42:628- 636. . AND . 1964. Relationship of the herpetofaunas of the non-dipterocarp communities to that of the diptero- 219 carp forest on southern Negros Island, Philippines. Senck- enberg. Biol. 45:591-611. . AND 1982. Reproductive biology of some species of Philautus (Rhacophoridae) and other Philippine anurans. Philipp Kaliksan. J. Biol. 11:203-226. . AND 1983. Modes of reproduction of Phil- ippine Anurans. Pages 415-428 in Advances in Herpetology and Evolutionary Biology. A. G. J. Rhodin and K. Mujata, eds. Harvard Univ. Press, Cambridge. CHANNING, A. 1989. A re-evaluation of the phylogeny of Old World tree frogs. S. Afr. Tydskr. Dierk. 24 (2):116-131. Drina, J. 1987. Bornean tree frogs of the genus Philautus (Rhacophoridae). Amphibia-Reptilia 8:19-47. DuBors, A. 1981. Liste des genres et sousgenres nominaux de Ranoidea (Amphibiens Anoures) du monde, avec iden- tification de leurs Espéces-Types: Conséquences nomencla- turales. Monit. Zool. Ital. N. S., Suppl. 15:225-284. . 1987. Miscellania taxinomica batrachologica (1). Al- ytes 5:7-95. DUELLMAN, W. E. AND L. TRuEB. 1985. Biology of Amphib- ians. McGraw-Hill Book Co., New York. 1-670. DumeriL, A. M. C. ANDG. Brsron. 1841. Erpetolgie générale ou histoire naturelle compléte des Reptiles. Librairie Roret, Paris 8:1-792. Emerson, S. B. 1991. The ecomorphology of Bornean tree frogs (family Rhacophoridae). Zool. J. Linn. Soc. 101:337- 357. FiscHErR, J.G. 1885. Uber eine Kollektion von Amphibien und Reptilien aus Sudost-Borneo. Archiv. Naturg. 1885:41- W2: Fitzincer, L. 1843. Systema Reptilium. Fasc. Amblyglossae. Braumiiller and Seidel, Wien. 1-106. GistTeEL, J. 1848. Naturgeschichte des Thierreichs fiir héhere Schulen. Stuttgart, Hoffmann. 1-216. Gosner, K.L. 1960. A simplified table for staging anuran embryos and larvae with notes on identification. Herpeto- logica 16:183-190. GRAVENHORST, J. L.C. 1829. Delicae Musei Zoologici Vrats- laviensis. Fasciculus primus continens chelonios et Batra- chia. Lipsiae 1829:1-106. GUnTHER, A. 1859. Catologue of the Batrachia Salientia in the collection of the British Museum. Taylor and Francis, London. 1-160. HALLowELL, E. 1860. Report upon the reptilia of the North Pacific exploring expedition under command of Capt. John Rodgers, U.S. N. Proc. Acad. Nat. Sci. Philadelphia 12: 480-510. Hamitton, W. 1979. Tectonics of the Indonesian Region. U.S. Geol. Surv. Prof. Pap. 1078:1-345. Heaney, L. R. 1985. Zoogeographic evidence for middle and late Pleistocene land bridges to the Philippine Islands. Mod. Quaternary Res. SE Asia 9:127-143. 1986. Biogeography of mammals in southeast Asia: estimates of rates of colonization, extinction and speciation. Biol. J. Linn. Soc. 28:127-165. 1991. An analysis of patterns of distribution and species richness among Philippine fruit bats (Pteropodidae). Bull. Amer. Mus. Nat. Hist. No. 206:145-167. Heaney, L.R. AND E. A. RickarT. 1990. Correlations of clades and clines: Geographic, elevational, and phylogenetic distribution patterns among Philippine mammals. Pp. 321- 332 in Vertebrates in the Tropics, G. Peters and R. Hutterer, eds. Mus. Alexander Koenig, Bonn. HorrMan, A.C. 1932. Researaches relating to the validity 220 of the South African Polypedatidae (Rhacophoridae) as an autonomous family of the anura. S. Afr. J. Sci. 29:562-583. ILLIGER, J. K. W.. 1798. Verzeichniss der Kafer Preussens. Johann Jacob Gebaux, Halle. 1-510. InceER, R. F. 1954a. Systematics and zoogeography of Phil- ippine amphibia. Fieldiana: Zool. 33:181-531. . 1966. The systematics and zoogeography of the am- phibia of Borneo. Fieldiana: Zool. 52:1-402. 1985a. Rhacophoridae of Asia. Pages 525-549 in Amphibian Species of the World; D.R: Frost, ed. Allen Press, Lawrence, Kansas. 1985b. Tadpoles of the forested regions of Borneo. Fieldiana: Zool. (new series) 26:1-89. . 1989. Four new species of frogs from Borneo. Ma- layan Nat. J. 42:229-243. INGER, R. F. AND R. B. SteuBING. 1992. The montane am- phibian fauna of northwestern Borneo. Malayan Nat. J. 46: 41-51. InceR, R. F. AND F. L. TAN. 1990. Recently discovered and newly assigned frog larvae (Ranida and Rhacophoridae) from Borneo. Raffles Bull. Zool. 38(1):3-9. Kampen, P.N. VAn. 1907. Amphibien des Indischen Ar- chipels. Pp. 383-416 in Zool. Ergebn. Reise Niederl. Ost- Indien, M. Weber, Vol. 4. 1923. The Amphibia of the Indo-Australian Archi- pelago. E. J. Brill, Leiden. 1-304. Kunt, H. AND J. C. VAN HasseELT. 1822. Uittreksels uit brieven van de Heeren Kuhl en van Hasselt aan de Heeren C-. J. Temminck, T. van Swinderen en W. de Haan. Algemeene Konst en Letter-Bode 7:99-104. Leviton, A. E., R. H. Grpss, Jr., E. HEAL, AND C. E. DAwson. 1985. Standards in herpetology and ichthyology: Part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia 1985:802-832. Liem, D. S.S. 1970. The morphology, systematics and evo- lution of the Old World tree frogs. (Rhacophoridae and Hy- peroliidae) Fieldiana Zool. 57:1-145. Marx, K. W. 1975. A substitute name, Edwardtayloria, for a genus of tree frogs from southeast Asia. Sci. Publ. Sci. Mus. Minnesota 2:1-3. McLean, R.F. 1980. Spatial and temporal variability of external physical controls on small ecosystems. Pages 149- 175 in Population-environment in tropical islands: the case of eastern Fiji, UNESCO 1980, MAB Technical Notes 13. Ocitsy, C.D. 1837. Remarks on some rare or undescribed ruminants in the Societies’ collection. Proc. Zool Soc. Lon- don 1837:119-121. Our, C.D. 1985. The geological background to prehistory in island Southeast Asia. Mod. Quaternary Studies in South- east Asia. 9:25-42. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 10 Peters, W. 1863. Mittheilungen tiber neue Batrachier. Mon- atsber. Akad. Wiss. Berlin. 1863:445-470. 1867. Herpetologische Notizen. Monatsber. Akad. Wiss. Berlin. 1867:13-37. . 1871. iiber neue Reptilien aus Ostafrica und Sarawak (Borneo), vorziiglich aus der Sammlung deo Hrn. Marquis J. Doria zu Genua. Monatsber. Akad. Wiss. Berlin 1871: 566-581. Rasor, D. S. AND A. C. ALCALA. 1959. Notes ona collection of amphibians from Mindanao Island, Philippines. Philipp. J. Sci. 88:335-344. Ripe, W. D.L. 1985. International Code of Zoological No- menclature. International Trust for Zoological Nomencla- ture. 1-338. ScHEGEL, H. 1826. Notice sur l’erpetologie de Vile Java; par M. Boie. Bull. Sci. nat. Geol. 9:233-240. 1837-1844. Abbildungen neuer oder unvollstandig bekannter Amphibien. . .Conservator des Niederlandischen Reichs-Museum. Verlag v. Arnz & Company 1837-1844: 1- 151+ SO pls. STEJNEGER, L.. 1905. Three new frogs and one new gecko from the Philippine Islands. Proc. U. S. Nat. Mus. 28:343- 348. 1907. Herpetology of Japan and adjacent territory. Bull. U. S. Nat. Mus. 58:1-577. Taytor, E.H. 1920. Philippine Amphibia. Philipp. J. Sci. 16:213-359. 1921. Amphibia and Turtles of the Philippine Is- lands. Dept. Agri. and Nat. Res. Bur. Sci., Pub. 15:1-193. 1922a. Additions to the herpetofauna of the Phil- ippine Islands I. Philipp. J. Sci. 21:161-206. 1922b. Additions to the herpetofauna of the Phil- ippine Islands. II. Philipp. J. Sci. 21:257-303. : Tscuupt, J.J. 1838. Classification der Batrachien, mit Be- rucksichtigung der fossilen thiere. Mem. Soc. Sci. Nat. Neu- chatel. 2:1-99. VILLADOLED, D. V. AND N. Dex Rosario. 1930. Studies on the development and feeding habits of Polypedates leuco- mystax (Gravenhorst), with a consideration of the ecology of the more common frogs of Los Banos and vicinity. Phi- lipp. Agriculturist 18:475-503. Voct, T. 1911. Beitrag zur Amphibienfauna der Insel For- mosa. Sitzsber. Gesell. Naturf. Berlin. 1911:179-184. Watcortrt, C. D. 1920. Cambrian Geology and Paleontology. IV. Middle Cambrian Spongiae. Smithson. Misc. Coll. 67: 261-364. Wo tr, S. 1936. Revision der Untergattlung Bull. Raffles Mus. No. 12:137-217. WiecmMan, A. F. A. 1835. Amphibien. Nova Acta Acad. Leop. Carol. 17:225-261. © CALIFORNIA ACADEMY OF SCIENCES, 1994 Golden Gate Park San Francisco, California 94118 PROCEEDINGS OF THE Marine Biological Laboratory/ — Woods Hole Oceanographic Institution Library OCT _ 7 1994 CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 11, pp. 221-237, 9 figs., 1 table. Woods Hole, MA 02543 October 3, 1994 REVISION OF THE GENUS KUMBA (PISCES, GADIFORMES, MACROURIDAE), WITH THE DESCRIPTION OF THREE NEW SPECIES By Tomio Iwamoto Department of Ichthyology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118, and Yuri I. Sazonov Department of Ichthyology, Zoological Museum, Moscow State University, Herzen Street 6, Moscow 103009, Russia Asstract: The genus Kumba Marshall, 1973 is revised to include nine species: K. calvifrons n. sp. from the mid-Atlantic Ridge; K. dentoni Marshall, 1973; K. gymnorhynchus n. sp. from the eastern Indian Ocean; K. hebetata (Gilbert, 1905); K. japonica (Matsubara, 1943); K. maculisquama (Trunovy, 1981); K. punctulata n. sp. from the Bismark and Coral seas; and two other species, here described but not named because of the immature state or questionable status of available specimens. Parakumba Trunov, 1981 is synonymized with Kumba. The characters in combination uniting the genus include the extensive naked areas on the dorsal and ventral snout surfaces and the underside of the head; the absence of snout scutes; a distinctive squamation on the head characterized by small scales with one to few comblike, ridge-rows of small, erect spinules; and no coarse scutelike scales on suborbital shelf. Received September 16, 1993. Accepted December 15, 1993. INTRODUCTION During examination of the Soviet deep-sea fish collections in the P. P. Shirshov Institute of Oceanology (IOAN) and the Zoological Museum of Moscow State University (ZMMGU) in 1988, we found three specimens of two undescribed species of macrourids that appeared to belong to the genus Nezumia Jordan, 1904. The extensive naked areas on the head and a distinctive squa- mation in these specimens suggested, however, that they represented a different genus. These findings spurred a subsequent reexamination of a juvenile macrourid (CAS 77314) from the Gulf of Mexico and another (ZMMGU P.17762) from off New Guinea, which also shared these peculiar characteristics. A mature female from New Cal- edonia, received after completion of the first draft of this paper, represented the adult stage of the species from New Guinea. A search of the lit- erature disclosed six other species that shared these characters of naked head areas and squa- mation: Nezumia hebetata, N. japonica, N. lio- lepis (Gilbert, 1890), Parakumba maculisquama, Asthenomacrurus victoris Sazonov and Shcher- bachev, 1982, and Macrosmia phalacra Merrett, Sazonov, and Shcherbachev, 1983. Although the original description of Kumba [221] 222 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 dentoni did not indicate whether the snout was scaled or not, Nigel Merrett of the BMNH (in litt. to YIS, 23 Feb. 1981, and to TI, 14 Jan. 1992) informed us that the holotype, and only known representative of that genus, has exten- sive naked areas distributed much as in P. ma- culisquama and other related species. The spi- nulation on scales of this species are reduced but suggestive of others in this group. The monotypic genus Haplomacrourus Trunov, 1980 is char- acterized by a completely naked snout, but it is very different in many other ways from all other genera. Placement of the new species into a known genus has proved troublesome, especially with the presence of at least five different genera to which one or more of these distinctive characters have been attributed. Nezumia Jordan, 1904 is excluded because the other species of the genus (excluding those here considered) lack the full complement of these characters. Asthenomacru- rus and Macrosmia are excluded because they lack a well-developed light organ, and the anus is situated immediately before the anal fin, far from the forwardly placed pelvic fins. This leaves only Kumba and Parakumba as possible genera in which to place our new species, but the species lack the inflated head common to those two gen- era and they have a weakly to strongly serrated spinous dorsal ray, as opposed to weak or ob- solete serrations. We contemplated erection of a new genus for the species but it was deemed pre- mature in the absence of a more comprehensive analysis of relationships within the tribe. For rea- sons discussed below, we include Parakumba in the synonymy of Kumba and treat our new spe- cies in an expanded concept of that genus. The purposes of this paper are to revise the genus Kumba, discuss its possible relationships to other related genera, describe the three new species, and record other unidentifiable repre- sentatives of the genus. METHODS Methods for making counts and measure- ments and the use of abbreviations follow Iwa- moto (1970) and Iwamoto and Sazonov (1988). Institutional abbreviations follow Leviton et al. (1985) and Leviton and Gibbs (1988). Refer- ences for species names are given in the LITER- ATURE CITED section. The reader is referred to Eschmeyer (1990) for complete references to ge- neric names. RELATIONSHIPS OF KUMBA The genus Kumba was erected to include a single species, P. dentoni, known only from the holotype taken in the northeastern Atlantic in the Bay of Biscay. Marshall (1973:616) consid- ered the genus most closely related to Parace- tonurus Marshall, 1973, differing only in the lat- ter having a serrated dorsal spine, fewer pelvic rays (5-7), and different location of anus (just before origin of anal fin). In describing Parakumba, Trunov (1981) con- sidered that its relationships lie with the group that includes Cetonurus Giinther, 1887, Para- cetonurus, and Kumba, with Kumba being the most similar. Among the diagnostic characters of the genus, Trunov included the comparatively large head lacking ridges of modified scales, the extensive naked areas on the head, and the small scales having divergent crestlike rows of small, conical spinules. In two papers, Sazonov and Shcherbachev (1982, 1985) provided their preliminary views of relationships among the Cetonurus group of macrourine grenadiers. They considered Kumba, Parakumba, Macrosmia Merrett, Sazonov, and Shcherbachev, 1983, and Asthenomacrurus Sa- zonov and Shcherbachev, 1982, as constituting ‘“‘a natural group with Nezumia hebetata”’ (Sa- zonov and Shcherbachev 1985:24), based on the following features: (1) no scales on anterodorsal and ventral parts of snout; (2) scale spinules in crestlike rows; (3) no modified terminal snout scales; (4) suborbital without ridges of modified scales; and (5) pores of cephalic-sensory system present. The two authors hypothesized that the group arose from an ancestor close to Nezumia ventralis Hubbs and Iwamoto, in Iwamoto 1979, and N. japonica (Matsubara, 1943). Although these five characters might seem to be synapomorphies defining a monophyletic group, all are found to various degrees in other related (sometimes remotely related) macrourid taxa. For example, a partially naked snout is found in numerous species of Nezumia, Mataeoce- phalus Berg, 1898, Sphagemacrurus Fowler, 1925, Caelorinchus Giorna, 1809, Coryphae- IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA noides Gunnerus, 1765, and others, although not in combination with the absence of snout scutes. It seems that loss of scales, particularly on the underside of the snout, is a common occurrence among grenadiers. The almost completely naked snout of P. maculisquama and others here con- sidered appears to be a reduction carried to the penultimate stage. The second character, scale spinules in crestlike rows, is much like that found in Mesobius Hubbs and Iwamoto, 1977 and some species of Caelorinchus. Characters 4 and 5 are also widely and randomly distributed among dif- ferent genera. We consider the relative positions of the pelvic fins, anal fin, and the anus, and the relative de- velopment of the light organ in Asthenomacrurus and Macrosmia to be of phylogenetic importance and evidence for a distant relationship of Kum- ba, Parakumba, and Nezumia. In the first two genera only the pelvic fins are in a forward po- sition, with the anus far removed and immedi- ately before the anal fin, and the light organ is apparently absent. In contrast, in Nezumia, Kumba, and Parakumba, the pelvic and anal fins are both anteriorly placed, with the anus in be- tween the pelvic and anal fins; the light organ is well-developed with a light gland (referring to the pouch-like structure housing the luminescent bacteria), a distinct lens-like body, and an an- terior dermal window (condition of these in K. dentoni not known). Previous authors have accorded considerable importance to the large head of Kumba, Par- akumba, and a number of other genera, usually erecting monobasic genera based primarily on this feature. But should so much weight be given to a feature that appears to have evolved nu- merous times in macrourids and other related groups? For example, Cetonurus, Cetonurichthys Sazonov and Shcherbachev, 1982, Coryphae- noides rupestris Gunnerus, 1765, Echinomacru- rus Roule, 1916, Kumba, Parakumba, Parace- tonurus, and Pseudocetonurus Sazonov and Shcherbachev, 1982 are all characterized by rel- atively large heads, and the condition seems to have reached an extreme in the peculiar ma- crouroidines Squalogadus Gilbert and Hubbs, 1916 and Macrouroides Smith and Radcliffe, 1912. Similarly enlarged heads are found in other deep-sea fishes, e.g., Acanthonus Giinther, 1878 and Typhonus Giinther, 1878 (both ophidi- iforms), and Momonatira Paulin, 1985 (Mori- 223 dae). With the exception of Paracetonurus, with more than four species, Cetonurus and Echino- macrurus, each with two species, and Coryphae- noides rupestris, which is one of more than 60 species (Iwamoto and Sazonov 1988), all other genera listed are monobasic. The diagnosis for Cetonurichthys is much the same as that for Ce- tonurus, except that the latter is characterized by its somewhat larger head. In a similar but op- posite vein, the diagnosis for Pseudocetonurus, except for the expanded head and the replace- ment of the grooved lateral line scales with free neuromasts, agrees quite well with that for Ven- trifossa. Perhaps relative head size has been over- emphasized as a generic character. If we ex- clude head size, our new species then fall readily into the circumscription of the genera Kumba and Parakumba, with the exception in the for- mer of a smooth spinous dorsal ray and a reduced light organ. These last two characters—the only tangible differences between Kumba and Par- akumba—we consider insufficient to continue recognition of the latter genus. We consider the smooth spinous dorsal ray of Kumba to be the end point of a transformation series beginning with the strongly serrated condition found in K. japonica and K. hebetata, to the weakly serrated spine of K. maculisquama and K. gymnorhyn- chus, to the totally smooth state in K. dentoni. Similarly, the small light organ in K. dentoni, externally manifested in a narrow periproct and small naked area before the anus, represents an- other reductive trait. A special note must be made of Nezumia lio- lepis, which we have excluded from Kumba be- cause of the small coarse scales along the leading snout edge, the lack of scales on the suborbital shelf, and the presence of scales on the ventral surface of the suborbital and mandible. The oth- erwise naked snout is very much like that in members of Kumba, the slender second spinous ray of the first dorsal is smooth or has much reduced serrations along the leading edge, and the periproct area is small, with only a rudimen- tary light organ, much as in K. dentoni. This particular combination of characters is similar to that in Kumba, yet it is different in its own way. It may well suggest some closer affinity that we have not been able to detect. We feel that a more detailed analysis of other characters (especially osteological, myological, and other internal features) will help remove the 224 uncertainty of our current ideas of relationships, but that will remain undone until more speci- mens become available for study. DESCRIPTIONS Kumba Marshall, 1973 Kumba Marshall, 1973:616 (type species Kumba dentoni Mar- shall, 1973, by original designation) Parakumba Trunov, 1981:27 (type species Parakumba ma- culisquama Trunov, 1981, by original designation). D1AGcnosis.— Branchiostegal rays 7. Anus in middle 1/3 of space between anal and pelvic fin bases, usually closer to latter. Luminescent organ with a single gland and lens in abdominal wall anterior to rectum. Most of dorsal surface of snout and almost entire ventral surfaces of snout, sub- orbital, and lower jaw naked. No terminal or lateral snout scutes present. Scales on upper sur- face of suborbital small, in several rows, each armed with small, erect spinules closely aligned in 1-3 comblike rows. Pelvic fin rays 8-12. Gill- rakers, inner series of first arch 10-14 total (8- 11 on lower limb). DISTRIBUTION. — The genus is known from rel- atively few specimens collected in widely scat- tered localities. Kumba japonica is known from at least 12 specimens from off Japan and the Kyushu-Palau Ridge (Okamura in Okamura et al. 1982:147, 349, and our data), but the re- maining eight species are known from only 10 specimens from Hawaii, the eastern and western Indian Ocean, off New Guinea and New Cale- donia, the northeastern Atlantic, the mid-South Atlantic, and the Gulf of Mexico. Depth distri- butions range from about 550 m to about 1,600 m. REMARKS. — We tentatively recognize nine spe- cies of Kumba. Four species are already de- scribed: K. dentoni, K. maculisquama, K. hebe- tata, and K. japonica. Three others are herein described and named. Descriptions and illustra- tions are provided for the other two, but they are not named because they are each represented by a single, small specimen in a mediocre state of preservation. KEY TO THE SPECIES OF KUMBA 1 (8) Scaleless areas on upper surface of snout extend posteriorly onto fore- head beyond lateral nasal angles ..... 2 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 2 (3) Head soft, without distinct ridges, sensory canals inflated, with open pores; interorbital width about 33% of HL, equal to snout length; barbel rudimentary, less than 3% HL; V. 10 i eis Bie See K. maculisquama Head firm, suborbital and nasal ridges distinct, sensory canals not inflated, pores present or absent; interorbital width shorter or longer than snout, both less than 32% HL; barbel 8-19% HL; V. 8210 2 eee 4 Upper jaw short, about 34% HL __ i aOR ae K. sp. A (Gulf of Mexico) Upper jaw long, 40% HL or more .... 6 Head sensory canals with enlarged pores; barbel about 16-19% HL; snout shorter than interorbital 0. COE TA. 2 tae 82 K. punctulata n. sp. Head sensory canals without enlarged pores; barbel short, about 9% HL; snout longer than interorbital fs 1th 51 1) Ae K. calvifrons n. sp. Scaleless areas on upper surface of snout extend only to level of lateral nasal angles... "6 Oe 9 Head soft, sensory canals inflated, snout longer than 33% HL; interor- bital about 40% HL; no separate ADW, periproct only slightly expand- edanteriorly = a K. dentoni Head firm, sensory canals not inflat- ed, snout shorter than 33% HL; in- terorbital less than 30% HL; ADW present, separate or within notable anterior projection of periproct ...... 11 V. 10-13; upper jaws longer than 33% cl i 12 V. 10-11; no large pores on head; bar- bel short, about 7% HL; orbit small, 26-27% HL, equal to or less than in- terorbital _.. K. gymnorhynchus n. sp. V. 12, rarely 11 or 13; pores in head present; barbel 21-33% HL; orbit large, more than 30% HL; greater than interorbital width K. japonica V. 8; upper jaw length less than 33% HL ww eS ee 15 1D. II,13; snout notably shorter than Orbity_ 0 Nii) ae K. hebetata 16 (15) 1D. II,9; snout about equal to orbit. we Se Ae WN K. sp. B (Mozambique) 3 (2) 4 (5) 5 (4) 6 (7) 7 (6) 8 (1) 9 (10) 10 (9) 11 (14) 12113) 13 (12) 14 (11) 15 (16) IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 225 Zz a7: Leseiaes IT DT es Tie, ee - ee Ficure 1. Kumba calvifrons new species. Holotype: ZMMGU P.17764, 270 mm TL, from Mid-Atlantic Ridge at equator, in 930-960 m. (a) Lateral view; (b) dorsal view of head to show extent of naked area on snout, indicated by dashed line. Fins and squamation partially reconstructed. Scale bar equals 25 mm. Kumba calvifrons new species (Figure 1) MATERIAL EXAMINED. — Holotype: ZMMGU P.17764 (75.5 mm HL, 370+ mm TL); Mid-Atlantic Ridge, 00°27’N, 16°52.2'W; 930-960 m; R/V Chronometer cr. 4, trawl 7; 20.VII.1979. Dracnosis.—Head firm, non-inflated; head sensory canals without enlarged pores; snout dor- sally scaleless posteriorly onto forehead as two projections lateral to median rostral ridge (Fig. 1b); V. 8; orbit 27% HL; snout equal to orbit, 1.2 times longer than interorbital; upper jaw about 40% HL; ADW and PDW within scaleless thumb- shaped projection of periproct. COUNTS AND MEASUREMENTS. —(see also D1- AGNOsIs and Table 1). Scales below 1D. 15, be- low 2D. 13.5, below mid-base 1D. 8, lat. 1. over predorsal length 41. The following in percent HL: postrostral 76.2; pre-A. 145; pre-anus 127; V.- A. 36; isthm.-A. 68; body depth 90; 1D.-2D. 44; length 1P. 50. DESCRIPTION. — General shape best seen in Fig- ure 1. Dorsal profile of nape highly arched, giving humpbacked appearance. Head relatively deep (about 1.2 in HL), moderately compressed lat- erally, length about 4 or more in TL (a large pseudocaudal developed). Interorbital region with subtriangular concavity or depression posterior to median rostral ridge. Snout blunt, forming obtuse angle in dorsal and lateral profiles. Sub- orbital region subvertical, without strongly an- gular ridge. Scaled surfaces of opercle and sub- opercle form deep inverted triangle; free margins of preopercle, subopercle, and interopercle finely crenulated. Mouth large, little restricted at lateral corners; upper jaw extends to below midorbit or slightly beyond. Chin barbel tiny, length slightly less than diameter of posterior nostril. Gill open- ings wide, membranes broadly connected to isth- mus with a posterior free fold. Opercular opening extends forward to below end of mandible. Sensory canals on head not broadened or swol- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 226 ‘ROL6I VINWILYO WOY Be , ‘T86[ AounI, Woy eG - - - weneenenenenennencneneneenesees QBCAST rere - = - - - 0S< OL< B09R9 “[Ad isell OiMaergilaere (Olareditiar ONsragelargG Wisergiiarll Working ANS WOM WilergAQseG (arises (WiartOlse6 Cartiaell warohyae — (sarro(iarce 1-aoD TI+1/6 OI+£/6 OI1+7/6 11+7/8 OI+£€/6 O1+Z/01 201 +7/6 OI+Z/II OI+E€/I1T TI +£€/6 6+1/8 8+7/6 OI1+7/8 luo 6/6 OI/0I 6/6 TI/ZI CI/ZI IT/TT 201/01 O1/01 I1/01 8/8 8/8 8/8 8/8 “A Iz! 07/77 (GAs Ava a! 1z/0c! 6/61t 17'/07! 1c'/0c! CCY/CCI 6I/6lt v7r/eT! 0c'/0z! 611/07! ‘dl 611 6Il II Old LUT Ol'll 26 II Oli Ol‘Tl ITT ro 6Il IVI ‘dl syunoD (oe) 601 (6 \(I Sy — OL cs 19 9S L’8 78 £9 bY [Lsou "sod 19 09 IZ ae LS 19 28 OL £9 L9 08 £6 OOT cil “A 9°07 9°07 SC ysl 06! 0°07 26 81 Lgl 6SI TOI vit 67TI £ vl WS TTD 09 Lt 61 LEG 8° V7 8°77 ec £ 9°9 = 68 = 601 08 Jeqieg OPE L6£ Iv 6 8E 18 69€ e0'8E bv Or 9D bv Or Ji NS JENS, 6£E mef sadd) bv 8P £°9¢€ 8€ £8E S rE 69£ 1 98 10s Cty a0) yey SLE ‘doasd-q19q10 9SP 1 0¢v 87 fee C 6£ 16£ 20'SP 7 8P CIS 9 8P VCP Ly a4 [B11G101SOq (SKE yl CL £6 67TI Oe! 2007 fey L9l Oe | 091 691 6LI [e1iqioqns IeLG L9¢€ tp CP WLS OSE et 62 192 9ST O'LZ VTE 6 87 £0€ qiO 6'6£ STE Sz LY OMLLG SET ef SE 9°97 p97 0°77 8° LZ LASG LYT [B11G1019} UT o bY 661 IZ (G(YG £97 LI 87 LEGG OPT 761 LS €°8C CLT [eseuto}uy £87 Trl 91 GAUGE SET 97 IZ S07 £61 8 rl O'V~ Dale £°0C [e1031g Ive LLZ 07 p87 6°67 6 £7 el OF £67 £87 697 877 bv 67 167 ynoug 9°6£ L97 vel T9I O17 pe eS 9¢ £°89 (Ste SSL 797 OSE 1S7z TH 077 +Crl +68 (aa! L71 €SI eL9T +7COP +96 +OLE £7I +691 +671 ae S]UDWIOINSeIPy adh} vE-P66l COLLId £9LLI'd 6177E £9LLI'd CO8tyr £lELl 99LLI'd POLLI'd 809TS TSTOS PIELl -ojoy NHNW NOWNZ NOWWNZ Nwsad NOWNZ NIZ SVO NOWWNZ NOWNWZ WNSn SVO SVO ae pipjnjound ‘uodol eae snyoudysouulds ge pane ance ee ‘paiuodol ‘y = ‘uodpl ‘pumnbsijnovu “y = ‘jnIDUL ‘DIDJaqay WY = Djaqay ‘suosfiajod “y = “Aj09 :Butmoyoy 10} 1d99x9 “(gg6]) OLOWRM] Pue AOUOZES pue (0/61) OLOWeM] MOTIOJ SUONLIAZIQqY ‘dds nquiny Jo sjuNOD puk s}UsUTOINSeaY “| «ATAV IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA len. Open pores absent in all canals except on preopercle, where single pore exits at dorsal edge of bone. This pore tubular (tube declined pos- teriorly), surrounded by narrow area of naked skin. Free neuromasts serially arranged (between scales) along scaled surfaces of infraorbital, su- praorbital, and preopercular canals; those on na- ked areas of above canals and along mandibular canal irregularly placed. Small, thin-walled black tubules possibly representing modified sensory pores present along horizontal portion (below orbit) of infraorbital canal (3 tubules) and along vertical arm of canal (1), along medial border of mandibular canal (3), and near ventral margin of preopercular canal (4). Small black papillae scattered on naked parts of snout and lower por- tion of infraorbital shelf, and on lips. Olfactory cavity small, its length about equal to pupil diameter. Anterior nostril round, small, bordered posteriorly by semitubular flap of skin about twice as high as diameter of anterior nos- tril. Posterior nostril large, oval, its longer axis 8.7% HL. Approximately 22-24 olfactory la- mellae. Premaxillary teeth uneven in size, those along outer margin notably enlarged, conical, slightly curved inwardly, tips somewhat arrowhead shaped. Largest of these emarginate teeth anter- iad, size gradually decreasing posteriad to where they are scarcely longer than those from inner rows. Left premaxillary has 21 enlarged teeth, right premaxillary 18. As many as 7 or 8 teeth in a section across premaxillary. Inner rows of teeth uniformly small and arranged in 7 subpar- allel, oblique strips, each strip making acute an- gle with bone margin (similar to premaxilla il- lustration of Caelorinchus jordani in Okamura 1970b:25, Fig. 14L). Dentary teeth uniformly small and arranged similar to those on premax- illary but more irregular near symphysis, where there are 4 or 5 teeth in transverse section; length of subparallel tooth strips lessens rapidly pos- teriorly. Head scaled except for following: undersides of snout and suborbital shelf to vertical through posterior margin of orbit; mandibular rami; low- er portion of preopercular flap; gular and bran- chiostegal membranes; upper surface of snout posteriorly along both sides of median nasal ridge about to level of anterior % of orbits (Fig. 1b). Supranarial ridges scaled, as well as area from lateral snout angle posteriorly along suborbital shelf to sides of head. Broad interopercle scaled 22 only at exposed posterior tip with small, non- spinulated scales. Fin membranes and narrow band around base of first dorsal naked. Scales on head with fine, short, needlelike spi- nules that are only slightly reclined from vertical and arranged, with bases confluent, in close, tight, comblike rows; no reticulate pattern. Usually only 1-6 comblike rows on smaller scales of head, but as many as 2-7 on large scales of interorbital, postorbital canal, cheek, and operculum. Almost all scales dorsally on snout small, elliptical, with single comblike file of spinules. Scales on sub- orbital shelf small, oval, scarcely imbricate, un- thickened, 4—6 irregular scale rows present across narrowest part of shelf (6-8 rows before or be- hind this point), most scales here with single, median, comblike row of spinules, some poste- riorly with 1 or 2 divergent spinule rows on either side of median row. Body scales small and similar to those of head except that spinules more reclined and the 8-10 rows more numerous and essentially parallel. Those under cleithrum, behind pectoral base, and behind and lateral to first dorsal base smooth or faintly ridged. Scales on chest with 3-6 slightly divergent rows of conical, reclined spinules. Scales dorsally on sides of body with strongly reclined spinules in 8-12 parallel ridge rows, none of spi- nules enlarged, those posteriormost extend be- yond scale margin. Lateral line gently curved from postorbital ca- nal to vertical from origin of second dorsal, be- coming straight caudally. Lateral line scales with 1 or 2 short rows of notably reclined conical spinules on each side of canal tube, rarely spi- nules absent. A slender, teardrop-shaped, black ADW be- tween pelvic fins (and behind insertions of pelvic fin bases) connected by narrow isthmus to broad, scaleless periproct region, the PDW immediately fronting anus. Anus closer to pelvic insertions than to anal origin, separated by about 15 scales from anal fin. Fin positions best seen in Figure la. Leading edge of second spinous dorsal ray with low, wide- ly spaced, non-overlapping, almost indistinct serrations. Interspace between dorsal fins about 1.4 times length base of first dorsal fin. Outer pelvic ray produced, extending to base of 13th anal ray. Color overall gray-brown. Belly, operculum, naked surfaces of head, gular and branchiostegal membranes, and iris dark brown to black. Bar- 228 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 Ficure 2. Diagrammatic lateral views of: (a) Kumba den- toni (adapted from Marshall 1973); (b) Kumba japonica, ZMMGU P.17663, 112 mm TL, probably from Kyushu-Palau Ridge; (c) Kumba maculisquama, holotype, ZIN 43803, 267 mm TL, from South Atlantic, in 1,350—1,600 m; (d) Kumba sp. A, CAS 77314, 129+ mm TL, from Gulf of Mexico, in 1,280 m. Scale bars equal 25 mm. bel, lining of oral cavity, maxillary, and lateral sides of dentary pale. Skin covering premaxillary and upper and lower lips brown. First dorsal blackish brown overall, pale distally behind third branched ray. Paired fins brownish black basally, dusky distally. Anal fin brownish black to dusky Over anterior two-thirds, pale posteriorly. DISTRIBUTION. — Known only from the holo- type taken at the equator on the Mid-Atlantic Ridge, in 930-960 m. ETyMOLoGy.—From the Latin calvus, bald, and frons, forehead, in reference to the naked snout and forehead in this species. ComPARISONS. — Kumba calvifrons shows no- table differences from most of its congeners in proportional measurements of the head, the val- ues usually being smaller for the preoral, inter- nasal, interorbital, and suborbital lengths, and slightly larger for the length posterior nostril (see Table 1). It further differs from K. gymnorhyn- chus in having fewer pelvic fin rays, a slightly shorter orbit-preopercle distance, and in having naked areas on the dorsal surface posterior to the snout. K. calvifrons further differs from K. he- betata, species A, and species B in having a longer upper jaw and longer outer gill slit. K. japonica has much larger orbits, longer barbel, and more pelvic fin rays. K. punctulata shows a number of morphometric differences from K. calvifrons as well as a higher pelvic ray count and enlarged head sensory pores. Kumba dentoni Marshall, 1973 (Figure 2a) Kumba dentoni Marshall, 1973:617-618, Fig. 36 (holotype BMNH 1961.1.30.6, 220 mm TL, Bay of Biscay, 47°37'N, 7°29'W; 1207 m). D1aGcnosis. — Head large, rather soft, inflated, the head sensory canals with small pores; scale- less areas dorsally on snout extend to level of anterior nostril; V. 9; orbit diameter about 28% HL; snout much longer than orbit, about 1.1 into broad interorbital; light organ poorly developed, visible only as a small, naked, black projection anterior to anus. Spinous second ray of first dor- sal fin smooth. COUNTS AND MEASUREMENTS. —(provided by Y. N. Shcherbachev; see also Table 1). Scales below 1D. 11, below 2D. 6, below mid-1D. 8. The following in percent of HL: postrostral 71; pre-A. 125; pre-V. 103; V.-A. 29; isthm.-A. 53; body depth 75, depth over A. origin 58; height 1D. 63; length 1P. 51; length V. 61; length base 11330. DESCRIPTION. — We have not examined the ho- lotype, but our colleague, Y. N. Shcherbachev of IOAN, provided the following information from his examination in 1991: scales over suborbital shelf with a single row of spinules; other scales on head with diverging rows of spinules, those on body with 4-8 almost parallel spinule rows. Premaxillary teeth with up to 4 rows anteriorly, outermost teeth enlarged, those on lower jaw in 2 or 3 rows. REMARKS. — We can add nothing to the rather brief original description of the species, which is known from a single specimen taken in the north- eastern Atlantic in the Bay of Biscay. IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 229 Ficure 3. Kumba gymnorhynchus new species. (a) Lateral view of holotype, CAS 77313 75.3 mm HL, 496+ mm TL, from West Australian Ridge in 1,260-1,370 m; (b) dorsal view of head showing extent of naked area on snout. Fins and squamation partially reconstructed. Scale bar equals 25 mm. Kumba gymnorhynchus new species (Figures 3, 4) Ho.otypre.—CAS 77313 (formerly ZMMGU P.17765)(68.3 mm HL, 402+ mm TL); West Australian Ridge (Broken Ridge), 30°46’S, 93°20’E; 1,260-1,370 m; R/V Zvezda Kryma cr. 6, trawl 162; 26.1X.1976. Paratype: ZMMGU P.17766 (75.3 HL, 496+ TL); same data as for holotype. Dracnosis.—Head firm, non-inflated head sensory canals lacking pores; scaleless areas dor- sally on snout extend to lateral nasal angles but not onto forehead; V. 10-11; orbit diameter about 26% HL; snout slightly longer than orbit; inter- orbital about equal to orbit; upper jaw 40-43% HL; ADW and PDW situated within scaleless anterior projection of periproct. COUNTS AND MEASUREMENTS.—(see also D1- AGNOsIs and Table 1; holotype marked with as- terisk if different from paratype). Scales below 1D. 11*, 13, below 2D. 11*, 12-13, below mid- base 1D. 8-9*, 13-14, lat. 1. over predorsal length 41*, 44. The following in percent HL: postrostral 74.9*), 73.2: pre-A. 132*, 131; pre-anus 117*, 116; V.-A. 30; isthm.-A. 65*, 58; body depth Ficure 4. Scanning electron micrograph of scale from dor- sum of holotype of Kumba gymnorhynchus. Scale bar (dotted line) equals 1.25 mm. 230 81*,-; 1D.-2D. 44*, 31; length 1P. 59*, 58; length Vi67 63: DESCRIPTION. — General shape best seen in Fig- ure 3. Dorsal profile of nape gently convex from level of orbit to origin of first dorsal. Head mod- erately deep, compressed laterally, maximum head width less than depth at mid-orbit, about equal to postorbital length, head length about 5.3 into TL. Interorbital space deeply concave except for median rostral ridge, but concavity may be a result of preservation-caused shrinkage. Snout moderately high, profile of upper and lower mar- gins form approximately a right angle in lateral view. Suborbital, nasal, and median rostral ridg- es conspicuous, probably more so from shrink- age. Suborbital region vertical, without a strongly angular ridge, although demarcation of broad, densely scaled dorsal shelf strongly set off from naked underside. Mouth large, subterminal, up- per jaw extends posteriorly to below mid-orbit. A minute chin barbel, scarcely visible, shorter than diameter of posterior nostril. Preopercle large, broadly rounded, the upper margin in- clined forward. Subopercle attenuated ventrally into subtriangular projection, its posterior mar- gin gently concave. Exposed posterior tip of in- teropercle subtriangular and weakly scaled. Free margins of subopercle, interopercle, and pre- opercle crenulated. Opercular opening broad, ex- tends forward under head to below posterior end of mandible or slightly forward of that. Gill membranes connect narrowly under isthmus, forming a free posterior fold. Sensory canals on head similar to those de- scribed for K. calvifrons, except that postorbital canal wider and no open pores present. Free neu- romasts scarcely visible on scaled parts of sen- sory canals or naked head surfaces. Small black papillae restricted to lips. Olfactory cavity small, its length about *4ths pupil diameter; anterior nostril round, small, bordered posteriorly by semitubular flap slightly higher than diameter of anterior nostril. Poste- rior nostril large, broadly oval, its long axis 5.6- 6.1% HL; 16 olfactory lamellae in holotype. Premaxillary and mandibular teeth as de- scribed for K. calvifrons except oblique strips of teeth of premaxillae more dense, more irregu- larly placed, and teeth increase in size toward outer margin. Emarginate premaxillary teeth en- larged, 21 on each side; smallest posteriormost teeth lacking arrow-head tips. Squamation of head similar to that of K. cal- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 vifrons except for absence of posterior extension of naked areas onto forehead. Scales covering suborbital shelf each with | or 2 diverging ridges of spinules; as many as 5 rows of scales on nar- rowest part of shelf. Scales on postorbital sensory canal notably enlarged. Reticulate pattern on scales well developed. Few body scales (Fig. 4) remain. Those on nape and chest with 3-7 di- verging rows of reclined spinules. Scales cover membranes of basal parts of pelvic, pectoral, and first dorsal fins; these scales lack ridges, occa- sionally have a single spinule. Similar scales pres- ent along both sides of anal fin base. Lateral line gently curved from postorbital ca- nal to vertical through origin of second dorsal fin, then straightens caudally. Lateral-line scales with one or no spinules, but often with trace of ridges above or below canal tubes. Light organ externally similar to that of K. calvifrons, although a PDW not distinguishable within subtriangular naked projection of peri- proct. About 10-12 scales between anal fin and rear margin of periproct. Stomach in mature female holotype everted. Swim bladder thin walled, narrowly oval, its pos- terior end extends to level of 8th anal fin ray; the abdominal cavity extends to 12th ray. Two gas glands in posterior '4 of bladder, with 2 long, narrow, and straight retia. Fin positions best seen in Figure 3. Leading edge of spinous second ray of first dorsal fin with 7 fine, greatly reclined spinules in holotype; tip filamentous but short. Interdorsal space 1.1—1.4 times length base of first dorsal. Outer pelvic ray produced, threadlike, extends to base of 9th or 10th anal ray. Color in alcohol dark brown overall. Dorsally and anteriorly on head lighter brown; anteriorly on lower jaw and entire barbel pale. Operculum, gill, abdominal and chest areas darker, more swarthy. Upper lips black. Fins pale to dusky; but pelvic base dark. Orbit narrowly ringed in black. Mouth lining gray. Gill-chamber walls blackish along outer and upper margins, pale me- dially and antero-ventrally over cleithrum and on epibranchial and ceratobranchial. Peritone- um pale with randomly scattered small chro- matophores. ETYMOLOGY.—From the Greek, gymnos, na- ked, and rhynchos, snout. The name is to be treated as a noun in apposition. DISTRIBUTION. — Known only from one collec- tion in the eastern Indian Ocean on the West IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 231 yee LLPETLIS DS i ipep MMP. FiGure 5, Australian Ridge (= Broken Ridge), in 1,260- 1,370 m. CompaArRIsons.— Kumba gymnorhynchus is closely similar in overall appearance to K. cal- vifrons but differs in having the naked area on the dorsal surface of the snout abruptly ending along a line between the lateral nasal angles, without posterior extensions. The mouth 1s also grayish overall (vs. completely pale), the man- dibular dentition is in a narrower band that be- comes slightly constricted anterolaterally (vs. moderately broad, uniformly tapered band), opercular opening extends farther anteriorly, and there are slight differences in morphometry (see D1iacnosis and Table 1). The 10 or 11 pelvic fin rays of Kumba gymnorhynchus distinguishes it from others of the genus except K. maculisqua- ma, K. punctulata (holotype only), and K. ja- ponica, but those species are highly distinctive in their own ways (see descriptions of those spe- cies). Kumba gymnorhynchus appears to have the smallest orbit, and commensurately, the lon- gest postorbital and orbit-to-preopercle distanc- es, but these differences may simply be size re- lated. Kumba hebetata (Gilbert, 1905) (Figure 5) Macrourus hebetatus Gilbert, 1905:671-672, Fig. 262 (Ha- waii). Kumba hebetata. Holotype, USNM 51608, from off southern coast of Oahu, Hawaii, in 497-591 m. From copy of original drawing by R. L. Hudson, published in Gilbert (1905, Fig. 262). Scale bar equals one inch (25.4 mm). (Note: chin barbel not drawn, but present in specimen.) MATERIAL ExAMINED.— Holotype: USNM 51608 (26.2 mm HL, 123 mm TL); off south coast of Oahu; Albatross st. 3925; 299-323 fms (547-591 m); 7.V.1902. D1aGcnosis. —Scaleless areas dorsally on snout extend only to level of lateral nasal angles; 1D. 11,13; 1P. i23-i24; V. 8; orbit about 32% HL; snout short, blunt, about 34 orbit, slightly less than interorbital; barbel small, 4 orbit diameter. CouNTS AND MEASUREMENTS. — (see also Table 1). Scale rows below 1D. about 13, below 2D. about 10, lat. 1. over pre-D. length 46. The fol- lowing in percent HL: pre-A. 131; pre-anus 119; isthm.-A. 59; anus-A. 30; height 1D. about 77; length 1P. 62; 1D.-2D. 32. DISTRIBUTION. — Hawaiian Islands. REMARKS. —Géilbert’s (1905) original descrip- tion and illustration of the holotype are excellent, and we can add little new. No other specimen has been collected, as far as we know. Iwamoto’s (1986) record of the species from the Indian Ocean off southern Africa was based on the specimen here reported as Kumba sp. B, but differences in counts of the first dorsal and pectoral fins, and the length of the snout, suggest that the two rep- resent separate species. Kumba japonica (Matsubara, 1943) (Figure 2b) Lionurus japonicus Matsubara, 1943:149-152, Fig. 9 (holo- type: FAKU 1951, 176 mm TL; Kumano-Nada, Japan; 232 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 Ficure 6. Kumba punctulata new species. Holotype: MNHN 1994-34, 145+ mm TL, from off New Caledonia in 530 m. (a) Lateral view; (b) dorsal view of head to show extent of naked area on snout, indicated by shading. Scale bar equals 25 mm. paratype FAKU 1938, 155 TL). Kamohara 1959:8 (Tosa Bay, Japan). Nezumia japonicus: Okamura 1970a:88-91, Pl. 19, Text-fig. 39 (descr.). Ventrifossa japonica: Okamura in Okamura et al. 1982:147, 349, Fig. 91 (10 spec.; s. Japan and Kyushu-Palau Ridge, 550-710 m). MATERIALS EXAMINED. —ZMMGU P.17763 (2, 16.2-23 mm HL, 112-153 mm TL); western Pacific, probably over Kyushu- Palau Ridge; Prof. Derjugin tr. 189; 13.V.1971. BSKU 32219 (21.0 HL, 127 TL); Tosa Bay, Japan. Dr1aGnosis.—Head firm, non-inflated head sensory canals with prominent pores; scaleless Sram Figure 7. Kumba punctulata new species. Paratype, ZMMGU P.17762, prejuvenile from off New Guinea, captured in a midwater trawl between O and 1,000 m. area on dorsal surface of snout extends only to lateral nasal angles; V. 11-12; orbit 36-43% HL, snout and interorbital much shorter than orbit, snout about 1.25-1.5, interorbital 1.4-1.7 into orbit; ADW and PDW lens-like and widely sep- arated. Three small black pigmented areas near mid-length of anal fin base. Pyloric caeca 40-52. REMARKS.—This species has been well de- scribed by Matsubara (1943) and Okamura (1970; 1982), with a good color figure provided in the last reference. It has the best-developed light or- gan and sensory head pores, longest barbel, larg- est eyes, and highest pelvic ray count of the ge- nus. Many features are suggestive of the genus Ventrifossa, to which it has been attributed by Okamura (in Okamura et al. 1982), but the ex- tensive naked areas on the head and the distinc- tive spinulation on the head scales support its placement in Kumba. Kumba japonica is prob- ably a small species, as all representatives cap- tured so far have failed to exceed 170 mm. The largest specimen in our collection was an im- mature male with small testes. Kumba maculisquama (Trunov, 1981) (Figure 2c) Parakumba maculisquama Trunov, 198 1:30-35, Figs. 14 (mid- South Atlantic). IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 233 MATERIAL EXAMINED. — Holotype: ZIN 43803 (56.6 mm HL, 267 mm TL); 37°09’S, 07°38'W; in 1,350-1,600 m; 5.1.1978. D1acnosis. — Head large, rather soft, inflated, the head sensory canals with small pores; scale- less areas dorsally on snout extend onto forehead; V. 10; orbit diameter about 29% HL; snout about equal to orbit, about 1.1 into broad interorbital; ADW and PDW well developed, at either end of long periproct region, the two separated by narrow isthmus. REMARKS. — Trunov’s (1981) original descrip- tion and illustration of the single representative are thorough and accurate. No other specimen has been found in collections we examined. The very broad, inflated head readily distinguishes this species from all others of the genus except K. dentoni, which has weaker scales, a poorly developed light organ, and a smooth dorsal spine. Kumba punctulata new species (Figures 6, 7) MATERIAL EXAMINED.— Holotype: MNHN 1994-34 (26.7 mm HL, 145+ mm TL); off New Caledonia, 20°54’S, 168°21'02”E; 530 m; trawl; Vauban SMiB3, st. CP.4; 20.V.1987. Paratype: ZMMGU P.17762 (13.4 mm HL, 89+ mm TL); Bismark Sea off New Guinea, 5°20.9’S, 146°16’E; 1,000-0 m; IKMWT, 3000m.w.o.; R/V Dimitry Mendeleev cr. 18, sta. 1549; 16.11.1977. DiaGcnosis.—Head firm, non-inflated head sensory canals, with conspicuous pores; config- uration of naked areas on snout as in K. calvif- rons, V. 9 or 10; orbit 37-43% HL; snout about 1,4 orbit, about 1.1—1.3 times into interorbital width; upper jaw 40-41% HL; barbel 16-19% HL. MEASUREMENTS. —(in percent of HL; see also Table 1) Postrostral 78; pre-A. 142; pre-anus 97; V.-A. 13; greatest body depth 71; 1D.-2D. in- terspace 37; height 1D. 86; length 1D. base 30; length 1P. 60; length V. 70. DESCRIPTION OF HOLOTYPE. — (paratype data in parentheses). General shape of adult seen in Fig- ure 6 (of prejuvenile in Fig. 7). Head 0.9 (1.4) into greatest body depth, greatest width 0.7 (1.5) into postorbital length; body compressed later- ally, gradually tapering posteriorly, tip of tail missing in holotype. Orbit large, greater than snout length, slightly greater than interorbital width. Suborbital width about 3 (6) in orbit. Up- per jaw extends posteriorly to below posterior '/2 of orbit. Chin barbel thin, tapered to a fine tip, its length about equal to snout length. Preopercle broadly rounded, not lobelike posteriorly. Premaxillary teeth in moderate band, 3-4 rows laterally (in juvenile paratype uniserial laterally, in 2 rows anteriorly); outer teeth larger than those of inner row. Mandibular teeth irregularly set, about 2 rows laterally, 4-5 rows near symphysis; none enlarged. Light organ well developed; a large oval an- terior expansion of periproct containing PDW has a narrow isthmus connected to large ADW, whose anterior border lies slightly forward of line between insertions of pelvic fins (well behind line in paratype). Fin positions of holotype best seen in Figure 6. Pelvic fin origin well in advance of vertical from pectoral origin. Outer pelvic ray elongated, extending to base of Sth anal fin ray. Leading edge of 2nd spinous ray of 1D. with about 14 denticles (about 5 in paratype); height of ray slightly greater than HL. Interdorsal space 1.25 times longer than 1D. base. Color medium brown overall, belly dark. First dorsal, pectoral, and pelvic fins all uniformly dark; anal fin dusky. Barbel pigmented basally, pale distally. Lower gill cover walls pale, otherwise cavity dark. S1zE.—To at least 150 mm TL ETYMOLOGY.—From Latin, punctulatus, dot- ted, in reference to the dense covering of mela- nophores on the head. DISTRIBUTION. —Currently known only from the Coral Sea off New Caledonia and off north- eastern New Guinea. REMARKS AND COMPARISONS. — Represented by a single pelagic-stage prejuvenile and a mature female with large ovaries. The new species is most similar to Kumba japonica in its overall physiognomy and in having prominent open pores of the head sensory canals. Kumba punc- tulata differs, however, in having a broader in- terorbital (31% HL vs. 24—28%), shorter preoral length (14% vs. 23-27%), fewer pelvic fin rays (9-10 vs. 11-12), and a shorter barbel (16% vs. 25-27%). It also lacks the three black pigment marks long the base of the anal fin that charac- terizes K. japonica. The Key To SPECIEs provides characters for distinguishing K. punctulata from other congeners. In many ways, K. punctulata appears very close to members of Lucigadus and probably would be assigned to that genus were it not for the extensive naked areas on the head. 234 Kumba sp. A (Figure 2d) MATERIAL EXAMINED.—CAS 77314 (25.1 mm HL, 129+ mm TL); western Gulf of Mexico off Brownsville, Texas, 25°05’N, 96°00'W; 1,280 m; R/V Oregon sta. 4803; 7.1V.1964. DiaGnosis.—Head firm, non-inflated head sensory canals without enlarged pores; naked area dorsally on snout extends posteriorly beyond lat- eral nasal angles about to level of anterior orbital margin; V. 8; orbit 30% HL; snout. about equal to orbit, about 1.2 times interorbital width; up- per jaw about 34% HL; ADW teardrop-shaped, connected to periproct by a narrow isthmus. COUNTS AND MEASUREMENTS. —(see also Dr- AGNOsIS and Table 1). Scales below 2D. 11 or 12, below mid-base 1D. 7.5, lat. 1. over predorsal length 42. The following in percent HL: post- rostral 73.3; pre-A. 122; pre-anus 113; V.-A. 30; isthm.-A. 68; body depth 72; 1D.-2D. 30; height 1D. 76; length base 1D. 32; length 1P. 61. DESCRIPTION. — General features seen in Figure 2d. Head length about 5 in TL, moderately com- pressed laterally, greatest depth about 2 in length; body long, gradually tapering, depth at point twice HL from tip of snout about equal to postorbital length of head. Snout protruding slightly beyond mouth; snout broad, width between lateral nasal ridges slightly more than interorbital width. Or- bit large, long axis slightly oblique. Suborbital region broad, about 0.6 into orbit. Preopercle large, broadly rounded posteriorly, the posterior margin inclined forward. Upper jaw extends pos- teriorly to below midorbit. Chin barbel small, thin, length less than twice diameter of posterior nostril. Gill filaments short, longest about equal to length chin barbel. Gill rakers more or less tubercular. Gill membranes narrowly united over isthmus, forming a free posterior fold. Opercular opening extends forward to posterior end of mandibles. Teeth in villiform tapered bands, 4-6 rows wide. Teeth short and rather bluntly tipped. Out- er premaxillary teeth slightly enlarged and spaced. Scales of suborbital as described for K. calvif- rons. Naked areas of snout, suborbital, lower jaw, and preopercle similar to that of K. calvifrons, including posterior extension of naked areas be- yond level of lateral nasal angles. Characteristics of spinulation on body scales not known because these scales almost completely missing. Pyloric caeca thick, short; about 70 distal el- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 ements counted. Light organ relatively small: ADW extends forward from periproct area, fall- ing short of level of pelvic fin insertions. Fin positions best seen in Figure 2d. Pelvic fin origin anterior to that of pectoral and first dorsal fins; anal fin origin below posterior '4 of base of first dorsal. Interspace between first and second dorsals about equal to length of base of first dor- sal. Second dorsal poorly developed throughout. First dorsal about equal to postrostral length of head; spinous second ray with 4 widely spaced spinules. Outer pelvic ray produced into a long, thin filament that extends posteriorly to base of 20th anal ray. Color mostly lost after almost 20 years in 40% isopropanol. Body and tail appears to have been relatively dark; narrow vertical dark bands still apparent on tail, reminescent of condition illus- trated by Gilbert (1905: Fig. 262) for P. hebetata. DISTRIBUTION. — Known only from the single specimen taken in the western Gulf of Mexico in 1,280 m. REMARKS AND COMPARISONS.—The only known specimen of this species is a juvenile with head and body almost completely denuded of scales. Characteristic scales remain on the sub- orbital shelf and on top of the head, however, to verify these diagnostic features of the genus. Scale pockets indicate the extent of scaled versus naked areas on the head. The species appears to be close to K. hebetata and K. calvifrons in having similar naked areas over the snout and similar pelvic fin ray counts. It further resembles the holotype of K. hebetata in having faint vertical bands on the tail—un- doubtedly a juvenile feature. But the longer snout (29.1% HL vs. 24.8%) with higher dorsal profile, the fewer pectoral fin rays (119-20 vs. 123-24), the fewer denticles on the leading edge of the spinous first dorsal ray, and other less-pro- nounced morphometric and meristic features serve to distinguish the two. Kumba sp. A is very similar to Kumba sp. B in most morphometric and meristic features, but differs in having slight- ly shorter measurements of the postorbital and orbit-to-preopercle lengths, a slightly higher first dorsal ray count, and slightly more gill rakers on the outer series of the first arch. The naked area on the dorsal surface of the snout also extends farther back, posterior to the lateral angles of the snout, compared with only to the lateral angles in Kumba sp. B. IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 235 ——| Ficure 8. Kumba sp. B. CAS 50152, 165+ mm TL, from off Mozambique, in 850-960 m. Fins and squamation partially reconstructed. Scale bar equals 25 mm. Kumba sp. B (Figure 8, 9) Nezumia hebetata: Iwamoto, 1986:338 (nec Macrurus hebe- tatus Gilbert, 1905) (brief descr. of CAS 50152 from Mo- zambique). MATERIAL EXAMINED.—CAS 50152 (35.0 mm HL, 165+ mm TL); off Mozambique; 22°30’S, 36°09’E; 850-960 m; otter trawl; R/V Anton Bruun cr. 8, sta. 399B; 1.X.1964. D1aGcnosis.—Head firm, non-inflated head sensory canals without enlarged pores; naked area dorsally on snout extends posteriorly to lateral angles but not beyond; V. 8; orbit 29% HL; snout about equal to orbit, about 1.1 times interorbital width; upper jaw about 32% HL; ADW teardrop- shaped, connected to periproct by a narrow isth- mus. COUNTS AND MEASUREMENTS. —(see also DI- AGNOsIs and Table 1). Scales below 2D. 13, be- low mid-base 1D. 8.5, lat. 1. over predorsal length 38. The following in percent HL: postrostral 75.4; pre-A. 127; pre-anus 111; V.-A. 31; isthm.-A. 64; body depth 74; 1D.-2D. 36; height 1D. 69+; length base 1D. 27; length 1P. 57. DESCRIPTION. — General shape seen in Figure 8. Head about 1.2 of greatest body depth, greatest width about equal to postorbital length; body moderately compressed laterally, gradually ta- pering; body depth at point twice HL behind tip of snout about equal to postorbital length of head; straplike tail with large part of tip missing, a pseudocaudal developed. Orbit large, about equal to snout length, more than interorbital width. Ventral profile of snout elevated, snout blunt, slightly protruding beyond mouth; least width Figure 9. Scanning electron micrograph of scale of Kumba sp. A, CAS 50152, from dorsum below origin of second dorsal fin. Dotted line equals 0.75 mm. 236 between lateral nasal ridges about equal to in- terorbital width. Suborbital region about 0.6 into orbit. Upper jaws extend posteriorly to below midorbit. Chin barbel small but distinct, tapered to a fine tip, its length about equal to pupil di- ameter. Preopercle somewhat lobelike poster- oventrally, the posterior margin inclined for- ward. Length of gill filaments about equal to diameter of posterior nostril. Gill rakers tuber- cular. Gill membranes rather narrowly united over isthmus, but lacking a free posterior fold. Opercular opening extends forward to below ver- tical arm of preopercular ridge. Premaxillary teeth in a broadly tapered band, 6 or more rows wide; outer teeth enlarged and spaced. Mandibular teeth also in a broad, tapered band, but no enlarged series. All teeth short and rather bluntly tipped. Snout naked dorsally to area slightly behind a line connecting lateral angles. Broad suborbital shelf completely scaled from lateral angles of snout to preopercle. Underside of snout and sub- orbital naked to above angle of lower jaw, the naked areas extend onto lower margin of pre- opercle and interopercle and all of lower jaw. Posterior tip of interopercle probably with few small scales, but none currently on specimen. Head scales as described for Kumba calvifrons except spinule rows fewer on occipital (5-8 rows) and suborbital (1-2 rows) scales—probably attri- butes of small size of specimen compared with adult holotype of K. calvifrons. Most scales of body missing, although some patches remaining over dorsum, along lateral line, over nape, and on chest anterior to pelvic fins. A series of spi- nuleless scales on shoulder girdle below oper- cular margin. Scales of dorsum (Fig. 9) with 5- 7 parallel rows of greatly reclined, short, slender, conical spinules. Scales over nape with spinules tightly aligned in 3-5 vertical, comblike rows, similar to those of top of head. Pyloric caeca thick, short; more than 50 distal elements. Light organ relatively small; light gland immediately anterior to anus; a well-developed lens. A faintly visible anterior dermal window forward of periproct area, between pelvic fin in- sertions. Fin positions best seen in Figure 8. Pelvic fin origin below posterior margin of gill cover, an- terior to pectoral and first dorsal fin origins; anal fin origin below anterior 1/3 of base of first dor- sal. Interspace betwen first and second dorsals slightly longer than length base of first dorsal. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 11 Second dorsal weakly developed throughout. Length of first dorsal about equal to postrostral length of head; spinous second ray with numer- ous short, conical spinules. Outer pelvic ray pro- duced into a long, thin filament that extends pos- teriorly to base of 16th anal ray. Color somewhat faded, overall light to medi- um brown; gill membranes, abdomen, and chest blackish. First dorsal fin basally dark, paler dis- tally; pectoral fin pale; pelvic fin dark; anal fin dark anteriorly but otherwise mostly pale. Gill cavity mostly pale over inner walls, but darker along distal margins. Barbel pale. DISTRIBUTION. — Known only from the single specimen taken in the western Indian Ocean off Mozambique in 850-960 m. REMARKS.— This specimen was originally re- corded by Iwamoto (1986) as Nezumia hebetata, but slight differences in several measurements and counts (see Table 1) suggest that it represents a different but closely related species. The spec- imen is slightly larger than the holotype and only known specimen of Kumba hebetata, and this size difference may account for some of the mor- phometric differences. The high dorsal-ray (II, 13) and pectoral-ray (i23-i24) counts of the K. he- betata holotype contrast sharply with the II,9 and i20 found in the Mozambique specimen. Oth- erwise, the two specimens agree well in other general features. The Mozambique specimen ac- tually agrees closer in most morphometric and meristic characters with Kumba sp. A than with K. hebetata. The primary contrasting characters with Kumba sp. A are the fewer dorsal fin rays, the slightly fewer gill rakers, the fewer pyloric caeca, and the less extensive dorsal naked areas on the snout (extends only to lateral nasal angles vs. posterior to angles in sp. A). ACKNOWLEDGEMENTS We thank the following for their help in var- ious ways: N. V. Parin, Y. N. Shcherbachev, and other staff at IOAN; A. Neyelov and staff at ZIN; N. R. Merrett (BMNH), O. Okamura (BSKU); C. D. Roberts (NMNZ) made available the New Caledonian specimen of Kumba punctulata; D. Ubick (CAS) made the SEMs of the scales; D. Catania, J. Fong, and P. M. Sonoda (CAS) pro- vided other technical assistance. The first author is indebted to the Scientific Exchange Program of the National Academy of Sciences/National Research Council, which enabled him to visit IWAMOTO AND SAZONOV: REVISION OF GENUS KUMBA 237 Russia and to begin the collaboration that led to this and other papers. RUSSIAN ABSTRACT AHHOTAILIIM4A: Pesu30BanH pox Kumba Mar- shall, 1973; B Hero BKJIOYUEHbI JEBATb BHIOB: K. calvifrons sp.n. cO cpeMHHHO-ATIIaHTHYeCc- KOro NomBORHOrO xpeba; K. dentoni Marshall, 1973; K. gymnorhynchus sp.n. 43 BOCTOUHOH yacTtH Muguiickoro oOKeaHa;' K. _ hebetata (Gilbert, 1905); K. japonica (Matsubara, 1943); K. maculisquama (Trunov, 1981); K. punctulata sp.n. H3 MOpeH BucMapKa Hu Ko- pammoporo, a TakxKe 2 [pyrHx Bua, He NONYUMBUIMX Ha3BaHHA. II puHBoaATCA omuca- HHA IByX NOCIeNHHX BHDOB, HO OHH OCTaBsle- HbI 6€3 Ha3BaHHH, TaK KaK MpefcTaBJIeHbl He3peJIbIMH OCOOAMH HJIM HMEIOT COMHHTEJIb- HbIM cTaTyc. Pon Parakumba Trunov, 1981 cBelleH B CHHOHHMbI Kumba. KomMOnHallHa WMarHocTwuecKHX MpH3HaKOB pola TaKOBa: MMeIOTCA OOWIHPHbIe OecuelIyHHble yuacTKH Ha BepXHeH WH HWMXKHEH CTOpOHAaXx Ppbila U HW2KHEM MOBEPpXHOCTH FOJIOBLI; OTCYTCTBYIOT YBeJIMYeHHble PbIJIbHble YelllyH; 4YelllyH Ha FONOBe CBOeOOpa3HOro CTpOeHHA: C OTHHM WJIM HEMHOrHMH rpeOeHKOBHHBIMH pxaMu HeOOJbUIMX MpAMOCTOALIMX WIKMHKOB, MelI- KHe; B NOLIa3HHYHOH OOMacTH OTCYTCTBYIOT yBeJIMYeHHble UIMTKOBHOHbIe YelllyH. LITERATURE CITED EsCHMEYER, W. N. 1990. Catalog of the genera of Recent fishes. Calif. Acad. Sci., San Francisco. 697 pp. GmBerT, C. H. 1890. A preliminary report on the fishes collected by the steamer Albatross on the Pacific coast of North America during the year 1889, with descriptions of twelve new genera and ninety-two new species. Proc. U. S. Natl. Mus. 13:39-126. . 1905. The deep-sea fishes of the Hawaiian Islands. Pp. 575-713, Figures 230-276, Pls. 66-101 in The aquatic resources of the Hawaiian Islands. D. S. Jordan and B. W. Evermann, eds. Bull. U. S. Fish Comm. 1903, 22 (Pt. 2, Sec. 2). Gunnerus, J. E. 1765. Efterretning om berglaxen, en rar norsk fishk, som kunde kaldes: Coryphaenoides rupestris. K. Norske Videnskabers Selskab Skrifter Trondhjem, 3(4):50- Sos ble 3) Pigs: 15 2: Iwamoto, T. 1970. The R/V Pillsbury Deep-Sea Expedition to the Gulf of Guinea, 1964-65. 19. Macrourid fishes of the Gulf of Guinea. Stud. Trop. Oceanogr. (4) Pt. 2:316-431. 1979. Eastern Pacific macrourine grenadiers with seven branchiostegal rays (Pisces: Macrouridae). Proc. Calif. Acad. Sci. 42(5):135-179, Figs. 1-18. . 1986. Family Macrouridae. Pp. 330-341 in Smith’s Sea Fishes. M. M. Smith and P. C. Heemstra, eds. Macmillan South Africa, Ltd., Johannesburg. 1047 pp. Iwamoto, T. AND Y. I. Sazonov. 1988. A review of the southeastern Pacific Coryphaenoides (sensu lato) (Pisces, Gadiformes, Macrouridae). Proc. Calif. Acad. Sci. 45(3):35- 82, Figs. 1-9. KamonarA, T. 1959. New records of fishes from Kochi Pre- fecture, Japan. Rept. Usa Mar. Biol. Sta. 6(2):1-8, 1 PI. Leviron, A. E. AND R. H. Gispss, Jr. 1988. Standards in herpetology and ichthyology: Standard symbolic codes for institutional resource collections in herpetology and ichthy- ology. Supplement No. 1: Additions and corrections. Copeia 1988(1):280-282. Leviton, A. E., R. H. Grass, Jr., E. HEAL, AND C. E. DAwson. 1985. Standards in herpetology and ichthyology: Part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia 1985(3):802- 832. MarsHaLt, N. B. 1973. Family Macrouridae. Pp. 496-665 in Fishes of the western North Atlantic, D. M. Cohen, ed. Mem. Sears Found. Mar. Res. (1)(Pt. 6). MatsusaRA, K. 1943. Ichthyological annotations from the depth of the Sea of Japan. VIII. Emendation of fish including three new species. J. Sigenk. Kenky. 1(2):131-143, Figs. 1-5. Merrett, N. R., Y. I. SAzONOv, AND Y. N. SHCHERBACHEV. 1983. A new genus and species of rattail fish (Macrouridae) from the eastern North Atlantic and eastern Indian Ocean, with notes on its ecology. J. Fish Biol. 22:549-561. OKAmMuRA, O. 1970a. Fauna Japonica. Macrourina (Pisces). Academic Press, Tokyo. 216 pp., 64 Pls. 1970b. Studies on the macrouroid fishes of Japan. Morphology, ecology and phylogeny. Rep. Usa Mar. Biol. Sta. 17(10-2):1-179. OKAMURA, O., K. AMAOKA, AND F. MITANI(EDS.). 1982. Fishes of the Kyushu-Palau Ridge and Tosa Bay. The intensive research of unexploited fishery resources on continental slopes. Japan Fish. Resource Conserv. Assoc., Tokyo. 435 pp. Sazonov, Y. I. AND Y. N. SHCHERBACHEV. 1982. A prelim- inary review of grenadiers related to the genus Cetonurus Giinther (Gadiformes, Macrouridae). Descriptions of new taxa related to the genera Cetonurus Giinther and Kumba Marshall. (In Russian, with English summary.) Vopr. Ikh- tiol. 22(5):707—721, Figs. 1-4. (Engl. transl., J. Ichthyol. 22(5): 1-15.) 1985. Preliminary list of macrourids related to the genus Cetonurus Giinther (Gadiformes, Macrouridae). II. Genus Cetonurus Giinther. Comparative taxonomical anal- ysis of recognized groups. (In Russian.) Vopr. Ikhtiol. 26(2): 179-195. (Engl. transl., J. Ichthyol. 25(3):12-17.) Trunov, I. A. 1981. Parakumba maculisquama gen. et sp. n. New fish of the macrourid family (Macrouridae) from the South Atlantic. (In Russian.) Vopr. Ikhtiol. 21(1):28-35. (Engl. transl., J. Ichthyol. 21(1):27-34.) © CALIFORNIA ACADEMY OF SCIENCES, 1994 Golden Gate Park San Francisco, California 94118 rare dhisry, et sick eg f r | gn rare AA vi: ame M lpticeal ( ar gopialt pany Sd bin Sta Qtrerry eth wes ip Sete wits? se (pre) (woeetTy aa oy eat Raids Me aie De "Sipping ’s SHS nowy! Wd OW Bae moor AGA A207 sy fas) © = i 7 Tr 404 VLAN ARE align Be Mui lhe wz howe 1 - 4 SS < ue = 1) 4 e . \s } v =—— bape st iw Ga US oi Ss P, 00 — PROCEEDINGS OF THE Marine Biological Laboratory/ CALIFORNIA ACADEMY OF SCIEN@ES &: Oceenographic Institution Vol. 48, No. 12, pp. 239-252, 8 figs. NEW SPECIES OF CHROMODORIS AND'NO Library OCT octhnde 34 994 Woods Hoie, UMEA*: ry/ Mi Fa Ucu-n ‘ oods H (NUDIBRANCHIA: CHROMODORIDIDAE) FROM THE WESTERN INDIAN OCEAN AND SOUTHERN AFRICAy jo9/ Terrence M. Gosliner Woods Hole, MA 02543 Department of Invertebrate Zoology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 AssTract: Four new species of Chromodoris and one species of Noumea are described. Chromodoris lekker sp. noy. and C. pruna sp. nov., are known from more than one locality in the western Indian Ocean, while C. kitae sp. noy. is known only from northwestern Madagascar. Chromodoris heatherae sp. nov. and Noumea protea sp. nov. are restricted to the temperate waters of South Africa and appear to be more closely related to sister taxa from temperate Australia than to other congeners. Received March 1, 1994. Accepted June 7, 1994. INTRODUCTION The chromodorid opisthobranch fauna of the tropical Indo-Pacific has received considerable attention in recent years (Rudman, 1982, 1983, 1984, 1985, 1986a, 1986b, 1986c, 1987, 1988, 1990; Baba, 1988; Bertsch and Gosliner, 1989). During the last decade, more than 70 new species have been described. Despite this resurgence in interest, knowledge of this speciose family re- mains far from complete. Gosliner (1987) re- corded almost 40 species of chromodorids from southern Africa; more than half were considered to be undescribed. Since my original work on the opisthobranchs of southern Africa, I have made additional field collections from the granitic Seychelles, Aldabra Atoll, and Madagascar. This paper includes sev- eral of these new species from southern Africa and other tropical localities in the western Indian Ocean. Descritions of four new species of Chromo- doris and one of Noumea are contained in this paper. Specimens deposited in the Department of Invertebrate Zoology and Geology of the Cal- ifornia Academy of Sciences, San Francisco, are designated by the abbreviation CASIZ. Speci- mens deposited in the Department of Marine Biology of the South African Museum, Cape Town, are designated by the abbreviation SAM. SPECIES DESCRIPTIONS Chromodoris heatherae sp. nov. (Figs. 1A, 2, 3A) Chromodoris sp. 1. Gosliner, 1987: 76, Fig. 112. Type MATERIAL. — Holotype: CASIZ 073919, one specimen, Onrus, Walker Bay, Cape Province, intertidal zone, 5 February 1980, T. M. Gosliner. Paratypes: CASIZ 086859, one speci- men, Philip’s Reef, Algoa Bay, Cape Province, 10 m depth, 21 May 1984, T. M. Gosliner. CASIZ 086869, one specimen, Hottentot’s Huisie, Atlantic coast of Cape Peninsula, 15 m depth, January 1981, T. M. Gosliner. CASIZ 073978, one specimen, Philip’s Reef, Algoa Bay, 10 m depth, May 1984, [239] 240 T. M. Gosliner. CASIZ 074653, one specimen. Philip’s Reef, Algoa Bay, 10 m depth, 21 May 1984, G. C. Williams. CASIZ 074097, one specimen, The Mill, Bakoven, Atlantic coast of Cape Peninsula, 16 September 1982, T. M. Gosliner. CASIZ 073470, three specimens, one dissected, Philip’s Reef, Algoa Bay, 11 m depth, 15 May 1984, T. M. Gosliner. CASIZ 074079, one specimen, Philip’s Reef, Algoa Bay, 10 m depth, 21 May 1984, T. M. Gosliner. SAM A 35436, five specimens, Castle Rocks, False Bay, and Cape Hangklip, no date given, T. M. Gosliner. SAM A 35435, one specimen, off Llandudno, Cape Peninsula, 20-23 m depth, 16 March 1989, T. M. Gosliner. SAM A 35442, one dissected specimen, Onrus, Walker Bay, intertidal zone, no date given, T. M. Gosliner. SAM A 35437, one specimen, Sea Forth, False Bay, 9 November 1981, W. R. Liltved. SAM A 35433, Rooi Els, False Bay, 10 m depth, 26 October 1980, T. M. Gosliner. SAM A 35432, two specimens, one dissected, Oudekraal, no depth date or collector given. SAM A 35434, five specimens, one dissected off Llandudno, Atlantic coast of Cape Peninsula, no depth given, 23 December 1979, W. R. Liltved. EtTyMOLOGY.—Chromodoris heatherae is named for my daughter, Heather Erica Gosliner, who was born in Cape Town, where this species is common. DISTRIBUTION. — This species is found around the temperate coast of the Cape Province from the Atlantic coast of the Cape Peninsula to Port Elizabeth. EXTERNAL MorpHo.toGy.—The living ani- mals (Fig. 1A) are 20-70 mm in length. The body is translucent white with white gills and rhin- ophores. A band of numerous, crowded, opaque white glands is present along the margin of the notum. In most specimens, a series of irregular blood-red spots is arranged randomly over the surface of the notum. Only a few specimens from Algoa Bay lacked red spots. Some specimens from Algoa Bay had an additional submarginal yellow band along the edge of the notum. The triangular foot extends posteriorly beyond the notum. The foot lacks any red or yellow markings in all spec- imens examined. On either side of the head is a short, digitiform oral tentacle. There are 12-22 unipinnate gills forming the branchial plume. The perfoliate rhinophores bear 14—20 lamellae. BuccaL Mass.—The anterior portion of the buccal mass is elongate and glandular. The glan- dular portion extends further posteriorly on the ventral surface than it does dorsally. The broad jaws bear dense chitinous rodlets (Fig. 2A), each bearing bifid denticles at their apex. Occasion- ally, some rodlets may be undivided or trifid. The radular formula is 46 x 36.1.36, 55 x 43.1.43,57 x 44.1.44, and 68 x 52.1.52, in four specimens examined. The rachidian tooth (Fig. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 2B) is reduced and devoid of denticles. The in- nermost lateral tooth (Fig. 2B) bears one to three denticles on the inner side of the triangular pri- mary cusp and two to five denticles on its outer side. The remaining lateral teeth lack denticles on their inner side but have denticles on their outer margin. The lateral teeth from the central portion (Fig. 2C) are arched with an elongate cusp. There are 5—14 denticles on the outer mar- gin of the teeth. The outermost laterals (Fig. 2D) are blunt and elongate with 2-4 denticles. REPRODUCTIVE SySTEM.—(Fig. 3A) The am- pulla is thin and elongate. It narrows distally and bifurcates into the short oviduct and the vas de- ferens. The oviduct enters the female gland mass in the region of the albumen gland. The mem- brane and mucous glands are larger than the al- bumen gland and are situated peripherally around the gland mass. The mucous gland exits at the nidamental opening. At this opening 1s a nodular vestibular gland. The highly convoluted proxi- mal portion of the vas deferens is the glandular, prostatic segment. The prostate narrows slightly and again expands into the ejaculatory portion. This coiled section terminates in the unarmed penis, which lacks armature and a distinct pa- pilla. Adjacent to the penis is the straight vaginal duct. Near the middle of the length of the vagina, the uterine duct emerges and joins with the fe- male gland mass. More proximally are the curved, muscular, digitiform receptaculum seminis aand the spherical, thin-walled bursa copulatrix. Discussion. — Chromodoris heatherae is sim- ilar in color pattern to several species of Chro- modoris that are found along the temperate coast of Australia (Rudman, 1983). It is most similar to Chromodoris splendida (Angas, 1864). How- ever, C. splendida has red lines on the gills and rhinophores that are absent in all material of C. heatherae examined. C. hunteri Rudman, 1985, also from Australia, lacks red on the gills and rhinophores. In C. hunteri the yellow border is marginal, while in C. heatherae it is submarginal, when present. Chromodoris hunteri rarely ex- ceeds 15 mm in length, while the length of C. heatherae may exceed 70 mm. Both C. splendida and C. hunteri have fewer denticles (a maximum of eight) on the lateral teeth than C. heatherae, which has 12-14 denticles on the teeth bearing the most denticles. Even in small specimens of C. heatherae that are the same size as those of C. hunteri, these differeces are consistent. GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA 241 < ey 4 Ficure 1. Living animals. A. Chromodoris heatherae sp. nov, from Algoa Bay, South Africa. B. Chromodoris kitae sp. nov., from Nosy Komba, Madagascar. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 Ficure 2. Chromodoris heatherae sp. nov., scanning electron micrographs. A. Jaw rodlets, scale = 15 u. B. Rachidian and inner lateral teeth, scale = 30 uv. C. Lateral teeth from central portion of half-row, scale = 25 yp. D. Outer lateral teeth, scale = 43 wu. Chromodoris kitae sp. nov. (Figs. 1B, 3C, 4) Type MATERIAL. — Holotype: CASIZ 092488, Nosy Tanike- ly, Madagascar, | m depth, 14 April 1989, T. M. Gosliner. Paratypes: CASIZ 092487, one specimen, dissected, Nosy Komba, Madagascar, | m depth, 21 April 1992, T. M. Gos- liner. CASIZ 092486, two specimens, Nosy Komba, Mada- gascar, | m depth, 12 April 1990, T. M. Gosliner. ETYMOLOGyY.—This species is named for my friend and colleague Katharine (Kit) Stewart, who made it possible for me to make two expeditions GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA 243 Ficure 3. Reproductive systems. A. Chromodoris heatherae sp. nov., scale = 2 mm. B. Chromodoris lekker sp. nov., scale = 1 mm. C. Chromodoris kitae sp. nov., scale = 1 mm. D. Noumea protea sp. nov., scale = 1 mm, a = ampulla; be = bursa copulatrix; ej = ejaculatory portion of vas deferens; fg = female gland mass; pr = prostatic portion of vas deferens, rs = receptaculum seminis; v = vagina; vg = vestibular gland. to survey the opisthobranch fauna of Madagas- car. EXTERNAL MorpHoLoGy.—The living ani- mals (Fig. 1B) are 5-18 mm in length. The body is translucent white. Along the margin of the notum and over the surface of the mantle are scattered spots of dark purple. Immediately in- side the area of outer purple spots are concen- trations of reflective yellow or orange-yellow granules. The gills and rhinophores are gray- brown. There are numerous small opaque white spots arranged in distinct rows along the edges of the rhinophoral lamellae and gill pinnae. The- branchial plume consists of eight unipinnate gills. The perfoliate rhinophores are composed of up to 15 lamellae. The oral tentacles are short and digitiform. BuccaL Mass.—The buccal mass is divided evenly into an anterior glandular portion and a posterior muscular one. At the posterior end of the mass is an elongate pair of salivary glands. Within the buccal mass, at the anterior end of the muscular portion of the buccal mass are the jaws. They contain numerous elongate, bifid rod- lets (Fig. 4A). The radular formula in one spec- imen is 45 x 35.1.35. The rachidian teeth (Fig. 4B) are small and vestigial, devoid of denticles. The innermost lateral teeth (Fig. 4B) have one PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 Figure 4. Chromodoris kitae sp. nov., scanning electron micrographs. A. Jaw rodlets, scale = 15 uw. B. Rachidian and inner lateral teeth, scale = 15 y. C. Lateral teeth from central portion of half-row, scale = 20 uw. D. Outer lateral teeth, scale = 30 u. or two denticles on their inner side of the larger cusp and two to three denticles on the outer side. The remaining outer lateral teeth lack denticles on the inside of the primary cusp and have a series of denticles along the outer edge. The outer laterals from the middle of the radular half-row (Fig. 4C) are elongate with 6-8 denticles along the margin while the more elongate outermost teeth (Fig. 4D) have three to four denticles, sit- uated at the tip of the teeth. REPRODUCTIVE SYSTEM.—(Fig. 3C) The am- pulla is moderately long and tubular and narrows into a long postampullary duct prior to bifur- cating into the oviduct and vas deferens. The GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA 245 oviduct is relatively thin, curved and elongate. It enters the female gland mass near the albumen gland. The proximal prostatic portion of the vas deferens is tightly coiled, consisting of several distinct loops. The duct narrows slightly into the muscular ejaculatory portion, which terminates at an unarmed penis and shares a short common duct with the vagina prior to the joint exit at the gonopore. The female gland mass consists of a large lobed mucous gland and smaller membrane and albumen glands. Near the exit of the mucous gland is a small, lobed vestibular gland. The va- gina is of moderate length. The uterine duct emerges from the common junction of the va- gina, the bursa copulatrix and the receptaculum seminis, and enters the female gland mass. The thin-walled, spherical bursa has an extremely short duct while the muscular, pyriform recep- taculum has an elongate, curved duct. DiscussIon.— Rudman (1983) reviewed spe- cies of chromodorids with large, purple spots. Chromodoris kitae is most similar in appearance to C. bimaensis Bergh, 1905. In C. bimaensis, the largest purple markings form rings rather than spots. In C. kitae, the yellow submarginal pig- ment is interrupted with areas of translucent white, while in C. bimaensis the yellow line is interrupted only by purple spots. The yellow of C. kitae is composed of reflective granules, while the pigment of C. bimaensis does not appear to be reflective. The rhinophores and gills of the four specimens of C. kitae are all ornamented with opaque white spots, that are absent in spec- imens of C. bimaensis. Chromodoris bimaensis has purple spots on the edge of the foot, which are absent in C. kitae. The radular morphology of the two species is similar. However, the out- ermost lateral teeth of C. kitae are elongate and denticulate, while those of C. bimaensis are short, curved, and lack denticles. The reproductive sys- tem of C. bimaensis has not been described. Al- though the two species have a similar color pat- tern, the differences noted above appear to be consistent. Chromodoris kitae is thus far known only from northern Madagascar, while C. bi- maensis has been found from Indonesia, New Caledonia, and Fiji, all in the Western Pacific. These differences strongly suggest that they rep- resent distinct species. Chromodoris kitae is also similar to western Indian Ocean and Red Sea specimens of C. as- persa (Gould, 1852) (Rudman, 1983). However, C. aspersa lacks purple spots outside of the or- ange marginal ring. The gills and rhinophores of C. aspersa are orange rather than gray-brown and lack opaque white spots. Chromodoris lekker sp. nov. (Figs. 3B, 5A, 6) Chromodoris sp. 4, Gosliner, 1987: 77, Fig. 115. Type MATERIAL. — Holotype, CASIZ 073265, Nosy Tanike- ly, Madagascar, 1 m depth, 11 April 1990, T. M. Gosliner. Paratypes, CASIZ 074127, one specimen, dissected, 1 km N of Mahé Beach Hotel, Mahé Island, Republic of the Seychelles, 1 m depth, 30 April 1984, T. M. Gosliner. CASIZ 073291, one specimen, | km N of main pass in barrier reef off Mora Mora Resort, 20 km N of Tulear, Madagascar, 20 m depth, 28 March 1990, T. M. Gosliner. CASIZ 073554, one specimen, fringing reef, Mangalimasa, S of Hotel Soanambo, Ile Sainte Marie, Madagascar, | m depth, 6 April, 1990, T. M. Gosliner. CASIZ 073540, five specimens, south end of cove at La Crique Hotel, Ile Sainte Marie, Madagascar, | m depth, 8 April 1990, T. M. Gosliner. CASIZ 073556, two specimens, off Ambatarao Village, NW end of Ile Sainte Marie, Madagascar, | m depth, 7 April 1990, T. M. Gosliner. SAM A 35466, one specimen, dissected, Adlam’s Point, Sodwana Bay National Park, Natal, South Africa, date depth and collector not known. ETYMOLOGY.—The epithet “lekker” comes from the South African slang meaning nice or sweet, referring to the attractiveness of this spe- cies. This name has not been latinized to preserve the vernacular meaning of the word in South Africa. DISTRIBUTION. — This species is widespread in the western Indian Ocean and has been recorded from Natal, South Africa, both coasts of Mad- agascar, the granitic Seychelles, and Reunion Is- land (Dr. Maurice Jay, photo and personal com- munication). EXTERNAL MorpHo.LoGy.—The living ani- mals (Fig. SA) are 10-22 mm in length. The body is translucent white. Around the margin of the notum is an interrupted line of opaque white. Immediately centrad to the opaque white line is a wider submarginal band of burnt orange, which may be entire or interrupted. In the specimen from South Africa, a line of small dark purple to black spots is present on the orange band. These spots were absent in all other material examined. Centrad to the orange band is a wide band of light purple. Within this band is a series of 16-30 large, dark purple to black spots form- ing a ring. In the central portion of the notum are some irregular, opaque white areas that bear small, scattered opaque white tubercles. The rhinophores and gills are transulcent white with a covering of opaque white. The branchial plume PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 Ficure 5. Living animals. A. Chromodoris lekker sp. nov., from Ile Ste. Marie, Madagascar. B. Chromodoris pruna sp. nov., from Nosy Tanikely, Madagascar. C. Noumea protea sp. nov., from Hottentot’s Huisie, Oudekraal, Cape Town, South Africa. GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA Figure 6. Chromodoris lekker sp. nov., scanning electron micrographs. A. Jaw rodlets, scale = 15 y. B. Rachidian and inner lateral teeth, scale = 20 u. C. Lateral teeth from central portion of half-row, scale = 30 yw. D. Lateral teeth from central portion of half-row, scale = 20 u. E. Outer lateral teeth, scale = 50 wu. is composed of 6-7 unipinnate gills. The perfo- liate rhinophores have about 14 lamellae. On either side of the head are digitiform oral ten- tacles. BuccaL Mass.—The buccal mass is large and elongate, The anterior half is glandular. The pos- terior halfis highly muscularized. Within the mass at the anterior end of the muscular portion of the buccal mass are the jaws, which contain nu- merous bifid rodlets (Fig. 6A). The radular for- mula is 35 x 35.1.35 and 49 x 33.1.33 in two specimens examined. The rachidian teeth (Fig. 248 6B) are reduced and lack a distinct cusp or den- ticles. The innermost laterals (Fig. 6B) have one or two small denticles on the inside of the tri- angular cusp and three to four denticles on the outer side. The outer laterals (Figs. 6C, D) in- crease in length towards the outer margin. They lack denticles on the inner side and possess six to eight denticles along the outer edge. The out- ermost teeth (Fig. 6E) are elongate with the den- ticles concentrated near the end rather than being evenly spread along the length of the tooth. REPRODUCTIVE SySTEM.—(Fig. 3B) The am- pulla is long and tubular and narrows into a short postampullary duct prior to bifurcating into the oviduct and vas deferens. The oviduct is rela- tively thin and long and enters the female gland mass near the albumen gland. The proximal por- tion of the vas deferens is tightly coiled, con- sisting of several distinct loops. The duct narrows slightly into the muscular ejaculatory portion, which terminates at an unarmed penis and shares a short common duct with the vagina prior to the joint exit at the gonopore. The female gland mass consists of a large lobed mucous gland and smaller membrane and albumen glands. The va- gina is of moderate length. The uterine duct emerges from the common junction of the va- gina, the bursa copulatrix and the receptaculum seminis, and enters the female gland mass near the albumen gland. The thin-walled, spherical bursa has an extremely short duct while the mus- cular, pyriform receptaculum has an elongate, curved duct. Discussion. —Chromodoris lekker is most similar in its coloration to C. decora (Pease, 1860), which is widespread in the Indo-Pacific tropics from numerous localities from Christmas Island in the Indian Ocean to Hawaii (Rudman, 1986b). Though the two species have similar color pat- terns (Kay and Young, 1969; Bertsch and John- son, 1981; Rudman, 1986b), there are consistent differences. Chromodoris lekker has a largely complete, opaque white marginal band, while C. decora has a few opaque white spots on the or- ange ring. Chromodoris lekker has a purple ring inside the orange one that is absent in C. decora. In C. decora, there is a thin, opaque white band and several lines that are absent in C. /ekker, which does have scattered, opaque white tuber- cles. The radular morphology of the two species is very similar. All of the records for the two species indicate similarity in distribution, dentition, and PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 number of radular teeth. One exception is that Kay and Young stated that a rachidian row of teeth is absent in C. decora. Specimens of C. decora from Hawaii (CASIZ 071564) and Papua New Guinea (CASIZ 086403) examined in this study did indeed have vestigial rows of rachidian teeth. All specimens of C. decora examined have a larger maximum number of denticles on their radular teeth (9-15) than do those of C. lekker (7-8), though there is considerable overlap in denticle number. The reproductive systems of the two species are also similar in their mor- phology. Both species have an unusually long duct of the receptaculum seminis, though this is not evident in the specimen described by Kay and Young (1969) from Hawaii. A long duct was present in both specimens of C. /ekker and in specimens of C. decora examined here. Despite the similarity of Chromodoris lekker and C. decora in their coloration, radular anat- omy, and reproductive morphology, the consis- tent differences described above suggest that the two species are distinct. Chromodoris lekker ap- pears to be restricted to the western Indian Ocean, while C. decora is known from the eastern Indian Ocean, throughout the western Pacific, and east- wards to Hawaii. Chromodoris pruna (Figs. 5B, 7) Chromodoris sp. 8, Gosliner, 1987:79, Fig. 119. Type MateriAL.— Holotype, CASIZ 092490, Nosy Tanike- ly, Madagascar, | m depth, 14 April 1989, T. M. Gosliner. Paratype, SAM A dissected, Park Rynie, Natal, South Africa, intertidal zone, 26 April 1982, T. M. Gosliner. EtTyMoLocy.— The name “pruna”’ comes from the rosaceous genus Prunus, referring to the plum- colored spots on the notum. DISTRIBUTION. — This species has been record- ed from tropical southern Africa and Madagas- car. EXTERNAL MorpHo.tocy.—The living ani- mals examined (Fig. 5B) were 4 and 6 mm in length. As indicated in the description of the re- productive system, these were immature indi- viduals. The body is opaque white with a narrow, translucent white margin. Irregular orange patch- es are present around the margin. A large patch of orange is present at the antero-medial end of the animal. Irregular, dark plum spots are scat- tered over the notum. These spots are irregular in shape and have a diffuse poorly defined out- GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA / \ P| uN Chromodoris pruna sp. nov., scanning electron micrographs. A. Jaw rodlets, scale = 7.5 u. B. Rachidian and inner lateral teeth, scale = 15 uw. C. Lateral teeth from central portion of half-row, scale = 15 yu. D. Outer lateral teeth, scale = 15 yu. FIGURE 7. line. The rhinophores are opaque white basally, while most of the clavus is orange. The gills are translucent white and are largely covered with orange pigment. The surface of the orange gills and rhinophores is ornamented by minute, opaque white spots. The branchial plume is com- posed of six simply pinnate gills. The perfoliate rhinophores have eight lamellae. The oral ten- tacles are short and digitiform. BuccaL Mass. — The anterior half of the buccal mass is glandular, while the posterior portion is highly muscular. Inside the buccal mass, at the 250 level of the junction of the muscular and glan- dular portions of the mass, are the jaws. They consist of numerous rows of short, bifid rodlets (Fig. 7A). The radular formula of the paratype specimen examined (A 35477) is 38 x 26.0.26. A rachidian row is absent. The inner lateral teeth (Fig. 7B) bear two denticles on the inside of the triangular cusp and three to four on the outside. The remaining laterals bear denticles only on the outside face of the teeth. In the middle of the half row, the teeth (Fig. 7C) are curved with an elongate cusp with 3-6 denticles. The outermost teeth (Fig. 7D) are more elongate, with four to five denticles situated at the end of the teeth. REPRODUCTIVE SYSTEM.— The reproductive system was entirely immature in the larger of the two specimens. DiscussIon.— This species can be compared to other taxa with dark purple spots reviewed by Rudman (1983). It is most similar in color to C. aspersa (Gould, 1852). That species has a com- plete marginal orange line rather than interrupt- ed patches. Also, there are more purple spots than in C. pruna, and they are lighter in color. The orange gills of C. aspersa are paler than those of C. pruna and lack opaque white spots. In C. aspersa a vestige of the rachidian plate is present, while in C. pruna, the rachidians are absent. Despite the fact that the two specimens were immature, Chromodoris pruna is the only known species of Chromodoris with orange rhinophores and gills ornamented with opaque white spots. The color pattern is sufficiently different from other purple-spotted species that differences in coloration could not simply be attributed to the immature state of specimens of C. pruna ex- amined here. Noumea protea sp. nov. (Figs. 3D, 5C, 8) Chromodoris sp. 2, Gosliner, 1987: 77, Fig. 113. Type MATERIAL. — Holotype, SAM A35431, Oudekraal, At- lantic coast of Cape Peninsula, 15 km S of Cape Town, 10 m depth, 18 October 1980, T. M. Gosliner. Paratypes, SAM A 35430, three specimens same locality and date as holotype. CASIZ 092489, one specimen, dissected, same locality and date as holotype. ETyMoLocy.— The name “‘protea”’ comes from the genus of flowering plants Protea. The color of the present species is the same as the floral bracts of the king protea, Protea cynaroides. EXTERNAL MorpHoLoGy.—The living ani- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 mals (Fig. SC) are 10-13 mm in length. The an- imals are uniformly rose-pink in color with a series of opaque white mantle glands along the margin of the notum. The gills and rhinophores are also uniformly pink. The branchial plume consists of seven unipinnate lamellae. The per- foliate rhinophores contain seven lamellae. The oral tentacles are short and digitiform. BuccaL MaAss.—The jaws consists of numer- ous irregular, multifid rodlets (Fig. 8A). The rad- ular formula in one specimen is 22 x 11.1.0.1.11 (CASIZ 092489). A rachidian row of teeth is en- tirely absent. The inner lateral teeth (Fig. 8B) are much broader than the succeeding inner laterals. The inner laterals bear one to two denticles on the inner side of the elongate primary cusp and two to four denticles on the outer side. The suc- ceeding laterals (Fig. 8C) all lack denticles on the inner side of the cusp and have 0-2 denticles on the outer side. The teeth are more elongate near the outer margin and lack denticles. REPRODUCTIVE SyYSTEM.—(Fig. 3D) The am- pulla is short and wide and narrows into a thin postampullary duct. The postampullary duct narrows distally and bifurcates into the short ovi- duct and the vas deferens. The oviduct enters the female gland mass in the region of the albumen gland. The membrane and mucous glands are larger than the albumen gland and are situated peripherally around the gland mass. The mucous gland exits at the nidamental opening. At this opening, is a large vestibular gland, consisting of numerous discrete lobes. The short, curved, proximal portion of the vas deferens is the glan- dular, prostatic segment. The prostate narrows slightly into the muscular ejaculatory portion. This coiled section terminates in the unarmed penis, which lacks a distinct papilla. Adjacent to the penis is the slightly curved vaginal duct. Near the middle of the length of the vagina, the uterine duct emerges and joins with the female gland mass. More proximally are the curved, muscular, club-shaped receptaculum seminis and the spherical, thin-walled bursa copulatrix, which have a common junction with the proximal end of the vagina. Discussion. — Noumea protea is similar in ap- pearance to Noumea haliclona (Burn, 1957) from southern and temperate eastern Australia. Nou- mea haliclona varies in color (Rudman, 1983) from white to pink to yellow. Deep-pink speci- mens resemble material of N. protea, except that additional spots of red and a submarginal band GOSLINER: NEW SPECIES OF CHROMODORIS AND NOUMEA 251 Ficure 8. Noumea protea sp. nov., scanning electron micrographs. A. Jaw rodlets, scale = 15 yw. B. Inner lateral teeth, scale = 25 uw. C. Outer lateral teeth, scale = 25 u. are present in the former. Both species are among the few members of the genus described as hav- ing a well-developed vestibular gland consisting of numerous lobes (Rudman, 1984). The two differ in the arrangement of the radular teeth. In N. protea there are 11 outer lateral teeth per half row, while in mature specimens of N. haliclona there are 17-20 teeth per half row. The second lateral teeth of N. haliclona have six denticles while those of N. protea have only one or two denticles, The remaining outer lateral teeth of N. haliclona have more denticles than the corre- sponding teeth of N. protea. NOMENCLATURAL CHANGES IN SOUTHERN AFRICAN CHROMODORIDS Since Gosliner (1987) recorded 39 species of chromodorid nudibranchs from southern Africa, several systematic changes have taken place that affect the identifications in that work. The species identified as Chromodoris marginata (Pease, 252 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 12 1860) must be regarded as C. verrieri (Crosse, 1875) (Rudman, 1985). Chromodoris vicina El- iot, 1904, is regarded as a junior synonym of C. tennentana (Kelaart, 1859) (Rudman, 1987). Specimens described from the western Indian Ocean as Chromodoris geometrica (Risbec, 1929) by Edmunds (1971) and Gosliner (1987) may represent a distinct, undescribed species owing to consistent differences in color between Pacific and western Indian Ocean specimens. Chromo- doris sp. 3 has been described as Chromodoris geminus Rudman, 1987, and C. sp. 7 as C. alius Rudman, 1987, based on specimens described from the coast of Tanzania. The specimens con- sidered by Gosliner (1987) to represent Noumea varians (Pease, 1871) are in fact Durvilledoris pusilla (Bergh, 1874). Glossodoris sp. 2 of Gos- liner (1987) is G. cincta (Bergh, 1888), and G. sp. 3 is G. hikuerensis (Pruvot-Fol, 1954), based on re-examination of material and comparison with the recent review of Glossodoris (Rudman, 1986a). Several species of Cadlina, Hypselodoris, Ceratosomaand Glossodoris remain undescribed and will be treated in forthcoming papers. ACKNOWLEDGEMENTS Bill Liltved and Gary Williams assisted in col- lecting specimens from southern Africa. Field work in the Seychelles was supported by the Smithsonian Institution as part of the Aldabra Research Program. Financial support for travel to Madagascar was provided through the gen- erous support of Katharine Stewart. The Pho- tography Department of the California Academy of Sciences kindly printed all of the black and white photographs used in this paper. LITERATURE CITED Basa, K. 1988. Comparative study on two species of Glos- sodoris from Japan (Nudibranchia: Chromodorididae). Ve- nus 47(3):158-166. BertscH, H. AND T. GosLiIneR. 1989. Chromodorid nudi- branchs from the Hawaiian Islands. The Veliger 32(3):247- 265. BERTSCH, H. AND S. JOHNSON. 1981. Hawaiian nudibranchs. Oriental Publishing, Honolulu. 111 pp. Epmunps, M. 1971. Opisthobranchiate Mollusca from Tan- zania (Suborder: Doridacea). Zoological Journal of the Lin- nean Society 50:330-396. GosLINER, T. 1987. Nudibranchs of southern Africa. Sea Challengers, Monterey. 136 pp. Kay, E. A. AND D. Younc. 1969. The Doridacea (Opistho- branchia: Mollusca) of the Hawaiian Islands. Pacific Science 23:172-231. RupMaNn, W. B. 1982. The Chromodorididae (Opisthobran- chia: Mollusca) from the Indo-West Pacific: Chromodoris quadricolor, C. lineolata and Hypselodoris nigrolineata col- our groups. Zoological Journal of the Linnean Society 76: 183-241. 1983. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: Chromodoris splen- dida, C. aspersa and Hypselodoris placida colour groups. Zoological Journal of the Linnean Society 78:105-173. 1984. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: a review of the genera. Zoological Journal of the Linnean Society 81:115-273. 1985. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: Chromodoris aureo- marginata, C. verrieriand C. fidelis colour groups. Zoological Journal of the Linnean Society 83: 241-299. 1986a. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: the genus Glossodoris Ehrenbergh (= Casella, H. & A. Adams). Zoological Journal of the Linnean Society 86: 101-184. 1986b. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: Noumea purpurea and Chromodoris decora colour groups. Zoological Journal of the Linnean Society 86:309-353. 1986c. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: Noumea flava colour group. Zoological Journal of the Linnean Society 88:377- 404. 1987. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: Chromodoris epicu- ria, C. aureopurpurea, C. annulata, C. coiand Risbecia tryoni colour groups. Zoological Journal of the Linnean Society 90: 305407. 1988. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: the genus Ceratosoma J. E. Gray. Zoological Journal of the Linnean Society 93: 133-185. 1990. The Chromodorididae (Opisthobranchia: Mollusca) from the Indo-West Pacific: further species of Glossodoris, Thorunna and the Chromodoris aureomargin- ata colour groups. Zoological Journal of the Linnean Society 100:263-326. © CALIFORNIA ACADEMY OF SCIENCES, 1994 Golden Gate Park San Francisco, California 94118 Vibh@apoe,” 2,2: terete 25 a ee A ee naa F.0R i Pee 1657r ae aes PET e7 Sa baat > ee eee Ae Bere Cate a a | =— ee Pe “ot a 4 - = _ » Sea. &, 1 a8 ohe Tye was er ‘ a - : he. te ye 4 i 8 7 e 7 . ~= mi’? 7 ~ 6 trae bay Ce? soe Re. , 5 7 : 6 » Grr ea pi a hk epee be ray ‘ nat ; Ere) ele t ? . : ma, ° . ie be ps aa ? — ° x a hae eee ‘ Sen K 2% 5 ER : : > — : pe is 1 = - a be = thy i | —_—, . . one? © e- F a 5 ory A. é oe he | o F y . 2 “5 7c : - = a ; , ' 7 ( i 4 7 : - » é ‘ > : . . r n De 7 Vol. 48, No. 13, pp. 253-284, 13 figs. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES NEW AND RECONSIDERED MEXICAN ACANTHACEAE. VI. CHIAPAS By Thomas F. Daniel, ; »™ 1006 VIF 6 IIS Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 Ar Asstract: Novelties and taxonomic discussions pertaining to miscellaneous species of Acanthaceae occur- ring in Chiapas, Mexico are presented prior to a treatment of the family for the Flora of Chiapas series. The New World species Barleria micans is treated as conspecific with the west African species B. oenotheroides. Blechum pyramidatum is shown to be the correct name for the species often treated as B. brownei. Blechum grandiflorum is shown to belong to Blechum rather than Ruellia. Buceragenia is shown to comprise species of Pseuderanthemum with cleistogamous flowers. Trybliocalyx is treated as congeneric with Chileranthemum, and the new combination, C. pyramidatum, is made for the species previously known as both C. violaceum and 7. pyramidatus. Eight new species of Justicia are described from Chiapas; two new combinations are made in Justicia for Chiapan species previously treated in Neohallia and Chaetothylax; and a new name is provided for the species previously known as either Beloperone aurea or Justicia flava D.N. Gibson. The species often treated as Teliostachya alopecuroidea is referred to Lepidagathis, and a discussion of the generic distinctions is provided. The species previously known as Ruellia longituba from Chiapas and Guatemala does not pertain to the type of that name and a new species, R. maya, is described to accommodate it. Habracanthus (including Hansteinia) is treated as congeneric with Stenostephanus; two new Chiapan species are described in the genus; and seven new combinations in Stenostephanus are proposed for the Chiapan species previously treated in Habracanthus and Hansteinia. Received April 21, 1994. Accepted September 15, 1994. March 3, 1995 INTRODUCTION Chiapas is the southernmost state of Mexico, and its vascular flora is estimated to comprise more than 8200 species (Breedlove 1981). There are 131 species in 29 genera of Acanthaceae so far recorded from the state. This is the most spe- cies of Acanthaceae presently known for any Mexican state. However, further studies may re- veal that the family is even more speciose in Oaxaca. Sixteen species of Acanthaceae are en- demic to Chiapas and 13 others are known only from Chiapas and neighboring Guatemala. The following novelties and conclusions are based on studies preparatory to treatment of Acanthaceae for the Flora of Chiapas series. Be- cause new taxa, new combinations, new names, and lengthy discussions are not included in the Flora, they are being published here in antici- pation of the imminent publication of the Flora of Chiapas, Part 4, Acanthaceae. BARLERIA MICANS vs. B. OENOTHEROIDES The sole New World species of the predomi- nately African genus Barleria L. has long been known as B. micans Nees. This species has been considered a neotropical endemic since its de- scription by Nees in 1846. A superficial exami- [253] 254 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 FiGuReE |. nation of several African species revealed that B. oenotheroides Dum. Cours. greatly resembles B. micans. Further examination of specimens of B. oenotheroides at CAS, K, and P revealed these species to be similar in all characters. In addition to sharing all of the diagnostic characteristics usually attributed to B. micans (including the unusual feature ofa yellow corolla that turns dark blackish-purple on drying), B. oenotheroides in Africa expresses the same types of character vari- ation (e.g., robustness of spikes, size of corollas, degree of bracteal serration) as seen among American plants. Therefore, B. micans is in- cluded in the synonymy of B. oenotheroides. Based on the apparent pre-Columbian presence of B. oenotheroides in both hemispheres, the phy- togeographical link between Barleria in Africa and America is elevated to the level ofa common species. Indeed, the tendency of identical taxa to occur under different names in Africa and South America was noted by Gentry (1993). The relationships among African (including Madagascan) and American (including West In- dian) Acanthaceae have been shown to be closer than previously believed. In addition to numer- ous pantropical genera that occur in both Africa Distribution of Barleria oenotheroides. and America, the genera Oplonia Raf., Mendon- cia Vell. ex Vand., and Stenandrium Nees are known only from these two regions. African spe- cies of Oplonia were previously referred to For- sythiopsis Baker (Stearn 1971); African species of Mendoncia were previously referred to Afro- mendonica Gilg ex Lindau and Monachochlamys Baker (Lindau 1895); and African species of Stenandrium were previously treated in Sten- andriopsis S, Moore (Vollesen 1992). The present phytogeographical links of Acanthaceae between Africa (including Madagascar) and South Amer- ica are suggestive of a Gondwanan origin for the family (Leroy 1978) or, at least, an early radia- tion in southern land masses. Interestingly, the present distribution of B. oenotheroides in trop- ical west Africa and northern South America (Fig. 1) coincides with near adjacent regions of South America and Africa in northern West Gondwa- naland (cf. maps in Behrensmeyer et al. 1992, Goldblatt 1993). While B. oenotheroides has a modern distribution evocative of Gondwanan ancestry, Acanthaceae are absent from the Cre- taceous fossil record, and there is no evidence that the species had a formerly continuous dis- tribution in northern West Gondwanaland. The DANIEL: MEXICAN ACANTHACEAE species may have arrived in South America in relatively recent geologic time. B. oenotheroides, which often occurs in disturbed habitats, un- doubtedly radiated northward into Mexico. BLECHUM BROWNEI vs. B. PYRAMIDATUM Two names are currently in use for the tropical weed first described in the Acanthaceae simply as Blechum by Patrick Browne (Browne 1756; for a discussion of the validity of Browne’s ge- neric names, see Dandy 1967). Both B. brownei Juss. (e.g., Long 1970, Gibson 1974, Croat 1978, Correll and Correll 1982, Proctor 1984, Durkee 1986, Fosberg et al. 1993) and B. pyramidatum (Lam.) Urb. (e.g., Adams 1972, Hsieh and Huang 1978, Durkee 1978, Steyermark and Huber 1978, Howard 1989, Smith 1991, Wasshausen 1991) have been used in recent floristic treatments with one name usually listed as a synonym of the other. In order to determine the correct name for this widespread taxon that occurs in Chiapas, the appropriate literature was reviewed. Browne (1756) described the taxon as Blechum and cited pre-Linnaean descriptions and illus- trations of Sloane and Rhedde. Linnaeus (1759) included Browne’s taxon in Ruellia L. and gave itan epithet, R. blechum L. (as “‘blechniv’), which resulted in the first publication of a name for the species. Linnaeus cited illustrations of Plumier (1756) and Sloane (1707). Lamarck (1785) de- scribed the species as Barleria pyramidata and cited the illustration of Plumier (1756). Jussieu (1807) accepted the genus Blechum; cited the illustrations of Plumier and Sloane; referred to previous treatments of Browne, Linnaeus, and others; and gave the species the name B. browne. This name was used by Nees (1847b) and Lindau (1895) in important works on the family and therefore became widely known. Millspaugh (1900) transferred Linnaeus’s epithet to Blechum and created the tautonym B. blechum (L.) Millsp. Urban (1918) transferred Lamarck’s epithet to Blechum, resulting in B. pyramidatum. Breme- kamp (1938:149) argued that the combination B. pyramidatum could not be accepted because its basionym (i.e., Barleria pyramidata) was ‘‘merely a binomial appellation for Patrick Browne’s Blechum, and as Linné had used al- ready for the same purpose the name Ruellia Blechum, the epithet pyramidata is invalid. In the genus Blechum the name chosen by de Jus- sieu is therefore correct.’’ Long (1970) referred 255 to this argument in accepting B. brownei, and Gillis (1974) arrived at the same conclusion. In assessing the correct name for this taxon, the question of typification becomes relevant. The type of Linnaeus’s name would have to be chosen from among those illustrations or spec- imens cited or used by him in drawing up his protologue. Browne sold his Jamaican herbari- um to Linnaeus in 1758, and it is now at LINN. There are presently two specimens of Blechum at LINN. Sheet 804.1 appears to have been cul- tivated in the Hortus at Uppsala. It is not known when it was cultivated there nor whether Lin- naeus had it in his possession by 1759. Sheet 804.2 was collected by Mutis and likely was not received by Linnaeus prior to 1773 (C. Jarvis, in litt., 1990). Ifthere was a specimen of Blechum in Browne’s herbarium, it does not appear to be extant at LINN now. The illustrations of Sloane and Plumier cited by Linnaeus both appear to pertain to our species and one of them would be a logical choice for a lectotype. It is clear from an examination of the literature that because on transfer to Blechum the earliest epithet results in a tautonym, the epithet pro- vided by Lamarck (1783) is the next one avail- able for the name pertaining to this species. Be- cause Lamarck cited only one of the syntypes utilized by Linnaeus, his name may not be con- sidered as superfluous under the stated require- ments of Article 63 of the International Code of Botanical Nomenclature (Greuter et al. 1988). Blechum pyramidatum (Lam.) Urban is the cor- rect name for this tropical weed. BLECHUM GRANDIFLORUM VS. RUELLIA MIRANDANA Based on pollen and floral morphology, Ra- mamoorthy and Hornelas (1988) transferred Blechum grandiflorum Oerst. to Ruellia L. and provided the new name, R. mirandana Rama- moorthy & Hornelas, for it. They noted that pol- len of Blechum is tricolporate and syncolpate at the poles and that the exine is reticulate. Al- though it also has a reticulate exine, pollen of Ruellia is triporate. Interestingly, their figure (Ramamoorthy and Hornelas 1988, Fig. 1A) of R. mirandana shows a polar view of a pollen grain that is tricolporate and syncolpate and agrees with my observations of other collections of B. grandiflorum (Fig. 2a,b). At a gross level, pollen of B. grandiflorum superficially resembles that 256 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 24.8um Figure 2. Scanning electron micrographs of Blechum pollen. (a) B. grandiflorum (Breedlove 50544), polar view; (b) B. grandiflorum, equatorial (colpal) view; (c) B. pyramidatum (Daniel et al. 5454), polar view; (d) B. pyramidatum, equatorial (intercolpal) view; (e) B. costaricense (Daniel et al. 6342), polar view; (f) B. costaricense, equatorial (colpal) view. a, b at same scale; c-f at same scale. of Ruellia by its coarser and more open reticulum compared with that of B. pyramidatum (as B. brownei in Ramamoorthy and Hornelas 1988, Fig. 1D; Fig. 2c,d above). However, other species of Blechum (e.g., B. costaricense Oerst., Fig. 2e, f) exhibit a reticulum somewhat intermediate be- tween that of B. grandiflorum and B. pyrami- datum. Pollen of all species of Blechum (includ- ing B. grandiflorum) examined so far are tricolporate and syncolpate whereas species of Ruellia are neither. Ultimately, pollen provides excellent micromorphological characters for dis- tinguishing between these genera. Ramamoorthy and Hornelas (1988:161) fur- ther noted that the “large strongly exserted flow- ers” of B. grandiflorum were more suggestive of Ruellia than Blechum. Indeed, most species of the large and morphologically diverse genus Ruellia have large flowers (usually more than 20 mm in length). In Chiapas, B. pyramidatum has corollas 10-20 mm in length whereas B. gran- diflorum has corollas 30-45 mm long. However, B. costaricense has corollas intermediate in length (i.e., 20-25 mm long in Costa Rica; Durkee 1986) between the two Mexican species. Thus, corolla size is not very useful for distinguishing the two genera. Probably the most reliable macromorpholcg- ical character for distinguishing these two genera is whether or not the septa, with attached reti- nacula, separate from the inner wall of the ma- ture capsule upon dehiscence. In Blechum septal separation occurs whereas in Rvellia it does not. Septal separation was observed in all fruiting col- lections of B. grandiflorum from throughout its range. DANIEL: MEXICAN ACANTHACEAE Based on these lines of evidence, B. grandiflo- rum is maintained in Blechum. BUCERAGENIA vs. PSEUDERANTHEMUM Buceragenia Greenm. comprises five species from Mexico and Central America that have been distinguished from other Acanthaceae by their small (less than 5 mm long), budlike flowers and androecium consisting of two staminodes and two dithecous to monothecous stamens. Vege- tative and fruiting organs greatly resemble those of Pseuderanthemum. Cleistogamous flowers, re- sembling typical flowers of Buceragenia, some- times occur in species of Pseuderanthemum. For example, specimens of both P. alatum (Nees) Radlk. (e.g., Daniel & Baker 3713, CAS) and P. cuspidatum (Nees) Radlk. (e.g., Breedlove 15414, CAS) have both chasmogamous and cleistoga- mous flowers. Examination of the androecium of cleistogamous flowers of Pseuderanthemum (both collections noted above) reveals an an- droecium comprising stamens that have one theca either greatly reduced or entirely suppressed. Thus, there are no distinctions between these genera; Buceragenia is a genus that merely com- prises species of Pseuderanthemum with cleis- togamous flowers. In Chiapas, there are two species referable to Buceragenia. One of these concurs with descrip- tions of B. glandulosa Leonard in the protologue and in Gibson (1974) and closely resembles a paratype from Costa Rica. This species appears to represent cleistogamous P. cuspidatum. Both Leonard (1938) and Gibson (1974) had noted that B. glandulosa resembles P. cuspidatum but differs by its minute, densely fasciculate flowers. Some plants of P. cuspidatum from Chiapas (e.g. Breedlove 7007) exhibit both cleistogamous and chasmogamous flowers. The other Chiapan spe- cies matches the type (from Mirador, Veracruz) of B. foliaceobracteata (Oerst.) V.M. Baum. My examination of the type (from Zacuapam, Ve- racruz) of B. ruellioides Leonard confirms that it is similar in all respects to that of B. foliaceo- bracteata. A collection of this species from Chia- pas (Daniel et al. 5875, CAS, K, MEXU, MICH) has cleistogamous flowers. Seed from this col- lection grown in a greenhouse in San Francisco initially yielded plants with cleistogamous flow- ers similar to those observed in the field. Even- tually chasmogamous flowers typical of Pseu- deranthemum were produced on these plants. 257 Based on my preliminary studies of these and other species of Pseuderanthemum in Mexico, I tentatively include B. foliaceobracteata and B. ruellioides within P. fasciculatum. While there is little doubt that types of the former names rep- resent the same taxon, there appear to be several differences between that taxon and representa- tives of P. fasciculatum from Chiapas. In the latter, cauline trichomes are bifariously (vs. evenly) disposed, the rachis and abaxial surface of the calyx are glandular (vs. rachis eglandular and calyx usually eglandular), bracteoles are shorter (1.5—4 vs. 4-13 mm long), and flowers are pedicellate with pedicels (1-) 2.5-4.5 mm long (vs. sessile or with pedicels to 1 mm long). The type of P. fasciculatum, which is also from Mirador in Veracruz, is somewhat fragmentary. It has cauline trichomes concentrated in (but not restricted to) two lines, pedicels to 1 mm long, and some calyces distinctly glandular while oth- ers are eglandular. Plants of Daniel 587 5gh (CAS) grown in a greenhouse and producing chasmo- gamous flowers have cauline trichomes some- times concentrated in (but not restricted to) two lines, pedicels to 2 mm long, and calyces varying from eglandular to distinctly glandular. Consid- ering these data, and pending further studies of Pseuderanthemum, plants resembling the types of B. foliaceobracteatum and B. ruellioides are treated as part of a variable P. fasciculatum. Further studies will be necessary in order to determine whether there are already names available in Pseuderanthemum for the remaining two Mexican species that were described in Bu- ceragenia. CHILERANTHEMUM VS. TRYBLIOCALYX Gibson (1970) discussed the delimitation of Trybliocalyx Lindau and recognized the genus based on its “inflated, cupular calyx.” Although the calyx is not truly inflated, it is decidedly cu- pular and has broad, triangular lobes. Gibson (1970, 1974) recognized two species of Tryblio- calyx: T. pyramidatus Lindau (including Clero- dendrum standleyi Moldenke) and T. albicaulis (Brandegee) D.N. Gibson (based on Jacobinia albicaulis Brandegee). Unfortunately the holo- type of the former name was destroyed at B and despite attempts by L.O. Williams (fide corre- spondence at F) and myself to locate isotypes, none have been found. Fortunately, Lindau’s de- scription corresponds well with extant speci- 258 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 mens. The characters used by Gibson (1970, 1974) to distinguish these two species vary with- in populations. For example, plants from the only known locality of the species in Chiapas (Mpio. La Trinitaria, ca. 8 km S of La Trinitaria along Hwy. 190) have either glabrous or pubescent pe- duncles and pedicels, calyces (during anthesis) 6-10 mm long that are abaxially either glabrous or pubescent, and corollas 16-22 mm long. Also the isotype of J. albicaulis at MO has pubescent peduncles, pedicels, and calyx lobes and calyces varying from 6-12 mm in length. There appears to be no basis for distinguishing these two spe- cies. Miranda (1950) described Chileranthemum violaceum Miranda, a species that is similar to T. pyramidatus in all respects and is herewith included as a synonym of that species. In spite of its cupular calyx with broad lobes, Trybliocalyx has all of the generic characteristics of Chileranthemum Oerst.: distyly; androecium of two bithecous stamens and two staminodes; thecae parallel, subequally inserted, and lacking basal appendages; pollen tricolporate, hexapseu- docolpate, and reticulate; and corollas purplish or pinkish with colored markings on the lower lip and with a relatively short tube. For this rea- son, Trybliocalyx is included in Chileranthemum and a new combination is made below for 7. pyramidatus in Chileranthemum. A key to the three species of Chileranthemum is provided by Daniel (1993). The generic distinctions among several New World genera of tribe Justicieae sub- tribe Odontoneminae (e.g., Chileranthemum, Odontonema Nees, Oplonia Raf., and Pseuder- anthemum Radlk.) are much in need of study. Chileranthemum pyramidatum (Lindau) T-.F. Daniel, comb. nov. Trybliocalyx pyramidatus Lindau, Bull. Herb. Boiss. ser. 2, 4: 401. 1904. Type. —GUATEMALA. Huehuetenango: “prope Nenton,” September 1896, C. Seler & E. Seler 3276 (ho- lotype: B, destroyed). Jacobinia albicaulis Brandegee, Univ. Calif. Publ. Bot. 4: 386. 1913. Trybliocalyx albicaulis (Brandegee) D.N. Gibson, Fieldiana, Bot. 32: 176. 1970. TypeE.— MEXICO. Veracruz: near Bafios del Carrizal, August 1912, C. Purpus 6049 (ho- lotype: UC!; isotypes: BM!, F!, GH!, MO!, US!). Clerodendrum standleyi Moldenke, Known Geogr. Distr. Ver- ben. 76. 1942. Type. —GUATEMALA. Zacapa: near divide on road between Zacapa and Chiquimula, 500-600 m, 9 October 1940, P. Standley 73793 (holotype: NY; isotype: F!). Chileranthemum violaceum Miranda, Ann. Inst. Biol. México 21: 315. 1950. Type. —MEXICO. Oaxaca: barrancas SE de Cuicatlan, cercanas al camino a Reyes Papalo, 1100-1300 m, 18 September 1948, F. Miranda 4710 (holotype: MEXU!; isotype: MEXU)). JUSTICIA Justicia L. is the largest and morphologically most diverse genus of Mexican Acanthaceae. Thirty-three species are known from Chiapas. Below, eight new species from Chiapas are de- scribed, two new combinations are proposed, and a new name is provided. Justicia breedlovei T.F. Daniel, sp. nov. (Fig. 3) Type.— MEXICO. Chiapas: Mpio. La Trinitaria, 10 km ENE of Dos Lagos above Santa Elena, 1170 m, 15 December 1981, D. Breedlove 56242 (holotype: CAS!; isotypes: CAS!, C!, ENCB!, K!, MEXU!, MICH!, MO!, US)). Frutex usque ad 1.2 m altus. Folia petiolata, laminae ellip- ticae vel obovato-ellipticae, (37—) 65-200 mm longae (17-) 21- 53 mm latae, (2.2-) 3.1-4.6-plo longiores quam latiores. In- florescentia floribus in spicas pedunculatas vel paniculas sp1- carum; dichasia alterna, sessilia, uniflora. Bracteae ovato-el- lipticae vel anguste ellipticae vel ellipticae vel obovato-ellipticae, 8-19 mm longae, 2—9.5 mm latae, apice rotundatae vel trun- catae (vel emarginatae), pagina abaxialis glabra. Flores sessiles. Calyx 5-lobus, 8—9.5 mm longus, lobis homomorphis. Corolla aurantiaca, 32-34 mm longa, extus pubescens trichomatibus eglandulosis. Stamina filamentis pubescentibus trichomatibus eglandulosis, thecis 2-2.3 mm longis, impariter insertis vel superpositis, basi calcaratis. Capsula 9.5 mm longa, glabra. Shrub to 1.2 m tall; young stems quadrate to quadrate-sulcate, nodes pubescent with flexuose eglandular trichomes 0.3-0.6 mm long, inter- nodes glabrous or bifariously pubescent for a few mm proximal to nodes with trichomes like those at nodes. Leaves turning + blackish on drying, petiolate; petioles to 25 mm long; blades elliptic to obovate-elliptic, (37—) 65-200 mm long, (17-) 21-53 mm wide, (2.2-) 3.1—-4.6 times longer than wide, acuminate at apex, attenuate at base, sur- faces glabrous, margin entire. Inflorescence of axillary and terminal, pedunculate, dichasiate spikes or panicles of spikes to 180 mm long (in- cluding peduncles and excluding flowers) from axils of leaves or bracts ( = inflorescence bracts), spikes or panicles alternate or opposite, 1 per axil, peduncles to 45 mm long, glabrous or pu- bescent like young stems, rachises of both spikes and panicles pubescent with antrorse to flexuose eglandular trichomes 0.2-0.7 mm long, tri- DANIEL: MEXICAN ACANTHACEAE 259 Ficure 3. Justicia breedlovei (Breedlove 56242). (a) node with leaf and inflorescence, x 0.6; (b) bract, bractlets, and calyx, x 3.2; (c) distal portion of stamen, x 11. Drawn by J. Speckels. chomes restricted to or concentrated in 2 lines, inflorescence bracts foliose, elliptic to obovate, 10-30 mm long, 1-13 mm wide; dichasia alter- nate, sessile, 1-flowered, | per axil. Bracts alter- nate, apically tinged with reddish purple, ovate- elliptic to narrowly elliptic to elliptic to obovate- elliptic, 8-19 mm long, 2—9.5 mm wide, apically rounded to truncate (to emarginate), abaxial sur- face glabrous, margin ciliate with + antrorse eglandular trichomes to 0.7 mm long. Bractlets colored like bracts, elliptic-oblanceolate, 5.5-11 mm long, 1-2 mm wide, abaxial surface gla- brous. Flowers sessile. Calyx 5-lobed, 8—9.5 mm long, abaxially glabrous, lobes homomorphic, linear, 7-8.5 mm long, 1—1.3 mm wide. Corolla orange, 32-34 mm long, externally pubescent with flexuose eglandular trichomes 0.2-0.5 mm long, tube gradually expanded distally, 17-19 mm long, 2.5-3.6 mm in diameter near midpoint, upper lip 15-16 mm long, 2-fid at apex, lobes 0.3 mm long, lower lip 14-15.5 mm long, lobes 0.8-2 mm long, 0.7-1.5 mm wide. Stamens inserted near apex of corolla tube, 15-16 mm long, fila- ments proximally pubescent with sparse eglan- dular trichomes, thecae 2—2.3 mm long, equal, subperpendicular to parallel, unequally inserted (i.e., overlapping by up to | mm) to superposed (i.e., contiguous), glabrous, both with a bulbous, rounded, basal appendage to 0.3 mm long (ap- pendage of lower theca larger than that of upper theca); pollen (Fig. 4a,b) 3-aperturate, apertures flanked on each side by 1 row of insulae, exine reticulate. Style 29-32 mm long, pubescent with eglandular trichomes; stigma lobes 0.1 mm long, equal. Capsule 9.5 mm long, glabrous, stipe 2.5- 3 mm long, head ellipsoid, 6.5-7 mm long. Seeds lenticular, 2.5 mm long, 2 mm wide, surface and margin covered with sparse glandular and eglan- dular trichomes less than 0.05 mm long. PHENOLOGY. — Flowering and fruiting: Decem- ber. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur on cliff faces in montane rain forests at an elevation of about 1170 m. This species differs from other Mexican spe- cies of Justicia by the combination of its alter- nate, sessile, and uniflorous dichasia; alternate, relatively large, and apically colored bracts; calyx with five homomorphic lobes; orange corolla with eglandular trichomes on the external surface; thecae with unequal basal appendages; triaper- turate pollen; and seeds with minute glandular and eglandular trichomes. It shares numerous 260 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 Lf fit j Ficure 4. Scanning electron micrographs of Justicia pollen. (a) J. breedlovei (Breedlove 56242), equatorial (colpal) view; (b) J. breedlovei, polar view; (c) J. chol (Cowan & Magara 3138), equatorial (intercolpal) view; (d) J. jitotolana (Thorne & Lathrop 41662), equatorial (intercolpal) view; (e) J. madrensis (Breedlove 38656, equatorial (colpal) view; (f) J. turipachensis (Breedlove 31242), equatorial (colpal) view. a—c at same scale; d—f at same scale. characteristics with Justicia section Plagiacan- thus (Nees) V.A.W. Graham but differs from that section by its larger bracts, triaperturate pollen, and pubescent seeds. The epithet honors Dennis Breedlove, collec- tor of this and many other interesting Chiapan Acanthaceae. Justicia chol T.F. Daniel, sp. nov. (Fig. 5a-c) Type.— MEXICO. Chiapas: Mpio. Palenque, near Cascada Mizola S of Palenque on road to Ocosingo, 300 m, 26 February 1981, D. Breedlove 49836 (holotype: CAS!; isotypes: C!, K!, MEXU)). Herba perennis usque ad 6 dm alta. Folia petiolata, laminae ovatae vel ovato-ellipticae, 28-110 mm longae, 8-35 mm la- tae, 2.1-3.5-plo longiores quam latiores. Inflorescentia floribus in spicas pedunculatas e folium axillis ortas; dichasia alterna, sessilia, uniflora. Bracteae heteromorphae; bracteae fertiles late obovatae vel spatulatae vel obdeltatae, 4.5-8 mm longae, 3.5— 6 mm latae, apice rotundatae vel truncatae et plerumque ap- iculatae, pagina abaxialis pubescens trichomatibus eglandu- losis et glandulosis; bracteae steriles bracteis fertilibus parvi- ores. Flores sessiles. Calyx 5-lobus, 2.5—3.5 mm longus, lobis homomorphis. Corolla alba vel cremea-lutea, 8.5-11.5 mm longa, extus pubescens trichomatibus eglandulosis. Stamina DANIEL: MEXICAN ACANTHACEAE 261 Ficure 5. Justicia chol (Breedlove 49836; a-c) and J. madrensis (Breedlove 38656; d-f). (a) habit, x 0.6; (b) inflorescence node with flower, x 5; distal portion of stamen, x 10; (d) vegetative node with leaf, x 0.75; (e) inflorescence with flower, x 2; (f) distal portion of stamen, x 6. Drawn by J. Speckels. filamentis pubescentibus trichomatibus eglandulosis, thecis 1- 1.5 mm longis, superpositis, inferiore calcare usque ad 0.7 mm longo instructa. Capsula 6-8 mm longa, pubescens trichomati- bus eglandulosis. Perennial herb to 6 dm tall; young stems quad- rate to quadrate-sulcate, pubescent with flexuose to antrorse eglandular trichomes 0.4—1 mm long concentrated in 2 lines, trichomes with maroon septae. Leaves petiolate; petioles to 35 mm long; blades ovate to ovate-elliptic, 28-110 mm long, 8-35 mm wide, 2.1-3.5 times longer than wide, acuminate at apex, attenuate at base, surfaces (especially midvein) and margin pubescent with cauline type trichomes, margin entire to crenate. 262 Inflorescence of axillary, pedunculate, dichasiate spikes to 65 mm long (including peduncles and excluding flowers), 8-11 mm in diameter near midspike, spikes 1-2 per axil, opposite or alter- nate at nodes, peduncles to 6 mm long, pubescent like young stems or with trichomes + evenly disposed, rachis usually + visible near midspike, pubescent like peduncles; dichasia 1-flowered, alternate, sessile, 1 per axil. Bracts opposite, het- eromorphic with fertile ones larger than sterile ones, fertile bracts subsessile or borne on petioles to 2 mm long, broadly obovate to spatulate to obdeltate, 4.5-8 mm long, 3.5-6 mm wide, rounded to truncate, often apiculate, at apex, ab- axial surface pubescent with cauline type tr- chomes along midvein and elsewhere pubescent with erect eglandular and glandular (sometimes inconspicuous) trichomes 0.1-0.3 mm _ long, margin ciliate with erect to flexuose glandular and eglandular trichomes to 2 mm long, sterile bracts linear (to sometimes obovate-spatulate near base of spike), 2-6 mm long, 0.2-3 mm wide, pubescent like fertile bracts. Bractlets lin- ear to linear-oblanceolate, 4.5—8 mm long, 0.4— 1 mm wide, pubescent like bracts. Flowers ses- sile. Calyx 5-lobed, 2.5—3.5 mm long, lobes ho- momorphic, lance-subulate, 2-3 mm long, 0.4— 0.7 mm wide, abaxial surface and margin pu- bescent with flexuose to antrorse eglandular tri- chomes to 0.5 mm long. Corolla white to cream- yellow with maroon markings on lower lip, 8.5- 11.5 mm long, externally pubescent with flex- uose eglandular trichomes to 0.5 mm long, tube + funnelform (i.e., + ampliate distally), 5.5-7 mm long, 1.4—2 mm in diameter near midpoint, upper lip 3-5 mm long, emarginate, lobes 0.1- 0.2 mm long, lower lip 3-5.5 mm long, lobes 0.8-1.9 mm long, 0.8-2 mm wide. Stamens in- serted near apex of corolla tube, 3.5-4.5 mm long, filaments very sparsely pubescent with flex- uose eglandular trichomes, thecae 1—1.5 mm long (including basal appendage), subequal to unequal in length, + parallel, superposed (up to 0.5 mm distant), pubescent with eglandular trichomes, lower theca with a clublike basal appendage to 0.7 mm long; 2 staminodelike protrusions of co- rolla present near midpoint of corolla tube, each with a cluster of flexuose eglandular trichomes; pollen (Fig. 4c) 3-aperturate, apertures flanked on each side by both a continuous band and a pseudocolpus, exine reticulate. Style 7-9 mm long, pubescent with eglandular trichomes; stig- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 ma subelliptic, 0.2 mm long, lobes not evident. Capsule 6-8 mm long, pubescent with flexuose to retrorse eglandular trichomes 0.1—0.4 mm long, stipe 2-3 mm long, head obovoid to subellipsoid to ovoid, 4-5 mm long. Seeds 4, lenticular 1.3- 1.6 mm long, 1—1.4 mm wide, surface and mar- gin roughened with low rounded papillae or ridg- es, lacking trichomes. PHENOLOGY.— Flowering: February—July; fruiting: February, May-June. DISTRIBUTION AND HABITAT. —Southern Mex- ico (Chiapas, Tabasco); plants occur along streams in lowland rain forests and lower montane rain forests at elevations from 50 to 300 m. PaRATYPES.— MEXICO. Chiapas: Mpio. Palenque, Agua Azul, D. Breedlove 35373 (DS); Mpio. Palenque, near Cascada Mizola south of Palenque on road to Ocosingo, D. Breedlove 49836 (CAS); Agua Azul between Palenque and Ocosingo, ca. 10 mi SW of Rio Tulija, 7. Daniel & B. Bartholomew 5011 (CAS); Mpio. Palenque, vicinity of Palenque archeological site, G. Davidse et al. 20340 (CAS); ca. 30 mi from Palenque toward Ocosingo, L. McDade 204 (DUKE). Tabasco: Mpio. Tacotal- pa, ca. 3 km E del Ejido Lazaro Cardenas, C. Cowan & Solano 2085 (CAS); Mpio. Tacotalpa, cerro arriba del Ejyido Zunt en el camino de la Est. Tacotalpa hacia Tapijulapa, C. Cowan & M. Magafia 3138 (CAS). With its axillary, conspicuously and hetero- morphically bracteate spikes, this species resem- bles both J. nevlingii Wassh. & T.F. Daniel from Mexico and J. costaricana Leonard from Costa Rica. It differs from the former by its narrower (vs. 1.2—2.2 mm wide) bractlets; longer (vs. 6.5- 7 mm long) and pubescent (vs. glabrous) style; more numerous (vs. 2), smaller (vs. 2.2-2.8 mm long), and roughened (vs. smooth) seeds; and ax- illary spikes (vs. mostly terminal panicles). It dif- fers from J. costaricana by the presence of longer (vs. absent or up to 0.2 mm long) trichomes of the bracteal margin, shorter (vs. 5-7 mm long) calyx, white to cream-yellow (vs. greenish yellow or greenish white) and externally eglandular (vs. glandular) corolla, superposed (vs. unequally in- serted and overlapping by up to | mm) thecae, flexuose to antrorse (vs. retrorse to flexuose) cau- line trichomes with maroon septae (vs. lacking maroon septae), and lowland rain forest (vs. cloud forest) habitat at elevations from 50 to 300 m (vs. 900 to 1550 m). Unfortunately, capsules and seeds are not known for J. costaricana. The epithet honors the Chol (Maya) people who inhabit the lowlands of Chiapas and Ta- basco. DANIEL: MEXICAN ACANTHACEAE 263 Ficure 6. Justicia jitotolana (a, b), J. tianguensis (Breedlove 7365; c-e), and J. turipachensis (Breedlove 31242; f, g). (a) dichaisum with dehisced capsule (Breedlove & Almeda 60297), x 3.3; (b) distal portion of stamen (Thorne & Lathrop 41 662), x 8: (c) inflorescence node with flower, x 1.5; (d) bract, bractlets, and calyx, x 2.5; (e) distal portion of stamen, x 5; (f) distal portion of shoot with inflorescence, x 0.75; (g) distal portion of stamen, x 12. Drawn by J. Speckels. September 1971, R. Thorne & E. Lathrop 41662 (holotype: Justicia jitotolana T.F. Daniel DS!; isotype: RSA!). (Fig. 6a, b) Type. — MEXICO. Chiapas: Mpio. Ray6én, 9 mi NW of Pueblo Herba perennis vel frutex usque ad | m altus. Folia petiolata, Nuevo Solistahuacan along rd. between Rincon Chamula and laminae ovato-ellipticae vel ellipticae vel obovato-ellipticae, Rayon, near Puerto del Viento, 17°30'N, 92°40'W, 1760 m, 31-115 mm longae, 14-62 mm latae, 1.5-2.8-plo longiores 264 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 quam latiores. Inflorescentia floribus in thyrsos spicoideos con- gestos pedunculatos terminales et axillares; dichasia in quoque spica 3-7, alterna vel subopposita, pedunculata, uniflora. Brac- teae obovato-spatulatae, (6—) 7.5-16 mm longae, 2-5.5 mm latae, apice truncatae (vel emarginatae), pagina abaxialis glabra vel pubescens trichomatibus eglandulosis. Flores sessiles vel subsessiles. Calyx 5-lobus, 6.5-10 mm longus, lobis homo- morphis vel heteromorphis (4+ 1). Corolla subrosea-purpurea, 23-27 mm longa, extus glabra. Stamina filamentis pubescen- tibus trichomatibus glandulosis, thecis 1.2—-2.5 mm longis, su- perpositis, inferiore calcare usque ad 1 mm longa instructa. Capsula 13-16 mm longa, glabra. Perennial herb or shrub to | m tall; young stems quadrate to quadrate-sulcate to quadrate- flattened, bifariously pubescent with retrorse, eglandular, conspicuously multi-septate (with maroon septae) trichomes to 0.5 mm long. Leaves petiolate; petioles to 34 mm long; blades ovate- elliptic to elliptic to obovate-elliptic, 31-115 mm long, 14-62 mm wide, 1.5—2.8 times longer than wide, acuminate at apex, attenuate at base, ad- axial surface sparsely pubescent with coarse, eglandular trichomes to 0.5 mm long, soon gla- brate, abaxial surface pubescent along major veins with antrorse, conspicuously multi-septate, eglandular trichomes to 0.5 mm long, punctate- pitted, margin entire to subsinuate. Inflorescence of congested, somewhat headlike, axillary (in ax- ils of distalmost pair of leaves) and terminal, pedunculate spikelike thyrses to 37 mm long (in- cluding peduncle and excluding flowers), (6-) 10- 20 mm in diameter near midspike, peduncles to 13 mm long, pubescent like young stems; rachis pubescent like young stems; dichasia 3-7 per spikelike thyrse, 1 per axil, subopposite to alter- nate (proximalmost pair sometimes opposite), pedunculate, 1-flowered, peduncles to 1.5 mm long. Bracts subopposite to alternate, obovate- spatulate (proximalmost sometimes petiolate), (6—-) 7.5-16 mm long, 2—5.5 mm wide, rounded to truncate (to emarginate) at apex, abaxial sur- face glabrous or sparsely pubescent with eglan- dular trichomes to 0.2 mm long along major veins, punctate-pitted. Bractlets spatulate, (5.5-) 7.5-13.5 mm long, 1-3 mm wide, apically rounded to truncate, abaxial surface glabrous or pubescent like bracts. Flowers sessile to subses- sile (i.e., pedicels to 1 mm long). Calyx 5-lobed, 6.5-10 mm long, lobes homomorphic to heter- omorphic (i.e., with posterior lobe + reduced in length), linear, 6-9.5 mm long, 0.8—1.1 mm wide, abaxially glabrous and punctate-pitted (some- times obscurely so). Corolla pinkish purple with white markings on lower lip, 23-27 mm long, externally glabrous, tube distally expanded, 14—- 18 mm long, |.6—2.5 mm in diameter near mid- point, upper lip 7-9.5 mm long, apically 2-lobed, lobes 1-1.5 mm long, lower lip 8-10 mm long, lobes rounded, 4-6 mm long, 4—5.7 mm wide. Stamens inserted near apex of corolla tube, 7-8 mm long, filaments pubescent with glands to 0.1 mm long, thecae 1.2—2.5 mm long (including bas- al spur), unequal in length (lower theca longer), subparallel to subperpendicular, unequally in- serted (overlapping by up to 0.3 mm) to super- posed (contiguous), glabrous, lower theca with a broad, rounded, basal appendage to 1 mm long; pollen (Fig. 4d) 4-aperturate, apertures flanked on each side by 4—5 rows of insulae, rows con- tinuous across mesocolpia, exine evident only near poles, reticulate. Style 16 mm long, gla- brous; stigma 0.3 mm long, lobes (if distinct) 0.2 mm long. Capsule 13-16 mm long, glabrous, stipe 5-6 mm long, head subovoid to ellipsoid (often with a slight medial constriction), 8-10 mm long. Seeds 4, lenticular, 3.2-3.4 mm long, 2.22.5 mm wide, surface and margin minutely rough- ened, sometimes covered with sparse glands to 0.05 mm long. PHENOLOGY. — Flowering: September—Octo- ber; fruiting: November—January. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur on steep slopes in montane rain forests and evergreen cloud forests at ele- vations from 1700 to 2030 m. PARATYPES. — MEXICO. Chiapas: Mpio. Rayon, in the Selva Negra 10 km above Rayon Mezcalapa along rd. to Jitotol, D. Breedlove & F. Almeda 60297 (CAS), D. Breedlove & R. Dress- ler 29824 (DS, MEXU), D. Breedlove & B. Keller 49312 (CAS); Mpio. Pueblo Solistahuacan, N of Clinica Yerba Buena near Pueblo Nuevo Solistahuacan, P. Raven & D. Breedlove 20032 (DS, US); Mpio Rayon, 9 mi NW of Pueblo Nuevo Solista- huacan along rd. between Rincon Chamula and Rayon, 17°30’N, 92°40’W, R. Thorne & E. Lathrop 46662 (RSA), H. Zuill 631 (DS); Mpio. Pueblo Nuevo Solistahuacan, 3 km NW of Pueblo Nuevo Solistahuacan, H. Zuill 415 (DS). This species resembles J. angustiflora D.N. Gibson from Oaxaca and J. silvicola D.N. Gib- son from Guatemala. It differs from J. angusti- flora by its more floriferous (vs. 2 to 3-flowered) inflorescences, glabrous or sparsely pubescent with eglandular trichomes (vs. pubescent with glandular trichomes) bracts and bractlets, shorter (vs. 48-49 mm long) and externally glabrous (vs. pubescent) corollas, and shorter (vs. 18-20 mm long) stamens with the filaments pubescent (vs. glabrous) with glandular trichomes. It differs from DANIEL: MEXICAN ACANTHACEAE J. silvicola by the pubescent (vs. glabrous or with a few remnant trichomes like those of J. jitoto- lana present just proximal to several nodes on the holotype) young stems, pubescent (vs. gla- brous) inflorescence peduncles, pubescent (vs. glabrous) abaxial leaf surfaces, pinkish purple (vs. white) and longer (vs. 18-19 mm long) corollas, glandular pubescent (vs. glabrous) filaments, and sparsely glandular (vs. pubescent with apically barbed eglandular trichomes 0.05—0.2 mm long) seeds. Pollen of these three species is unusual, although not unique, in Justicia by having four apertures. Pollen of J. jitotolana (Fig. 4d) appears nearly identical to that of J. silvicola (cf. Gibson 1972, Fig. 10b). Pollen of J. angustiflora (cf. Dan- iel 1993, Fig. 7a, b) differs by having spines rather than insulae and no continuous exine evident. The epithet is in reference to the Jitotol Ridge in a region of the northern highlands of Chiapas known as the Selva Negra from which all of the collections came. Justicia madrensis T.F. Daniel, sp. nov. (Fig. Sd-f) Type.— MEXICO. Chiapas: Mpio. Angel Albino Corzo (Jal- tenango), 3—5 km above Jaltenango toward Finca Prusia, 900 m, 11 October 1974, D. Breedlove 38656 (holotype: DS!; iso- types: C!, K!, MEXU!). Frutex usque ad 2 m altus. Folia petiolata, laminae ovatae, 23-75 mm longae, 9-39 mm latae, 1.9-2.6-plo longiores quam latiores. Inflorescentia floribus in spicas pedunculatas e fo- liorum axillis ortas; dichasia alterna, + secunda, sessilia, un- iflora. Bracteae subulatae, 1.8-2.5 mm longae, 0.9 mm latae, apice attenuatae, pagina abaxialis pubescens trichomatibus eglandulosis. Flores sessiles. Calyx 4-lobus, 5.5—9 mm longus, lobis homomorphis. Corolla subrosea-aurantiaca, 34-43 mm longa, extus pubescens trichomatibus eglandulosis. Stamina filamentis pubescentibus trichomatibus eglandulosis, thecis 2— 2.2 mm longis, impariter insertis vel superpositis, basi calcar- atis. Capsula 20 mm longa, pubescens trichomatibus glandu- losis et eglandulosis. Shrub to 2 m tall; young stems subquadrate to quadrate, evenly pubescent with flexuose to re- trorse eglandular trichomes (some internodes with glandular trichomes as well) 0.2-1.3 mm long. Leaves petiolate; petioles to 12 mm long; blades ovate, 23-75 mm long, 9-39 mm wide, 1.9-2.6 times longer than wide, acuminate at apex, cor- date at base, surfaces pubescent with flexuose to antrorse eglandular trichomes to 1 mm long, margin entire to subsinuate. Inflorescence of ax- illary, pedunculate, dichasiate spikes to 55 mm long (including peduncle and excluding flowers), 3-4 mm in diameter near midpoint of fertile 265 portion, spikes alternate or opposite in leaf axils, 1 per axil, peduncles to 26 mm long, evenly pu- bescent with flexuose to retrorse eglandular tri- chomes to 1.3 mm long, rachises pubescent like peduncles; dichasia alternate, + secund, sessile, 1-flowered, | per axil. Bracts opposite, subulate, 1.8-2.5 mm long, 0.9 mm wide, attenuate at apex, abaxial surface pubescent with flexuose to an- trorse eglandular trichomes to 0.7 mm long. Bractlets subulate, 1.8—2.2 mm long, 0.6-0.7 mm wide, abaxial surface pubescent like bracts (sometimes with a few flexuose glandular tri- chomes to 0.5 mm long as well). Flowers sessile. Calyx 4-lobed, 5.5-9 mm long, lobes homo- morphic, lanceolate, 5-7 mm long, 1.5-1.8 mm wide, abaxially pubescent with flexuose glan- dular and eglandular trichomes 0.2-0.9 mm long. Corolla pinkish orange, 34-43 mm long, exter- nally pubescent with flexuose eglandular tri- chomes 0.2-0.7 mm long, tube very gradually (if at all) expanded from near base, 18-22 mm long, 2.8-4 mm in diameter near midpoint, upper lip 17-21 mm long, entire, lower lip 15-22 mm long, lobes 2.5—4 mm long, 2.9-4.5 mm wide, apically entire. Stamens inserted near apex of corolla tube, 19-22 mm long, filaments proximally pubescent with eglandular trichomes, thecae 2-2.2 mm long, equal in length, subperpendicular to perpendic- ular, unequally inserted (overlapping by up to 1.5 mm) to superposed (up to 1.2 mm distant), glabrous or pubescent with eglandular trichomes, lacking basal appendages; pollen (Fig. 4e) 2-aperturate, apertures flanked on each side by 2 rows of insulae, outer rows of insulae often intergrading into peninsulae, exine reticulate. Style 33-40 mm long, proximally pubescent with eglandular trichomes; stigma lobes 0.1-0.2 mm long, unequal. Capsule 20 mm long, pubescent with flexuose to retrorse eglandular and glan- dular trichomes 0.1-0.6 mm long, stipe 9 mm long, subellipsoid with a medial constriction, 11 mm long. Seeds 4, not seen. PHENOLOGY.— Flowering and fruiting: Octo- ber. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur on slopes in pine-oak for- ests at an elevation of about 900 m. In general appearance (i.e., axillary inflores- cences with large pink and orange corollas) and pollen morphology, J. madrensis looks some- what like J. macrantha Benth. of Oaxaca, Gua- temala, Costa Rica, and Panama. Justicia ma- crantha differs by its glabrous young stems (except 266 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 FiGure 7. Justicia mirandae. (a) habit (Breedlove 50163), x 0.4; (b) leaf (Breedlove 50163), x 0.5; (c) bract (Breedlove 50163), x 2.25; (d) bractlet (Breedlove 50163), x 2.25; (e) flower (Neill 5570), x 2.2; (f) anthers (Neill 5570), x 7.5; (g) capsule (Laughlin 285), x 3. Drawn by E. del Valle. in J. macrantha var. piliformis D.N. Gibson), basally attenuate leaf blades, pedunculate and often opposite dichasia, five-lobed calyces, long- er (45-48 mm long) corollas with apically fringed lobes, and parallel thecae. Rhytiglossa latifolia Nees (the species has not yet been transferred to Justicia; not J. latifolia Vahl) from Tabasco also superficially resembles J. madrensis. It differs from the Chiapan species by its glabrous or bifariously pubescent vegeta- tive internodes, basally rounded to acute leaf blades, glandular pubescent rachises, longer ca- lyces, and longer, red corollas. The epithet is in reference to the Sierra Madre de Chiapas where the species occurs. Justicia mirandae T.F. Daniel, sp. nov. (Fig. 7) Type.— MEXICO. Chiapas: Mpio. Chiapa de Corzo, above El Chorreadero, 800 m, 18 March 1981, D. Breedlove 50163 (holotype: CAS!; isotypes: C!, K!, MEXU!). Frutex usque ad 3 m altus. Folia petiolata; laminae anguste DANIEL: MEXICAN ACANTHACEAE ellipticae vel ellipticae vel ovatae vel oblanceolato-ellipticae, 65-250 mm longae, 12-67 mm latae, 3.3-6.9-plo longiores quam latiores. Inflorescentia floribus in spicas pedunculatas terminales et axillares (vel in paniculam); dichasia (opposita vel) alterna, aliquando + secunda, sessilia, uniflora. Bracteae lanceolatae vel lanci-ovatae vel ellipticae vel obovato-ellipti- cae, 15-22 mm longae, (2—) 3.5—9.5 mm latae, apice acumi- natae, pagina abaxialis glabra vel pubescens trichomatibus eglandulosis. Flores sessiles. Calyx 5-lobus, 5-10 mm longus, lobis homomorphis. Corolla lutea punctis rubris notata, 30— 34 mm longa, extus pubescens trichomatibus eglandulosis et glandulosis. Stamina filamentis pubescentibus trichomatibus eglandulosis (et aliquando glandulosis), thecis 1.8—2.4 mm lon- gis, impariter insertis, basi calcaratis. Capsula 9-11 mm longa, glabra. Shrub to 3 m tall; young stems quadrate to quadrate-alate, internodes glabrous or some- times evenly to bifariously pubescent with erect to flexuose to retrorse to antrorse to appressed eglandular trichomes to 0.6 mm long, nodes sometimes pubescent with flexuose eglandular trichomes to 0.5 mm long. Leaves sessile; blades narrowly elliptic to elliptic to ovate to oblance- olate-elliptic, 65-250 mm long, 12-67 mm wide, 3.3-6.9 times longer than wide, acuminate to attenuate at apex, attenuate to node and often somewhat amplexicaule at base, surfaces gla- brous (or with eglandular trichomes on plants with pubescent stems), margin entire to shallow- ly crenate. Inflorescence of axillary and terminal pedunculate dichasiate spikes (or panicles of spikes) to 120 mm long (including peduncles and excluding flowers), 18-40 mm in diameter near midspike, sometimes forming a terminal panicle with spikes or panicles in axils of subfoliose in- florescence bracts, inflorescence bracts (if pres- ent) tinged with red, ovate, 28-50 mm long, 10- 15.5 mm wide, spikes (or panicles) 1-2 per axil of leaf or inflorescence bract, alternate or op- posite, peduncles to 80 mm long, glabrous or pubescent like young stems, rachis glabrous or pubescent with cauline type trichomes; dichasia (opposite to) alternate, sometimes + secund, 1-flowered, 1 per axil, sessile. Bracts (opposite to) alternate, tinged with red, often drying black- ish, lanceolate to lance-ovate to elliptic to obo- vate-elliptic, 15-22 mm long, (2-) 3.5-9.5 mm wide (the proximalmost pair often subfoliose and larger), apically acuminate, abaxial surface gla- brous (or with eglandular trichomes on plants with pubescent stems), margin usually ciliate with flexuose eglandular trichomes to 0.7 mm long. Bractlets tinged with red, linear to linear-lanceo- late to oblanceolate-elliptic, 7-19 mm long, 0.9- 2.1 (-3) mm wide, abaxial surface glabrous (or pubescent like bracts), margin with flexuose 267 eglandular trichomes to 0.7 mm long. Flowers sessile. Calyx 5-lobed, 5—10 mm long (accrescent in fruit and up to 13 mm long), lobes homo- morphic, lanceolate to linear-lanceolate, 4.5—9 (— 12) mm long, 1.2—2 mm wide, abaxially glabrous (or pubescent like bracts), margin eciliate or cil- iate like bractlets or glabrous. Corolla yellow speckled with red, subfusiform in bud, 30-34 mm long, externally pubescent with straight to flexuose glandular trichomes to 0.4 mm long and flexuose eglandular trichomes to 0.8 mm long, tube gradually expanded distally, 17-19 mm long, 3-3.5 mm in diameter near midpoint, upper lip 11-14 mm long, apically emarginate, lobes 0.1- 0.2 mm long, lower lip 10-16 mm long, lobes 1- 2.8 mm long, 0.8-1.5 mm wide. Stamens in- serted in distal '2 of corolla tube, 15—20 mm long, filaments distally glabrous, proximally pubescent with eglandular (and sometimes glandular as well) trichomes, thecae 1.8-2.4 mm long (including basal appendages), equal in length, parallel to subperpendicular, unequally inserted (1.e., over- lapping by up to 1.2 mm), upper theca pubescent with eglandular trichomes, both thecae with a bulbous, rounded, basal appendage 0.2-0.4 mm long (appendage of lower theca larger than that of upper theca); pollen (Fig. 8a, b) 3-aperturate, apertures flanked on each side by | row of in- sulae, insulae sometimes nearly fused into a band with only 1-2 distinct, exine reticulate. Style 30 mm long, distally glabrous, proximally pubes- cent with eglandular (and sometimes glandular trichomes as well); stigma 0.2-0.3 mm long, asymmetrically funnelform to unequally 2-lobed. Capsule 9-11 mm long, glabrous, stipe 2-3 mm long, head subovoid to ellipsoid, 7-8.5 mm long. Seeds 4, lenticular, 2.5-3.5 mm long, 2.5-2.8 mm wide, surfaces minutely roughened, lacking trichomes, entire to + crenate. PHENOLOGY. — Flowering: December—April; fruiting: December-—July. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur along streams in tropical deciduous and subdeciduous forests at elevations from 600 to 900 m. PARATYPES. — MEXICO. Chiapas: Mpio. Ocozocoautla de Espinosa, 13-15 km S of Ocozocoautla along rd. to Villa Flores, D. Breedlove 24580 (DS); Mpio. La Trinitaria, along Hwy. 190, 13 mi S of La Trinitaria, D. Breedlove & P. Raven 8446 (DS, F, MICH, US); Mpio. Ocozocoautla de Espinosa, head of Rio de la Venta at Chorreadero near Derna, D. Breed- love & R. Thorne 30309 (DS, MICH); El Aguacero, 13 km NO de Ocozocoautla, E. Cabrera & H. de Cabrera 7895 (CAS, MEXU); along Hwy. 190, ca. 20 mi SE of Comitan, R. King 3036 (MICH); barranca between S. Fernando and Plan de 268 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 Ayala, I. Langman 3918 (US); Mpio. Venustiano Carranza, Rancho Nacimiento along rd. between Chiapilla and San Lu- cas, R. Laughlin 285 (DS, US); Mpio. Ocozocoautla, Cascada El Aguacero, Rio La Venta, 16°45’N, 93°31’°W, E. Martinez S. & A. Reyes 22017 (MEXU); arriba Chacona (NO Tuxtla G.), F. Miranda 5262 (MEXU), 7844 (MEXU); Mpio. Oco- zocoautla, canyon of Rio de la Venta at Cascada El Aguacero, 16°46°N, 93°33’W, D. Neill 5570 (CAS); Mpio. San Fernando, Cafiada Muniz, G. Rodriguez-Guillén 103 (CAS); road to San Fernando from Tuxtla Gutiérrez, B. Schubert & A. Gémez- Pompa 1764 (US). This species superficially resembles certain species of Lophostachys Pohl by its colored bracts and more or less secund dichasia. Several spec- imens of J. mirandae have been misidentified with the name Beloperone comosa Nees. This name applies to J. fulvicoma Schltdl. & Cham., a related species from northeastern Mexico (Daniel 1989). Both species have bracts colored with red or maroon, five homomorphic calyx lobes, yellow to orange corollas with reddish spots within, and 3-aperturate pollen. They may be distinguished by the following couplet: Leaves sessile, blades 3.3-6.9 times longer than wide; corolla externally pubescent with glandular and eglandular trichomes; capsule Mla DTOUS): 22 osm eee J. mirandae. Leaves petiolate, blades 1.6-3 times longer than wide; corolla externally pubescent with eglandular trichomes only; capsule pubescent with eglandular trichomes |... J. fulvicoma. Two sprigs of Cabrera & H. de Cabrera 7895 (CAS) differ from other specimens of J. mirandae and from another sprig on that sheet by having eglandular hairs on the young stems, leaves, and abaxial surface of bracts. They appear to repre- sent a pubescent form of the species. The epithet honors Faustino Miranda, student of the vegetation and flora of Chiapas who col- lected a paratype of this species. Justicia teletheca T.F. Daniel, sp. nov. (Fig. 9) Type.—MEXICO. Chiapas: Mpio. Arriaga, La Mina Mi- crowave Station, 915 m, 21 December 1981, D. Breedlove 56314 (holotype: CAS!; isotypes: C!, K!, MEXU!, MO!, US!). Herba perennis usque ad 1.1 m alta. Folia petiolata, laminae ovatae vel ellipticae, 19-185 mm longae, 6-77 mm latae, 2- 3.7-plo longiores quam latiores. Inflorescentia floribus in spicas pedunculatas terminales et axillares vel in paniculam; dichasia opposita, sessilia, uniflora. Bracteae lineares vel oblanceolatae vel obovatae, 6-14 mm longae, 1.2-7 mm latae, apice acutae, bracteae distales pagina abaxiali pubescenti trichomatibus glandulosis et eglandulosis. Flores sessiles. Calyx 4-lobus, 9- 11 mm longus, lobis homomorphis. Corolla atrorosea-rubra, 30-35 mm longa, extus pubescens trichomatibus glandulosis et eglandulosis. Stamina filamentis pubescentibus trichoma- tibus eglandulosis, thecis heteromorphis, superpositis, theca supera fertilis, 2—2.5 mm longa, basi ecalcarata, theca inferna sterilis, 1-1.2 mm longa, basi calcarata. Capsula (immatura) 8 mm longa, pubescens trichomatibus glandulosis. Erect perennial herb to 1.1 m tall; young stems subquadrate, internodes glabrous or bifariously pubescent (sometimes only for a short distance proximal to only certain nodes) with flexuose to antrorsely appressed eglandular trichomes to 0.7 mm long, nodes usually with at least a few an- trorse eglandular trichomes. Leaves subsessile to petiolate; petioles to 37 mm long; blades ovate to elliptic, 19-185 mm long, 6-77 mm wide, 2- 3.7 times longer than wide, acute to acuminate at apex, acute to attenuate at base, surfaces pu- bescent (mostly along major veins) with antrorse eglandular trichomes, margin entire to subcren- ate. Inflorescence of axillary (in leaf axils) and terminal, pedunculate dichasiate spikes or pan- icles of spikes to 110 mm long (including pe- duncle and excluding flowers), 11-18 mm in di- ameter near midpoint of fertile portion, spikes or panicles mostly opposite, 1-3 per axil, pe- duncles to 75 mm long, nearly glabrous or + evenly to + bifariously pubescent with antrorse to antrorsely appressed eglandular trichomes 0. 1-— 0.9 mm long (strigose), rachises strigose proxi- mally and pubescent with erect glandular and eglandular trichomes 0.05-0.1 mm long (glan- dular puberulent) distally, inflorescence bracts (if present) similar to leaves but sometimes smaller; dichasia opposite, sessile, 1-flowered, 1 per axil. Bracts opposite, linear to oblanceolate to obo- vate, 6-14 mm long, 1.2-7 mm wide, acute at apex, abaxial surface of proximal bracts glabrous or with antrorse eglandular trichomes 0.1-0.5 mm long, abaxial surface of distal bracts glan- dular puberulent and often with a few longer, antrorse eglandular trichomes as well, margin cil- iate with flexuose to antrorse eglandular tn- chomes and (on distal bracts) glandular puber- ulent. Bractlets subulate to linear, 4-6 mm long, 0.3-0.5 mm wide, pubescent like bracts. Flowers sessile. Calyx 4-lobed, 9-11 mm long, lobes ho- momorphic, linear to linear-lanceolate, 8—10.3 mm long, 0.9-1.2 mm wide, abaxially pubescent like bracts or sometimes lobes of proximal ca- lyces glandular puberulent like lobes of distal calyces. Corolla dark pink-red with white mark- ings on lower lip, 30-35 mm long, externally DANIEL: MEXICAN ACANTHACEAE 5.4um 269 Ficure 8. Scanning electron micrographs of Justicia pollen. (a) J. mirandae (Breedlove 50163), equatorial (intercolpal) view; (b) J. mirandae, equatorial (colpal) view; (c) J. teletheca (Breedlove 5631 4), equatorial (intercolpal) view; (d) J. teletheca, equatorial (colpal) view; (e) J. tianguensis (Breedlove 7365), equatorial (intercolpal) view; (f) J. tianguensis, equatorial (colpal) view. b-f at same scale. pubescent with erect to subflexuose glandular and eglandular trichomes 0.05-0.3 mm long, tube + funnelform, 17-19 mm long, 1.9—2.5 mm in di- ameter near midpoint, upper lip 1 1-16 mm long, entire to emarginate, lobes 0.2 mm long, lower lip 14-18.5 mm long, lobes 3-5 mm long, 1.5— 3 mm wide. Stamens inserted near apex of co- rolla tube, 13.5-16 mm long, filaments distally glabrous, proximally pubescent with eglandular trichomes, thecae subparallel to subperpendi- cular, superposed (1.5—3 mm distant), dimor- phic, distal theca fertile, 2-2.5 mm long, pubes- cent with eglandular trichomes, proximal theca borne on a projection of the connective to 0.5 mm long, sterile, 1-1.2 mm long (including a basal spur to 0.4 mm long); pollen (Fig. 8c, d) 2-aperturate, apertures flanked on each side by 3-4 rows of insulae, exine reticulate. Style 25- 30 mm long, sparsely pubescent at base (if at all) with eglandular trichomes; stigma 0.2 mm long, 270 lobes not evident. Capsule (immature) 8 mm long, glandular puberulent. PHENOLOGY. — Flowering: November—Decem- ber; fruiting: December. DISTRIBUTION AND HABITAT. — Southern Mex- ico (Chiapas, Oaxaca); plants occur on slopes in the ecotone between tropical deciduous forests and oak forests and in pine-oak forests at ele- vations from 900 to 915 m. PARATYPES.— MEXICO. Oaxaca: Mpio. San Miguel Chi- malapa, El Pedregal del Rio Portamonedas, ca. 3 km S de Benito Juarez, 16°42’N, 94°08’W, S. Maya J. 864 (CAS). This species is referable to Justicia sect. Chae- tothylax (Nees) V.A.W. Graham and conforms to plants previously treated in the genus Chae- tothylax Nees (see discussion below under J. rze- dowskii). Justicia teletheca is similar to C. phyl- lostachyus Nees from Tabasco in most respects but differs from the holotype (i.e., Linden 188, K!) of that species in characters of the inflores- cence and calyx, and in habitat preference. In C. phyllostachyus the dichasia are solitary in the leaf axils (or if this arrangement is construed to rep- resent a dichasiate spike, then the bracts are subfoliose (16-22 mm long, 5.5-10 mm wide) and the rachises lack glandular trichomes); and the calyx lobes lack glandular trichomes. Chae- tothylax phyllostachyus occurs at lower eleva- tions on the Caribbean escarpment in a region of rain forest. Justicia teletheca differs from the description of C. cuspidatus D.N. Gibson (Gib- son 1974) by its axillary inflorescences (vs. ter- minal and subcapitate), larger anthers, and lack of a cusp (1 mm long or more) on the bracts. There is a collection from Chiapas at BM that greatly resembles J. teletheca (Mpio. Frontera Comalapa, 12 km W de Frontera Comalapa, carr. 211 hacia Motozintla, 810 m, 6 Feb 1990, A. Reyes G. et al. 1577). It does not have any opened corollas and therefore I cannot be certain that it represents the species. The epithet is derived from the Greek ele- ments tele (far) and theca (case) in reference to the remote anther thecae. Justicia tianguensis T.F. Daniel, sp. nov. (Fig. 6c-e) Type.—MEXICO. Chiapas: Mpio. Tenejapa, Yochib, Pa- raje of Kotol Te’, 1300 m, 21 November 1964, D. Breedlove 7365 (holotype: DS!; isotypes: F!, MICH!, US!). Herba perennis. Folia petiolata, laminae ovatae, 58-130 mm longae, 18-57 mm latae, 2-3.2-plo longiores quam latiores. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 Inflorescentia floribus in spicas pedunculatas terminales; di- chasia opposita, sessilia, uniflora. Bracteae lanceolatae vel lan- ci-ovatae, 4.5—-6 mm longae, 1.6-2.5 mm latae, apice acumi- natae vel attenuatae, pagina abaxialis pubescens trichomatibus eglandulosis. Flores sessiles. Calyx 5-lobus, 8-10 mm longus, lobis heteromorphis (4+ 1). Corolla rubra, 39-51 mm longa, extus pubescens trichomatibus eglandulosis. Stamina filamen- tis pubescentibus trichomatibus eglandulosis, thecis 2—2.5 mm longis, impariter insertis vel superpositis, basi ecalcaratis. Ovarium pubescens trichomatibus eglandulosis. Capsula ig- nota. Perennial to 3 (probably considerably more) dm tall; young stems quadrate-compressed, in- ternodes unifariously to bifariously pubescent with retrorse eglandular trichomes 0.2-1 mm long, soon glabrate, trichomes sometimes sparse or absent along proximal portions of internodes, nodes pubescent with straight (not erect) eglan- dular trichomes to | mm long. Leaves petiolate; petioles to 10 mm long; blades ovate, 58-130 mm long, 18-57 mm wide, 2-3.2 times longer than wide, acuminate to subfalcate at apex, rounded to acute at base, adaxial surface pubes- cent with antrorse eglandular trichomes along midvein, abaxial surface glabrous or with a few antrorse eglandular trichomes along midvein, margin entire to subcrenate. Inflorescence of ter- minal (sometimes appearing axillary), pedun- culate, dichasiate spikes to 200 mm long (in- cluding peduncle and excluding flowers), 6-8 mm in diameter near midspike (excluding flowers), peduncles to 55 mm long, + evenly pubescent with erect to retrorse to flexuose to antrorse eglandular trichomes 0.1-0.5 mm long, rachises evenly pubescent with erect eglandular tri- chomes 0.1—0.2 mm long; dichasia 1-flowered, opposite, | per axil, sessile. Bracts lanceolate to lance-ovate, 4.5—6 mm long, 1.6-2.5 mm wide, acuminate to attenuate at apex, abaxial surface pubescent with erect eglandular trichomes 0.05- 0.1 mm long. Bractlets lanceolate to lance-su- bulate, 4-5 mm long, 0.7—1.1 mm wide, abaxial surface pubescent like bracts. Flowers sessile. Ca- lyx 5-lobed, 8-10 mm long, pubescent like bracts, lobes heteromorphic, 4 lobes homomorphic, lan- ceolate, 6.5—9 mm long, 1—1.5 mm wide, widest at base, posterior lobe greatly reduced, 1.8—2.5 mm long, margins neither thickened nor discol- ored. Corolla red, 39-51 mm long, externally pubescent with erect eglandular trichomes 0.05- 0.2 mm long, tube gradually expanded distally, 22-27 mm long, 4.3—5.2 mm in diameter near midpoint, upper lip 17-22 mm long, entire, low- er lip 17-24 mm long, lobes 2.5-5.5 mm long, DANIEL: MEXICAN ACANTHACEAE 1.8-5 mm wide. Stamens inserted near apex of corolla tube, 17-20 mm long, filaments proxi- mally pubescent with eglandular trichomes, the- cae 2—-2.5 mm long, subequal, perpendicular, un- equally inserted (overlapping by 0.5—1 mm) to superposed (up to 0.4 mm distant), glabrous, lacking basal appendages; pollen (Fig. 8e,f) 2-aperturate, apertures flanked on each side by 2 rows of insulae, exine reticulate. Style 37-44 mm long, proximally pubescent with eglandular trichomes; stigma unequally 2-lobed, | lobe 0.3- 0.4 mm long, other lobe 0.1 mm long. Ovary densely pubescent with erect eglandular tri- chomes 0.1—0.2 mm long. Capsule not seen. PHENOLOGY. — Flowering: November, Febru- ary. DISTRIBUTION AND HABITAT. —Southern Mex- ico (Chiapas) and Guatemala; plants occur on moist slopes in pine-oak-Liguidambar forests at elevations from 1300 to 1500 m. PARATYPE.—GUATEMALA. Alta Verapaz: along Rio Co- ban, ca. 5 km SE of Tactic, ca. 15°19’N, 90°15°W, L. Williams et al. 40603 (F). Gibson (1974) included the above cited spec- imens in J. inaequalis Benth. (based on her an- notations and inclusion of Chiapas within the range of that species). They differ from J. inae- qualis by the characteristics in the following cou- plet: Calyx 5-lobed (with posterior lobe greatly re- duced), 8-10 mm long, lobes widest at base, margins neither thickened nor discolored; in- florescences not secund, dichasia opposite (= paired) at nodes; peduncle, rachis, and ab- axial surfaces of bracts, bractlets, and calyx pubescent; style and ovary pubescent a eee Re eae J. tianguensis. Calyx 4-lobed, 11-18 mm long, lobes widest near middle, margins thickened and discol- ored; inflorescence secund, dichasia alternate (= solitary) at nodes; peduncle, rachis, and abaxial surfaces of bracts, bractlets, and calyx glabrous; style and ovary glabrous J. inaequalis. The epithet is derived from the Spanish word tiangue, which means small market. Yochib is a sacred place that serves as a common market- place for Tzeltal peoples living in three munic- ipalities. Justicia turipachensis T.F. Daniel, sp. nov. (Fig. 6f, g) Type.— MEXICO. Chiapas: Mpio. Berriozabal, 13 km N of Berriozabal near Pozo Turipache and Finca El Suspiro, 900 m, | Jan 1973, D. Breedlove 31242 (holotype: CAS!; isotypes: C!, DS!, K!, MEXU!, -US!). Frutex usque ad 4.5 m altus. Folia petiolata, laminae ellip- ticae, 121-260 mm longae, 31-90 mm latae, 2.9-3.9-plo long- iores quam latiores. Inflorescentia floribus in paniculam ter- minalem complexum; dichasia alterna vel opposita, sessilia vel subsessilia, uniflora. Bracteae subulatae vel lance-subula- tae, 1.5-3 mm longae, 0.4-0.7 mm latae, apice acutae vel subattenuatae, pagina abaxialis pubescens trichomatibus eglandulosis (et aliquando inconspicuo glandulosis). Flores pedicellati. Calyx 5-lobus, 4-6 mm longus, lobis homomor- phis. Corolla lutea, 31-35 mm longa, extus pubescens tri- chomatibus glandulosis et eglandulosis. Stamina filamentis pu- bescentibus trichomatibus eglandulosis, thecis 2.3-2.6 mm longis, subpariter vel impariter insertis, basi calcaratis. Capsula ignota. Shrub to 4.5 m tall; young stems subterete, glabrous. Leaves petiolate; petioles to 65 mm long; blades elliptic, 121-260 mm long, 31-90 mm wide, 2.9-3.9 times longer than wide, acu- minate at apex, attenuate at base, surfaces gla- brous (or with a few antrorsely appressed eglan- dular trichomes along midvein on abaxial surface), margin entire to subcrenulate. Inflores- cence of axillary and terminal, pedunculate, di- chasiate racemes or panicles of racemes from axils of distal leaves or bracts (= inflorescence bracts) forming a complex terminal panicle to 200 mm long (excluding flowers) and 150 mm in diameter near midpoint, commonly both a raceme and a panicle per axil, inflorescence bracts subfoliose, petiolate, 5.5—8 mm long, 1.3-—2.1 mm wide, main rachis + evenly pubescent with flex- uose-antrorse to antrorsely appressed eglandular trichomes 0.1—0.3 mm long, peduncles to 15 mm long, pubescent like main rachis, raceme rachises pubescent like main rachis (or with trichomes denser); dichasia 1-flowered, alternate or oppo- site, 1 per axil, sessile to subsessile (i.e., pedun- cles to 0.5 mm long). Bracts opposite, subulate to lance-subulate, 1.5-3 mm long, 0.4-0.7 mm wide, acute to subattenuate at apex, abaxial sur- face pubescent with antrorsely appressed eglan- dular trichomes 0.05—0.2 mm long and some- times with a few inconspicuous erect glandular trichomes to 0.1 mm long. Bractlets subulate, 1.2—-1.5 mm long, 0.3-0.4 mm wide, abaxial sur- face pubescent like bracts. Flowers pedicellate, pedicels 1.5-2.5 mm long, pubescent like ra- chises. Calyx 5-lobed, 4-6 mm long, lobes ho- 272 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 FiGureE 9. Justicia teletheca (Breedlove 56314). (a) vegetative node with inflorescences, x 0.75; (b) inflorescence node with flower, x 2.5; (c) distal portion of stamen, x 6. Drawn by J. Speckels. momorphic, lance-subulate, 3.2-5 mm long, 0.7- 0.9 mm wide, abaxially and marginally pubes- cent like bracts. Corolla yellow, 31-35 mm long, externally pubescent with flexuose eglandular tri- chomes 0.1-0.3 mm long and distally with erect glandular trichomes to 0.2 mm long as well, tube 17-18 mm long, + gradually expanded distally, upper lip 14-15 mm long, 2-fid at apex, lobes to 0.3 mm long, lower lip 14.5—16 mm long, lobes 1.5 mm long, 0.5-0.7 mm wide. Stamens in- serted near apex of corolla tube, 15-16.5 mm long, filaments proximally pubescent with eglan- dular trichomes, thecae 2.3-—2.6 mm long, equal in length, subparallel to subsagittate, subequally to unequally inserted (overlapping by 1.5—2 mm), glabrous, each theca with a bulbous, rounded, basal appendage to 0.3 mm long (appendage of lower theca longer than that of upper theca); pol- len (Fig. 4f) 2-aperturate, apertures flanked on each side by 2 (—3) rows of insulae, exine retic- ulate. Style 28.5-31 mm long, glabrous; stigma lobes 0.2-0.3 mm long. Capsule not seen. PHENOLOGY. — Flowering: January. DISTRIBUTION AND HABITAT.—Endemic to Chiapas where plants occur on limestone ridges in lower montane rain forests at elevations of about 900 m. Justicia turipachensis resembles the wide- spread J. aurea Schltdl. in numerous features including its expansive terminal panicle and flowers with large, yellow corollas. These species can be distinguished by the following couplet: Corolla 48-70 mm long, buds prominently curved near apex; filaments glabrous, thecae 2.5-3.6 mm long, usually dorsally pubescent, lacking basal appendages; pollen lacking dif- ferentiated trema areas, surfaces smooth .. eA AACR TONE J. aurea. Corolla 31-35 mm long, buds not promi- nently curved near apex; filaments proxi- mally pubescent with eglandular trichomes, thecae 2.3-2.6 mm long, glabrous, each theca with a basal appendage; pollen with differ- entiated trema areas, surfaces reticulate A Mee DN ok ioe sh. J. turipachensis. The epithet is derived from the type locality ‘Pozo Turipache,” near which at least 16 species of Acanthaceae have been collected. Justicia kanal T.F. Daniel, nom. nov. Beloperone aurea Leonard, Publ. Carnegie Inst. Wash. 461: 233. 1936, non Justicia aurea (Rose) Lindau (1897) nec Justicia aurea Schlecht. (1832). Justicia flava D.N. Gibson, Fieldiana, Bot. 34:69. 1972, non Justicia flava (Vahl) Vahl DANIEL: MEXICAN ACANTHACEAE (1791) nec Justicia flava Kurz (1873). Type. —GUATE- MALA. Petén: Fallab6n-Yaxha Road, 22 March 1933, C. Lundell 2189 (holotype: US!; isotype: MICH!). Gibson (1972) provided the new name J. flava for the species previously known as Beloperone aurea. Unfortunately, her new name represented a later homonym and is therefore illegitimate according to Article 64 of the International Code of Botanical Nomenclature (Greuter et al. 1988). Accordingly, another name is proposed for this species in Justicia. | have chosen one that, like the previous names for this species, highlights a conspicuous feature (i.e., the bright-yellow tri- chomes evident on the young growth). The epi- thet is derived from k’anal, which means yellow in the Tzeltal (Maya) language. Justicia borrerae (Hemsl.) T.F. Daniel, comb. nov. Neohallia borrerae Hemsl. Biol. Centr. Amer. Bot. 2:519. 1882. Type.— MEXICO. Chiapas: Rancho de la Montana, 6 leagues from Tuxtla, November 1864-70, 4. Ghiesbreght 722 (ho- lotype: K; isotype: GH!). Neohallia Hemsl. was described as a probable relative of Justicia with large, leathery or fleshy, cup-shaped involucres enclosing several large, sessile flowers. Based on floral features (i.e., ru- gulate corolla, unequally inserted thecae with basal appendages, and 2-aperturate pollen with the apertures flanked by rows of insulae), the obvious affinities of this monotypic genus are with Justicia in the broad sense in which that genus is presently treated (Graham 1988). Two features of N. borrerae readily distinguish it from species of Justicia in North and Central America: the partially connate bracts that form the floral involucre and the exceptionally large and woody capsules. Given the diversity of bracteal size and fusion and capsule size within both Justicia and other genera of Acanthaceae, these features are not considered to be sufficient for recognition of a distinct genus. Pollen of J. borrerae corre- sponds to Graham’s “Type 5,” which is found in several sections of the genus and in several species of uncertain affinities (Graham 1988). Justicia rzedowskii (Acosta) T.F. Daniel, comb nov. Chaetothylax rzedowskii Acosta, Acta Bot. Méx. 5:5. 1989. Type.— MEXICO. Chiapas: Mpio. Huehuetan, Rio Cha- mulapa, 50 m, E. Ventura & E. Lépez 1074 (holotype: ENCB; isotypes: CAS!, CHAPA, IEB, MEXU, OAX). 273 Chaetothylax Nees is treated as a distinct ge- nus by some (e.g., Gibson 1974, Durkee 1986) and is included in Justicia by others (e.g., Gra- ham 1988). The genus supposedly differs from Justicia by its dense inflorescence, four-lobed ca- lyx, and stamens with the thecae conspicuously unequal (the smaller one sometimes sterile or vestigial). All of these features are known in spe- cies of Justicia. In species treated in Chaetothy- lax the corollas have a rugula, the thecae often have a basal appendage, and the pollen is typical of that found in Justicia. Thus, there appear to be no mutually exclusive distinctions between these occasionally recognized genera. LEPIDAGATHIS VS. TELIOSTACHYA Lepidagathis alopecuroidea (Vahl) R. Br. ex Griseb., a species known from Chiapas, and about 10 other strictly American species are sometimes treated as Teliostachya Nees. Bremekamp (1938) distinguished Teliostachya from Lepidagathis Willd. by its radially symmetric and terminal spikes, unfused anterior calyx lobes, unequally inserted anther thecae, and non-reticulate pollen. With the exception of its cylindric inflorescences, none of these character states that are diagnostic of Teliostachya apply to the widespread L. alo- pecuroidea. Subsequently, Bremekamp (1960) included Teliostachya in his tribe Lepidagathi- deae that he characterized as having partially connate anterior calyx lobes. Our species, usually treated in American literature as 7. alopecuro- idea, has all of the diagnostic characteristics of Lepidagathis (except for its radially symmetric inflorescences) as indicated by Bremekamp (1938). It remains to be seen whether the other American species, all of which are South Amer- ican, treated by Bremekamp (1938) in Telio- stachya, are worthy of being maintained as dis- tinct from Lepidagathis. RUELLIA Ruellia is the second largest genus of Acan- thaceae. It is represented in Mexico by about 65 species. Twenty of these occur in Chiapas. The species previously treated as R. longituba D.N. Gibson (Gibson 1974) does not conform to the type of that name but represents a new taxon that is described below. Ruellia maya T.F. Daniel, sp. nov. (Fig. 10) 274 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 Ficure 10. Ruellia maya (Daniel & Bartholomew 5000). (a) habit, x 0.35; (b) leaf, x 1.25; (c) flower, x 0.5, with enlargement showing pubescence of external surface of corolla; (d) calyx following dehiscence of capsule, x 1.1, with enlargement showing pubescence of abaxial surface of lobes; (e) androecium, x 2.7; (f) stigma, x 7.5; (g) capsule, x 2.3, with enlargement showing pubescence of external surface; (h) seed, x 7. Drawn by J. Speckels. Type.— MEXICO. Chiapas: Agua Azul between Palenque and Ocosingo, ca. 10 mi SW of Rio Tulia, ca. 300 m, 15 March 1987, 7. Daniel & B. Bartholomew 5000 (holotype: CAS!; isotypes: MEXU!, MICH!) Herba perennis vel frutex usque ad | m altus. Folia petiolata, laminae lanceolatae vel lanceolato-ellipticae, 33-150 mm lon- gae, 6—SO mm latae, (2.8—) 3.9-7.1-plo longiores quam latiores. Dichasia in axillis foliorum distalium, sessilia, uniflora. Brac- teolae lanceolato-ovatae vel anguste ellipticae, 17-46 mm lon- gae, 4-9 mm latae, glandulosae. Calyx 14-27 mm longus, extus glandulosus. Corolla caerulea-purpurea, 65-90 mm longa, ex- tus pubescens trichomatibus glandulosis et eglandulosis. Sta- mina inclusa thecis 5—5.5 mm longis. Capsula substipitata, 1 3- 18 mm longa, pubescens trichomatibus glandulosis et eglan- dulosis. Semina usque ad 16, 3.5-4 mm longa, 2.5—3 mm lata, pagina laevis vel substriata, margo pubescens trichomatibus hygroscopicis. Erect to + diffuse perennial herb to shrub to DANIEL: MEXICAN ACANTHACEAE 1 m tall. Young stems quadrate to quadrate-sul- cate, pubescent with flexuose eglandular and glandular (sometimes absent) trichomes 0.5-1.3 (-2) mm long, trichomes often concentrated in 2 lines. Leaves petiolate; petioles to 34 mm long; blades lanceolate to lance-elliptic, 33-150 mm long, 6-50 mm wide, (2.8—) 3.9—7.1 times longer than wide, acute to acuminate at apex, attenuate at base, surfaces pubescent with scattered glan- dular (sometimes absent) and eglandular tn- chomes 0.2-1.5 mm long or becoming glabrate, margin entire to subsinuate. Inflorescence of ses- sile dichasia in axils of distal leaves; dichasia alternate or opposite at nodes, | per axil, 1-flowered. Flowers sessile to subsessile (i.e., with pedicels to | mm long). Bractlets petiolate, lance- ovate to narrowly-elliptic, 17-46 mm long, 4-9 mm wide, pubescent like young leaves (i.e., glan- dular). Calyx 14-27 mm long, tube 2-3 mm long, lobes lanceolate to elliptic to oblanceolate, 12- 24 mm long, 5.4—8 times longer than tube, 1.5- 4 mm wide, abaxial surface pubescent like brac- tlets, margin ciliate with erect to flexuose to an- trorse glandular and eglandular trichomes 0.3- 1.6 mm long. Corolla blue-purple, 65-90 mm long, externally pubescent with flexuose glan- dular and eglandular trichomes 0.2—1 mm long, tube 56-70 mm long, narrow-proximal portion 32-50 mm long, arched or curved near apex, abruptly expanded into throat, throat 16-25 mm long, shorter than narrow-proximal portion of tube, 9-15 mm in diameter near midpoint, limb 32-60 mm in diameter, lobes 15-25 mm long, 14-25 mm wide. Stamens included, longer pair 13-15 mm long, shorter pair 11-12 mm long, thecae presented at 2 heights, 5-5.5 mm long, connective often with an apical elongation. Style 50-55 mm long, pubescent with eglandular tri- chomes + throughout, stigma unequally 2-lobed, 1 lobe 1.7—2.8 mm long, other lobe 0.2-0.5 mm long or not evident. Capsule substipitate, 13-18 mm long, pubescent with scattered erect to flex- uose eglandular and glandular (rarely becoming + entirely eglandular with age) trichomes 0.1- 0.3 mm long, stipe 2—2.5 mm long, head ellipsoid to ellipsoid-obovoid. Seeds up to 16, 3.5—-4 mm long, 2.5—-3 mm wide, surface smooth to sub- striate, margin with a prominent band of hygro- scopic trichomes. PHENOLOGY. — Flowering: September, Decem- ber—March; fruiting: December—March. DISTRIBUTION AND HABITAT.— Southern Mex- ico (Chiapas) and Guatemala; plants occur along 275 streams in lowland rain forests, lower montane rain forests, montane rain forests, and seasonal evergreen forests at elevations from 280 to 1700 m. PARATYPES.— MEXICO. Chiapas: Mpio. Palenque, near Agua Azul, D. Breedlove 47419 (CAS), D. Breedlove & F. Almeda 57264 (CAS, MEXU), D. Breedlove & B. Keller 49578 (CAS, MEXU); Mpio. Ocosingo, 5 km NE of Ocosingo toward Pal- enque, D. Breedlove 49094 (CAS); Mpio. La Libertad, 10-20 km toward Chancala on road to Bonampak from Palenque- Ocosingo road, D. Breedlove 49113 (CAS), D. Breedlove & F. Almeda 57394 (CAS); Mpio. Peltalcingo, slope of Ahk’ulbal Nab above Peltalcingo, D. Breedlove 49918 (CAS), 50450 (CAS), 56133 (CAS, MEXU); Mpio. Ocosingo, near El Real, E of Ocosingo, D. Breedlove 56347 (CAS); Mpio. La Independencia, Santa Elena Valley, 40-42 km E of Lagos de Montebello Natl. Park near Rio Santa Domingo, D. Breedlove & F. Almeda 57688 (CAS); Mpio. Bachajon, 3 km N of Bachajon-Ocosingo road toward Palenque, D. Breedlove & B. Keller 49404 (CAS); Mpio. Palenque, near Cascada Mizola, 25 km S of Palenque toward Ocosingo, D. Breedlove & J. Strother 46906 (CAS); ruins of Palenque, N. Diboll s.n. (WIS). GUATEMALA. Alta Verapaz: Coban, H. von Tuerckheim II 832 (8569) (US); Cob- an, H. Johnson 648 (US); between Finca Chimoté near Ru- beltein and Finca Cubilgiiitz, J. Steyermark 44173 (F, US); near Rio Icvolay and Hacienda Yaxcabnal, 5 mi NW of Cub- ligiiitz, J. Stevermark 44688 (F); Pansamala, H. von Tuerck- heim 857 (K, US). Gibson (1972) provided the new name Ruellia longituba D.N. Gibson for the species originally described as Cryphiacanthus macrosiphon Nees. Because of the prior existence of R. macrosiphon Kurz for a different species, the later name for C. macrosiphon in Ruellia (i.e., R. macrosiphon (Nees) Donn. Sm.) is illegitimate according to Article 64 of the International Code of Botanical Nomenclature (Greuter et al. 1988). Plants from Chiapas and Guatemala conform to Gibson’s (1974) description of R. /ongituba, a species that she indicated was known only from southern Mexico and Guatemala. Interestingly, the syn- types of R. longituba (1.e., syntypes of C. macro- siphon) are from Texas (J. Berlandier 316, K!; J. Berlandier 1586, K!) and Oaxaca (G. Andrieux 132, G, K!, M), and they are described as having white corollas. Hemsley (1882) treated C. ma- crosiphon and one syntype (Andrieux 132) under R. lactea Cav. Examination of the syntypes at K reveals that Andrieux 132 is a perennial herb to 2.2 dm tall that further differs from our plants by having leaves with the blades elliptic, 14-41 mm long, and 6.5-16 mm wide; flowers borne in a viscid terminal thyrse; corollas to 50 mm long; and dichasia 3-flowered. In these features it is suggestive of either R. nudiflora (Engelm. & A. Gray) Urb. or R. lactea. Berlandier’s collec- 276 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 tions are mounted on a sheet with three labels and seven plants. The label at lower left notes, “No 1586 = 316, fl. albi, Bejar Julio 1828.” Above the label, ““Berlandier”’ is written and the insignia for Hooker’s herbarium is printed. Nees annotated several plants on the sheet (which also includes a collection of Drummond) as “Cry- phiacanthus barbadensis var. tr.” The two plants at the lower left were apparently annotated by Nees as C. macrosiphon. Both are small herbs with ovate to elliptic leaf blades that are consid- erably shorter than those of R. maya. One re- sembles Andrieux 132 in having a viscid terminal thyrse but differs by having corollas up to 70 mm long. The other specimen has dichasia borne on short peduncles from the leaf axils. While the identities of the three syntypes remain in ques- tion, it is certain that none of them corresponds to R. maya, which inhabits rain forests consid- erably to the south and east of either Texas or Oaxaca. Gibson’s attribution of C. macrosiphon to this species perhaps resulted from Donnell Smith’s new combination in Ruellia with which he annotated a Guatemalan collection of von Tuerckheim that pertains to R. maya. The epithet honors the many Mayan people who inhabit Chiapas and adjacent regions. STENOSTEPHANUS VS. HANSTEINIA VS. HABRACANTHUS Wood (1988) reviewed the taxonomic and no- menclatural history of Habracanthus Nees and several other genera that he treated as congeneric with it. Based on his extensive comparison of the characters used to distinguish these genera, a sin- gle genus appears to be justified for species pre- viously treated in Glockeria Nees, Habracan- thus, Hansteinia Oerst., Kalbreyeracanthus Wassh., and Syringidium Lindau. Most Mexican species in this assemblage were described in Hansteinia. Several other related genera, not dis- cussed by Wood, that are in need of study in order to determine whether they should be main- tained include Cylindrosolenium Lindau, Kal- breyeriella Lindau, Razisea Oerst., and Steno- stephanus Nees. The former three genera were described subsequent to Habracanthus but the latter was described earlier (Nees 1847a) than Habracanthus (Nees 1847b) and would have no- menclatural priority if it were treated as conge- neric. Seven species of Stenostephanus have been de- scribed from Mexico and South America. Sev- eral were examined at K and BM, and these ap- peared to resemble species of Habracanthus (including Hansteinia) in all respects. In the pro- tologue of Stenostephanus (Nees 1847a) the an- droecium was noted as comprising two stamens and two minute staminodes. Nees (1847b) sub- sequently noted that the staminodes were lacking in some species. Bentham (1876) noted that Stenostephanus had all of the characters of Han- steinia except for the inflorescence. Lindau (1895) grouped four genera in his tribe Isoglosseae, sub- tribe Isoglossinae with monothecous stamens and girdled pollen: Oreacanthus Benth., Habracan- thus, Hansteinia, and Stenostephanus. These were distinguished as follows (translated from Ger- man): Flowers in loose, diffuse panicles. Panicles with stout pedicels, lax. Tube'‘short. broad = = Oreacanthus. Tube cylindric, straight or bent, hardly expanded \! (27) 121 ie Habracanthus. Panicles with filiform pedicels, few-flow- ered, very lax See Hansteinia. Flowers in contracted panicles (2eeaeeeoam Stenostephanus. Leonard (1953) noted that Stenostephanus, Hansteinia, and Habracanthus, were closely re- lated but distinguishable by the following char- acteristics: Corolla tube relatively long, narrow, and sub- ventricose, lips relatively long, upper lip nar- rowly linear and often curled or contorted, lower lip cuneiform and flat; panicles loose and few-flowered to dense and many-flow- (2) 76 EE ce Habracanthus. Corolla tube relatively short, broad, and ven- tricose, lips very short (scarcely more than lobes extending from rim of corolla tube). Inflorescence lax, === Hansteinia. The degree of contraction and the density of the inflorescence varies among the currently rec- ognized species of Stenostephanus, Habracan- thus, and Hansteinia and does not appear to pro- vide viable generic distinctions for these genera. Based on Leonard’s distinctions, it would be il- logical to consider Hansteinia as congeneric with Habracanthus and to exclude Stenostephanus. DANIEL: MEXICAN ACANTHACEAE 277 FiGureE 11. Stenostephanus breedlovei (Breedlove 49995). (a) leaf, x 0.75; (b) inflorescence, x 0.7; (c) inflorescence node with flower, x 2.3, and with enlargement of pubescence on peduncle. Drawn by J. Speckels. Therefore the two Chiapan species of Habracan- thus (one of which was also described in Sten- ostephanus) and the five Chiapan species of Han- steinia are transferred into Stenostephanus. In addition, two new species of the genus are de- scribed from Chiapas. Other Mexican species that would be included in Stenostephanus but that have not yet been sufficiently studied are not treated here. Thus, combinations in Stenoste- phanus have not been made for all Mexican spe- cies likely to belong in the genus. Oreacanthus, a genus of four species from Cen- tral Africa (Friis and Vollesen 1982), might also prove to be congeneric with New World Sten- ostephanus. Stenostephanus breedlovei T.F. Daniel, sp. nov. (Fig. 11) Type.— MEXICO. Chiapas: Mpio. Tenejapa, near paraje Yashanal, 2400 m, 5 March 1981, D. Breedlove 49995 (ho- lotype: CAS!; isotypes: C!, K!, MEXU!, MO!, US!). Frutex usque ad 1.2 m altus. Folia petiolata, laminae ovato- ellipticae vel ellipticae vel obovato-ellipticae, 32-140 mm lon- gae, 11-45 mm latae, 2.4—4.3-plo longiores quam latiores. Flo- res in racemum (vel thyrsum) terminalem angustum pedun- culatum dispositi; rachis pubescens trichomatibus glandulosis et eglandulosis; dichasia plerumque sessilia, uniflora; flores pedicellati. Corolla rubra, 25-29 mm longa, extus glabra; faux 17-19 mm longa et 5-7 mm diametro; labium superiorus 4— 4.7 mm longum; labium inferiorus 4-5 mm longum lobis 3- 3.5 mm longis. Capsula ignota. Shrub to 1.2 m tall. Young stems quadrate to quadrate-sulcate, bifariously pubescent with re- trorsely appressed, conspicuously septate eglan- dular trichomes 0.2-0.4 mm long. Leaves peti- olate; petioles to 36 mm long; blades ovate-elliptic to elliptic to obovate-elliptic, 32-140 mm long, 11-45 mm wide, 2.4—4.3 times longer than wide, acuminate to subfalcate at apex, acute to subat- tenuate at base, surfaces pubescent with antrorse to antrorsely appressed eglandular trichomes along major veins, margin entire to subcrenate, ciliate with closely appressed trichomes. Inflo- rescence of terminal, narrow, pedunculate ra- cemes (to thyrses) to 200 mm long (including peduncles), peduncles to 25 mm long, rachis subquadrate-flattened to somewhat ridge-an- gled, pubescent with an understory of erect most- ly eglandular trichomes 0.05—0.2 mm long and an overstory of flexuose glandular trichomes 0.2- 0.5 mm long; dichasia opposite or alternate, ses- sile (or borne on peduncles to 6 mm long at proximalmost nodes), + contracted, 1-flowered; flowers pedicellate, pedicels to 5.5 mm long, pu- 278 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 e8.8um FiGure 12. Scanning electron micrographs of Stenostephanus pollen. (a) S. breedlovei (Breedlove 49644), apertural view; (b) S. chiapensis (Breedlove 34374), apertural view; (c) S. chiapensis, interapertural view. a—c at same scale. bescent like rachis. Bracts triangular-subulate to subulate, 1.4-2.2 mm long, 0.5—0.8 mm wide, abaxial surface glabrous or with a few antrorsely appressed eglandular trichomes or flexuose glan- dular trichomes to 0.3 mm long. Bractlets tri- angular-subulate to subulate, 1.3—2.2 mm long, 0.3-0.4 mm wide, abaxial surface pubescent like rachis. Calyx 4.5—7.5 mm long, abaxially pubes- cent like rachis, lobes lance-subulate, 3.5-6.2 mm long, 0.6—0.9 mm wide. Corolla linear to subfus- iform in bud, red, 25-29 mm long, externally glabrous (margins of lobes with a few flexuose eglandular trichomes evident at apex of buds), tube + gradually expanded into throat, narrow proximal portion 4-8 mm long, 2-3 mm in di- ameter, throat 17-19 mm long, 5—7 mm in di- ameter, widest near midpoint, upper lip erect to spreading, 44.7 mm long, 1.9-2.5 mm wide, lower lip 4-5 mm long, 3-lobed, lobes 3-3.5 mm long, 2-2.5 mm wide. Stamens inserted near base of throat (i.e., in proximal 1/2 of corolla tube), 22-31 mm long, thecae red, 3—3.4 mm long; pol- len (Fig. 12a) compressed, 2-porate, exine echi- nate, encircling peripheral band lacking spines. Style 29-33 mm long, glabrous; stigma 0.2-0.3 mm long. Capsule not seen. PHENOLOGY. — Flowering: January—March. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur in evergreen cloud forests and pine-oak-Liquidambar forests at elevations from 1600 to 2460 m. PARATYPES.— MEXICO. Chiapas: Mpio. Tenejapa, near pa- raje Yashanal, D. Breedlove 49644 (CAS); Mpio. Jitotol, 5 km SE of Jitotol toward Bochil, D. Breedlove & B. Keller 49368 (CAS, MEXU, MICH, MO). This species differs from all others by the com- bination of its retrorsely appressed cauline tr- chomes; sessile, one-flowered dichasia (except at the proximalmost nodes); pedicellate flowers; and red corollas with the throat 17 to 19 mm long and the lobes of the lower lip 3 to 3.5 mm long. The epithet honors Dennis Breedlove, curator of botany at the California Academy of Sciences, editor of the Flora of Chiapas, ethnobotanist, dillegent collector, and longtime student of the Mexican flora. Stenostephanus chiapensis T.F. Daniel, sp. nov. (Fig. 13) Tyre.— MEXICO. Chiapas: ridge above Ejido Berriozabal near Cerro Boqueron, 2440 m, 29 November 1991, D. Breed- love & C. Burns 72688 (holotype: CAS!; isotypes: C!, K!, MEXU!, MICH!, MO!, US!). Frutex usque ad 4 dm altus. Folia petiolata, laminae ovatae vel ovato-ellipticae, 13-190 mm longae, 7-80 mm latae, 1.4— 3-plo longiores quam latiores. Flores in racemum (vel thyrsum) vel paniculam racemoideam (vel thyrsoideum) axillarem et terminalem dispositi; rachis pubescens trichomatibus glan- dulosis et eglandulosis; dichasia subsessilia, 1-3 (-multi)-flora; flores pedicellati. Corolla rubra et lutea, 18-23 mm longa, extus glabra vel puberula; faux 12-16 mm longa et 5.5-7.5 mm diametro; labium superiorus 2.5—4 mm longum; labium in- feriorus nullum vel lobis minus quam 0.5 mm longis. Capsula 12-16 mm longa, glabra. Shrub to 4 dm tall. Young stems quadrate- sulcate to ridge-angled, bifariously pubescent (for varying distances proximal to nodes) with flex- DANIEL: MEXICAN ACANTHACEAE 279 Figure 13. Stenostephanus chiapensis (Nelson 3771). (a) habit, x 0.5; (b) inflorescence node, x 5; (c) flower, x 2.25; (d) anthers, x 5; (e) style, x 10; (f) capsule, x 3.12; (g) seed, x 9. Drawn by E. del Valle. uose to antrorse to antrorsely appressed eglan- dular trichomes to 0.5 mm long. Leaves petiolate (distal pair often sessile); petioles to 5S mm long; blades ovate to ovate-elliptic, 13-190 mm long, 7-80 mm wide, 1.4—3 times longer than wide, acuminate to abruptly acuminate at apex, acute to attenuate at base (distal, sessile pair rounded to cordate at base), surfaces glabrous or pubes- cent with antrorse multicelled eglandular tri- chomes to 0.8 mm long on major veins, margin entire to subcrenate, ciliate. Inflorescence of ax- illary and terminal, + narrow, pedunculate ra- cemes (to thyrses) or panicles of racemes (to thryses) to 23 cm long, peduncles to 75 mm long, rachis ridge-angled, densely pubescent with an understory of erect to flexuose multicelled eglan- 280 dular trichomes 0.2-0.5 mm long or erect to subflexuose subglandular to glandular trichomes 0.05-0.2 mm long and an overstory of flexuose glandular trichomes 0.3—2 mm long; dichasia op- posite or alternate, subsessile (i.e., borne on pe- duncles to 1 mm long), + contracted, 1-3 (-many)-flowered; flowers pedicellate, pedicels 1-4 mm long, pubescent like rachis, lateral flow- ers borne on secondary peduncles to 1 mm long, pedicellate. Bracts triangular to subulate, 1.3-2.5 mm long, 0.6—1 mm wide, abaxial surface pu- bescent like rachis or nearly glabrous. Bractlets and secondary bractlets triangular to subulate, 1-2 mm long, 0.2-1 mm wide, abaxial surface pubescent like bracts. Calyx 4.5—13 mm long (ac- crescent in fruit; e.g., during anthesis 4.5-5 mm long, in fruit 9-13 mm long), pubescent like ra- chis, lobes linear-lanceolate to lance-subulate, 4.3-12 mm long, 0.6-0.8 mm wide. Corolla + c-shaped in bud, red and yellow, 18-23 mm long, externally glabrous or often appearing glabrous but inconspicuously puberulent with trichomes less than 0.1 mm long and sometimes with a few flexuose eglandular trichomes to 0.4 mm long, tube abruptly expanded into throat, narrow proximal portion 2-4 mm long, 2—2.5 mm in diameter, throat 12—16 mm long, 5.5—7.5 mm in diameter, widest near base or midpoint, + nar- rowed distally, upper lip erect to spreading, 2.5-- 4 mm long, 1-1.4 mm wide, lower lip absent (corolla truncate there) or with lobes less than 0.5 mm long and wide. Stamens inserted at base of throat (i.e., in proximal 1/3 of corolla tube), 17-27 mm long, filaments red, glabrous, thecae red, 2.8-3.6 mm long; pollen (Fig. 12b,c) com- pressed, 2-porate, exine echinate to bacculate, encircling peripheral band lacking spines or bac- culae. Style red, 22-29 mm long, glabrous; stig- ma 0.2 mm long. Capsule 12-16 mm long, gla- brous, stipe 3-5 mm long, head ellipsoid. Seeds lenticular, 2.5-3.5 mm long, 1.6-2.1 mm wide, surface roughened. PHENOLOGY. — Flowering and fruiting: No- vember—May. DISTRIBUTION AND HABITAT.—Endemic to Chiapas; plants occur in montane rain forests and evergreen cloud forests at elevations from 1500 to 2400 m. PARATYPES. — MEXICO. Chiapas: Mpio. Villa Corzo, E base of Cerro Tres Picos near Cerro Bola along road SW of Colonia Agronomos Méxicanos, D. Breedlove 24984 (DS, ENCB), D. Breedlove & R. Thorne 30229 (DS, ENCB); SE side of Cerro Tres Picos and ridges near summit, D. Breedlove 34374 (DS, PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 ENCB); from Chicharras [= Las Chicharras, ca. 23 mi NE of Tapachula], E. Nelson 3774 (GH, US). Stenostephanus chiapensis appears most sim- ilar to S. tacanensis (Acosta & R. Fernandez) T.F. Daniel, both of which differ from other Chiapan species of the genus by their flexuose- retrorse to flexuose to antrorse cauline trichomes, sessile to subsessile dichasia with mostly 1-3 flowers, lateral flowers of the dichasia (if present) sessile to subsessile, and corollas with the throat 8-16 mm long and lobes of the lower lip absent or less than 0.5 mm long. They may be distin- guished by the following couplet: Bracts lanceolate to ovate to ovate-elliptic, 2.5-11 mm long; corolla red, externally pu- bescent with flexuose eglandular trichomes 0.2-1 (-2) mm long, upper lip 4.5-8.5 mm long, lower lip 3-6.5 mm long; capsule pu- bescent with eglandular trichomes 0.05-0.2 mm long; pollen spheric S. tacanensis. Bracts subulate, 1.3—2.5 mm long; corolla red and yellow, externally glabrous or inconspic- uously puberulent with trichomes less than 0.1 mm long (rarely with a few flexuose eglan- dular trichomes to 0.4 mm long), upper lip 2.5-4 mm long, lower lip not evident or less than 0.5 mm long; capsule glabrous; pollen compressed’... = 2) eae S. chiapensis. The holotype from Boqueron differs from the other Chiapan collections of S. chiapensis (all from Cerro Tres Picos) in the understory pubes- cence of the rachis. The epithet is in reference to the Mexican state of Chiapas. Stenostephanus glabrus (Leonard) T.F. Daniel, comb. nov. Glockeria glabra Leonard in Lundell, Contr. Univ. Michigan Herb. 6:60. 1941. Hansteinia glabra (Leonard) D.N. Gibson, Fieldiana, Bot. 34:62. 1972. Type. —MEXICO. Chiapas: “Mt. Pasitar” [Mt. Paxtal], 4 August 1937, E. Matuda S-212 (ho- lotype: US!; isotypes: GH!, MICH!, NY!, US!). Stenostephanus gracilis (Oerst.) T.F. Daniel, comb. nov. Hansteinia gracilis Oerst. Vidensk. Meddel. Dansk Naturhist. Foren. Kjobenhavn 1854:143. 1855, non Hansteinia gracilis (Nees) Lindau (1893). Hansteinia oerstedii Lindau, Engl. Bot. Jahrb. 18:58. 1893, nomen illegit. (Article 63, Inter- national Code of Botanical Nomenclature, Greuter et al. 1988). Type.—COSTA RICA. San José: Mt. Jaris (fide pro- DANIEL: MEXICAN ACANTHACEAE tologue), prope San José (fide specimens), November 1846, A. Oersted 10660 (holotype: C!; isotype: CAS!). Stenostephanus latilabris (D.N. Gibson) T.F. Daniel, comb. nov. Habracanthus latilabris D.N. Gibson, Fieldiana, Bot. 34:60. 1972. TypE.—GUATEMALA. El Quiché: Cerro Putul, “Zona Reyna,” 1640 m, 3 December 1934, 4. Skutch 1836 (US!). Stenostephanus monolophus (Donn. Sm.) T.F. Daniel, comb. nov. Glockeria monolopha Donn. Sm. Bot. Gaz. 27:439. 1899. Han- steinia monolopha(Donn. Sm.) D.N. Gibson, Fieldiana, Bot. 34:62. 1972. Type.—GUATEMALA. Zacatepéquez: Cape- tillo, 1500 m, November 1889, E. Heyde & E. Lux 4556 (holotype: US!; isotypes: GH!, US!). Glockeria moralesii Standl. Field Mus. Bot. 8:47. 1930. Type. — GUATEMALA. Chimaltenango: San Martin, 1800 m, No- vember 1928, J. Morales R. 1237 (holotype: F!). Stenostephanus purpusii (Brandegee) T.F. Dan- iel, comb. nov. Hansteinia purpusii Brandegee, Univ. Calif. Publ. Bot. 6:67. 1914. Type.—MEXICO. Chiapas: Cerro de Boqueron, Au- gust 1913, C. Purpus 6842 (holotype: UC!; isotypes: BM!, GH!, NY!, US!). Stenostephanus silvaticus (Nees) T.F. Daniel, comb. nov. Habracanthus silvaticus Nees in A. DC. Prodr. 11:312. 1847. Type. — MEXICO. Oaxaca: Sierra San Pedro Nolasco, Tal- ea, etc., 1843-1844, C. Jiirgensen 902 (lectotype, designated here: K!; isolectotype: CGE)). Stenostephanus lindenii Baill. Bull. Mens. Soc. Linn. Paris 2:855. 1890, as “lindeni.”” Type.—MEXICO. Chiapas: entre San Bartolo et Pueblo-Nuevo, 1840, J. Linden s.n. (holotype: P!; probable isotypes, i.e., Linden 181: G!, K!). Stenostephanus tacanensis (Acosta & R. Fernan- dez) T.F. Daniel, comb. nov. Hansteinia tacanensis Acosta & R. Fernandez, Novon 3:221. 1993. Type.—MEXICO. Chiapas: Mpio. Union Juarez, SE side of Volcan Tacana above Talquian, 23 November 1980, D. Breedlove & F. Almeda 47714 (holotype: MEXU; iso- types: CAS!, US!). ACKNOWLEDGEMENTS I am grateful for the field assistance of W. An- derson, B. Bartholomew, D. Breedlove, S. Jes- sup, and A. Ton; artistic renderings of J. Speckels and E. del Valle; and electron microscopy of D. Ubick. Dennis Breedlove assisted with many as- pects of this study, including the creation of spe- cific epithets from Mayan words and places. It is a pleasure to dedicate this paper to him in the year of his retirement from the California Acad- emy of Sciences. Richard Brummitt, R. Howard, C. Jarvis, and D. Nicholson provided informa- tion or otherwise assisted with nomenclatural matters in the genus Blechum. The curators of the following herbaria permitted access to their collections either through visits or loans: BM, C, CAS, DS, DUKE, ENCB, F; GH, K, L, MEXU, MICH, MO, NY, P, RSA, U, UC, US, WIS. I am especially grateful for the hospitality extend- ed to me by Dick Brummitt (K), Paul and Hiltje Maas (U), Mike and Diane Dillon (F), and Bill and Chris Anderson (MICH) during recent visits to their respective institutions. RESUMEN Esperando un tratamiento taxondmico de las Acanthaceae del estado de Chiapas, México, se presentan novedades y discusiones taxonoémicas sobre la familia en Chiapas. Se incluye Barleria micans, una especies del Nuevo Mundo, en B. oenotheroides, una especies anteriormente co- nocido solamente del oeste de Africa. Se reco- noce Blechum pyramidatum como el nombre correcto para la especies muchas veces tratado como B. brownei. Se reconoce Blechum gran- diflorum (= Ruellia mirandana) en Blechum an- tes que Ruellia. Se trata Buceragenia como es- pecies varias de Pseuderanthemum con flores cleistogamas. Se trata Tribliocalyx como con- genérico con Chileranthemum y se propone la combinacion nueva, C. pyramidatum, para la especies anteriormente conocido como C. viola- ceum y T. pyramidatus. Se describen ocho es- pecies nuevas de Justicia en Chiapas; se pro- ponen dos combinaciones nuevas en Justicia para especies anteriormente tratado en Neohallia y Chaetothylax; y se provee un nombre nuevo para la especies anteriormente conocido como Belo- perone aurea o Justicia flava D.N. Gibson. Se refiere Teliostachya alopecuroidea a Lepidaga- this y se discuten las diferencias entre los dos géneros. La especies anteriormente conocido como Ruellia longituba no pertenece al tipo de este nombre; asi es que se describe para ella la especies nueva, R. maya. Se trata Habracanthus (incluyendo Hansteinia) como congenérico con Stenostephanus; se describen dos especies nue- vas de Stenostephanus de Chiapas; y se proponen siete combinaciones nuevas en Stenostephanus 282 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 13 para especies de Chiapas anteriormente tratado HeEms.ey, W.B. 1882. Acanthaceae. Pp. 500-526 in Biologia en Habracanthus y Hansteinia. LITERATURE CITED Apams, C.D.. 1972. Flowering plants of Jamaica. University of the West Indies, Mona. BEHRENSMEYER, A.K. ET AL., EDS.. 1992. Terrestrial ecosys- tems through time. University of Chicago Press, Chicago. BENTHAM, G. 1876. Acanthaceae. Pp. 1060-1122 in Genera plantarum, Vol. 2. G. Bentham and J.D. Hooker. Reeve & Co., Covent Garden. BREEDLOVE, D.E. 1981. Introduction to the Flora of Chiapas. Pp. 1-35 in Flora of Chiapas, Pt. 1. D.E. Breedlove, ed. California Academy of Sciences, San Francisco. BREMEKAMP, C.E.B. 1938. Notes on the Acanthaceae of Su- rinam. Rec. Trav. Bot. Neerl. 35:130-176. 1960. New Bornean Acanthaceae. Blumea 10:151- WS» Browne, P. 1756. The civil and natural history of Jamaica in three parts. London. CorRELL, D.S. AND H.B. CorreLit. 1982. hama Archipelago. J. Cramer, Vaduz. Croat, T.B. 1978. Flora of Barro Colorado Island. Stanford University Press, Stanford. Danpy, J.E. 1967. Index of generic names of vascular plants 1753-1774. Regnum Veg. 51:1-130. DantEL, T.F. 1989. Taxonomic notes on two cultivated spe- cies of Justicia (Acanthaceae). Baileya 23:47-SO. 1993. New and reconsidered Mexican Acanthaceae. V. Contr. Univ. Michigan Herb. 19:271-291. Durkee, L.H. 1978. Acanthaceae. Pp. 155-284 in Flora of Panama. R.E. Woodson et al., eds. Ann. Missouri Bot. Gard. 65. Flora of the Ba- . 1986. Acanthaceae. Pp. 1-87 in Flora Costaricensis. W. Burger, ed. Fieldiana, Bot. n.s. 18. FosBerG, R.F., M. SACHET, AND R.L. Oxtver. 1993. Flora of Micronesia, 5: Bignoniaceae—Rubiaceae. Smithsonian Contr. Bot. 81:1-135. Frus, I. AND K. VOLLESEN. 1982. New taxa from the Imatong Mountains, South Sudan. Kew Bull. 37:465-479. Gentry, A. H. 1993. Diversity and floristic composition of lowland tropical forest in Africa and South America. Pp. 500-547 in Biological Relationships between Africa and South America. P. Goldblatt, ed. Yale University Press, New Ha- ven. Gipson, D.N. 1970. Studies in American plants, II. Fieldi- ana, Bot. 32:173-177. . 1972. Studies in American plants, III. Fieldiana, Bot. 34:57-87. 1974. Acanthaceae. Pp. 328-461 in Flora of Gua- temala. P.C. Standley et al., eds. Fieldiana, Bot. 24(Pt. 10, Nos. 3,4). Gris, W.T. 1974. Name changes for the seed plants in the Bahama flora. Rhodora 76:67-138. GoLpDBLATT, P., ED. 1993. Biological Relationships between Africa and South America. Yale University Press, New Ha- ven. GrauaM, V.A.W. 1988. Delimitation and infra-generic clas- sification of Justicia (Acanthaceae). Kew Bull. 43:551-624. GREUTER, W. ET AL., EDS. 1988. International code of bo- tanical nomenclature. Regnum Veg. 118:1-328. centrali-americana: botany, Vol. 2. F.D. Godman and O. Slavin, eds. R.H. Porter and Dulau and Co., London. Howarp, R.A. 1989. Flora of the Lesser Antilles, Vol. 6. Arnold Arboretum, Jamaica Plain. Hsien, C. AND T. HUANG. 1978. Acanthaceae. Pp. 620-663 in Flora of Taiwan, Vol. 4. H. Lietal., eds. Epoch Publishing Co., Taipei. Jussieu, A.L. DE. 1807. Mémoire sur le Dicliptera et le Ble- chum, genres nouveaux de plantes, composés de plusiers espéces auparavant réunies au Justicia. Ann. Mus. Hist. Nat. Paris 9:251-271. Leroy, J. 1978. Composition, origin, and affinities of the Madagascan vascular flora. Ann. Missouri Bot. Gard. 65: 535-589. Linpau, G. 1895. Acanthaceae. Pp. 274-354 in Die Natiir- lichen Pflanzenfamilien, Vol. 4 (3b). H.G.A. Engler and K.A.E. Prantl, eds. Wilhelm Engelmann, Leipzig. Linnaeus, C. 1759. Systema naturae, ed. 10, Vol. 2. Holmiae, Stockholm. LaMaARCK, J. 1785. Encyclopédie méthodique. Botanique, Vol. 1. Chez Panckoucke, Paris. Lona, R.L. 1970. The genera of Acanthaceae in the south- eastern United States. J. Arnold Arbor. 51:257-309. MILLsPAUGH, C.F. 1900. Plantae Utowanae. Field Col. Mus. Bot. Ser. 11:1-110. MiranpbA, F. 1950. Fouqueria fasciculata y dos nuevas gam- opetalas de México. Anal. Inst. Biol. Méx. 21:309-317. Nees VON EsENBECK, C.G. 1846. Acanthaceae. Pp. 145-149 in The botany of the voyage of H.M.S. Sulphur, Pt. 6. G. Bentham. Smith, Elder and Co., London. 1847a (Jun). Acanthaceae. Pp. 1-164 in Flora bras- iliensis, Vol. 9. C.F.P. von Martius, ed. Munich. 1847b (Nov). Acanthaceae. Pp. 46-519 in Prodro- mus systematis naturalis regni vegetabilis, Vol. 11. Alph. de Candolle, ed. Victoris Masson, Paris. Piumier, C. 1756. Plantarum americanarum fasciculus pri- mus [—decimus], fasc. 2. Leiden. Procror, G.R. 1984. Flora of the Cayman Islands. Royal Botanic Garden, Kew. RAMAMOORTHY, T.P. AND Y. HorneELAS U. 1988. A new name and a new species in Mexican Rwuellia (Acanthaceae). Pl. Syst. Evol. 159:161-163. SLOANE, H. 1707. A voyage to the islands Madera, Barbados, Nieves, S. Christophers and Jamaica. London. SmituH, A.C. 1991. Flora Vitiensis Nova, Vol. 5. National Tropical Botanical Garden, Lawai. STEARN, W.T. 1971. A survey of the tropical genera Oplonia and Psilanthele (Acanthaceae). Bull. British Mus. (Natur. Hist.) Bot. 4:259-323. STEYERMARK, J.A. AND O. Huser. 1978. Flora del Avila. Sociedad Venezolana de Ciencias Naturales y Ministerio del Ambiente y de los Recursos Naturales Renovables, Caracas. Urpan, I. 1918. Sertum antillanum VI. Repert. Spec. Nov. Regni. Veg. 15:305-323. VoLLESEN, K. 1992. The Old World species of Stenandrium (Acanthaceae: Acantheae). Kew Bull. 47:169-202. WassHausen, D.C. 1991. Acanthaceae. Pp. 15-18 in Flora of Dominica, part 2: Dicotyledoneae. D.H. Nicholson, ed. Smithsonian Contr. Bot. 77. Woop, J.R.I. 1988. Colombian Acanthaceae—some new dis- coveries and some reconsiderations. Kew Bull. 43:1-51. © CALIFORNIA ACADEMY OF SCIENCES, 1995 Golden Gate Park San Francisco, California 94118 i 4 > a “ee . y } oa " y t) ‘9 ran 74 PAY lec eee im ia" sipfl yeblugy 44 Fe wie VL ivan U the Vol. 48, No. 14, pp. 285-314, 8 figs., 3 tables. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES INDIAN OCEAN GRENADIERS OF THE SUBGENUS COR YPHAENOIDES, GENUS COR YPHAENOIDES (MACROURIDAE, GADIFORMES, PISCES) By Yuri N. Shcherbachev P.P. Shirshov Institute of Oceanology, Academy of Sciences, 23 Krasikova Street, Moscow 117218, Russia and Vesrd Tomio Iwamoto Woods Hole, iia California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 Asstract: The subgenus Coryphaenoides of the large macrourid genus Coryphaenoides Gunnerus, 1765, is represented in the Indian Ocean by at least nine species: C. asprellus (Smith and Radcliffe, 1912); C. castaneus n. sp.; C. hextii (Alcock, 1890); C. hoskyni (Alcock, 1890); C. macrolophus (Alcock, 1889); C. rudis (Gunther, 1878); C. sibogae Weber and de Beaufort, 1929; C. woodmasoni (Alcock, 1890); and an unnamed long-barbeled species. C. castaneus is described from four specimens taken on the Ninety East Ridge; it is most similar to C. longicirrhus Gilbert, 1905 of the mid-Pacific but differs in having a naked underside of snout and larger orbits. C. asprellus, previously known only from the holotype taken in Indonesian waters, is now recorded from five specimens taken in the Indian Ocean. C. aequatoris (Smith and Radcliffe, 1912) is questionably synonymized with C. asprellus. An undescribed, widely distributed species, notable in having a long barbel, relatively well-developed outer gill rakers, a long upper jaw, and low blunt snout is left unnamed. The species is somewhat intermediate in its characters with subgenera Coryphaenoides and Chalinura. Three species of peripheral waters are redescribed after evaluation of their status from examination of type specimens: C. microps (Smith and Radcliffe, 1912); C. semiscaber Gilbert and Hubbs, 1920; and C. tydemani (Weber, 1913). They may eventually be found in the area. A provisional key to all Indian Ocean species of the genus is provided. Received May 2, 1994. Accepted October 13, 1994. March 3, 1995 INTRODUCTION The Indian Ocean has a diverse and interesting deep-sea fish fauna that was first brought to light at the turn of the century through expeditions of the Challenger, Investigator, and Valdivia. Few subsequent expeditions investigated the deep-sea fauna until the former Soviet Union expanded its world-wide fishery and oceanographic explo- rations beginning in the mid-1960s and extend- ing through the 1980s. The Indian Ocean was widely sampled by Soviet trawlers, and at con- tinental slope depths, grenadiers formed one of the chief components of the fauna. The extensive collections of deep-sea fishes from these Soviet expeditions are largely deposited at the Shirshov Institute of Oceanology (IOAN) and the Zoolog- ical Museum of Moscow State University [285] 286 (ZMMGU), both in Moscow, and at the Zoo- logical Institute (ZIN) in St. Petersburg. This is the fourth ofa series of papers on Indian Ocean grenadiers that we and Yuri I. Sazonov (ZMMGU) have collaborated on. We treated species of the subgenus Chalinura of genus Cor- yphaenoides in a previous paper (Iwamoto and Shcherbachev 1991), Kuronezumia in a second paper (Shcherbachev, Sazonov, and Iwamoto 1992), and Kumba in our most recent paper (Iwa- moto and Sazonov 1994). The present paper treats nine species of the subgenus Coryphae- noides (as defined by Iwamoto and Stein, 1974) known from the Indian Ocean and adjoining In- donesian and Philippine seas. Three species that appear to form a natural group (see Gilbert and Hubbs 1920:413-414) with C. macrolophus (Al- cock, 1889) are included even though they have yet to be recorded from the Indian Ocean proper: C. microps (Smith and Radcliffe, 1912) and C. semiscaber Gilbert and Hubbs, 1920, from the northern Philippines; and C. tydemani (Weber, 1913), known from captures in the Arafura, Bali, Ceram, and Flores seas and off Celebes. Proper characterization of these three species has been difficult, and we hope that our fresh view of them will help clarify identification problems and in- dicate the areas needing more study and more specimens. The genus Coryphaenoides poses some of the most difficult taxonomic problems in the family for anumber of reasons, including: (1) specimens often attain large size with consequent ontoge- netic changes; (2) samples are small because they are, for the most part, deeper living than most other genera; (3) their distributions are often broad, resulting in difficulty in assessing the ex- tent of variation; and (4) the species often lack good externally visible apomorphies. METHODS AND MATERIALS Methods for taking measurements and counts follow procedures elaborated in Iwamoto (1970) and Iwamoto and Sazonov (1988). Institutional abbreviations follow Leviton et al. (1985) and Leviton and Gibbs (1988). Most of the previ- ously unreported material for this study were collected by former Soviet Union vessels and are deposited in Moscow at the Zoological Museum of Moscow State University (ZMMGU) and the P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences (IOAN). Other institutions PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 from which specimens were examined include: AMS, BMNH, CAS, ISH, RUSI, SAM, USNM, ZMA, ZMUC, and ZSI. Key TO INDIAN OCEAN SPECIES OF THE GENUS CORYPHAENOIDES The following key is provided as a guide to identifying all species of the genus found in In- dian Ocean waters. Three Indonesian and Phil- ippine species are also included as they may eventually be found in the area, but four others are not, owing to the lack of adequate study ma- terial to characterize the species. (An asterisk fol- lowing a species name indicates that it has been keyed out twice.) la. Upper jaw extends to posterior of orbit or beyond . —. ee lb. Upper jaw falls short of reaching to pos- terior '3.0f Orbit. ae 18 2a. V7 2 Eee 4 5 ramen Fits GS ae 5 3a. Inner gill rakers on first arch 16"1OnaEs 3b. Inner gill rakers on first arch 10 ... hextii* 4a. Chin barbel rudimentary; a greatly elon- gated ray in pectoral and pelvic fins _. Penn. 1052000 0000 4b. Chin barbel well developed; pectoral and pelvic fins without greatly elongated rays ae nee nN He 2500710005 5a. Barbel rudimentaty —_..- = aaa 6 5b. Barbel well developed, short to long _ 8 6a. Inner gill rakers on first arch 16-19 total; VS (rarely 9) 2 memillani 6b. Inner gill rakers on first arch 8-13; V. 9- WD poco et 7 7a. Inner gill rakers on first arch 8-10; in- terorbital width 31-33% of HL Ie 6 aR ln Re filicauda* 7b. Inner gill rakers on first arch 11-13; in- terorbital width 35-42% HL ... carapinus* 8a. V. 8 (rarely 9); barbel notably long, 21- 34% HL; a pronounced heightening of nape profile beginning somewhat behind mid-length of head, producing a strongly humpbacked appearance =e ety SA ere sp. | [longbeard grenadier] 8b. V. 9-13 (rarely 8); barbel short to long, 4—26% HL; humpback beginning over orbits or essentially lacking 9 9a: Underside of snout naked __) 22 = 10 9b. Underside of snout completely scaled _ 15 SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 10a. 10b. lla. 11b. 12a. 12b. 13a. 13b. 14a. 14b. 15a. 15b. l6a. 16b. 17a. ib: 18a. 18b. 19a. 19b. 20a. Inner gill rakers on first arch 14-16; bar- bel long, 23-26% HL; snout bluntly pointed with subvertical ventral profile nk SE eee murrayl Inner gill rakers on first arch 8-14; barbel short to moderately long, 4-23% HL; snout distinctly pointed Scales all coarsely spinulated, relatively 11 adherent, especially on head 12 Scales thin, weakly or not at all spinu- lated and highly deciduous 14 Mandibular teeth in | row; underside of head almost wholly naked 13 Mandibular teeth in narrow band; un- derside of head almost all scaled, except for a median swath on snout ... castaneus Premaxillary teeth in | or 2 distinct rows; outer teeth enlarged, stout, bluntly pointed armatus Premaxillary teeth in narrow band 3 or 4 rows wide; all teeth rather narrow and sharply pointed = ferrieri Inner gill rakers on first arch 8-10; in- terorbital width 31-33% of HL SE A filicauda* Inner gill rakers on first arch 1 1-13; in- terorbital width 35-42% HL .. carapinus* Inner gill rakers on first arch 11-16 16 Inner gill rakers on first arch 8-10 Spikelike processes on preopercle; pos- terior nostril about 9% HL; interorbital width 29-31% HL grahami No spikelike processes on preopercle; posterior nostril 3-6% HL; interorbital width) 23—30% HE striaturus Snout distinctly conical in lateral profile, protruding beyond mouth; mandibular teeth in narrow band about 3 or 4 teeth wide laterally; upper jaw 33-38% HL _. asprellus* Snout blunt, barely protruding beyond mouth; mandibular teeth in 1-3 irregular rows laterally; upper jaws 39-44% HL oe a AR MED EA OO rudis Second spinous ray of first dorsal fin elongated, 1.2 or more of HEL __. No greatly elongated ray in first dorsal, height about equal to or less than HL .. 22 Preopercle produced posteroventrally forming an acutely angulated corner ...... | WER. sed ee macrolophus Preopercle margin broadly rounded _... 20 Elongated spinous ray of first dorsal fin 20b. Dakar 21b. DD 22b. 23a. 23b. 24a. 24b. DS5ae 25b. 287 about 2 or more of HL; first dorsal with a blackish distal tip; a thin naked ventral edge along snout and suborbital region Elongated spinous ray of first dorsal fin 1.2-1.5 of HL; first dorsal lacking a black tip; snout and suborbital completely SCalemiewm eee fie aS semiscaber Postorbital length 49-56% HL; orbit- preopercle distance 44-46% HL _ microps Postorbital length 45-49% HL; orbit- preopercle distance 38-41% HL... PA keen Jie MORI CY tydemani Underside of snout naked; V. 10; snout 27-28% HL; upper jaw 38-42%; teeth in lower jaw in | or 2 irregular series lat- CkAlive sae Bel eee = castaneus* Underside of snout completely scaled; V. 7-10; snout 28-32% HL; upper jaw 27- 39% HL; teeth in lower jaw in narrow to broad bands, usually 3 or more teeth wide laterallyeee.% 2../!.3° ototy 7a eet eee! 23 V. 7; upper jaw about 39% HL; postor- bital length 53%; internasal 17%; barbel [GDG RS Lore eM ie) on OD ee hextii* V. 8-10; upper jaw 27-38% HL; post- orbital length 45-53%; internasal 21- 31%; barbel 3—20% 24 Barbel 10-20% HL; V. 9-10; preoral 13- Rwy. Gl eee ee ee asprellus Barbel 3—9% HL; V. 8-9; preoral 13- 21% HL 2S Teeth in lower jaw in broad band, band falls short of end of rictus; upper jaw 27— 30% HL; scale rows below midbase of hoskyni Teeth in lower jaw in narrow band, 3 or 4 teeth wide laterally, extends to or be- yond end of rictus; upper jaw 31-38% HL; scale rows below midbase of 1D. AS—6.0 oe es. Set Swoodmasoni DESCRIPTIONS Genus CORYPHAENOIDES Gunnerus, 1765 The genus as used here follows definitions giv- en by Iwamoto and Stein (1974) and Iwamoto and Sazonov (1988). Coryphaenoides asprellus (Smith and Radcliffe, 1912) (Fig. 1) PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 288 ‘ParoNNsuodad A[sNOlIVA Sa[eos PUL SUL “WOOL ‘1-009'T Ul BI[eISNY Jo sed IsoMYIIOU YO WO (TL WW +8/p) LEPIL SVO ‘(QYTOpey pue yyuss) snjjadsv sapio SGA sip Saye hnlaaleensira “wu ¢Z sjenba 1eq [BS uapyddioy) *| FANS SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES TABLE |. and 6 specimens). Asterisk (*) denotes holotype. 289 Selected measurements of Coryphaenoides asprellus (holotype), C. aequatoris (holotype), and C. hoskynii (holotype hoskynii asprellus aequatoris ZSI ZSI IOAN TLS ZMUC ZMUC CAS *USNM *USNM 1061 1057 uncat. 123833 373083 373082 66462 72935 72937 TL (mm) 245+ 266+ 333+ 350+ 321+ 309+ 410+ 413+ 188 HL (mm) 50.3 60.5 70.3 72.0 2D BD 89.7 90.8 43.4 The following in percent of head length Snout 32.4 30.3 30.6 30.0 31.9 ay 7/ 30.1 28.3 28.6 Preoral 19.9 18.2 19.1 15.6 19.7 19.9 19.9 15.4 16.1 Internasal 32.8 30.1 28.3 28.9 25.9 26.1 24.2 25m 24.7 Nostril 9.5 9.1 Te NED 7.6 Syl 5.0 7.9 Sp Interorbital 28.2 24.5 26.3 28.9 24.2 25.3 23.6 28.3 24.2 Orbit 27.8 30.3 DiS 24.4 24.9 22.8 27.4 23.9 25.8 Postorbital 44.9 46.3 46.7 45.8 48.5 49.5 47.2 Sy, 46.3 Orb-preopercle 45.7 46.3 43.4 47.2 46.5 45.9 46.4 46.8 44.9 Upper jaw 28.4 30.3 27.0 Piel 28.5 29.1 28.5 33.0 Sie3 Rictus NS — 20.8 22.3 72310) 22.1 24.2 _ — Barbel a 6.9 5a 4.4+ 5.5 4.8 9.1 9.9 15.7 Gill slit - - 9.1 8.3 6.9 6.7 8.1 _ 8.5 Macrourus asprellus Smith and Radcliffe in Radcliffe, 1912: 118-119, pl. 26, fig. 1 (holotype, USNM 72935; Dutch East Indies, se. of Bachian Island, 1°00'00’S, 127°50'00’E; 845 fm [1,545 m]) ?Macrourus aequatoris Smith and Radcliffe in Radcliffe, 1912: 120-121, pl. 26, fig. 3 (holotype, USNM 72937; Celebes, Gulf of Tomini, 0°08'00’S, 121°19'00”E; 1,089 fms [1,992 m)]). DraGcnosis.—V. 9-10; snout and mandible uniformly and completely scaled; length upper jaw '3 or more of HL, extends to or beyond midorbit; barbel long, slender, about 10-20% HL, equal to or less than orbit diameter and much less than snout length or interorbital width; py- loric caeca 8-10; premaxillary teeth in broad bands, outer series slightly enlarged, mandibular teeth band about 3-4 teeth wide laterally. MATERIAL ExAMINED.—USNM 72935 (holotype of M. as- prellus, 90.8 mm HL, 413+ mm TL; see synonymy for locality data). USNM 72937 (holotype of M. aequatoris, 43.4 HL, 188 TL; see synonymy for locality data). ZMMGU P.19256 (110 HL, 514+ TL) and CAS 71487 (102 HL, 478+ TL); Exmouth Plateau off NW Australia, 16°55’S, 114°53’E, 1600-1700 m; Prof. Mesiatzev cr. 7, tr. 45; 24.1V.1979. ZMMGU P.19257 (2, 117-102 HL, 515-497+ TL); Ninety East Ridge, 11°14’S, 88°48’E, 1500-1600 m; Fiolent cr. 9, tr. 22; 13.VI1.1977. ZMMGU P.19258 (90 HL, 428 TL); Mascarene Ridge; 11°17’S, 62°33’E; 2020-2290 m; Vityaz’ cr. 17, sta. 2806; 7.1.1989. COUNTS AND MEASUREMENTS (see also Table 1 and Diagnosis).— 1D. IIJ,9-11; 1P. i118-i21; to- tal GR-I (outer/inner) 3-5/8-9, GR-II 6-7/7-9; scales below 1D. 5.0-7.5, below midbase 1D. 4.5-5.5, below 2D. 4—6, lat. line 30-39. Total length 428+-515 mm, head length 90- 117 mm. The following in percent HL: postros- tral 73-78; snout 28-31; preoral 13-16; inter- nasal 21-27; interorbital 22-30; orbit 21-26; suborbital 13-16; postorbital 46-53; orbit—preop. 45-50; upper jaw 31-38; barbel 7-20; gill slit 8- 9; pre-A. 146-159; V.-A. 34-47; isthm.-A. 74— 84: body depth 68-75; depth over A. origin 54—- 63; 1D.-2D. 12-24; height 1D. 83-92; length 1P. 46-62; length V. 72-88; post. nostril about 3 (female), 6-8 (male). DEscRIPTION.— Head broad, greatest width about equal to postorbital length, length about 5 in total length; body long and gradually tapered from anal fin origin to tip of tail. Snout broadly conical in lateral view, broad over internasal space, protrudes beyond mouth a distance less than lens diameter. Orbit diameter less than snout and interorbital, greatest diameter less than 5 in HL. Interorbital space broad, flat to slightly con- vex. Suborbital region smoothly and gently con- vex, without a sharp, angular ridge. Preopercu- lum broadly rounded, its posterior margin slightly inclined, its ridge forming a strong club-shaped lobe at posteroventral angle; interopercle broadly triangular at exposed posterior end. Mouth wide, subterminal, upper jaw extends to slightly be- yond vertical through midorbit, rictus falls short of that vertical. Barbel slender from base to tip, length less than orbit diameter. Free neuromasts on head prominent, especially so over surfaces of snout, suborbital, mandible, interorbital and 290 occipital region, their blackish color contrasting with light to medium brown ground color. Gill membranes broadly attached to isthmus with no free posterior fold; gill openings extend forward close to point below juncture of pre- opercle and dentary. Outer gill slit short, greatly restricted; rakers along outer side of first arch few and rudimentary, other rakers tubercular. Most of body and head uniformly covered with moderately large, finely and densely spinulated scales. Scale size highly variable on head; those on snout and on ventral surfaces generally much smaller; scales along suborbital, supranasal, and supraorbital slightly thickened and more adher- ent; median and lateral angles of snout with small, inconspicuous, scutelike scales. Entire underside of head scaled; interopercle naked, as are gill membranes, lips, and nasal fossa. Spinules on scales below interspace of dorsal fins extremely fine, short, spikelike, and in densely packed, ir- regularly convergent rows. More ventrally and forward on trunk, scales have spinule arrange- ment less well defined, and some spinules flat- tened, approaching lanceolate shape. In 497+ mm male specimen from Ninety East Ridge, squamation differs slightly; spinules appear coarser, rows better defined and fewer, and me- dian row slightly, but noticeably enlarged. Premaxillary dentition consists of a broad band of small teeth flanked along outer edge with a row of slightly enlarged, sharp conical teeth. Mandibular teeth all small, in a rather uniformly narrow band (3-4 teeth wide; slightly wider at symphysis). Fins well developed except second dorsal ru- dimentary throughout. Long spinous ray of first dorsal fin finely and sparsely toothed, the greatly reclined spikes not overlapping, the ray distally terminating in a hair-fine tip. Outer pelvic ray thickened and elongate (probably more so in males than in females) and extending in one spec- imen (102 mm HL from Ninety East Ridge) to 11th anal ray. Pectoral origin slightly in advance of pelvic origin, first dorsal origin slightly behind that. Interspace between dorsals short, less than length base of first dorsal; anal origin below in- terspace. Pyloric caeca in 110 mm HL male from Ex- mouth Plateau about 10, relatively short and slender, length of longest caecum about equal to orbit diameter. Gonads moderately developed. Swim bladder large, with a thick spongy layer and 4 very slender retia, each terminating in a PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 small gas gland. The 90-mm HL male from Mas- carene Ridge had 8 long, slender caeca and rel- atively small gonads. Color in Exmouth Plateau specimens and fe- male from Ninety East Ridge light brown overall; male from latter area dark brown. Fins all black- ish, except outer pelvic ray paler. Barbel light brown. Oral cavity gray, paler towards periphery, darker on tongue; gill cavity black. Branchios- tegal membranes black; gular membrane dark grayish brown. DISTRIBUTION. — Indonesia; northwestern coast of Australia; and Indian Ocean over Saya de Mal- ha and Ninety East ridges, in 1,500-2,290 m. SizE.—To at least 52 cm TL. REMARKS AND COMPARISONS.— The holotype of C. asprellus is badly deteriorated, with the head having come apart from the rest of the body. Nonetheless, most of the important head features remain for comparison with our material. The barbel in the holotype may have been longer than the 10% of HL measured, and in closer agree- ment with the 13—20% we measured in our fresh- er material. The original figure suggests that part of the belly, including the pelvic fins, was already gone when the illustration was rendered. We were unable to locate a pelvic fin in the specimen jar, but in the original description the count was giv- en as questionably eight. That count is low for our material (9 or 10) and raises some doubt as to whether we have correctly identified our spec- imensas C. asprellus. Other features, on the other hand, like the jaw size, dentition, squamation, barbel length, and general physiognomy, support our identification. The holotype of Macrourus aequatoris, which we here treat as a junior synonym of C. asprellus, is substantially smaller than other specimens, and measurement differences may in part be attrib- utable to size. Thus, the preoral length (16% HL), orbit diameter (26%), suborbital width (16%), and outer pelvic ray (88%) were slightly longer, and the lengths of postorbital (46%), orbit-pre- opercle (45%), upper jaw (31%), and preanal (146%) were slightly shorter. Meristic characters, however, were in complete agreement. The large female from the Ninety East Ridge showed some differences from the four male specimens studied in width of internasal (21% HL cf. 23-27%), interorbital (22% cf. 27-30%), orbit diameter (24% cf. 21-23%), postorbital length (50% cf. 52-53%), and length posterior nostril (3% cf. 6-8%). The lateral-line scale count SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 291 over a distance equal to predorsal length was also higher (39 cf. 30-37). In addition, the outer pel- vic fin ray, though missing its tip on both fins, appeared shorter and thinner in the female, com- pared with those of the males. We presume these differences to be sexual, as the female was in every other way so similar to the males, es- pecially the Exmouth Plateau specimens, that there is no reason to doubt their conspecificity. The male from Ninety East Ridge was trou- bling in that it had a different overall appearance. The snout appeared somewhat blunter, the head contours smoother; squamation coarser, the spi- nule rows fewer, and with a distinct, enlarged median row; the underside of head had smaller, more adherent scales; teeth in the mandible ap- peared to be in a wider band; and the overall color was darker. In all these characters except the snout shape and mandibular teeth, the dark male from the Mascarene Ridge was similar. Nonetheless, all counts and measurements for both specimens agreed with the others (excluding sexually dimorphic female characters). With so little material available, we do not speculate on these variants. Coryphaenoides hextii appears to be closely related, but a pelvic fin ray count of seven is well below the 9 or 10 of our specimens, and there were 13 pyloric caeca in a BMNH specimen (1892.6.17.4) examined. Also, the snout was shorter (25%) and internasal width narrower (18%). Neither the holotype nor any other spec- imen of C. hextii was found ‘in the ZSI, but the one BMNH specimen we examined was proba- bly sent there by Alcock. (We also examined a 133-mm juvenile from Vityaz’ cruise 17 station 2832.) Coryphaenoides asprellus is similar in many respects to C. castaneus but can be distinguished by its completely scaled underside of snout (cf. naked medially in C. castaneus), broader man- dibular dentition (about four teeth wide laterally cf. about one row), and its shorter upper jaw (33- 38% HL cf. 39-41%). Coryphaenoides hoskynii is readily differenti- ated from C. asprellus in having a greater inter- nasal width (28-30% HL), larger orbit (24-30%), shorter postorbital length (45-46%), shorter or- bit-preopercle length (45-47%), shorter upper jaw (21-30%), and shorter barbel (about 5—7%). We examined numerous specimens, including the holotype, of C. woodmasoni for comparison with C. asprellus. The size of the chin barbel was the primary distinguishing character, it being much larger in C. asprellus (usually 13-20% HL cf. 3—9%). The pyloric caeca count was somewhat lower (8-10) than in C. woodmasoni (10-14). Coryphaenoides asper Giinther, 1877, known only from the holotype taken off Japan, is dis- tinguishable from C. asprellus by its blunter snout profile, and coarser scale spinulation (spinules rather long, hindmost extending well beyond posterior edge of scale, dispersed in about five divergent rows). Coryphaenoides castaneus new species (Fig. 2) No literature applies to this species, as far as we could determine. Diacnosis.— V. 10, outer ray elongated, about 61-123% HL, extending well posterior to A. or- igin to between 6th and 16th ray; a broad median naked area on underside of snout; upper jaw 39- 42% HL, maxilla extends to below mid-orbit or beyond; greatest orbit diameter 21-23% HL, slightly less than interorbital width (24-27% HL); barbel 16-23% HL; premaxillary teeth band 5- 6 teeth wide, with a distinctly enlarged outer se- ries; mandibular teeth in a narrow band that ta- pers to | row; spinules on body scales in irregular, slightly convergent rows, spinule tips extend be- yond posterior scale margins. Type SPecIMENS. — HOLOTYPE: CAS 71486 (female, 95 HL, 420 TL); Ninety-Degree East Ridge, 14°42.4’S, 86°49.4’E; 1760 m; Prof. Mesiatzev cr. 7, tr. 12; 19.III.1979. PARATYPES: ZMMGU P.19259 (155 mm HL, 690+ mm TL); 14°51’S, 86°49’E; 1760 m; Prof. Mesiatzev cr. 7, tr. 11; 19.III.1979. ZMMGU P.19260 (159 HL, 740+ TL); same data as for ho- lotype. Three specimens of Coryphaenoides longicirrhus Gilbert, 1905, were examined for comparison: BPBM 3429 (198 HL, about 800 TL); Honolulu Fish Market; 10.11.1927. SIO 68- 457 (136 HL, 650 TL); 17°54.5’N, 174°14.9'W, Hess Guyot; 2.1X.1968. SIO 68-460 (116 HL, 502+ TL); 18°32.0'N, 178°13.4’'W, Hamilton Guyot; 4.1X.1968. COUNTS AND MEASUREMENTS (see also Table 3 and Diagnosis).—II,10—11; 1P. 118-120; total GR-I (outer/inner) 3-4 / 9, total GR-II 7-8 / 8- 10; scales below 1D. 5.5-—7.5, below midbase 1D. 4.5-6, below 2D. 7-8, lat.line 34-39. Total length 412-740+ mm; HL 95-159. The following in percent HL: postrostral 75—76; snout 27-28; preoral 12-14; internasal 22-24; post. nostril 5—6; suborbital 13-14; postorbital 54-56; orbit-preop. 51-53; gill slit 6-8; pre-A. 163-164; 292 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 7 ides castaneus new species. Holotype, CAS 71486 (420 mm TL) from Ninety Degree East Ridge, 1,760 m. Fins and scales partially reconstructed. Scale bar Coryphaeno FiGure 2. equals 25 mm. SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 293 TaBLE 2. Measurements and counts of specimens of Coryphaenoides microps, C. tydemani, and C. semiscaber. C. semiscaber C. microps C. tydemani USNM USNM USNM_ CAS-SU CAS-SU USNM_ CAS-SU CAS-SU ZMA ZMA ZMA *83625 +14882 +14883 +25452 +25452 *72933 25440 25440 *111.491 *111.493 *111.493 MEASUREMENTS: IME 212+ DNS 200+ 187+ 124+ 413+ 265 263 265 174+ 160+ HL 48.3 41.8 45.4 39.9 2307 82.0 59.9 48.9 53" 37.0 34.4 The following in mm: Snout 29.0 28.9 27.8 30.1 31.2 28.4 31.9 SiS 28.2 29.5 SylE7/ Preoral Sa 14.6 12.6 14.8 14.3 11.0 13.3 13.9 12.8 13.2 14.5 Internas. [Sjsit 15.6 16.1 17.0 19.8 Sez 16.2 16.6 16.6 15.4 17.4 Interob. 19.9 Pill 17.6 19.3 23D. 20.7 18.8 24.9 DVT 22.4 23.0 Orbit 26.3 25.6 28.0 20.8 24.9 17.4 2355 25.6 26.4 DSI 25.3 Suborb. 12.6 12.4 ES 11.8 13.1 S45) 12.9 13.1 14.1 1315) 13.4 Postorb. 46.4 47.6 47.6 50.1 49.4 56.1 47.9 45.8 49.3 49.2 45.3 Orb-preop. 37.9 35.6 39.0 44.4 43.5 46.3 41.0 39.5 38.4 40.8 37.8 Up. jaw 29.4 29.4 34.8 30.8 32.5 30.4 29.0 Silat Syl 33.8 32.8 Barbel WS 10.3 16.7 7.8 9.7 9.1 9.3 6.5 10.9 9.7 11.0 Gillslit 10.6 9.6 6.6 6.0 We? 6.7 13.8 8.4 ca6 — - Len. 1D. 119 144 119+ 293 388 190 282 262 266 319 297 Counts: 1D. 11,10 11,9 11,9 11,10 11,9 11,9 11,10 11,10 11,9 11,9 11,9 Be i20/i21 i—-/Ai19 i20/—- i19/119 i20/120 119/119 120/121 120/120 121/121 121/121 119/118 V. 8 8 = 8/8 7/7 7/7 8/9 8/8 8/8 8/8 8/8 GR-I(in) 10 10 8 11 11 10 8 9 9 9 9 GR-II(out) 10 8 8 10 8 10 10 9 9 9 9 GR-II(in) 9 10 8 8 9 9 9 9 9 9 9 Scales: ANDY 8 10 8.5 8.5 TS 9.7 WS 6.5 — 8 8.5 mid1D. 525 6.5 3) TES — 6.5 6.5 6.5 fe 6.5 6.5 31 37 40 — — _ lat.l. 34 35 32 35 37 * denotes holotype or lectotype. + denotes paratype or paralectotype. isthm.-A. 85-94: body depth 81-89; 1D.-2D. 18-27; height 1D. 77 (1 spec.); length 1P. 50-54. DEscRIPTION. — Head rather large and broad, length about 44.5 in TL; body deep in trunk region, but tapers rapidly behind vent. Snout bluntly pointed and slightly protruding, rather broad when viewed from dorsally, its width at lateral angles about equal to that of interorbital; median tip and lateral angles weakly tipped with small thickened scales. Orbits small, slightly more than % HL. Interorbital broad, slightly convex, width greater than that between supranarial ridg- es. Preopercle margin rather broadly rounded, hind margin only slightly inclined from vertical; preopercle ridge forming a pronounced lobe at posteroventral corner; interopercle broadly tr- angular at posterior end, where it is slightly ex- posed beyond preopercle. Mouth large, rictus scarcely restricted, extends to below anterior of orbit, hind end of maxillary below mid-orbit or beyond; lips rather thick and papillaceous. Ridges of head not strongly developed, although suborbital and preopercle ridges well marked. Barbel moderate in length and slender, some- what less than, to slightly more than, orbit di- ameter. Head pores not well developed, but free neuromasts numerous on head, occipital region and nape (although not especially prominent be- cause of dark ground color). Gill membranes broadly attached to isthmus, the gill openings extending forward only to level of vertical ridge of preopercle. Outer gill slit small, greatly restricted, only 3 or 4 small rakers along distal side of arch. Gill rakers low, tubercular; those on upper arm flattened and platelike. Body and head uniformly covered with rather deciduous scales except for naked area on un- derside of snout, narrow strip of lower edge of 294 suborbital region, interopercle, anteriormost tip of mandibular rami, and usual places like gill membranes, lips, nostril region, and fins. Scales along ridges of head not especially large or thick- ened into coarsely spined scutes, although those present more adherent than adjacent scales. Body scales moderately large, scale-pocket margins darkly outlined, spinules sharp and more or less conical, in irregularly convergent rows to irreg- ular quincunx pattern. In 420+ mm specimen, spinules in about | 2 or more rows in largest scales; middle row consisting of 5 or 6 spinules, with posteriormost spinule slightly enlarged and broadly overlapping posterior edge of scale. Premaxillary teeth in a tapered band 5 or 6 teeth wide at widest part; an enlarged outer series of conical, slightly recurved, evenly but rather widely spaced teeth. Lower jaw teeth in a narrow band that tapers to 1 row posteriorly. Relative fin positions somewhat difficult to de- termine because of dorsal flexure in all 3 speci- mens, but pectoral apparently farthest forward, followed by pelvic and first dorsal. Origin of sec- ond dorsal close behind first dorsal, about on same vertical as anal. Second spinous ray of first dorsal rather thin, length slightly more than post- rostral length of head, ending in a hair-fine fil- amentous tip; weak teeth along leading edge widely spaced and non-overlapping, becoming obsolete in 2 largest specimens. Pyloric caeca of female holotype slender, sim- ple, about 20 mm at the longest. Swim bladder in this specimen large, with a tough white tunica externa and a spongy internal layer; 4 slender retia folded once over themselves, each termi- nating in a small globular gas gland. The stomach contained a single, shrimp-like crustacean. Overall ground color dark chocolate brown; fins all blackish; oral, branchial, and peritoneal linings, and branchiostegal and gular membranes all blackish or dark gray. Lips black. Naked areas on parts of head, under snout, and on mandible with a violet hue. Barbel in smallest specimen dark brown basally, but distal 4 or so whitish; in larger two specimens, barbel uniformly me- dium brown. DISTRIBUTION. — Known only from the holo- type and two paratypes taken on the Ninety East Ridge in the Indian Ocean in 1,760 m. SizE.—To more than 740 mm TL. EtyMoLoGcy.—From the Greek, castaneus, of chesnut color. REMARKS AND COMPARISONS. — Coryphae- noides castaneus is most similar to C. longicir- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 rhus, differing only in the presence of a naked area under the snout, a coarser squamation, with spinules on body scales overlapping the posterior margin, and having a slightly narrower interor- bital (24-27% HL vs. 28-30% in C. /ongicirrhus). Additionally, the new species appears to have a slightly smaller orbit that goes 0.75—0.90 into the interorbital width, compared with about 0.63- 0.78 in C. longicirrhus. These proportional dif- ferences should be substantiated with more spec- imens representing a broader size range. The new species is also closely similar to C. asprellus, but is distinguishable mainly by the dentition on the lower jaw (narrow band becoming uniserial pos- teriorly, vs. in a broad band in C. asprellus), by the completely scaled underside of head in C. asprellus, the shape of the suborbital region (gently convex, almost flat in C. asprellus, angular in C. castaneus), and the squamation (spinules of body scales somewhat lanceolate in C. asprellus, con- trasted with distinctly spikelike and conical in C. castaneus). The nape of C. castaneus is also high- er, with a humpbacked appearance. Color of C. asprellus is somewhat paler, medium brown over most of body, dark brown to swarthy on head and over abdomen. Coryphaenoides castaneus can be distinguished from C. woodmasoni by its larger barbel, higher pelvic fin ray count, and generally longer outer pelvic ray. Coryphaenoides hextii (Alcock, 1890) Macrurus Hextii Alcock, 1890: 299-300 (one specimen, ho- lotype, female, “*22 inches” [ca. 56 cm]; /nvestigator st. 104; 1,000 fathoms [1,829 m]). Alcock 1892: 351, 353 (female, “nearly 23 inches long,” Jnvestigator st. 122, 865-880 fath- oms). Alcock 1894: pl. XII, fig. 3. Diacnosis.—V. 7; underside of head com- pletely scaled; upper jaw 34-39% HL; maxillary extends to below posterior '2 to 4 of orbit, rictus to below middle third of orbit; barbel 16-21% HL; pyloric caeca 12-15; premaxillary teeth in broad band with outer enlarged series, mandib- ular teeth in narrow band about 3 teeth wide, inner series slightly larger than outer two. MATERIAL EXAMINED.— NEOTYPE (here designated) BMNH 1892.6.17.4 (120 mm HL); Laccadive Sea; 12°05'55’N, 71°33'30’E; 865-880 fm [1,581-1,609 mJ]; “Investigator” st. 122. ZMMGU uncat. (22.6 HL, 133 TL); off Somalia; 12°43'N, 52°44’E: 2,380-2,300 m; VITYAZ’ cr. 17, st. 2832; 16.1.1989. COUNTS AND MEASUREMENTS (BMNH speci- men first, followed by ZMMGU).—1D. IL,9, 11,10; 1P.-/i19, i21/i121; GR-I (outer/inner)-/ 1+1+8,ca. 6/2+8;GR-II (outer/inner) | +1+7/ SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS COR YPHAENOIDES 295 1+1+8, 2+8/2+8; scales below 1D. 7.5/7; be- low mid-base 1D. 5.5/ca. 6; below 2D. 5.4/ca. 8; lat. 1. 34/45. The following in percent HL: snout 24, 31; preoral 14, 19; internasal 17, 24; interorbital 24, 25; orbit 24, 30; suborbital 13, 15; postorbital 53, 46; orb.-preop. 49, 47; outer gill slit 10-14; pre-A. 182, 147; V.-A. 72, 44; greatest body depth 87, 80; 1D.-2D. interspace 47, 21; height 1D. 76,-; length V. 67,-; post. nostril 7, 10. DEescriPTION. — Head and body shape much as described and illustrated for C. castaneus. Snout conical in lateral profile, terminal and lateral an- gles tipped with small scutelike scales. Suborbital shelf broad, forming a shallow angle with ven- tral part of suborbital region. Posterior margin of preopercle broadly rounded; preopercle ridge forming a broad lobe at posteroventral angle. Interopercle broad, scarcely exposed beyond pre- opercle. Mouth relatively small for genus, sub- inferior. Barbel moderately long and slender, much shorter than interorbital width, about 7 of orbit. Squamation complete over head of BMNH specimen from Laccadive Sea, but juvenile from Somalia with underside of snout and suborbital, and anterior half of mandible naked. Scales of suborbital shelf strong and adherrent, but not enlarged. Body scales covered with slender, con- ical, densely packed spinules in no particular ar- rangement, so far as we can tell. First dorsal height about equal to postrostral length of head; leading edge of second spinous ray with small denticles almost rudimentary near base of BMNH specimen, but few and prominent in ZMMGU juvenile. Second dorsal fin rudi- mentary anteriorly, and origin difficult to ascer- tain; fin low throughout. Anal fin well developed. Outer pelvic ray slightly prolonged but falling short of anal fin origin. Pyloric caeca in BMNH specimen 13, slender, much longer than orbit diameter (longest about 40 mm); in ZMMGU juvenile 12, relatively thick and short, much shorter than orbit diameter (about 2.3 mm). Color described by Alcock (1890) as chocolate with blackish fins; mouth, gill cavity, and peri- toneum black. Abdomen and gill cover blackish in ZMMGU juvenile; barbel dark. REMARKS AND COMPARISONS. — The ZMMGU juvenile is questionably referred to this species. The naked undersides of snout and suborbital contrast with the fully scaled condition in the large Laccadive Sea female and Alcock’s (1894: pl. XII, fig. 3) figure of the species. Certain pro- portional measurements also show marked dif- ferences between the two examined specimens, notably those of the snout, orbit, internasal, post- orbital, upper jaw, barbel, and gill slit. The count of lateral line scales over a distance equal to pre- dorsal length is considerably higher in the ZMMGU specimen. Whether these differences are attributable to size is not known, and addi- tional specimens of more representative size se- ries are necessary to properly evaluate our iden- tification. A search by one of us (YNS) at the Zoological Survey of India in 1987 failed to find the holo- type of C. hextii, but the BMNH specimen, re- ported on by Alcock (1892) and examined by us, is in good shape. Menon and Yazdani (1968), in their catalog of type-specimens in the Zoological Survey of India, did not list the type of C. hextii, nor did Menon and Rama-Rao (1970). To assure stability of the name, we here designate BMNH 1892.6.17.4 as the neotype of Macrurus hextii. Coryphaenoides hextii differs from other members of the genus reported in this work by its seven pelvic fin rays (compared with eight or more). In addition, although closely similar to C. woodmasoni in overall appearance, it differs in having a longer barbel; compared with C. cas- taneus it has a somewhat shorter upper jaw, shorter outer pelvic ray, smaller premaxillary teeth, and different scale spinulation; compared with C. asprellus it has more pyloric caeca. Coryphaenoides hoskynii (Alcock, 1890) (Fig. 3) Macrurus (Macrurus) Hoskynii Alcock, 1890a:214-215 (Bay of Bengal near Madras); 1894a:126 (in key); 1894b, pl. 9, fig. 4, 1899:116-117. D1acnosis.— V. (8) 9; undersides of snout and mandible fully scaled; upper jaw length 27-30% HL, posterior end of maxillary does not extend beyond midorbit; rictus restricted laterally, its length 18.4-22.3% HL; barbel small, less than 10% of head length; pyloric caeca 6-9; teeth in broad bands in both jaws, none enlarged. MATERIAL EXAMINED.—ZSI 12833 (holotype, 72 mm HL, 357+ mm TL); Bay of Bengal, 18°26’N, 85°24'E; 2,396 m; Investigator sta. 97; 14.1II.1890). ZSI N1057 (60.5 HL, 266+ TL), ZSI 1061 (50.3 HL, 245+ TL) and ZSI 1062 (58 HL, 378+ TL); “West Ceylon.” ZMUC 373082-83 (2, 72.2-73.2 HL, 321+-309+ TL); Bay of Bengal, 17°10'N, 84°30’E; 2,820 m:; Galathea sta. 299; 24.1V.1951. CAS 66462 (89.7 HL, 410+ TL) and ZMMGU P.19261 (70.3 HL, 333+ TL); off Socotra I., 12°43'N, 52°44’E; 2,380-2,300 m; Vityaz’ cr. 17, sta. 2832; 16.1.1989. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 296 “WU ¢Z sfenba eq a[edg “pojonnsuodas Ayjersed sayeos pue sul “Ww OOE‘Z—YRE'Z UI BNOD0S Yo woy (TL WW +€E€) 19761'd NOWWZ 09] y) 1Aysoy saprouanyddu0y “¢ aFANOL] SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 297 COUNTS AND MEASUREMENTS (see also Tables 1 and 3).—1D. II,9-10; 1P. 117-22; total GR-I (outer/inner) 0-3/7-8, total GR-II 7—-8/7-9; scales below 1D. 5—7.5, below mid-1D. 3.5-4.5, below 2D. 4-5, lat.1. 31-34. The following in percent of head length: pre- A. 158-167; pre-V. 120-131; V.-A. 44-50; isthm.-A. 32-95; greatest body depth 71-80; 1D.- 2D. 17.1-34.3; height 1D. 82—92; length 1P. 48- 51; length V. 62-82. DESCRIPTION. — Head 4.5—5.0 into total length; greatest depth more than postrostral length of head. Snout distinctly conical in lateral profile, tipped with a small but stout conical scute; small- er scutes at lateral angles of snout. Posterior nos- tril small to moderate in size, about 2-3 into least suborbital width. Orbits round to somewhat oval in outline, diameter shorter than snout length. Interorbital space broad, about equal to or more than orbit diameter. Suborbital shelf prominent and angular in cross section (more so in larger specimens), forming a distinct division between upper and lower parts of head. Posteroventral margin of preopercle broadly rounded; ridge on preopercle forms a prominent clublike head ven- trally. Interopercle exposed beyond preopercle as a triangular fleshy tab. Mouth small for a Cor- yphaenoides, rictus falls below anterior one-third of orbit; posterior end of maxillary below mi- dorbit. Chin barbel small, length about one-half or less of suborbital width. Free neuromasts of cephalic sensory system small but relatively prominent over most of head and on nape. Gill membranes broadly attached to isthmus, without a free posterior fold; opercular opening extends forward to below or behind vertical por- tion of preopercle ridge. Outer gill slit greatly restricted, length of slit about equal to barbel length. Outer gill rakers on first arch rudimen- tary, sometimes absent; other rakers tubercular. Relatively small scales cover all of head sur- faces including entire underside of snout, sub- orbital, and lower jaw, but absent on gill and gular membranes and interopercle. Scales of body generally larger than those of head; fine, conical, greatly reclined spinules arranged in subparallel rows on scales below origin of second dorsal fin. Scales in other areas may have rows divergent or convergent. Median row of spinules on scales of dorsum usually with slightly enlarged spinules. Spinules generally reduced along posterior mar- gin of dorsal body scales (not known for scales over ventral areas). In holotype the enlarged me- dian spinule rows rather prominent, but ina large (female) Socotra specimen (CAS 66462) these enlarged spinules not present on those scales cur- rently remaining. Teeth small, in broad short bands in both jaws, none noticeably enlarged. In small Socotra spec- imens (ZMMGU P.19261), teeth bands about 6 or 7 rows wide in premaxillary, 5 or 6 rows wide in dentary; in larger specimen (CAS 66462), bands 8-10 rows wide in premaxillary, 6 or 7 in den- tary. Teeth bands of both jaws scarcely tapered posteriorly, fall short of end of rictus. Second spinous ray of first dorsal fin long, more than postrostral length of head; leading edge armed with slender, reclined, relatively wide- spaced denticles. Second dorsal rudimentary throughout; anal fin well developed. Pelvic fins well developed, outer ray stout, tapered to a pro- longed tip that extends well beyond anal fin or- igin to about 7th anal ray (in small Socotra spec- imen). Pyloric caeca of Socotra specimens relatively short and simple. Five caeca, longest about 5mm in male, CAS 66462; 7 caeca, 14-17 mm in fe- male, ZMMGU P.19261 from same trawl (count may have been higher in CAS 66462; a caecum or two may have been destroyed in attempt to make count). Testes of CAS 66462 relatively large; swim bladder with well-developed drum- ming muscles on anteroventral surface. Ovaries of ZMMGU P.19261 packed with spherical eggs, the largest 0.8—1.0 mm in greatest diameter; in- testines long, folded three times (with 6 major bends). Color in alcohol light grayish-brown, slightly more grayish over abdominal cavity; black over operculum, blackish over branchiostegal mem- branes. Fins dusky; tip of anal and spinuous sec- ond ray of first dorsal blackish. Mouth and gill cavities dark gray; peritoneum blackish. Lips and nostril membranes blackish. SizeE.—To more than 410 mm TL. DISTRIBUTION. — Northern Indian Ocean (Bay of Bengal and W. Arabian Sea) in 2,380- 2,820 m. REMARKS AND COMPARISONS.—Three speci- mens of C. hoskynii are deposited in the ZSI, including the holotype (ZSI 12833) taken in the Bay of Bengal in 2,396 m. The two others (ZSI cat. no. 1061 and 1057), trawled west of Sri Lan- ka and mislabeled ‘‘Stomias affinis,““ were dis- covered by one of us (YNS) during a visit to that museum in Dec. 1987. Alcock (1894:pl. 9, fig. 298 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 TABLE 3. Comparison of characters of four species of Coryphaenoides from the Indian Ocean: C. woodmasoni, C. hoskynii, C. asprellus and C. castaneus. Figures in parentheses refer to exceptional values for one specimen. woodmasoni hoskynii asprellus castaneus MEASUREMENTS (in percent head length): Snout 28-33 30-32 28-31 27-28 Preoral 13-19 16-20 13-15 12-14 Internasal 21-27 24-33 21-27 22-24 Interorbital 21-27 24-29 22-30 24-27(29) Orbit 22-27(29) 23-30 21-24 (18)20-23 Postorbital 46-52 44447 50-53 54-56 Rictus 22-31 21-24 ca. 26 — Upper jaw 30-37 27-30 33-38 38-42* Barbel 3-9 5-9 13-20 16-23 Posterior nostril 2-9 5-12 3-7 5-6 Outer V. ray 46-72 2-82 72-82 61-122 COUNTS: V. 8—9(10) 8-9 9-10 10 Scales below mid-1D. 4.5-7.0 3.54.5 4.5-5.5 4.5-6.0* Pyl. caeca 10-14 6-9 8-10 11 (1 spec.) OTHER: 3-4 teeth wide lat- erally; extends to Teeth: mand. broad band; falls short of end of narrow band; | row laterally extends narrow band; 3-4 teeth wide later- or beyond rictus rictus ally to or beyond ric- tus Underside of snout scaled scaled scaled naked 4) figured the body scales of C. hoskyniias having a prominent median spinule row, but in our spec- imens this row is not so noticeably enlarged. Ex- amination of the holotype revealed that Alcock’s figure was somewhat exaggerated in this regard, and in reality, the median spinule row on most scales were only slightly larger than those on the scales of most of our specimens, suggesting some individual variation. Coryphaenoides hoskynii is most similar to C. woodmasoni, the two are compared in the de- scription of the latter species. As with C. wood- masoni, C. hoskynii differs from C. asprellus in having a shorter barbel and somewhat smaller orbit diameter. Additional differences include a broader posterior nostril, shorter upper jaw, broader, shorter teeth bands, and somewhat larg- er scales (see Table 3 for comparison of these characters). Coryphaenoides macrolophus (Alcock, 1889) (Fig. 4) Macrurus macrolophus Alcock, 1889:394 (type locality An- daman Sea, SE by S of Ross Island; 265 fm [485 m]); 1891 in Wood—Mason and Alcock: 121 (2 spec., Bay of Bengal; 1892a:351, fig. 1 (Bay of Bengal 410 fm [750 m]); 1894a (in key): 126; 1894b: pl. 12, fig 1; 1899:115. Brauer 1906:266- 267 (23 spec., e. African coast, 628—1362 m) (in part?). We- ber 1913:159 (comparison of Macrurus tydemani with M. macrolophus). Macrurus lophotes Alcock, 1889:385 (type locality Bay of Ben- gal, in 285-405 fm [521-741 m]); 1894a (in key):126; 1892b: pl. 3, fig. 2; 1899:116. Brauer 1906:266 (considered M. /o- photes a synonym of Macrurus macrolophus Alcock). Coryphaenoides macrolophus: Gilbert and Hubbs 1916:144; 1920:417 (concurred with Brauer’s synonymizing of M. lo- photes with M. macrolophus). Coryphaenoides lophotes?: Norman 1939:51 (Gulf of Aden, 1,06 1—1,080 m; questionably identified small, 110 mm spec. as this species). REMARKS ON SYNONYMyY.— Brauer (1906:266) stated that the branchiostegal ray count was sev- en as a rule, not six, thus suggesting that another genus was intermixed with his specimens. This was noted by Gilbert and Hubbs (1920:417). Al- cock’s two syntypes of Macrurus lophotes were **.. far too much spoilt for complete descrip- tion” (Alcock 1889:385). Our examination of the specimens (TI in 1970, YIS in 1987) were in- conclusive because of their deteriorated condi- tion (fragments in a vial). The physiognomy of the fish as illustrated by Alcock (1894b: pl.3, fig. 2), its proportions and counts, and particularly the angular outline of the preopercle bone and associated ridge generally support the synonymy. D1acnosis.— V. 8, rarely 9; head completely scaled, without naked ventral margins on snout SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 299 Ficure 4. Coryphaenoides macrolophus (Alcock). CAS 66479 (263 mm TL) from off Socotra in 1,000—1,120 m. Fins and scales partially reconstructed. Scale bar equals 25 mm. and suborbital regions; first dorsal fin high, sec- ond spine 1.3-2.1 times HL, adjoining seg- mented ray longer than postrostral length of head; barbel small, less than posterior nostril; pyloric caeca 9-12; preopercle slightly produced poster- oventrally; mouth opening restricted, rictus ex- tends to below anterior margin of orbits; upper jaw less than '4 HL; outer pelvic ray filamentous at tip, extends beyond base of 8th or 9th anal ray. MATERIAL EXAMINED. — Holotype, ZSI 11776 (56.5 mm HL, about 230 mm TL); Andamans, 7 miles SE by S of Ross L., Port Blair. BMNH 1891.9.2.8 (50.1 HL, 222 TL); Bay of Ben- gal; Investigator collection received from Indian Museum. BMNH 1939.5.24.690 (25.4 HL, 110 TL); Gulf of Aden; John Murray st. 193. ZSI uncat. (5 spec., 47.2-67 HL, 210-280+ TL); locality not known, but probably Bay of Bengal. ZMMGU P.19262 (5, 49.5-70.3 HL, 209-321 TL) and CAS 66479 (3, 39.4-60.3 HL, 178-263 TL); off Socotra, 12°22.8'N, 53°02.7'E; 1,000-1,120 m, Vityaz cr. 17, st. 2565; 28.X.1988. ZMMGU P.19263 (2, 65.6-78.6 HL, 304-322+ TL); 13°11'N, 46°24’E; 600 m; Dmitry Stephanovy cr. 5, tr. 60. ZMMGU P.19264 (52.6 HL, 257+ TL); Timor Sea; 9°01.2'S, 130°59.5’E; 540 m; Aka- demik Berg. ZMMGU P.19265 (47.7 HL, 235+ TL); S. of Java; 8°39’S, 111°55'E; 560-600 m; Prof. Mesiatzev cr. 7, tr. 46. COUNTS AND MEASUREMENTS (see also Diag- nosis).—1D. IJ,9-11; 1P. 120-124; total GR-I (outer/inner).4—6/10-12, total GR-IT 11-12 / 9- 10; scales below 1D. 7.5—9, below mid-base 1D. 5.5-6.5, below 2D. 6-7, lat. line 33-36. Total length 178-321 mm, HL 35.4-70.3 mm. The following in percent of HL: snout 28-31; preoral (9) 12-14; internasal 1 7—26; interorbital 18-21; orbit 21-26; suborbital (10) 12-15; post- orbital 46-49 (51); orbit-preop. 41-44 (46); up- per jaw 28-31; rictus 19-23; post. nostril 8-12; barbel 7-8 (11); gill slit (8) 10-12; pre-1D. 109- 118; pre-A. 140-163; pre-V. 100-119; isthm.-A. 74-88; V.-A. 43-50; body depth 68-79; depth over A. origin 56-66; 1D.-2D. 21-40; height 1D. 129-213; 1P. 51-60; V. 54-74; DESCRIPTION.— Head rather low and broad, greatest width about equal to greatest height, about equal to postorbital length, HL about 4.2- 5.0 into TL; body tapers rapidly behind anal or- igin; greatest depth under | D. origin, about equal to postrostral length of head in 60 mm HL spec- imen. Snout bluntly angled when viewed dor- sally, broadly conical in lateral profile, extends beyond mouth a distance less than pupil diam- eter, tipped with a prominent conical tubercle with radiating spiny ridges; lateral angles with one or two smaller tubercles. Orbit less than snout length; interorbital space shallowly concave, 1.4— 1.5 into orbit diameter. Suborbital region angular in cross section, ventral half angled strongly me- dially. Upper jaws extend posteriorly to below anterior edge of pupil, mouth opening restricted posterolaterally, the rictus extends to below pos- terior nostril or anterior margin of orbit. Lips plicate anteriorly, somewhat papillaceous most 300 elsewhere. Barbel small, thin, about equal to or less than posterior nostril. Posterior margin of preopercle inclined forward, forming an acute angle with lower horizontal margin; in Socotra specimens, posteroventral corner not particular- ly extended or lobelike, but in Bay of Bengal specimen (BMNH 1891.9.2.8) corner produced into a prominent lobe (Fig. 4). Preopercle ridge forms a lobelike posteroventral corner. Intero- percle exposed as a fleshy narrow tab postero- ventrally. Gill membranes broadly attached to isthmus, the two sides not connected; no posterior free fold. Gill filaments long; gill rakers small, gen- erally tubercular. Outer gill slit small, usually less than least width of suborbital region; 3-6 rudi- mentary spiny rakers on outer side of first arch; 10-12 tubercular rakers on inner series, usually arranged (1-2) + (8-10). Fins well developed, second spinous ray of 1D. greatly prolonged, usually more than 1.5 of HL, anterior edge with low, closely set spinules prox- imally. First segmented ray longer than postros- tral length of head, succeeding rays rapidly short- er. Pectoral fins pointed, without prolonged rays, its length somewhat more than postorbital length of head. Outer pelvic ray produced into a hair- fine filament that extends well beyond anal fin origin to 8th or 9th anal ray. Scales densely covered with slender, greatly reclined, needlelike spinules that increase in length posteriorly on scale field, the tips of pos- teriormost spinules greatly overlap scale edge; spinules arranged in slightly convergent rows on most scales of body. Median nasal ridge and sub- orbital ridge with somewhat thickened, coarser scales; these thickened scales to a lesser degree developed along supranarial ridges and along dorsal edge of orbits. Head essentially complete- ly covered with scales; no ventral naked margins of snout and suborbital as in C. microps. Lower jaw rami lacking scales only at anteriormost one- quarter. Gular and branchiostegal membranes completely naked. Intestines long, with 4 bends (a double “S”’); stomach walls rather thin, 9, 9, 11, 11, and 12 short and slender pyloric caeca in five specimens. Swim bladder large, a pair of drumming muscles in males; two kidney-shaped gas glands, each connected to a short, broad rete mirabile in 60 mm HL Socotra specimen (CAS 66479); its stomach contained many copepods, polychaete PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 worm fragments, small shrimplike remains, and a tanaidacean. Color swarthy overall, fins all black, mouth dark gray, gill cavity black, lips and gums pale. SizE.—To more than 32 cm TL. DIsTRIBUTION. — Northern Indian Ocean, from off Socotra and the Gulf of Aden to the Bay of Bengal and the Andamans in 628-1,362 m. REMARKS AND COMPARISONS. — Coryphae- noides macrolophus belongs to a group of species that Gilbert and Hubbs (1920:413) characterized as having “‘a produced dorsal spine, a deep and sharply compressed body, and a dorsal contour horizontal behind the first dorsal fin.”” The group includes C. marginatus Gilbert and Hubbs, 1916, C. macrolophus, C. microps, C. semiscaber, and C. tydemani. The species share other characters such as a relatively small, restricted mouth open- ing; low, slightly protruding snout; small teeth in bands lacking a prominently enlarged series; small barbel; and highly restricted gill openings. Cor- yphaenoides macrolophus is readily distin- guished from C. tydemani, C. microps, and C. marginatus by its almost completely scaled un- derside of head (the other three species have ven- tral margin of snout, suborbital, and edge of mandibular rami naked), and the shape of its preopercle, which is strongly lobed posteroven- trally with the ridge forming a sharp, prolonged, posteroventral angle. The latter character also distinguishes the species from C. semiscaber, which, in addition, has a narrower internasal (15— 16% HL cf. 17-26%). Coryphaenoides microps (Smith and Radcliffe, 1912) Macrourus microps Smith and Radcliffe, 1912:410-414, fig. 6 (E of Masbate I., Philippines; 604 fm [1,105 m]). Coryphaenoides microps: Gilbert and Hubbs, 1920 We include a diagnosis and brief description of this species even though it is not known from the Indian Ocean, because it is closely similar to C. macrolophus and C. tydemani, and it may eventually be found in the area. Diacnosis. —A Coryphaenoides with upper jaw not extending beyond vertical through mid-or- bit, its length 30-33% HL; V. 7 or 8; length elon- gated spinous ray of first dorsal fin more than 2 times HL; ventral aspects of snout and suborbital mostly scaled, but with ventral margin naked; orbit diameter 17-25% HL, less than, to about SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 301 equal to interorbital width, much less than snout length; orbit-preop. distance 44-46% HL; free margin of preopercle broadly rounded, preoper- cle ridge forming a short, angular lobe at pos- teroventral corner; dorsal, pelvic, and anal fins generally pale basally, blackish distally. MATERIAL EXAMINED.—HOLOTYPE: USNM 72933 (82.0 mm HL, 413+ mm TL); Philippines, near Atulayan [.; 13°37'39”"N, 123°41'09”E; 560 fm [1,024 ml]; Albatross sta. 5470; 18.VI.1909. PARATYPES: CAS-SU 25446 (2, 23.7- 39.9 HL, 124+187+ TL); Philippines off northern Luzon; 18°34'15”"N, 121°51'15”E; 224 fm [410 m]; Albatross sta. 5325; 15.XI.1908. COUNTS AND MEASUREMENTS.—See Table 2. Other measurements in percent HL: length post. nostril 7-10; pre-A. 160-172; isthmus to A. 89- 97; body depth 68-79; 1D.-2D. interspace 22- 79; length 1P. 58-73; length V. 48-96. There were 14 slender pyloric caeca in a paratype. DESCRIPTION. — The description of C. macro- lophus closely applies also to this species. Cor- yphaenoides microps has a blunter snout and smaller orbits than do any other species of this group. Bluntness of snout most prominent in large holotype, where ventral profile low and almost vertical. Orbit proportionately smaller than in other species. In concert with small orbits, post- orbital and orbit-preop. distances long. Free margin of preopercle broadly rounded, but preo- percular ridge forms short but rather sharp lobe at posteroventral corner. Lips in holotype thick and papillaceous, but in smaller paratypes, lips more slender and smooth. Suborbital region no- tably broad in holotype and relatively flat in cross- section, but narrower and more angular in para- types, and more similar in those respects to other related species. Premaxillary teeth in broad bands, the out- ermost series slightly larger than others in ho- lotype, but essentially not enlarged in paratypes. Mandibular teeth in a broad band, tapering to a narrow band posteriorly. Scales like those in C. semiscaber, with many subparallel rows (about 18 in holotype) of short, slender, greatly reclined spinules, each row with 8-10 spinules. Underside of snout with moder- ately broad margin of naked skin; naked margin becoming narrower posteriorly on suborbital and ventral margin of preopercle. Mandibular rami have naked margins, broader anteriorly and merging with margins of suborbital and pre- opercle. Elongated spinous dorsal ray decreases pro- portionally in length with increase in size: in large holotype (82 mm HL), 1.9 times HL; in 40 mm HL and 24 mm HL paratypes, 2.9 and 3.9 times HL, respectively. Origins of first dorsal and pel- vic fins about on same vertical (or pelvics slightly anterior); pectoral origin anterior to that vertical. Color of specimens examined faded from long preservation, but recorded in original description (Radcliffe 1912:117) as: ‘““Ground color, sepia, scales with grayish cast; fins dark seal-brown.” DIsTRIBUTION.— Known only from the Phil- ippines, in 410-1,024 m. REMARKS.—Gilbert and Hubbs (1920:418- 419) provided some additional descriptive in- formation, as well as counts and measurements, for the type specimens of C. microps that were not given in the original description. This infor- mation allowed us to correct a blunder made presumably at the Stanford Natural History Mu- seum during the cataloging procedure, where two paratypes of the species were inadvertently switched with a paratype of C. semiscaber. (See Remarks in description of C. semiscaber.) Our measurement and count data agreed closely with Gilbert and Hubbs’s data and, fortunately, the specimen tags with the Stanford catalog numbers also corresponded with the correct species and not with the bottle labels. Coryphaenoides rudis Giinther, 1878 Coryphaenoides rudis Giinther, 1878:24 (Kermadec Islands). Gilbert and Hubbs, 1916:144 (“The largest specimen, the one figured [by Giinther, 1887, pl. 27], may be designated as the type... .”) Macrurus (Macrurus) rudis: Giinther, 1887:131 (description; additional information on type specimens). Nematonurus macrocephalus Maul, 1951:17 (Madeira). Macrourus paradoxus Smith and Radcliffe in Radcliffe, 1912: 115-116, pl. 25, fig. 1 (eastern Palawan, Philippines). Coryphaenoides paradoxus: Sazonov and Iwamoto, 1988:52- 53 (discussion of possible synonymy of C. rudis and C. par- adoxus). Dracnosis.—A large species (120+ cm) with 1D. II, 9-11; V. 9-10 (usually 10, rarely 8 or 11). Snout low, scarcely protruding; orbits small, about 1.5 into snout, 1.8—1.9 into interorbital width, 5.0-6.0 into HL. Suborbital region vertical; no strong sharp ridges on head. Mouth large, sub- terminal; upper jaw long, extends beyond ver- tical through hind margin of orbits, rictus ex- 302 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 tends about to under posterior '4 of orbit. Teeth prominent, an outer enlarged series on premax- illary behind which a narrow villiform band; mandibular teeth in about 3 irregular series near symphysis, becoming uniserial posteriorly. Head fully scaled except lips, gill membranes, and in- terorpercle. (After Sazonov and Iwamoto 1992.) MATERIAL EXAMINED. — LECTOTYPE: BMNH.1889.12.7.74 (male with large testes, 170 mm HL, about 840 mm TL); N of Kermadec Islands; 600 fm [1,097 m]; Challenger st. 171. Other specimens: ZMMGU P.19266 (196 HL, 935+ TL); West Australian Ridge, 28°32’S, 98°29’E; 1120 m; Prof. Mesiatzev cr. 7, tr. 42; 18.1V.1979. ZMMGU P.19267 (150 HL, 713 TL); West Australian Ridge, 31°33’S, 95°40’E; 1,400-1,420 m; Zvezda Kryma cr. 6, tr. 138; 19.1X.1976. AMS 29340-001 (144 HL, 710 TL); Australia, off Sydney; 33°45’S, 152°03’E; 1,120-1,170 m; FRV Kapala st. K89-13-02; 30.VI.1989. SIO 68-479 (255 HL, 1200 TL); 22°07'N, 171°36’E; 16.1X.1968. ZMMGU P.19268 (162+ HL, 775 TL); off Socotra, 12°43'N, 52°44’'E; 2,300—2,380 m; Vityaz’ cr. 17, st. 2832; 16.1.1989. COUNTS AND MEASUREMENTS. — LECTOTYPE: D. 11,9 + 92; 1P. 119/119; V. 10/10; GR-I (outer/ inner) ?- 1+9, GR-II 0+7/1 + 8; scales below 1D. 9, below 2D. 6.5, below mid-base 1D. 7.5, lat. line over pre-1D. length 37; pyloric caeca 14. Measurements in millimeters, percent of HL in parentheses: 170 mm HL, postrostral 131 (77); snout 44 (26); preoral 19 (11); internasal 33 (19); post. nostril 12 (7); interorbital 53 (31); orbit 30 (18); suborbital 25 (14); postorbital 102 (60); or- bit-preop. 91 (54); upper jaw 68 (40); barbel 30 (17); gill-slit 10 (6); preanal 270 (159); V.-A. 100 (59); isthm.-A. 177 (104); body depth 148 (87); 1D.-2D. 53 (31); height 1D. 120 (71); 1P. 95 (56); W275): SizE.—To at least 120 cm. DIsTRIBUTION.— Broadly distributed: Pacific (Kermadecs, Philippines; Darwin Seamount in central Pacific, Tasman Sea, Nazca Ridge); In- dian Ocean (off Socotra Island and West Aus- tralian Ridge); subtropical to tropical Atlantic, including the Gulf of Mexico and Caribbean Sea. REMARKS. — Gilbert and Hubbs (1916), in their list of grenadier species, designated the 33-inch specimen figured in Giinther (1887: pl. 27) as “the type.”’ Because Giinther (1878) referred to more than one specimen in his original descrip- tion, Gilbert and Hubbs’s type designation is for a lectotype, and the other specimens of the type series become paralectotypes. Our measurement and count data for the lectotype, most of which have not been previously recorded, are provided above. Descriptions of the species by Maul (1951, as Nematonurus macrocephalus), Marshall and Iwamoto (1973, as Coryphaenoides macrocepha- lus), and Iwamoto and Sazonov (1988, as C. par- adoxus) are more than adequate and need not be duplicated here. The illustration of the lec- totype (Giinther 1887: pl. 27) is exceptional in its accuracy and artistry. Three BMNH paralectotypes we examined were representives of two species of Nezumia. Other records of C. rudis by Brauer (1906) and McCann and McKnight (1980) represent other species. Sazonov and Iwamoto (1992:53) were reluc- tant to synonymize Macrourus paradoxus and Nematonurus macrocephalus with this species because of the absence of good comparative se- ries from throughout the distribution area, es- pecially of smaller individuals. Our experience with this genus, particularly over the past few years, has demonstrated the sorts of variation one might expect with a large, widely distributed species. Our specimens appear to confirm in C. rudis these kinds of variation. Coryphaenoides semiscaber Gilbert and Hubbs, 1920 (Fig. 5) Coryphaenoides semiscaber Gilbert and Hubbs, 1920:410-414, fig. 6 (E of Masbate I., Philippines; 604 fm [1,105 m]) We include a diagnosis and brief description of this species even though it is extralimital, be- cause of its close similarity to C. macrolophus, C. microps, and C. tydemani, and because it eventually may be caught in the Indian Ocean. Diacnosis. — A Coryphaenoides with upper jaw not extending beyond vertical through mid-or- bit, its length 29-35% HL; V. 8; spinous dorsal ray elongated, 119-144% HL; ventral aspects of snout and suborbital completely scaled; orbit di- ameter 26-28% HL; orbit-preop. distance 36- 39% HL; free margin of preopercle broadly rounded, preopercle ridge not forming a sharp, angular lobe at posteroventral corner. MATERIAL EXAMINED.— HOLOTYPE: USNM 83625 (48.3 mm HL, 251 mm TL); E of Masbate I., Philippines; 12°31'30’N, 123°35'24”E; 604 fm [1,105 m]; A/batross sta. 5215. PARA- TYPES: USNM 148882 (41.8 HL, 213 TL); same data as for holotype. USNM 148883 (45.4 HL, 200+ TL); off E coast Mindoro, Philippines; 12°52'N, 121°48'30°E; 218 fm [399 m]; Albatross sta. 5124. The small, 93+ mm paratype (CAS-SU 25452) from Al/- batross sta. 5534 represents another species (see Remarks). SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES FiGure 5. Island, in 1,105 m. Scale bar equals 25 mm. COUNTS AND MEASUREMENTS.—See Table 2. Other measurements in percent HL: length post. nostril 7-10; pre-A. 160-172; isthmus to A. 93- 97; body depth 68-79; 1D.-2D. interspace 44— 79; length 1P. 60-73; length V. 95-96. COMPARATIVE DESCRIPTION.—The descrip- tions for C. macrolophus and C. microps ade- quately apply also to this species. Coryphae- noides semiscaber has a somewhat blunter snout profile than does C. macrolophus and is similar in that respect to C. tydemani and smaller in- dividuals of C. microps. The elongated spinous ray of the first dorsal fin in C. semiscaber is dis- tinctly shorter and more heavily serrated than in the other three species. The completely scaled underside of the snout and suborbital region are like that in C. macrolophus. The preopercle bone is broadly rounded posteriorly with the chord of the posterior margin vertical; the preopercular ridge has a small, rounded, posteroventral angle (Fig. 5). The shape of the bone and associated ridge is similar to that in C. tydemani and C. microps and contrasts sharply with that in C. macrolophus, where the posterior margin of the preopercle is inclined forward to form a distinct lobe or acute angle at the posteroventral corner, 303 XLS ex Coryphaenoides semiscaber Gilbert and Hubbs. After Gilbert and Hubbs (1920: fig. 6). Philippines, east of Masbate and the vertical and horizontal ridges meet at a sharp angle or acute lobe. Color of the specimens after long preservation is medium brown, with fins darker except anteriorly at the base of the anal and the base of the pelvic fins, where the fins are somewhat paler. Unlike C. microps and C. tydemani, the species lacks a distinctly black- ish tip of first dorsal. Gilbert and Hubbs (1920: 412) counted six and eight pyloric caeca in two of the paratypes, counts that are lower than the 14 we obtained from a paratype of C. microps. DIsTRIBUTION.—The species appears to be confined to the Philippines region, but we suspect that more thorough sampling in the seas off the Malayan Archipelago will reveal its presence over a wider area. REMARKS ON Type MATERIAL.—In their orig- inal description, Gilbert and Hubbs (1920:410) provided a table of counts and measurements for three paratypes—the two listed above in the Ma- terials Examined section, and a third specimen from Albatross sta. 5534. That third paratype, cataloged as CAS-SU 25452, was apparently switched with two paratypes of C. microps, CAS- SU 25446. Our measurements and counts of these switched specimens correspond almost perfectly 304 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 Ficure 6. Coryphaenoides tydemani (Weber). CAS-SU 25440 (263 mm TL) from Buton Straits near Celebes in 1,600-1,700 m. Fins and scales partially reconstructed. Scale bar equals 25 mm. with the data provided for the two species by Gilbert and Hubbs (1920). Metal tags attached to the specimens also confirm the inadvertent switch and subsequent mislabeling of the spec- imen bottles. The specimens are now reunited with their proper identification and jar labels. The small paratype, CAS-SU 25452, appears to represent a different species from C. semis- caber in having an acute, lobelike preopercle an- gle, with a preopercle ridge that is prolonged and pointed, as in C. macrolophus. The small size and mediocre condition of the specimen, and the lack of adequate comparative material, leave its proper identity undetermined. Coryphaenoides tydemani (Weber, 1913) (Fig. 6) Macrurus tydemani Weber, 1913:158, pl. 1, fig. 6 (syntypes from Bali Sea, Flores Sea, Ceram Sea, and Arafura Sea; 538— 1,310 m). Coryphaenoides tydemani: Gilbert and Hubbs 1920:414—417 (Buton Strait near Celebes; 559 fms [1,022 m]). Weber and de Beaufort 1929:31-32, fig. 6 (compiled). DiaGnosis.— A Coryphaenoides with upper jaw not extending beyond vertical through mid-or- bit, its length 29-34% HL; V. 8 (rarely 9); length elongated spinous ray of first dorsal fin more than 2.5 times HL; ventral aspects of snout and sub- orbital almost completely scaled (a thin, naked ventral margin only); orbit diameter 23-26% HL, slightly more than interorbital width, but less than snout length; orbit-preop. distance 38-41% HL; free margin of preopercle broadly rounded, preopercle ridge not forming a sharp, angular lobe at posteroventral corner; dorsal, pelvic, and anal fins generally pale basally, blackish distally (some individuals may have overall dark fins). MATERIAL EXAMINED.—LECTOTYPE (here designated): ZMA 111.493 (34.4 mm HL, 160+ mm TL); Flores Sea, 7°36.0'S, 117°30.8'’E; 694 m; Siboga sta. 314; 17.11.1900. PARALECTOTYPES: ZMA uncat. (37.0 HL, 174+ TL), same data as for lectotype; ZMA 111.491 (53.1 HL, 265 TL); Ceram Sea, 3°37.7'S, 131°26.4’E; 324 m; Siboga sta. 170; 26. VIII.1889. Other specimens: CAS-SU 25440 (2:48.9-54.9 HL, 263-265 TL) and USNM 99475 (2:52.4-58.8 HL, 258+288+ TL); Bu- ton Straits near Celebes, 5°35’S, 122°20’E; 559 fm [1,022 ml]; Albatross sta. 5648; 16.XII.1909. COUNTS AND MEASUREMENTS (see also Diag- nosis and Table 2).—The following in percent HL: post. nostril 8—11; pre-A. 163-181; isthmus to A. 81-96; body depth 78-95; 1D.-2D. inter- space 16-19; length 1P. 60-71; length V. 61-85. DESCRIPTION (primarily of lectotype, supple- mented with comments on other specimens). — Head more than 4.5 of TL (tail tip missing), SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES 305 slightly longer than greatest body depth, its great- est width over opercles about equal to distance snout tip to hind rim of orbit. Body behind ab- dominal cavity greatly compressed and deep, depth over anal origin about equal to postrostral length of head. Snout pointed, slightly projecting beyond mouth, tipped with a prominent median spinous tubercle; no similar lateral tubercles de- veloped. Orbit small, about 4 in HL, less than snout length but more than least interorbital width. Mouth small, restricted laterally, poste- rior edge of maxilla below anterior edge of pupil, posterior extent of rictus below posterior naris. A small, but rather stout-based, chin barbel. Free margin of preopercle broadly rounded, the low preopercular ridge forming a slight lobe at pos- teroventral corner. Tip of interopercle scarcely exposed beyond preopercle margin. Gill membranes broadly attached to isthmus at level approximately under opercle (difficult to gauge because of expanded gill cover and bran- chiostegal rays). Gill filaments rather short, about as long as diameter posterior nostril; rakers tu- bercular, armed with several recurved needlelike spinules; first gill slit greatly restricted. Paired fins well developed; first dorsal fin with short, steep base; first spinous ray spikelike, sec- ond greatly elongated, rather stout basally with 14 reclined, spaced denticles along leading edge, the ray becoming filamentous distally, extending most of distance to end of tail. Outer pelvic ray filamentous, extends to base of 4th or 5th anal ray. Second dorsal fin rudimentary throughout, but anal well developed. Pectoral fin origin an- terior to pelvic origins. Scales densely covered with long, slender, nee- dlelike, greatly reclined and overlapping spi- nules, arranged in 9—12 more or less parallel rows (smaller paralectotype from same locality had 8— 10 rows, larger of CAS-SU 25440 had as many as 16), the tips of posteriormost spinules over- lapping posterior edge of scale. Spinules generally so small that no overall striated pattern pro- duced. Head essentially completely covered with scales except lips, nasal, branchiostegal, and gu- lar membranes. A thin margin along ventral edge of snout and suborbital naked, but scarcely no- ticeable in this small specimen. In larger para- lectotype (ZMA 111.491) and in both from CAS- SU 25440, this naked margin more prominent. Scales and scale spinules along suborbital and nasal ridges slightly thickened, but not promi- nently so. Teeth in narrow bands in both jaws, about 4 or 5 teeth wide on mandible and premaxillary. Bands appear slightly wider in ZMA 111.491. Color in alcohol overall light brown to grayish. Branchial and opercular regions and area over abdominal cavity blackish. Lips blackish. First dorsal, pectoral, and pelvic fins blackish distally, but rather pale at base. Anal mostly pale ante- riorly, but distal margin black anteriorly, the fin becoming overall blackish toward posterior end of tail. (In paralectotype ZMA 111.491, all fins completely dark.) Gill cavity blackish, mouth grayish. Chin barbel pale except dark at base. SizeE.—Attains at least 27 cm. DIsTRIBUTION.—Apparently confined to In- donesian waters in 324—1,022 m. REMARKS.— We examined the three syntypes of C. tydemani through the courtesy of I.J.H. Isbriicker of ZMA. We have chosen one of these (ZMA 111.493, 160 mm TL) from the Flores Sea as lectotype, as it is in relatively good condition and agrees well with the original description of the species. Paralectotype ZMA 111.491 from Ceram Sea appeared somewhat different from the other two specimens in having a blunter snout, broader head, and completely dark fins without pale bases. The first two differences may be at- tributable to distortions caused by preservation. The fin-color difference may be individual vari- ation. We also examined two specimens collect- ed by the Albatross off Buton Straits in 1,022 m. These agree in essentially all respects with the type specimens. We compared these specimens of C. tydemani with the holotype (USNM 72933) and two para- types (CAS-SU 27452) of C. microps. They were similar in most features, notably in having an exceptionally long spinous ray in the first dorsal (190, 293 and 388% HL in C. microps), a thin naked ventral margin along the snout and sub- orbital, and a black-margined anal fin. The two smaller (CAS-SU) paratypes of C. microps had a black-tipped first dorsal fin, but the large ho- lotype had a uniformly dark first dorsal. Small differences in length of postorbital (49-56% HL in C. microps vs. 45-49% HL in C. tydemani) and orbit-preopercle distance (44-46% vs. 38- 41%) were the only proportional measurements of note. The orbit diameter may be smaller in C. microps—the large holotype and larger para- type from CAS-SU 25452 had measurements of 17% and 21% HL, compared with 23-26% in the five C. tydemani we examined. The smaller para- type of C. microps from CAS-SU 25452 had an orbit measurement of 23% HL, but relative orbit PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 306 WW +06¢ Woy paronnsuosas ApUeg “WI OZL‘I-OZS‘T Ul o8pry ouareose| ‘YURg EYL ap BALE Yo woy (TL wu +0¢¢) O£99 § ee <5 i ee Ae “Wu ¢Z sfenba req afedg “UOTda][09 auIes ‘UsUIIOads 7]. VO ‘poopy) mospuipoom sapiouanyddio) *1, AANOLA SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS CORYPHAENOIDES size in most species usually decreases with age, which would be in agreement with the small size of that specimen. Examination of adequate size series of both species are necessary to confirm these differences. It appears that C. tydemani and C. microps are so similar that additional material from through- out the Indo-Malayan and Philippines regions may eventually show them to be conspecific, and the minor differences we have observed are sim- ply populational or size-related. Because such material is not now available to us, we continue to recognize the two as distinct species. Coryphaenoides woodmasoni (Alcock, 1890) (Fig. 7) Macrurus Wood-Masoni Alcock, 1890b:301 (Arabian Sea; 1,000 fms [1,829 m]). Alcock in Wood-Mason and Alcock 1891: 121 (Gulf of Manaar, “‘a male nearly 18 inches long”; Jn- vestigator sta. 109, 738 fms [1,350 m]). Alcock 1892:363 (Laccadive Sea, 559 and 902 fm [1,022 and 1,650 m]); 1894a: 12 (in key); 1894b: pl. 13, fig. 1); 1899:114-115. D1aGcnosis. — V. 8-9 (rarely 10); underside of snout and mandibles completely scaled; upper jaws 30-35% HL, premaxilla extends to slightly beyond midorbital level, rictus not restricted lat- erally; barbel short, 3—9% HL; pyloric caeca 10- 14; teeth in moderately wide bands 34 teeth wide laterally, outer premaxillary slightly en- larged. MATERIAL EXAMINED.—Laccadive Sea: HOLOTYPE: ZSI 12867 (108 mm HL, about 480 mm TL); 11°12'47’N, 74°25'30"E; 1,829 m; Investigator sta. 104. Gulf of Manaar?: BMNH 1892.6.17.3 (94.4 HL, 435 TL) (accession from ZSI, probably specimen reported by Wood-Mason and Alcock 1891). Ninety East Ridge: ZMMGU P.19269 (77.7 HL, 387 TL); 11°44’S, 88°47’E; 1,600 m; Fiolent cr. 9, sta. 24; 14.VI.1977. Mascarene Ridge, Saya de Malha Bank: ZMMGU P.19270 (3, 48.2-72.5 HL, 214-349 TL); 9°48.5’'S, 60°42’E; 1,650 m; Fiolent cr. 9, sta. 51; 1.X1.1977. ZMMGU P.19271 (3, 86.1- 98.0 HL, 415-452+ TL) and CAS 71480 (85.4 HL, 447 TL); 8°26’S, 59°29’E; 1,300-1,260 m; Fiolent cr. 9, sta. 52; 2.XI.1977. ZMMGU P.19272 (2, 73.2-90.9 mm HL, 368-437+ mm TL); 8°07'S, 59°18’E; 1,300-1,240 m; Fiolent cr. 7(9), tr. 53; 3.1X.1977. ZMMGU P.19273 (2, 78.2-93.2 HL, 384-446 TL); 8°08’S, 59°37'E; 1,247-1,259 m; Zvezda Kryma cr. 6, tr. 210; 3.X1.1976. ZMMGU P.19274 (10, 76.4-114 HL, 367-494+ TL) and CAS 66430 (5, 75.8-119 HL, 350+525+ TL); 09°40’S, 60°31'E; 1,520-1,720 m; Vityaz’cr. 17, st. 2815; 9.1.1989. CAS 66475 (81.0 HL, 390+ TL); 8°32’S, 59°14’E; 960-1,130 m; Vityaz’ cr. 17, st. 2820; 10.1.1989. No specific locality: ZSI F361/1, 358-60/1 (4, 65.5-95.3 HL, 240+387+ TL). COUNTS AND MEASUREMENTS (see also Table 3 and Diagnosis).—D. II,9-10 + 100-115; 1P. 307 i19-i24; total GR-I (outer/inner) usually 3-4/ (7)8-9, total GR-II 7—8(9)/(7)8-9; scales below 1D. usually 7.5—9.0, below mid-1D. 4.5-7.0, be- low 2D. 5.5-8.5 (usually 6.0—7.5), lat. line 39- 46. Total length 214-525 mm; HL 48.2-119 mm. The following in percent HL: postrostral 69-75; snout 28-33; preoral 13-19; internasal 21-29; interorbital 21-27; orbit 22-27; suborbital 13- 16; postorb. 46-52; orbit-preop. 45-50; gill slit 5-8; post. nostril 2-10; pre-A. 146-174; isth.-A. 70-9 1: body depth 68-84; 1D.-2D. 14-41; height 1D. 57-79; length V. 46-72. DESCRIPTION (based primarily on Mascarene Ridge specimens).—Head about 4.5-5 in TL; greatest body depth about equal to postrostral length of head. Snout pointed, broadly conical in lateral view, protruding less than lens diameter beyond mouth. Orbit shape circular to somewhat oval, about 4 into head, about equal to or slightly more than interorbital width. Interorbital space almost flat. Suborbital region with a distinct shelf, strengthened by 3 or 4 rows of slightly thickened scales. Preopercle margin broadly rounded, slightly lobate, the inner ridge forming a strong club-shaped process at lower angle. Jaws subter- minal, premaxillary extends to about mid-orbit, rictus unrestricted, extends to below anterior of orbit. Barbel small, thin, length about one- half of least suborbital width. Cephalic-sensory pores little developed, free neuromasts rather conspicuous in most specimens, evident over snout, suborbital, interorbital, and on surfaces of nape, occiput, and operculum. Gill membranes broadly attached to isthmus, without a free posterior fold; gill openings extend forward to below vertical portion of preopercular ridge. Outer gill slit strongly restricted by folds of skin, length of slit usually less than barbel length, about equal to length of gill filaments. Gill rakers rudimentary and few along outer side of first arch, tubercular or platelike on inner side of arch. Head and body surfaces covered with mod- erate-sized scales, those over head ridges slightly (if at all) thickened, only those on suborbital and supraorbital ridges heavier; terminal and lateral angles of snout tipped with small tubercular scales. Underside of head almost entirely covered with small scales except lips, gill membranes, and small area immediately above symphysis of upper jaws. Body scales densely covered with small, spike- like, recumbent spinules, generally aligned in 308 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 Ficure 8. Coryphaenoides sp. 1. CAS 71488 (573+ mm TL) from off Chatham Rise, east of New Zealand in 1,200 m. Fins and scales partially reconstructed. Scale bar equals 25 mm. a — SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS COR YPHAENOIDES slightly convergent rows, the median row in each scale somewhat more elevated than adjacent rows. Teeth all small and fine except for slightly en- larged outer series on upper jaw flanking a broad inner band of small teeth; in one specimen (CAS 71480), teeth in inner band arranged in some- what diagonal rows, but teeth more uniformly scattered in others. Mandibular teeth uniformly small and in a narrow band 2 or 3 rows wide laterally, but band wider at symphysis. Fin origins best seen in Figure 7. Outer pelvic ray weak, hair thin distally, extending at most only past bases of first few anal rays. Second spinous ray of first dorsal serrated along leading edge with small non-overlapping, reclined, spikelike teeth. Pyloric caeca (of 12 specimens examined) well developed, 10 to 14 in distal counts, lengths from about half orbit diameter to more than orbit di- ameter. Intestine long; coiling pattern with 4 or 6 bends. Color in alcohol light brown overall, ventral aspects of head, operculum, and trunk darker, somewhat swarthy. Mouth, gill chamber, bran- chiostegal membrane, and lips black; gular mem- brane dark gray. Orbital rim and all fins except second dorsal blackish; barbel brown basally, but pale distally. SizE.—To more than 53 cm TL. DISTRIBUTION. — Northern portion of Indian Ocean off Indian subcontinent and on Mascarene and Ninety East ridges, in 1,240-1,829 m. The species is fairly abundant at these mid-slope depths, which lie above the depths at which C. hoskynii has been found (2,380-2,820 m). REMARKS AND COMPARISONS. — Coryphae- noides woodmasoni is closely similar in several important features to C. hoskynii (Alcock, 1890), including squamation, dentition, a small barbel, eight or nine pelvic fin rays, a broadly conical snout, and fin pattern. Major differences lie in the count of pyloric caeca, width interorbital, length posterior nostril, length upper jaw, length rictus, and number of scale rows below midbase of first dorsal fin. These features are compared in Table 3. Coryphaenoides woodmasoni may also be con- fused with C. castaneus, but the naked underside of snout and the longer barbel (13-20% HL) of that species is distinguishing. Other features are compared in Table 3. 309 Coryphaenoides sp. | (Fig. 8) Coryphaenoides (Coryphaenoides) sp. indet. Iwamoto 1986: 355 (2 spec., Agulhas Bay and off Mozambique; 696-960 m). Coryphaenoides (Coryphaenoides) rudis (nec Giinther, 1887): McCann and McKnight, 1980:35, figs. 16, 17, 18 (up. rt.) (1 spec., off Cape Palliser, New Zealand, 1,205—-1,293 m). Coryphaenoides sp.: Okamura in Amaoka et al. 1990:184, fig. 126 (6 spec., New Zealand, 100-763 m). Coryphaenoides sp. B long barbel rattail: Paulin et al. 1989: 129, fig. 56.30b (in key). D1aGcnosis.—V. 8 (9); naked surfaces on un- derside of head confined to a median swath be- low snout tip, a narrow strip along ventral mar- gins of snout and suborbital, and anterior end of mandible; snout low, barely protruding in adults; mouth large, upper jaw extends to end of orbit; chin barbel long, usually about 4 to 4 HL; outer GR-1 relatively well developed, lappet-shaped; a distinct hump in profile of nape. MATERIAL ExAMINED.— New Zealand area: ISH 4009/79 (157 mm HL, 715+ mm TL); 48°46’S, 172°33’E, 705-695 m; Wes- ermunde sta 411 (157)/79. ISH 4046/79 (126 HL, 515+ TL); 42°50'S, 176°45'W, 900-913 m; Wesermunde sta. 557 (137)/ 79: 6.VII.1979. ISH 4041/79 (170 HL, 770+ TL); 42°53’S, 176°07'W, 895-905 m; Wesermunde sta. 154/79; 10.VII.1979. ZMMGU P.19275 (114 HL, 533 TL) and CAS 71488 (122 HL, 573+ TL); 44°04’S, 178°19’E, 1,020 m; Poseidon trawl 75; 3.11.1978. Tasman Sea (off New South Wales, Australia): AMS uncat (5, 30.4-54.0 HL, 151+ 283+ TL); E of Broken Bay; FRV Kapala; 1979. AMS 1.17866-007 (61.7 HL, 280+ TL); off Sydney, 33°57'S, 151°45’E, 720 m; FRV Kapala; 6-7.X1.1977. AMS I.17867-007 (2:51.6-64.3 HL, 210+280+ TL); off Syd- ney, 33°41'S, 151°57’E, 720 m; FRV Kapala; 9.X1.1972. AMS 1.20452-018 (46.6 HL, 225+ TL); off Broken Bay, 33°38’S, 152°04’E, 822 m; FRV Kapala; 19.VIII.1975. AMS 1.24059- 016 (2, 142-154 HL, 690+730 TL); off Norah Head, 33°32’S, 152°09’E, 942-978: FRV Kapala st. 83-09-02; 23.VIII.1983. AMS 1.20477-002 (88 HL, 355 TL); SE of Newcastle, 33°11’S, 152°24'E, 732 m; FRV Kapala; 7.X11.1977. AMS 1.20485-006 (59.7 HL, 285+ TL); SE of Broken Bay, 33°40’S, 151°56’E, 731 m; FRV Kapala; 1977. AMS 1.21724-006 (48.9 HL, 233+ TL); E of Broken Bay, 33°37'S, 152°07’E, 1005 m; FRV Ka- pala; 1979. AMS 1.24613-005 (43.3 HL, 343 TL); off Broken Bay, 33°32’'S, 152°04’E, 823 m; FRV Kapala st. 75-05-05; 19.VIII.1975. AMS 1.24771-002 (470 TL); off Broken Bay, 33°43’S, 152°03’E, no depth; FRV Kapala st. 84-10-08; 1984. AMS 1.24992-002 (66.2 HL, 340+ TL); Sydney, Gabo Is., 35°29’S, 150°56’E, 1,116m; FRV Kapala; 1984. AMS1.25056- 004 (2, 127-149 HL, 620+730+ TL). AMS 1I.28717-003 (90 HL, 383 TL); E of Newcastle, 32°53'S, 152°47’'E, no depth; FRV Kapala; 1988. Coral Sea: ZMMGU P.19276 (2, 143-148 HL, 635-610+ TL); 25°07'S, 162°51’E, 1,100-1,130 m; Mis Tikhi; IV. 1978. Indian Ocean: CAS 50151 (36 HL, 130+ TL); off Mozam- bique, 22°30'S, 36°09’E, 960 m; Anton Bruun cr. 8, sta. 399B; 1.X. 1964. SAM 27707 (1 spec.); off Natal, South Africa, 310 27°17.5'S, 32°54.1'E; 780-720 m; RV Meiring Naude, st. SM 66; 20.V.1976. SAM 31568 (1 spec.); 27904.6'S, 33°04.0'E; 680 m; RV Benguela, st. G13493/27-08; 12. VIII.1988. SAM 27711 (1 spec.); off Natal, 28°09.8’S, 32°47.4’E; 940 m; RV Meiring Naude, st. SM 90; 23.V.1976. SAM 27712 (1 spec.), off Natal, 28°37.8'S, 32°38.4'E; 1,200-1,000 m; RV Meiring Naude, st. SM 107; 25.V.1976. SAM 32735 (2 spec.); off Natal, 28°21.9’S, 32°34.6'E; 775-825 m; RV Meiring Naude, st. SM 38; 28.V.1975. ZMMGU P.19277 (111 HL, 470+ TL); Mada- gascar Ridge, 29°59’S, 45°57’E, 1,285-1,360 m; Zvezda Kryma cr. 6, tr. 15; 2. VII.1976. ZMMGU P-16027 (70 HL, 363 TL); Madagascar Ridge, 32°25’S, 35°04’E, 1,230-1,260 m. ZMMGU P.19278 (158 HL, 735+ TL); Madagascar Ridge, 32°28.0’S, 43°29.3'E, 1,060 m; Prof. Mesiatzev cr. 7, tr. 136; 24.V1.1979. ZMMGU P.19279 (134 HL, 620+ TL); Madagascar Ridge, 33°30'S, 44°00’E, 940-1,100 m; Zvezda Kryma cr. 6, tr. 20; 7.VII.1976. ZMMGU P.19280 (138 HL, 665 TL); Madagascar Ridge, 32°31.6'S, 43°36.0’E, 1,050 m; Heroevka cr. 1, tr. 206; 12.VI.1980. Atlantic Ocean: ZMMGU P.19281 (93 HL, 432 TL); Gulf of Guinea off Congo R. mouth, 6°29’S, 11°13’E, 1,000 m; Fiolent tr. 99; 1.111.1976. SAM 29026 (1 spec.); off Namibia, 19°59'S, 11°27.4'E; 805 m; st. 804-6. CAS 71481 (93 HL, 460 TL); South Africa, 27°03’S, 13°55’E, 1,000 m; Poltava tr. 487; 5.11.1970. ZMMGU P.19282 (57 HL, 265+ TL); South Africa, 31°47'S, 15°27’E, 1,500-1,600 m; Fiolent cr. 4, tr. 70; 18.111.1974. ZMMGU P.19283 (129 HL, 585+ TL); South Africa, 33°40’S, 17°16’E, 900 m; Poltava tr. 429; 1.1970. ZMMGU P-15643 (109 HL); Walvis Ridge, 33°17.7’S, 2°14.3’E, 928-1,115 m; Prof. Mesiatsev cr. 8, tr. 22. ZMMGU P-15690 (67 HL, 375+ TL); Walvis Ridge, 31°54’S, 2°06’E, 940-990 m; Prof. Mesiatsev cr. 8, tr. 24; 3.1X.1979. SAM 32937 (4 spec.); 34°55.5’S, 18°12.8’E; 894 m; RV Africana cr. 60, st. A7024,; 12.111.1988. SAM 32938 (1 spec.); 34°53.6’S, 18°13.1’E; 901 m; RV Africana cr. 60, st. A7027; 13.111.1988.SAM 32946 (2 spec.); 33°18.6'S, 17°28.8'E; 480 m; RV Africana cr. 60, st. A6990; 5.111.1988. SAM 32947 (2 spec.); 34°53.6'S, 18°13.1'E; 901m; RV Africana cr. 60, st. A7027; 13.11T.1988. SAM 32948 (1 spec.); 34°55.6’S, 18°11.7’E; 903 m; RV Africana cr. 60, st. A7037; 14.1II.1988. SAM 32960 (1 spec.); 26°43.6’S, 13.36.0’E; 900 m; RV Africana cr. 69, st. A8413; 21.1.1989. SAM 32937 (4 spec.); 34°55.5'S, 18°12.8'E; 894 m; RV Africana cr. 60, st. A7024; 12.111.1988. RUSI 25768 (156 HL, 740+ TL); off Cape Point, 33°25.8'S, 17°02.1'E; 826 m; 14.1.1986. COUNTS AND MEASUREMENTS.—D. II,9-10 + 95-115; 1P. 117-121; V. 8 (rarely 7 or 9); total GR-I (outer/inner) 7—9/11—13, GR-II 9-12/11- 13; scales below 1D. 9.5-11, below 2D. 7-12., below mid-1D. 7-10, lat.l. 38-48 Total length 265-770 mm, HL 57-170 mm. The following in percent HL: snout 24.5-29; preoral 7-11 (15); interorbital (14) 16-20; orbit (bony) 19-23; suborbital (fleshy) 9-13; postor- bital about 48-57; orbit to preopercle 44-50; up- per jaw 40-45; barbel 21-34; outer gill slit 17- 24; pre-A. 143-179; pre-1D. 118-129; isthmus to A. 82-111; body depth 62-89; 1D.-2D. in- terspace 21-40; height 1D. 60-84; length V. 43- 70. DESCRIPTION. — Head long, shallow, 4.2—5.0 in PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 TL (5.6 in one spec.), width 2 or less in HL; body more slender in young, becoming deep in mature adults; an elevated nape, very pronounced in large individuals (greater than about 130 mm HL) giv- ing severe humpbacked appearance. Orbit small, 4-5 in HL; in larger adults (e.g., 93 mm HL from Gulf of Guinea, ZMMGU P.19281) anteroven- tral margin fleshy, especially so in larger speci- mens, where upper edge of suborbital difficult to determine. Snout low, bluntly pointed and slight- ly protruding beyond mouth in small specimens, but in larger specimens snout lower, less pro- truding (e.g., 93 mm HL specimen) or virtually nonprotruding (e.g., 138 mm HL specimen from the Madagascar Ridge, ZMMGU P.19280). In- terorbital space flat or slightly convex; width less than orbit diameter. Mouth large, rictus extends posteriorly to below hind '4 of orbit; maxillary notably wide, posterior edge about equal to least suborbital width, and extends to below hind edge of orbit. Barbel long, slender; length varies from about 4 to 4 HL. Suborbital region lacking a heavily scuted ridge; ridge low, poorly devel- oped, especially in largest specimens (> 130mm HL); upper shelf area broader than that below ridgeline. Sensory pores generally large and well developed along ventral surfaces of head, es- pecially along ventral margin of snout, subor- bital, preopercle, and mandibles. Preopercle with somewhat prolonged posteroventral margin, forming a shallow lobe. Interopercle broad, but almost entirely hidden behind preopercle, ex- posed posterior tip somewhat flaplike. Gill opening wide, extending forward ventrally about to posterior edge of maxillary. Gill fila- ments long; outer rakers of first arch well devel- oped, characterically (for subgenus) flaplike, shaped somewhat like a flattened spiny-headed club, usually one at or above angle; other rakers of more tubercular form. Scales generally cover most of head and body. Naked areas limited on head to gill membranes, lips, median swath below snout tip in most spec- imens, ventral edge of suborbital and preopercle in some individuals, and anterior part of man- dible. In a specimen 635+ mm from the Coral Sea (ZMMGU P.19276), virtually entire under- side of snout, suborbital, preopercle, and man- dible scaled; also a few small scales at tip of interopercle, but this specimen an exception. Scales on underside of head small; those on sub- orbital also small, but a file of small, thickened scales aligned above ridge; these, however, not SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS COR YPHAENOIDES 311 forming a longitudinal ridge. Scales dorsally on head have spinules aligned in longitudinal rows, which give surface a somewhat striated texture. Spinules on body scales numerous and closely packed, individual spinules greatly reclined, small, fine, aligned in either discrete subparallel or slightly convergent rows. Premaxillary teeth with a broad band of small inner teeth and a single series of moderately en- larged and widely spaced, conical teeth forming outer boundary. Outer teeth rather broad tipped. Lower jaws with teeth clustered in a band an- teriorly, but narrowing laterally to 3 or 4 irregular series; teeth larger along outer edge and similar in shape to outer premaxillary series but smaller. Paired fins moderately developed; in largest specimens, barely or not at all extending past level of anus. Pelvic and pectoral origins about on same vertical, first dorsal behind this. Anal origin behind vertical through hind margin of first dorsal a distance about half length of first dorsal base. Interspace between first and second dorsal fins variable, somewhat greater than length first dorsal base in some, less than base in others. Second dorsal rudimentary over most of length, rays becoming well developed only near poste- rior quarter or so. Pyloric caeca long, twice internasal width, bifid branching at base in some; distal count 10-18. Retia and gas gland numbers in one Mozam- bique Plateau specimen 5 or 6 (somewhat de- teriorated); retia long, slender, convolutely coiled. Color in alcohol in large Coral Sea and New Zealand specimens light brown, overall, lips, jaws, and underside of head generally pale. Mouth and gill cavities dark gray or black (including rakers and arches, but not filaments). All but dorsal part of orbital rim blackish. Paired fins dusky. Distal margin of anal fin blackish, otherwise fin dusky. Indian Ocean and Atlantic specimens darker, medium to dark brown overall, fins blackish, underside of snout, lips, lower jaws, and gill membranes blackish; orbital rim entirely black. SizE.—To more than 85 cm TL. DISTRIBUTION. — Throughout most of the In- dian Ocean, east to New Zealand, and west to the Atlantic coast of Africa from South Africa in the south, to the Gulf of Guinea (6°S) in the north. Depth range 696-1,600 m, but most com- mon around 900-1,200 m. REMARKS AND COMPARISONS.— McCann and McKnight (1980) first recorded this species as Coryphaenoides rudis for a specimen collected off New Zealand. Iwamoto (1986:335) recog- nized two small specimens from southern Africa as probable representatives of “‘an undescribed species related to Coryphaenoides macrolophus,” but because they were juveniles, he refrained from describing them as new. Subsequently, he and Shcherbachev examined numerous specimens in collections of ISH, FSFRL, AMS, IOAN, ZMMGU, and NMN2Z, and were planning to describe the species. However, in the meantime, it was learned that Peter McMillan of New Zea- land was also planning to describe the species from New Zealand and Australian collections. It was thought that McMillan’s description would long precede ours, but other commitments pre- vented him from completing his descriptions as hoped. Recent communication with McMillan suggests that his description of the longbeard grenadier will be published soon. The species is similar to Coryphaenoides du- bius Smith and Radcliffe, 1912, described from a single 425-mm specimen taken in the Philip- pines. They share in common a distinctive head profile, eight pelvic fin rays, a hump-backed nape, small eyes, and low blunt snout. However, the holotype and only known specimen of C. dubius has a much smaller mouth (the maxillary falling well short of the posterior margin of the orbit), cardiform teeth in a moderately broad band in the lower jaw, and a shorter barbel. Relationships with a second undescribed spe- cies from the Tasman Sea that will be described by McMillan are apparently even closer, and we have had difficulty distinguishing the two. That species has been given the common name slender rattail (Paulin et al. 1989). (In keeping with our desire to avoid the term rattail for these fishes, we henceforth call it slender grenadier, and Cor- yphaenoides sp. | longbeard grenadier, in con- trast to “long barbel rattail”’ of those authors.) So far as we can tell, all counts are similar and only a few proportional measurements can be used for comparisons. The slender grenadier has somewhat longer posterior nostrils (7-10% HL, cf. 4-8% [but only 3 of 21 specimens had a mea- surement exceeding 6.7%]), a wider internasal (15-17% HL, cf. 12-16%, usually 13-15%) and interorbital (18-21% HL, cf. 14-21%, usually less than 19%), and shorter head relative to total length. These features, however, overlap to a large degree and generally cannot be used alone for identification of the species. The slender grena- dier has more extensive naked areas on the un- 312 derside of head, a smaller maximum size (about 43 cm TL in specimens we have examined, cf. 85 cm for the longbeard rattail), a more-slender body lacking the humped back, and the snout is generally more pointed and protruding. The longbeard grenadier is highly distinctive in having a series of characters that sets it apart from most of its congeners, including the long, low head; broad maxillary; lobelike preopercle; elevated lappet-shaped gill rakers on first arch; wide gill openings that extend forward to nearly below posterior end of maxillary; long barbel; and fairly numerous gill rakers on either side of the first arch. Relationships with other members of the genus (aside from the slender grenadier) are obscure. The five or six retia and gas glands and the dentition pattern place it closest to mem- bers of subgenera Chalinura and Nematonurus, but the extent of squamation on the underside of the head, the spinulation on the scales, and the physiognomy of the head (especially its shal- low, long shape) is quite different from those of Chalinura and Nematonurus. The species also lacks the characteristic row of scales along the leading edge of the snout, behind which are two lunate naked patches, a feature of most Chal- inura (in this regard the species is similar to C. striaturus). The elongated head resembles to some extent that of C. (Nematonurus) ferrieri (Regan, 1913), and the rather extensive squamation on the underside of the head is reminescent of the condition in C. (N.) lecointei (Dollo, 1900). But, in addition to the dentition differences, the dis- tinctive gill rakers on the first gill arch and the low pelvic fin ray count set the species apart from members of subgenus Nematonurus. ACKNOWLEDGMENTS We thank the many curators and assistants at the following institutions who helped with loans, services, and hospitality: AMS, J.R. Paxton, D.F. Hoese, J. Leis, M. McGrouther, et al.; BMNH, N. R. Merrett, G. Howes, O. Crimmen; IOAN, N.V. Parin, E. Karmovskaya et al.; ISH, M. Steh- mann; LACM, R. Lavenberg, J. Seigel, D.M. Co- hen, et al.; RUSI, M.E. Anderson, P.C. Heem- stra, B. Ranchod; SAM, L.J.V. Compagno, M. Roeleveld-Compagno, C. Goliath; SIO, R. Ro- senblatt, H.J. Walker, C. Klepadlo; USNM, V.G. Springer, J. Williams, et al.; ZSI, P.K. Talwar; ZMA, I.J.H. Isbriicker; ZMUC, [the late] E. Ber- telsen, J. Nielsen; ZMMGU, Y.I. Sazonov, I.A. Verighina. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 14 The National Academy of Sciences/National Research Council made possible TI’s visit to Moscow in 1988 through its Soviet-East Euro- pean Program for exchange scientists. The Cal- ifornia Academy of Sciences In-House Research Fund provided funds for two museum visits by TI. The Australian Museum provided TI a vis- iting scientist fellowship, wich made possible a three-month visit to Australian museums in 1993. LITERATURE CITED Atcock, A. 1889. Natural history notes from H.M. Indian Marine Survey Steamer ‘Investigator, Commander Alfred Carpenter, R.N., D.S.O., commanding. No. 13. On the bath- ybial fishes of the Bay of Bengal and neighbouring waters, obtained during the seasons 1885-1889. Ann. Mag. Nat. Hist., ser. 6, 4 (23): 376-399; (24):450-461. 1890a. Natural history notes from H. M. Indian Marine Survey Steamer ‘Investigator,’ Commander R. F. Hoskyn, R.N., commanding. No. 16. On the bathybial fishes collected in the Bay of Bengal during the season 1889-1890. Ann. Mag. Nat. Hist., ser. 6, 6: 197-222, pls. 8-9. 1890b. Natural history notes from H. M. Indian Marine Survey Steamer ‘Investigator,, Commander R. F. Hoskyn, R.N., commanding. No. 18. On the bathybial fishes of the Arabian Sea, obtained during the season 1889- 90. Ann. Mag. Nat. Hist., ser. 6, 6: 295-311. . 1892a. Natural history notes from H.M. Indian Ma- rine Survey Steamer ‘Investigator,’ Lieut. G. S. Gunn, R.N., commanding. Series 2, No. 5. On the bathybial fishes col- lected during the season of 1891-92. Ann. Mag. Nat. Hist., ser. 6, 10 (59):345-365, pl., 2 figs. 1892b. Illustrations of the zoology of the Royal In- dian Marine Survey Ship ‘Investigator.’ Fishes. Part I, pls. I-VII. . 1894a. Natural history notes from H.M. Indian Ma- rine Survey Steamer ‘Investigator,’, Commander C. E. Old- ham. R.N., commanding. Series 2, No. 11. An account of a recent collection of bathybial fishes from the Bay of Bengal and from the Laccadive Sea. J. Asiat. Soc. Bengal 63 (pt. 2)(2):115-137, pls. 6-7. 1894b. Illustrations of the zoology of the Royal In- dian Marine Survey Ship ‘Investigator.’ Fishes. Part II, pls. VITI-XIII. 1899. A descriptive catalogue of Indian deep-sea fishes in the Indian Museum, collected by the Royal Indian Marine Survey Ship ‘Investigator.’ Indian Museum, Cal- cutta. ii1+211 pp., 8 pls., 1 map. AMAOKA, K., K. MATsuurRA, T. INADA, M. TAKEDA, H. Ha- TANAKA, AND K. OKADA. 1990. Fishes collected by the R/ V Shinkai Maru around New Zealand. Japan Marine Fishery Resource Research Center, Tokyo. 410 pp. Brauer, A. 1906. Die Tiefsee-Fische. I. Systematischer Teil. Wissenschaftliche Ergebnisse der deutschen Tiefsee-Expe- dition auf dem Dampfer “Valdivia,” 1898-1899, 15:1-432, text-figs. 1-176, pls. 1-18. Dotto, L. 1900. Macrurus lecointei, poisson abyssal nouveau recueilli par cette expédition Antarctique Belge. Bull. Acad. Roy. Belgique (Sci.)(6): 1-20. GitBert, C. H. 1905. The deep-sea fishes of the Hawaiian Islands. Pp. 575-713, figs. 230-276, pls. 66-101 in The aquatic SHCHERBACHEV AND IWAMOTO: GRENADIER SUBGENUS COR YPHAENOIDES 313 resources of the Hawaiian Islands. D.S. Jordan and B. W. Evermann, eds. Bull. U.S. Fish Comm. 1903, 22(pt. 2, sect. 2). GicserT, C. H. AND C. L. Husss. 1916. Report on the Jap- anese macrouroid fishes collected by the United States fish- eries steamer “Albatross” in 1906, with a synopsis of the genera. Proc. U.S. Natl. Mus. 51: 135-214, pls. 8-11. GILBERT, C. H. AND C. L. Husss. 1920. The macrourid fishes of the Philippine Islands and the East Indies. U.S. Natl. Mus. Bull. 100, 1 (pt. 7): 369-588, figs. 1-40. Guwnnerus, J. E. 1765. Efterretning om berglaxen, en rar norsk fishk, som kunde kaldes: Coryphaenoides rupestris. K. Norske Videnskabers Selskab Skrifter Trondhjem, 3(4):50- Sosiples. fig: 1, 2. Guntuer, A. 1877. Preliminary notes on new fishes collected in Japan during the expedition of H.M.S. “Challenger.” Ann. Mag. Nat. Hist., ser. 4, 20: 433-447. GunTuer, A. 1878. Preliminary notices of deep-sea fishes collected during the voyage of H.M.S. “Challenger.” Ann. Mag. Nat. Hist., ser 5, 2: 17-28. 447. GUnTHER, A. 1887. Report on the deep-sea fishes collected by H.M.S. Challenger during the years 1873-76. Rep. Sci. Res. Challenger 22(Zool.)(pt. 1)[text]: 1-335; (pt. 2)[plates]: pls. 1-73. Iwamoto, T. 1970. The R/V Pillsbury Deep-Sea Biological Expedition to the Gulf of Guinea, 1964-65. 19. Macrourid fishes of the Gulf of Guinea. Stud. Trop. Oceanogr. (4)(pt.2): 316-431. Iwamoto, T. 1986. Family Macrouridae. Pp. 330-341 in Smiths’ sea fishes. M.M. Smith and P. C. Heemstra, eds. Macmillan South Africa, Johannesburg. Iwamoto, T. AND Y. I. SAzONov. 1988. A review of the southeastern Pacific Coryphaenoides (sensu lato) (Pisces, Gadiformes, Macrouridae). Proc. Calif. Acad. Sci. 45(3): 35- 82, figs. 1-9. Iwamoto, T. AND Y. N. SHCHERBACHEV. 1991. Macrourid fishes of the subgenus Chalinura, genus Coryphaenoides, from the Indian Ocean. Proc. Calif. Acad. Sci. 47(7): 207-233, figs. 1-17, tabs. 1-7. IwAmMoTo, T. AND D. L. Stern. 1974. A systematic review of the rattail fishes (Macrouridae: Gadiformes) from Oregon and adjacent waters. Occas. Pap. Calif. Acad. Sci. (111):1- 79. Leviton, A. E. AND R. H. Gras, Jr. 1988. Standards in herpetology and ichthyology. Standard symbolic codes for institution resource collections in herpetology and ichthy- ology. Supplement No. 1: additions and corrections. Copeia 1988(1):280-282. Leviton, A. E., R. H. Grass, Jr., E. HEAL, AND C. E. DAwson. 1985. Standards in herpetology and ichthyology: Part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia 1985(3):802- 832. Marswa_t, N. B. 1973. Family Macrouridae. Pp. 496-665 in Fishes of the western North Atlantic. Daniel M. Cohen, ed. Mem. Sears Found. Mar. Res. (1)(pt.6). MarsHaLt, N. B. AND T. IwAMoTo. 1973. Genus Coryphae- noides. Pp. 565-600, in Fishes of the western North Atlantic. Daniel M. Cohen, ed. Mem. Sears Found. Mar. Res. (1)(pt.6). Maut, G. E. 1951. Monografia dos peixes do Museu Mu- nicipal do Funchal. Familia Macrouridae e Merlucciidae. Bol. Mus. Munic. Funchal (5) (art. 12):5—55, figs. McCann, C. AND D. G. McKnicut. 1980. The marine fauna of New Zealand: macrourid fishes (Pisces: Gadida). New Zealand Oceanogr. Inst. Mem. 61:1-91. Menon, A.G.K. AND K.V. RAMA Rao. 1970. Type-speci- mens of fishes described in the R.I.M.S. “Investigator” col- lections (1884-1926). Copeia 1970(2):377-378. MeEnon, A.G.K. AND G.M. YAZDANI. 1968. Catalogue of type specimens in the Zoological Survey of India. Part 2. Fishes. Rec. Zool. Surv. India 61: 91-190. Norman, J.R. 1939. Fishes. Sci. Rep. John Murray Exped. 1933-1934, 7(Zool.)(1): 1-116. Pautin, C., A. STEWART, C. ROBERTS, AND P. MCMILLAN. 1989. New Zealand fish, a complete guide. Natl. Mus. New Zeal. Miscell. Ser. 19. 279 pp. RapcutFFe, L. 1912. Descriptions of a new family, two new genera, and twenty-nine new species of anacanthine fishes from the Philippine Islands and contiguous waters. Proc. U.S. Natl. Mus. 43: 105-140, pls. 22-31. REGAN, C. T. 1913. The Antarctic fishes of the Scottish Na- tional Antarctic Expedition. Trans. Roy. Soc. Edinburgh 49(pt. 2)(2): 229-292, pls. 1-11. Sazonov, Y.I. AND T. IwAMoTo. 1992. Grenadiers (Pisces, Gadiforems) of the Nazca and Sala y Gomez ridges, south- eastern Pacific. Proc. Calif. Acad. Sci. 48(2):27-95, 37 figs., 7 tabs. SHCHERBACHEYV, Y.N., Y.I. SAZONOV, AND T. IwAMoTo. 1992. Synopsis of the grenadier genus Kuronezumia (Pisces: Gad- iformes: Macrouridae), with description of a new species. Proc. Calif. Acad. Sci. 48(3): 97-108, 9 figs., 1 tab. Weber, M. 1913. Die Fische der SIBOGA-Expedition. Si- boga Exped. 57: 1-719, pls. 1-12. Weber, M. AND L. F. DE BEAUFORT. 1929. The fishes of the Indo-Australian Archipelago, 5. E. J. Brill. Leiden. 458 pp. Woop-Mason, J. AND A. Atcocx. 1891. Natural history notes from H.M. Indian marine survey steamer ‘Investiga- tor, Commander R.F. Hoskyn, R.N., commanding. Series 2, No. 1. On the results of deep-sea dredging during the season 1890-91. Ann. Mag. Nat. Hist., ser. 6, 8: 16-34; 119- 138, 2 pls. © CALIFORNIA ACADEMY OF SCIENCES, 1995 Golden Gate Park San Francisco, California 94118 ,) usit) «h ae ek derens «Suing alin dT aeWiiy Aarineeek a ee ee td wt ently As (eo pil’ ele eT wR pa be ry psegsth-t lad S seis: $i ed @ a fa ee fur ja ’ OOF psa? athe Ms ny Say bf re et Sait Sai T er wl ALA fyi Ab, O74 eat bhues 9 heap-4 oer MP el A Ce ee ee ANT WH (mt $2 ~~ ’. we oli al } 7 , Gy — : iy a -°* 1) @Sn8 @ 4 74 ( . \ iy t f J ¢ 1 iy ‘ a j ‘ i) / j : : iat a © Jah : é L a i% * iia U r il | ‘ re i’ iP tae = / | i * — J — +4 { ¢ cAlrsh ; — 1 i” rs | PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 15, pp. 315-328, 7 figs., 1 table. August 1, 1995 THE ENIGMATIC SEA PEN GENUS GYROPH YLLUM— A PHYLOGENETIC REASSESSMENT AND DESCRIPTION OF G. SIBOGAE FROM TASMANIAN WATERS (COELENTERATA: OCTOCORALLIA) Gary C. Williams sods Deedee: Department of Invertebrate Zoology, California Academy of Sciences, Marine Biological Labora} + Leboratory/ apni Institution noie Oc Golden Gate Park, San Francisco, California 94118 AUG U7 1995 Woods Hole, MA 02543 Asstracr: 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, and Pteroeides spinosum. Gyrophyllum is shown to have characters intermediate between the Pennatulidae and Pteroeididae, thereby obviating the separation of the two families. It is therefore proposed that only one family be recognized, the Pennatulidae, comprising the six genera Pennatula, Ptilosarcus, Gyrophyllum, Sarcoptilus, Crassophyllum, and Pteroeides. Received August 10, 1994. Accepted March 8, 1995. INTRODUCTION At least fifteen species in nine genera of sea pens have been recorded or are known to occur in southern Australia (Williams, in press). The recent discovery of Gyrophyllum sibogae off Tas- mania represents a new record to the Australian fauna. Southern Australia is defined here as the continental shelf region of southern Western Australia, South Australia, Victoria, Tasmania, and southern New South Wales (i.e., the southern coastal waters extending from Perth to Sydney). Hickson (1916) originally described Gyro- phyllum sibogae from the Malay Archipelago and differentiated it from the type species G. hiron- dellei of the northern Atlantic, the only other species known for the genus. Hickson differen- tiated the two species as follows—G. sibogae with two calyx teeth per polyp and larger colony size (up to 295 mm in length); and G. hiron- dellei without calyx teeth and smaller colony size (up to 127 mm long). In addition, he noted distributional and bathymetric differences—G. sibogae (Indonesia at 567 m), compared to G. hirondellei (vicinity of the Azores at 1222- 2220 m). The enigmatic nature of the genus Gyrophyl- lum has resulted in some disputation by previous authors regarding the systematics of the taxon. It is shown here that Gyrophyllum represents a morphological intermediate between the Pen- natulidae and Pteroeididae thereby making the distinction between the two families tenuous. [315] 316 MATERIALS AND METHODS The following material was examined for this study. Pennatula inflata Kiikenthal, 1910— SAFRM H MAD 28 KKK, Subantarctic, be- tween Prince Edward and Marion Islands (37°55°90” E; 46°43’56°‘S), 225-243 m, 28 April 1987, SA Agulhas dredging cruise, | partial spec- imen (peduncle missing). SAFRM H3238, South Africa, northwest of Lambert’s Bay (31°43.6’S; 16°13.2°E), 457 m, 9 June 1947, two specimens trawled, University of Cape Town Ecological Survey. Ptilosarcus undulatus (Verrill, 1865)— CASIZG 013015, Mexico, Baja California Sur, La Paz, intertidal, 22 March 1940, coll. Edward F. Ricketts, 1940 Steinbeck/Ricketts Sea of Cor- tez Expedition, FV Western Flyer, six whole specimens. Gyrophyllum sibogae Hickson, 1916—see “‘Material’’ section under Systematic Account below. Sarcoptilus grandis Gray, 1848 — SAUSM H11921, South Australia, Spencer Gulf, approximately 18.5 km northwest of Point Riley (35° 53’S; 137°24’E), 25 m, 3-4 May 1989, K. Gowlett-Holmes and P. Tudorovic, FV Angela Kaye, two whole specimens. Crassophyllum cris- tatum Tixier-Durivault, 1961—NMNHP St. 51, Angola, off Moita Seca, 44 m, 25 October 1948, one whole specimen. Pteroeides spinosum (Ellis, 1764)—CASIZG 028063, France, Mediterra- nean Sea, Banyuls-sur-Mer, 35 m, 15 October 1992, R. J. Moci and B. David, RV Nereis, one whole specimen. Abbreviations used in the text refer to the fol- lowing: SAFRM (South African Museum, Cape Town), SAUSM (South Australian Musuem, Adelaide), CASIZG (California Academy of Sci- ences, Department of Invertebrate Zoology and Geology, San Francisco), NMNHP (National Museum of Natural History, Paris). Terminol- ogy used in the present work conforms to that of Bayer, Grasshoff, and Verseveldt (1983). All figures and scanning electron micrographs are by the author. SYSTEMATIC ACCOUNT Gyrophyllum Studer, 1891 Gyrophyllum Studer, 1891: 94. 1901: 34. Roule, 1905: 454. Kiikenthal, 1915: 120. Hickson, 1916: 252. ?Bathypenna Marion, 1906: 147. Two species of the northern Atlantic, Indo- West Pacific, and southeastern Australia. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 Gyrophyllum sibogae Hickson, 1916 (Figs. 1-4, 5C) Gyrophyllum sibogae Hickson, 1916: 253, fig. 45, pl. 3 fig. 17, pl. 8 fig. 48. Tixier-Durivault and D’Hondt, 1973: 263. MATERIAL.—SAUSM H11844, Tasmania, 46 km west of Richardson Point (41°15’S; 144°08’E), 520 m, 20 October 1984, W. Zeidler CSIRO “Soela” Stn. 51, 7 whole and 3 partial specimens. 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 por- tion of the rachis distinctly wider than the rest of the colony. The rachis comprises roughly 30%- 40% of the total colony length, while the pedun- cle makes up approximately 60%-70%. The maximum width of the rachis varies from 12- 17 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 15-50 mm in breadth). The autozooids are confined to the outer margin of the leaves, where they are congested and dis- posed 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 usu- ally 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 di- ameter, and appears as an inconspicuous hemi- spherical 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 some- what longitudinally grooved but not three- flanged. 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 WILLIAMS: PHYLOGENETIC SYSTEMATICS OF GYROPHYLLUM Figure 1. Gyrophyllum sibogae (SAUSM-H1 1844). Photographs of a single specimen, 185 mm in length. A. Dorsal view. B. Ventral view. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 ia Ficure 2. Gyrophyllum sibogae (SAUSM-H11844). Scanning electron micrographs of sclerites. A-F. Sclerites from the tentacles. A. 0.16 mm. B. 0.16 mm. C. 0.16 mm. D. 0.21 mm. E. 0.21 mm. F. 0.24 mm., G-I. Sclerites from the polyp leaves. G. 0.43 mm. H. 0.34 mm. I. 0.07 mm. FY Fi | » $ a * geKine Ficure 3. Gyrophyllum sibogae (SAUSM-H1 1844). Scanning electron micrographs of sclerites. A, B. Rachis surface. A. 0.41 mm. B. 0.07 mm. C, D. Rachis interior. C. 0.25 mm. D. 0.09 mm. E-G. Peduncle surface. E. 0.26 mm. F. 0.29 mm. G. 0.30 mm. H-J. Peduncle interior. H. 0.29 mm. I. 0.28 m. J. 0.09 m. 320 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 WILLIAMS: PHYLOGENETIC SYSTEMATICS OF GYROPHYLLUM 321 er erate epee ce ana hr gry eS tap oo i Lae 4 > a: * + _ nw ae % me an ov Se, . - FiGure 5. Comparison of polyp leaf morphology in six pennatulid and pteroeidid taxa. A. Pennatula inflata, 33 mm in length. B. Ptilosarcus undulatus, 6 mm in length. C. Gyropohyllum sibogae, 35 mm in length. D. Sarcoptilus grandis, 31 mm in length. E. Crassophyllum cristatum, 32 mm in length. F. Pteroeides spinosum, 30 mm in length; a- autozooid, ct- spiculated calyx teeth, In- large leaf needle, m- mesozooid, p- siphonozooid plate, r- polyp leaf ray, s- siphonozooid, sp- siphonozooid pad. from the surface and interior of the peduncle are 0.17-0.38 mm long. Color of the colonies varies from tan to light brown. DISTRIBUTION (Fig. 4).—Indo-West Pacific and southeastern Australia, 520-585 m in depth: Ce- ram (Malay Archipelago), 567 m depth (Hick- — son, 1916); Nosy Bé (Madagascar), 580-585 m depth (Tixier-Durivault and d’Hondt, 1973); Tasmania, 520 m depth (present study). REMARKS. — Hickson (1916: 254) reported nu- merous but inconspicuous siphonozooids on the dorsal surface of the rachis of the type specimen from Indonesia—these were discerned only as a Ficure 4. A. World map showing geographical distribution for the genus Gyrophyllum (@ = collecting stations for G. hirondellei, * = collecting stations G. sibogae). B. Map of Tasmania showing collecting station of Gyrophyllum sibogae for material studied in the present paper (*). 322 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 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 sibo- gae, although from disparate geographical regions of the southern hemisphere (Madagascar, Indo- nesia, and Tasmania), show a remarkable simi- larity and restriction of depth distribution (520- 585 m). DISCUSSION — PHYLOGENETICS Gyrophyllum sibogae is morphologically in- termediate between the pennatulids (Pennatula and Ptilosarcus) and the pteroeidids (Sarcoptilus, Crassophyllum, and Pteroeides), based on the presence of three-flanged sclerites (as in the pen- natulids) and the presence of siphonozooids on the polyp leaves (as in the pteroeidids). This lat- ter character is considered derived and defines a clade separating Sarocoptilus, Crassophyllum, and Pteroeides from other pennatulaceans. Williams (1993: 732-733) provides a summary of previ- ous work on pennatulacean phylogeny, and sum- marizes the difficulty regarding analyses of phy- logenetic systematics in such groups as sponges and coelenterates. HIsTORICAL CONTEXT (Fig. 6).— K6lliker (1869) originally unified the subfamilies Pennatulinae and Pteroeidinae into one family. Subsequently, KOlliker (1880) elevated the status of the sub- families to separate families. Studer (1901: 34) placed the new genus Gyrophyllum in the family Pteroeididae. Kiikenthal and Broch (1911: 394) stated that Gyrophyllum should not be placed in the Pteroeididae due to the presence of three- flanged sclerites, which are characteristic of the family Pennatulidae. Kiikenthal and Broch (1911: 463) and Kiikenthal (1915: 120) considered the taxon to be a systematic enigma and consequent- ly labeled it as “‘incertae sedis”’ at the end of their accounts of the Pennatulacea (Williams, 1995). Hickson (1916: 252-253), disagreed with Ki- kenthal, 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 addi- tion, K6lliker (1880) placed Halisceptrum (a syn- onym of Virgularia) in the Pennatulidae. Scy- talium, 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 Virgu- lariidae (Williams, 1995). OuTGRovuP.— Williams (1993) has shown that the presence of polyp leaves is considered a syn- apomorphic feature in the Pennatulacea and serves to distinguish a derived clade composed of the virgulariids (such as Virgularia and Scy- talium), the pennatulids (Pennatula and Ptilo- sarcus), 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 si- phonozooids 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 con- taining 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 devel- oped as the pennatulids and pteroeidids. The vir- gulariid polyp leaves are often small, thin, and translucent, being composed of the fused polyp — Ficure 6. Summary of classification schemes for pennatulid and pteroeidid taxa. A. KGlliker, 1869. B. KGlliker, 1880. C. Kiikenthal, 1915. D. Hickson, 1916. E. Present study with cladogram showing overlapping characters that unify the six taxa. Williams (1995) considers Leioptilum and Leioptilus as synonyms of Ptilosarcus, Halisceptrum as a synonym of Virgularia, Sarcophyllum as a synonym of Sarcoptilus, Godeffroyia and Struthiopteron as synonyms of Pteroeides. 323 WILLIAMS: PHYLOGENETIC SYSTEMATICS OF GYROPHYLLUM PTEROEIDIDAE PENNATULIDAE PENNIFORMES WATTAHdOOYVS VIAOYAAAGOD saqgivOuald WNYLdSOSITIVH SNOYVSOTILd WATILdOI31 VINLVNNAd WNTTAHdOOYVS ViAO¥SISI0O9 saqiwzOusald PTEROEIDINAE WNTILdOI3T VINLVNNAd PENNATULINAE PTEROEIDIDAE PENNATULIDAE | | PTEROEIDIDAE WNTIAHdOYAD WANTTAHdOODYVS saqig0u¥sdld WAITIWLAIS WNTILdOHLNVOV WAN TILdOIST VINLVNN&Ad TAE SEDIS ac re [ wor1AHdouAo =z WNTTAHdOOYVS NOYALdOIHLNYLS w SAqIaOuAld SsN1LdOl31 VINLVNNAd PENNATULIDA PENNATULIDAE * polyp leaf siphonozooids * fleshy polyp bases * smooth sclerites ob saqigousld [o} N 12) e NJWN11AHdOSSVWHO 6 bs 2 ” m SNTILdOOYVS wo Lua rey AS © Oo} NNITAHdOYAD © O a) [= TD Oo Ore Dal SNOYVSOTILd & wo 5 2 2s =e VINLVNN&Ad *x * 324 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 TasLeE 1. Morphological diversity of pennatulid and pteroeidid pennatulaceans (0 = ancestral state, 1 = derived state). 1. Polyp leaf mesozooids 0 = absent 1 = present 2. Spiculated calyx teeth O = present 1 = reduced 3. Siphonozooids restricted to rachis dorsal tract 0 = absent = present 4. Polyp leaf siphonozooids 0 = absent 1 = present 5. Siphonozooids on both sides of polyp leaves 0 = absent 1 = present 6. Siphonozooid zones 0 = absent 1 = present 7. Smooth sclerites 0 = absent 1 = present 8. Distal rachis mesozooids 0 = absent 1 = present 9. Siphonozooid pads 0 = absent 1 = present 10. Large leaf needles 0 = absent 1 = present 11. Siphonozooid plates 0 = absent 1 = present 12. Rachis & peduncle sclerites 0 = numerous 1 = reduced 13. Polyp leaf rays 0 = absent 1 = present Character 1 2 3 4 5 6 7 8 9 10 11 12 13 Pennatula (Outgroup) 1 0 0 0 0) 0 0 0 0 0 0 0 0 Ptilosarcus 0 1 1 0 0 0 0 0 0) 0 0) 0 0 Gyrophyllum 0) 1 0 1 1 0 0 0 0 0 0 0 0 Sarcoptilis 0 1 0 1 0 1 1 1 1 0 0 0 0 Crassophyllum 0 1 0 1 0 1 1 0 0 1 1 1 0 Pteroeides 0 1 0 1 0 1 1 0 0) 1 1 0 1 bodies of a single layer of the autozooids them- selves. Consequently, the pennatulids are con- sidered the pennatulacean taxa closest to the pteroeidid clade. CHARACTERS AND PoLariry (refer to Table 1). Ie Polyp leaf mesozooids: Pennatula inflata has conspicuous mesozooids, each with two spi- culated 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 char- acter 1s unique to Pennatula, and is therefore considered an autapomorphy. . Calyx teeth: In the genus Pennatula, eight distinctly spiculated teeth are present on the distal portions of each autozooid calyx (Fig. SA). In the other pennatulid genus, Prilo- sarcus, 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 pres- ence of eight distinct calyx teeth is consid- ered a plesiomorphic character, while a re- duction of the number of teeth or of the degree of spiculation is considered the de- rived state. . 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 pol- yp leaves. In Gyrophyllum, siphonozooids may be scattered on the dorsal surface of the rachis in addition to the polyp leaves (Hick- son, 1916: 254). In Sarcoptilus, Crasso- phyllum, and Pteroeides, the siphonozooids are restricted to the polyp leaves—being ab- sent from the rachis altogether (Figs. 5D-F). The restriction of the sipnonozooids to the dorsal region of the rachis is here considered autapomorphic for Pti/osarcus while the dis- position of siphonozooids on the lateral por- tions as well as the dorsal surface of the ra- chis, as in Pennatula, is considered plesiom- orphic. 4. Polyp leaf siphonozooids: In species of the pennatulid genera Pennatula and Prtilosar- cus, 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 cov- ered 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 penna- tulacean genera with polyp leaves such as Virgularia, Scytalium, and Stylatula, si- phonozooids are present on the rachis. Con- sequently, 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. . Distribution of leaf siphonozooids: Gyro- phyllum is the only pteroeidid taxon to have siphonozooids present on both surfaces of each polyp leaf. In Pennatula and Ptilosar- cus, siphonozooids are not present on the polyp leaves and in Sarcoptilus, Crasso- phyllum, 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. . Siphonozooid zones: The restriction of si- phonozooids 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 or- ganized into distinct zones on the leaves. The presence of siphonozooid zones is con- sidered 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. . Sclerites: Three-flanged sclerites, composed of three longitudinal grooves along the sur- face 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 WILLIAMS: PHYLOGENETIC SYSTEMATICS OF GYROPHYLLUM 10. 1 te 12 13% 325 presence of three-flanged sclerites is conse- quently considered to be a plesiomorphic character state, while the presence of smooth sclerites is considered the derived state. . 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. . 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 ra- chis, is unique to one pennatulacean taxon— Sarcoptilus (Fig. 5D). This character is con- sidered a further modification of the zona- tion and restriction of siphonozooids in the proximal portions of the polyp leaves. The presence of siphonozooid pads is conse- quently considered autapomorphic for Sar- coptilus, while the absence of these pads is considered the plesiomorphic state. Large leaf needles: Of the eleven genera of sea pens with well-developed polyp leaves, only Crassophyllum and Pteroeides have large needle-like sclerites contained on the polyp leaf surface (Figs. 5E, F). They are scattered in Crassophyllum and are arranged into distinct rays in Pteroeides (Figs. SE, F). The presence of large leaf needles is therefore considered a synapomorphy for Crasso- phyllum and Pteroeides, while their absence is considered a plesiomorphy. Siphonozooid plates: In Crassophyllum and Pteroeides, the siphonozooids form a flat plate covering the proximal portion of each polyp leaf (Figs. SE, F). This character is found only in these two genera and is there- fore considered a synapomorphy. Rachis and peduncle sclerites: In the six pen- natulid and pteroeidid genera, sclerites are abundant in the rachis and peduncle of all except Crassophyllum, where these sclerites are very sparse or absent. The reduction in the number of rachis and peduncle sclerites is considered an autapomorphy for Cras- sophyllum, while the presence of abundant densely-set sclerites in these regions is con- sidered the plesiomorphic state. Polyp leaf rays: One to several polyp leaf PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 15 NS Sami a Lis Ny Table | and the text (see Discussion). 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 par- allel and in close proximity. They act as sup- porting rays to give a state of rigidity to the PENNATULA GYROPHYLLUM CRASSOPHYLLUM PTILOSARCUS SARCOPTILUS PTEROEIDES Ficure 7. Illustrated cladogram of six pennatulid and pteroeidid taxa. Numbered synapomorphies correspond to those in 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 con- spicuous ray is present along the dorsal mar- WILLIAMS: PHYLOGENETIC SYSTEMATICS OF GYROPHYLLUM 327 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 Prteroeides, and its absence is considered the plesiom- orphic state. CONCLUSION The cladogram in Figure 7 was produced from the character matrix in Table 1 using PAUP (Phylogenetic Analysis Using Parsimony) ver- sion 3.1. A single tree was generated with a Cl (Consistency Index) of 0.923. A detailed comparative examination of the taxa Pennatula, Ptilosarcus, Gyrophyllum, Sarcopti- lus, 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 KGlliker, 1880, to be synonymous with the Pennatulidae Ehrenberg, 1834, which has priority. The six gen- era listed above are here considered members of the family Pennatulidae. The family Pennatulidae is here defined as fol- lows: 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 conspicu- ous 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 three- flanged or smooth needles, rods, plates, or ovals. ACKNOWLEDGMENTS I am grateful to Shane Parker, late Curator of Lower Marine Invertebrates at the South Aus- tralian Museum in Adelaide, for originally sug- gesting the study of the southern Australian pen- natulacean fauna and for kindly lending the ma- terial used in this study. I also thank Karen Gow- lett-Holmes, Wolfgang Zeidler, and Eric Matthews of the South Australian Museum for their cooperation. I thank Dr. Stephen D. Cairns (United States National Museum, Washington, D.C.) for his helpful comments regarding the manuscript. LITERATURE CITED Bayer, F. M., GRASSHOFF, M., AND VERSEVELDT, J. 1983. Illustrated trilingual glossary of morphological and anatom- ical terms applied to Octocorallia. Leiden: E. J. Brill. Hickson, S.J. 1916. The Pennatulacaea of the Siboga Ex- pedition, with a general survey of the order. Siboga Expeditie Monogr. 14 (Livr. 77): 1-265. KOLLIKER, R. A. von. 1869-72. Anatomisch-Systematische Beschreibung der Alcyonarien. I. Die Pennatuliden. Abhan- dlungen von der Senckenbergischen naturforschenden Ge- sellschaft 7: 111-2555 (1869); 487-602 (1870); 8: 85-275 (1872). 1880. Report on the Pennatulida dredged by HMS Challenger during the years 1873-1876. Report on the Sci- entific Results of the Voyage of HMS Challenger during the years 1873-76, Zoology 1(2): 141. KUKENTHAL, W. 1915. Pennatularia. Das Tierreich 43: 1- 132. Berlin: Verlag von R. Friedlander. KUKENTHAL, W. AND Brocu, H. 1911. Pennatulacea. Wis- senschaftliche Ergebnisse der Deutschen Tiefsee-Expedition auf dem Dampfer ’Valdivia’ 1898-1899 13 (1) Lieferung 2: 113-576. Marion, A. F. 1906. Etude des coelentérés atlantiques re- cueillis par la Commission de dragages de l’aviso le “Tra- vailleur” durant les campagnes 1880-1881. Oeuvres pos- thumes du A. F. Marion. Reunies par Paul Gourret. IV. Les Alcyonaires de la premiére expedition du Travailleur. Ex- pedition scientifique Travailleur et Talisman: 103-151. Par- is: Masson. Route, L. 1905. Notice préliminaire sur les pennatulides recueillis par le Travailleur et le Talisman, dans l’?Océan Atlantique, au large du Maroc. Bulletin du Muséum national d’Histoire naturelle 11: 454-458. STuDER, T. 1891. Note préliminaire sur les alcyonaires prov- enant des campagnes du yacht “l’Hirondelle” 1886, 1887, 1888. Part 2. Alcyonacea and Pennatulacea. Mémoires de la Société zoologique de France 4: 86-95. . 1901. Alcyonaires provenant des campagnes de I’Hi- rondelle (1886-1888). Res. Camp. Sci. Monaco 20: 1-64. TrxtER-DurRIVAULT, A. AND D’Honpt, M.-J. 1973. Nou- velles Récoltes D’Octocoralliaires a Madagascar. Tethys 5 (2-3): 251-266. Wixuiams, G. C. 1993. Biotic diversity, biogeography and phylogeny of pennatulacaean octocorals associated with coral reefs in the Indo-Pacific. Proceedings of the Seventh International Coral Reef Symposium, Guam, 1992, 2: 729- 735. 1995. Living genera of the sea pens of world seas- illustrated key and synopses (Coelenterata: Octocorallia: Pennatulacea). Zoological Journal of the Linnean Society 113: 93-140. . (In press). Revision of the pennatulacean genus Sar- coptilus (Coelenterata: Octocorallia), with descriptions of three new species from southern Australia. Records of the South Australian Museum. © CALIFORNIA ACADEMY OF SCIENCES, 1995 Golden Gate Park San Francisco, California 94118 Ne 7) a ae oe » — —_ PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 48, No. 16, pp. 329-342, 6 figs. August 1, 1995 NOTES ON PANAMA PALMS By logical Laboratory aphic Institution Greg de Nevers California Academy of Sciences, Golden Gate Park, San Francisco, CA 94118, U.S.A. SUG U7 1995 Woods Hole, MA 02543 Asstract: The four species of Pholidostachys are reviewed. Pholidostachys kalbreyeri is recognized as a species distinct from P. pulchra, described, and a neotype designated. These species are contrasted with P. synanthera and P. dactyloides, and a diagnostic key is provided. Pholidostachys dactyloides is a new record for Panama. Calyptrogyne pubescens and C. kunorum are described as new and illustrated. Geonoma costa- tifrons and G. allenii are transferred to Calyptrogyne. Geonoma concinna, a new record for Panama, is described, illustrated, and discussed. Geonoma triandra is illustrated, and discussed in the context of Geonoma species with other than six stamens. Geonoma chococola is a new record for Panama. Received February, 9, 1994. Accepted April 12, 1995. INTRODUCTION Recent field work in Panama, continuing mon- ographic work and routine determination of palms for the Flora Mesoamericana project have revealed the following novelties and range ex- tensions, and necessitate the following name changes in the Palmae. In order to simplify the descriptions the con- vention of length x width is used in preference to cm long, cm wide. PHOLIDOSTACHYS Wessels Boer (1968) recognized three species of Pholidostachys. A clearly delineated species of Pholidostachys with short, spicate inflorescences represented by specimens in many herbaria has for years remained unrecognized. The species is distinct from P. pulchra H. Wendl. ex Burret, the only other species with a spicate inflorescence. Pholidostachys kalbreyeri H. Wendl. ex Burret from Antioquia, Colombia has been treated as a synonym of P. pulchra (Galeano & Bernal 1987; Wessels Boer 1968) largely because the type can- not be located and is presumed to have been destroyed at Berlin. Only one collection from Colombia resembling the Panamanian material is available for study (Gentry et al. 30155, MO). These specimens match the protologue of P. kal- breyeri and are associated here with that concept. A full description is provided and a neotype des- ignated. Bernal et al. (1989) have discussed the status of Kalbreyer’s Colombian collections, including the lack of duplicates of types of palm species based upon his collections. A search of European herbaria (K, B, GOET) has failed to locate ad- ditional material of the type of P. kalbreyeri. A search in Antioquia (Bernal et al. 1989; Galeano & Bernal 1987) has failed to locate a Pholido- stachys with a spicate inflorescence. Another possibility is that Kalbreyer clipped one of the rachillae off of an inflorescence of P. synanthera (Mart.) H. E. Moore, a species that has been col- lected in the Mpio. Frontino of the Department of Antioquia, the area from which the type of P. [329] 330 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 Ficure 1. Pholidostachys kalbreyeri. A. Habit, x ca. 4) (de Nevers 7310). B. Infructescence showing tightly clustered fruits on proximal portion of rachilla and fruitless distal portion, x ca. ', (de Nevers 6220). kalbreyeri comes. I reject this idea, since in the protologue Wendland described both the pedun- cle (““pedunculus brevis”) and the peduncular bract (““Spatha secunda superne lacerata, 27 cm longa’). Pholidostachys kalbreyeri H. Wendl. ex Burret, Bot. Jahrb. Syst. 63: 130. 1930. Fig. 1. Tyre: Kalbreyer 1881, COLOMBIA. Antioquia: Cinegetas. Destroyed at B. Neotyrpe: PANAMA. San Blas: El Llano-Carti Rd., km 19.1, 350 m, 9 Jan. 1985, de Nevers et al. 4467 (MO, NY, PMA). Stem solitary, prostrate and rooting or erect, 1-3 m, 8-12 cm diameter; leaves 10-13, irreg- ularly pinnate, arching; sheath 15-20 cm; petiole 65-115 cm, channeled above, rounded below; rachis 1.4—1.6 m with 5-7 pinnae per side, these usually subopposite, 80-90 cm x 4-18 cm; in- florescence hidden among leaf bases, 16-26 cm, at anthesis enclosed within the disintegrating, fi- brous peduncular bract, spicate or sparingly branched, if branched the rachis short, the ra- chillae digitate; prophyll fibrous, 10-15 cm x 3- 5 cm; peduncular bract inserted close to the in- sertion of the prophyll, 15-22 cm x 2-4 cm; pe- duncle ca. 2.5 cm, flattened; rachilla 12- 18cm x 1-2cm, with floral pits spirally ar- ranged in 5-8 lines, bracts overlapping those of adjacent pits, pits of proximal half of rachilla with two staminate and one pistillate flower per pit, pits of distal half of rachilla not developing pistillate flowers; flowers subtended by three short, ovate, hyaline, irregular bracts; staminate flowers with sepals briefly connate basally, im- bricate apically, hyaline, irregular, petals enclos- ing staminal tube and ovary in bud, splitting upon expansion of staminal tube, valvate, hyaline, in- curved at apex, ca. 4 mm; stamens six, the fila- ments thick, fleshy, united in a tube for two- thirds their length, free and narrowly acute api- cally, anthers sagittate, dorsifixed, introrse, pis- tillode minute; pistillate flowers with sepals and DE NEVERS: NOTES ON PANAMA PALMS Ficure 2. Map of Panama showing provincial boundaries and localities for: Pholidostachys kalbreyeri (triangles); P. dac- tyloides (solid circles); Calyptrogyne pubescens and Geonoma chococola (hexagon); C. kunorum (solid square); G. triandra (open circle). petals as in the staminate flowers, these sur- rounding a ring of six to nine fleshy, narrow, basally connate, apically free and acute stami- nodes; staminodial tube adnate to the petals ba- sally; gynoecium trilocular, triovulate (with only one ovule maturing), style central, stigmas three, spreading, the inner surface receptive; fruit 1.5- 3.5 cm diameter, flattened proximally by com- pression against adjacent fruits, rounded apical- ly, maroon, with basal stigmatic residue; epicarp smooth; mesocarp with thick anastamosing fi- bers; seed 1.2—1.5 cm diameter, endosperm ho- mogeneous. ADDITIONAL SPECIMENS EXAMINED. COLOMBIA. Choco: Quibd6-Tutunendo Road ca. 3 km W of Tutunendo, pluvial forest, 80 m, 5'46’”N, 76'35”W, 5 Jan. 1981, Gentry et al. 30155 (MO). PANAMA. Cocle: Continental Divide above El Copé, 8'38”N, 80'39”W, 650-750 m, 27 Nov. 1985, de Nevers et al. 6371 (BH, MO, PMA); 27 April 1985, Hammel 13,661 (MO); Read et al. 81-9 (US); Read & Watson 84-53 (US); Read & Watson 84-49 (US). San Blas: El Llano-Carti Rd., km 19.1, 350 m, de Nevers et al. 9021 (CAS, PMA); de Nevers & Hammel 8552 (CAS, MO, PMA); 26 March 1973, Liesner 1255 (MO); de Nevers et al. 7310 (MO, PMA); de Nevers & Archibold 6220 (MO, PMA); Cerro Obu, 400-500 m, 25 June 1986, de Nevers et al. 8043 (CAS, MO, PMA); Rio Taindi 6 km above conflu- ence with Rio Mandinga, 30-100 m, 9’25’N, 79'11”W, 5 April 1986, de Nevers & Herrera 7644 (MO, PMA); Yar Bired (Cerro San José), Continental Divide, 9’20’N, 79'8”W, 400-500 m, 5 Feb. 1986 de Nevers & Herrera 6964 (MO, NY, PMA); Rio Cangandi at confluence with Rio Titamibe, 60 m, 9'24’N, 79'7’W, 8 Feb. 1986, de Nevers & Herrera 7031 (MO, PMA). Veraguas: trail from Bajo Chitra to Rio Gatu, 650-750 m, 14 Jan. 1986, de Nevers & Mc Pherson 6767 (MO, NY, PMA). DISTRIBUTION AND Hasitat. Pholidostachys kalbreyeri (Fig. 1) is known from the Atlantic slope of Panama from Veraguas Province east to the Colombian Department of Choco (Fig. 2). Its historic range apparently included the wet lowland forests of the Colombian Department of Antioquia. There are no known extant collec- tions of this species from Antioquia. In Panama, P. kalbreyeri is sympatric only with P. pulchra. In Colombia, in the area of Murri (the type lo- cality of P. kalbreyeri) both P. dactyloides H. E. Moore and P. synanthera occur. In Panama Pholidostachys kalbreyeri is un- common in wet forests from near sea level to 650 m. Pholidostachys kalbreyeri occurs along rivers, or in swampy flats and depressions on mountain slopes. In swamps this species occurs in dense shade in the understory and has a long, often bent, procumbent trunk which roots where it is buried or in contact with the soil. Along the lower, flat courses of rivers in shaded or partly sunny situations the trunk is erect and up to 2.5 m tall. The lower leaves are removed by floodwa- ters producing a very different aspect from the upland version. The genus Pholidostachys is composed of four species, three of which occur in Panama. They can be distinguished as follows: 332 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 |. }inflorescencespicaiewe< => seonn nes 2 2. Fruits tightly crowded, maroon, pro- duced on proximal half of rachilla; in- florescence interfoliar; stem 1—4 m, 8— 12 cm diameter, often bent or decum- bent, abscission scars obscure; spike 16— DOGMA soe P. kalbreyeri 2. Fruits distant, brown, produced throughout the rachilla; inflorescence infrafoliar; stem 4—7 m, 4-6 cm diam- eter, straight, erect, canelike (with clear abscission scars); spike 40 cm or longer se WM Ne sina ee Rie (2. P. pulchra 3. Infl. rachis short, rachillae more or less digitate; peduncular bract fibrous, en- closing the infructescence __ P. dactyloides 3. Infl. rachis long, the rachillae remote; peduncular bract not fibrous, the ra- chillae well exserted P. synanthera Discussion. Pholidostachys pulchra occurs from Costa Rica through Panama to Colombia (Dept. Valle), from sea level to 850 m. It grows in Tropical Wet Forests (Holdridge et al. 1971) on the Atlantic slope in Central America and in Tropical Rain Forest (Holdridge et al. 1971) on the Pacific slope of Colombia. In Panama it is a common component of wet forest understories and is often sympatric with P. kalbreyeri. Wessels Boer (1968) reported Pholidostachys dactyloides only from “‘the Pacific coastal area of Colombia near sea level, in forests flooded by fresh water during the greater part of the year.” Galeano and Bernal (1987) list this species as endemic to Colombia. Recent collections have shown this palm to occur on the Pacific slope from Ecuador (Esmeraldas, Pichincha) through Colombia (Antioquia, Choco, Narifio) to eastern Panama (de Nevers et al. 8402, CAS, MO, PMA; Folsom et al. 6639 MO), from sea level to 1,450 m. In Panama Pholidostachys dactyloides occurs on Cerro Tacarcuna and Cerro Mali, and has an apparently isolated outpost in the Cafiazas Mountains (Fig. 2). On Cerro Mali, Pholidosta- chys dactyloides occurs on mountain slopes and flats in very wet forests at 1,250-1,450 m ele- vation. Pholidostachys synanthera occurs in Colombia (Amazonas, Antioquia, Narifio, Santander), Ec- uador (Carchi, Napo, Pastaza, Pichincha), Peru (Amazonas, Cuzco, Junin, Loreto, Puno, San Martin), and Brazil (Amazonas) between 130 and 1,750 m. The pollination mechanisms of Pholidostachys are unknown (Henderson 1986). The inflores- cences of both P. kalbreyeri and P. dactyloides are completely enclosed in the only partially de- composed peduncular bract at anthesis, suggest- ing tunneling beetles. The inflorescences of P. synanthera and P. pulchra are exserted at anthe- sis. This genus would be an appropriate candi- date for intensive pollination study. The pro- duction of flowers enclosed in bracts by Pholi- dostachys has parallels among palms in the gen- era Manicaria, Pinanga and Ceratolobus (Dransfield 1982). The flowers of Pholidostachys show little vari- ation among species, a situation which occurs in many other palm genera, e.g. Calyptrogyne (Wes- sels Boer 1968), Ceratolobus (Dransfield 1979), Cryosophila (Evans 1992), Geonoma (Wessels Boer 1968) and Prestoea (Henderson & de Ne- vers 1988). A diversity of habits, leaf morphol- ogies and inflorescence types occur in genera with striking conservatism in floral morphology. An- derson (1979) ascribed floral conservatism in neotropical Malpighiaceae to specialized re- wards (oil glands) attracting only a narrow suite of pollinators (Hymenoptera only, principally oil- bees). Floral conservatism in palms seems to be a very different situation, with strikingly different pollination mechanisms developing in even closely related genera. An example among closely related genera of the tribe Geonomeae is: Aster- ogyne is pollinated by flies and bees (Henderson 1986), Calyptrogyne is pollinated by bats (de Ne- vers & Henderson 1988; Kress & Beach 1994), and Pholidostachys may be beetle-pollinated. If there is an adaptive explanation for this absence of a pattern (Gould & Lewontin 1979) it remains to be discovered. GEONOMA The following represent ecological or range ex- tensions, or new records for Panama. Geonoma triandra (Burret) Wess. Boer (Fig. 3) is the only species of Geonoma known to have only three stamens. Wessels Boer (1968) knew of only one locality in Colombia and one in west- ern Panama for this species. On the slopes of Cerro Mali, in eastern Panama, the plants are common from 700 to 1,400 m. Although com- parable habitat in central and western Panama has been well collected in the last decade, col- lections from the intervening gap have not been DE NEVERS: NOTES ON PANAMA PALMS 333 Ficure 3. A. Geonoma triandra, habit, showing the characteristic deflexed peduncle, x '4 (de Nevers 8392). B. G. concinna, habit, showing thin canes and simple to irregularly divided leaves, x % (de Nevers 9029). made. This disjunction may be real and not an artifact of collecting. When Wessels Boer (1968) described Geon- oma chococola it was known only from Colom- bia. It has recently been collected in western Pan- ama (de Nevers et al. 8823 CAS, MO, NY, PMA). Wessels Boer (1968) and Moore & Uhl (1982) described the genus Geonoma as having three or six stamens. Wessels Boer did not report the number of stamens for Geonoma chococola in the original description, which was based on fruiting material. Uhl and Dransfield (1987) and Henderson (1988) reported Geonoma to have three, six or rarely more stamens, but didn’t mention which species may have more. Bernal et al. (1991) discussed variability of stamen num- ber in palms and mentioned G. triandra, but did not mention Geonoma species with more than six stamens. The western Panama collection of G. chococala (one plant only) invariably has 9 stamens, and recent collections from Colombia have 9-12 stamens (R. Bernal, pers. comm.). Only two species of Geonoma are known to have more than 6 stamens, G. chococola and G. polyandra from Ecuador (Skov 1994). Both have spicate inflorescences and unequally divided leaves. Geonoma polyandra differs from G. chococola in its smaller, brown, apically attenuate, rugose in- dehiscent fruits vs. the larger, black, spherical, dehiscent fruits of G. chococola. Bernal et al. (1989) have recently neotypified Geonoma concinna Burret (Fig. 3). Examination of the isoneotype (NY) reveals that it is identical with a large series of previously unidentified col- lections from Panama. A description of the spe- cies is provided based on the neotype and the Panamanian specimens, which are cited below. Geonoma concinna Burret, Bot. Jahrb. Syst. 63: 216. 1930. Neotyre: COLOMBIA. Antioquia: carretera Granada-San Luis, 5.5 km adelante de El Choco, 1750 m, Bernal & Tobén 1385 (Neotype COL n.v., isoneotype NY!). Fig. 4. Stems solitary or (usually) caespitose with up to 15 stems per plant, slender, 2—3.5 m high, 5- 8 mm diameter at nodes; leaves simple and bifid or rarely with irregular divisions or trijugate; sheath 3.5-9 cm; petiole 5.5-19 cm, narrowly 334 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 > “% pu SS Syual 7 Lyf y sol Z% = v1 Ficure 4. Calyptrogyne pubescens. Acaulescent habit, individual with the leaves regularly divided, x '4) (de Nevers et al. 8825). channeled above, rounded beneath; rachis (9-) 18-29 (-37) cm; blade narrowly cuneate in out- line, glabrous on both surfaces, notched apically about one-third its length, with acuminate lobes (9-) 14-23 cm on inner margin, (8.5-) 11-19 cm wide at apex of rachis; primary veins (17-) 26- 35, at an angle of 25-35 degrees with the rachis, prominently upraised on both leaf surfaces, with DE NEVERS: NOTES ON PANAMA PALMS a less prominent unraised vein between each main vein; inflorescence infrafoliar, 2-4 nodes below current leaves, paniculate, (10.5-) 16-28 cm x (9-) 18-32 cm, with a distinctive pattern of trans- verse surface wrinkles, lightly tomentose with straight brown hairs when young, usually gla- brous in fruit; peduncle 3.5—6.5 cm; prophyll and peduncular bract inserted close together at base of peduncle; prophyll 4.2-7 cm x 1-1.2 cm, swollen in bud, striate; peduncular bract slightly smaller than prophyll, included in it; inflores- cence rachis (2.8-) 3.5-8 cm, with 3-5 primary branches, these again branched; rachillae 8-22, orange, (3.5-) 6.5-8 cm x 1 (-2) mm, bearing spirally arranged, dispersed flower pits, the lips exserted from the rachillae, the lower lip entire or bifid, the orifice ca. 1 x 1 mm; staminate flow- ers ca. 3 mm, sepals and petals subequal, free, erect at anthesis, slightly imbricate below; sta- mens 6, filaments spreading, anthers sharply re- flexed from the filaments; pistillate flowers ca. 2mm, sepals and petals subequal, barely ex- serted, styles 3, reflexed, staminodial tube trun- cate; fruit oblong to globose, green, drying brown, 5-6 mm x 6-8 mm, young fruits with minute columnar papillae between the tubercles, these lost at maturity leaving a minutely verrucose (pebbled) surface; seed black, oblong, 5-6 mm. ADDITIONAL SPECIMENS EXAMINED. PANAMA. Cocle: Continental Divide above El Copé, 650-750 m, 8°38’N, 80°39’W, 27 Nov. 1985, de Nevers et al. 6372 (CAS, MO, NY); 20 April 1978, Hammel 2615 (MO); 8 April 1988, Thompson 4743 (CM); 710-800 m, 22 June 1988, Croat 68743 (CAS); 9 Jan. 1981, Read et al. 81-11 (US); south of Cascajal along Continental Divide, 800-900 m, 8’45”N, 80'25’W, Knapp 1985 (CAS, MO); Coclecito Rd., trail along Continental Divide, 500 m, 8'42’N, 80'28’”W, 11 Jan. 1986, de Nevers et al. 6728 (CAS, MO, NY). Herrera: Dist. Las Minas, Chepo, sendero a la cima de El Higo, ca. 900 m, 7’43”N, 80'50’W, Galdames et al. 1744 (PMA, US, SCZ). Panama: El Llano-Carti Rd., km 10.1, 1100-1200 ft., 27 Dec. 1974, Mori et al. 4119 (BH); 16 July 1987, Croat 67,363 (CAS); Rancho Chorro, mountains above Torti Arriba, Cafiazas Mountains, 400-700 m, 3 Dec. 1977, Folsom et al. 6603 (BH, F); Campo Tres, 3 mi. NE of Altos de Pacora, 500-800 m, 10 March 1973, Croat 22695 (MO, US, WIS); Cerro Jefe, 1,000 m, 9'15”N, 79’30”W, 16 Aug. 1981, Knapp 893 (MO); 13 Sept. 1981, Knapp 1203 (MO); same loc.: 25 Aug. 1975, Mori & Joly 7933 (MO); 3 July 1971, Croat 15299 (CAS). San Blas: Yar Bired (Cerro San José), Continental Divide between Cangandi and San José, 400- 500 m, 9'20’N, 79’8”W, 5 Feb. 1986, de Nevers & Herrera 6942 (CAS, MO); Cerro Brewster, 800-850 m, 9'18”N, 79'16”W, 20 Nov. 1986, de Nevers et al. 6286 (MO, NY); 21 April 1985, de Nevers et al. 5381 (CAS, MO, NY); El Llano-Carti Rd., km 19.1, 350 m, 9'19”N, 78'55’W, 2 Nov. 1985, de Nevers et al. 6147 (CAS, MO, NY); 2 Feb. 1989, km 16, de Nevers et al. 9029 (CAS); 14 April 1985, de Nevers et al. 5367 (MO, NY); 335 km 16-18.5, 28 March 1974, Nee & Tyson 10967 (CAS, MO); 5 March 1985, de Nevers et al. 4974 (MO, NY); mi. 12, 26 March 1973, Liesner 1169 (CAS, MO); trail from Rio Esadi to Cerro Banega, 300-530 m, 9’23’N, 78'51”W, 21 Dec. 1985, de Nevers & Herrera 6616 (MO, NY); trail to Cerro Obu (Habu of maps) from Rio Urgandi (Rio Sidra), 100-300 m, 9'23’N, 78'48”W, 24 June 1986, de Nevers et al. 8006A (CAS, MO). DISTRIBUTION. Geonoma concinna is known from El Copé (Coclé), the Llano-Carti Road (San Blas), east in San Blas to Cerro Habu, and thence to the Colombian Choco and the middle valley of the Rio Magdalena. In Panama it occurs in Tropical Wet Forest (Holdridge et al. 1971) on the Atlantic slope between 100 and 1,000 m, with an isolated occurrence on the Azuero peninsula. Geonoma concinna, like most taxa which occur predominantly on the wetter Atlantic slope of Panama up to the Continental Divide, occurs sparingly on the drier Pacific slope near the con- tinental divide, especially where the topography is flat, as where the El Llano-Carti road crosses from Prov. Panama into San Blas. Discussion. Geonoma concinna fits into the group of Geonoma characterized by thin canes and branched, broomlike inflorescences with thin rachillae, prominently exserted floral pits, and a truncate or crenate staminodial tube (Grayum & de Nevers 1988). The leaves of Panamanian plants are consistently simple and bifid, but de Nevers & Herrera 6616 has gaps on one side of the blade and Mori & Joly 7933 has entire, ir- regularly divided and trijugate leaves on the same plant. All of the Colombian material I have seen, including the neotype, has irregularly divided leaves. Leaves that vary from entire to pinnate on a plant, in a population, or in a species is a common pattern of variation in palmae genera, e.g. Areca (Dransfield 1980), Bactris (Henderson 1995), Calyptrogyne, Chamaedorea (Hodel 1992), Geonoma (Wessels Boer 1968), Hyospathe (Skov & Balslev 1989), Iguanura (Dransfield 1980), Pinanga (Dransfield 1991) and Prestoea (Henderson & de Nevers 1988). Geonoma concinna differs from G. scoparia Grayum & de Nevers in its caespitose habit, lon- ger peduncle (3.5-6.5 cm vs. 2.3-3.8 cm), the degree of ramification of the inflorescence (2nd order vs. 3rd order), and the shape of the leaves (usually simple, bilobed vs. trijugate). Geonoma concinna shares the caespitose habit and simple, bilobed leaves of G. tenuissima H. E. Moore, but differs in its less branched inflorescence (2nd or- der vs. 3rd order), shorter rachis (4.2-8 cm vs. 336 11-13 cm), fewer primary inflorescence branch- es (3-5 vs. 13-16), thicker rachillae (1 mm vs. 0.5 mm) and pebbled fruit (vs. smooth on dry- ing). Geonoma concinna shares the pebbled fruit of G. leptospadix Trail, but not its solitary habit, pendulous inflorescence, small stature (4-10 dm), and long peduncle (15-20 cm vs. 3.5-6.5 cm). Geonoma gastoniana Glaz. ex Drude differs from G. concinna in its larger, divided leaves, pedun- cular bract inserted 1-2 cm above the prophyll, and longer peduncle (10 cm vs. 3.5-6.5 cm). The population on Cerro Jefe (Knapp 1203 and Knapp 893) has remarkably short leaf rachises and small inflorescences. CALYPTROGYNE Wessels Boer (1968) transferred Geonoma cos- tatifrons L. H. Bailey and G. allenii L. H. Bailey to Calyptrogyne. The former he regarded as a synonym of C. ghiesbreghtiana (Linden & H. Wendl.) H. Wendl., the latter as a synonym of C. brachystachys H. Wendl. ex Burret. Fieldwork in Panama has shown these two species to be distinct, making the following transfers neces- sary. Calyptrogyne costatifrons (L. H. Bailey) de Ne- vers, comb. nov. BasionyM: Geonoma costatifrons L. H. Bailey, Gentes Herb. 6: 206, fig. 105, 1943. Type: PANAMA. Canal Area: Rio Pequini, Fairchild & Jobbins 2639 (Holotype BH!, isotype MO!). Calyptrogyne allenii (L. H. Bailey) de Nevers, comb. nov. BasionyM: Geonoma allenii L. H. Bailey, Gentes Herb. 6: 204, fig. 104, 1943. Type: PANAMA. Cocle: El Valle de An- ton, Allen 2947 (Holotype BH!, isotype MO!). Both of these species are characterized by straight, erect, thin, canelike stems 2-5 m tall, with arching to pendent inflorescences, an ex- ceptional condition in the genus. They share this habit only with C. kunorum, described below. Calyptrogyne pubescens de Nevers, sp. nov. Calyptrogyne ghiesbreghtiana similis sed pedunculo arcte pannoso-tomentoso bractea pedunculi persistentique. Type: PANAMA. Bocas del Toro: low hills east of the Gualaca-Chiriqui Grande Rd., 10 miles N of the Continental Divide, | mile E along side road. 8'55’”N, 82'5”W, 120 m, 19 January 1989, G. de Nevers, F. Almeda & G. McPherson 8825. Holotype CAS; isotypes BH, COL, K, MO, NY, PMA. Figs. 4&5. Stem solitary, subterranean; leaves basal, 6- 10, divided laterally into about 30 equal or un- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 equal pinnae per side, these (22-) 36-52 cm x 7- 29 mm, glabrous, with 5 main veins prominent below, straight or slightly sigmoid, with very long drip-tip, distal pair of pinnae like the proximal ones, or much wider; sheath 6-8 cm, disinte- grating laterally into a brush of fibrous hairs; petiole 59-114 cm, narrowly channeled above, rounded beneath, densely tomentose; rachis 89- 142 cm, densely tomentose; inflorescence spi- cate, erect, arching, (166-) 215-374 cm x 0.8- 1.8 cm, covered in a dense layer of flattened and branched trichomes 2—3 mm thick, flattened par- allel to the stem distally, less so apically; prophyll 22-39 cm x 1.4-2.7 cm, flattened, with a winged margin 1-2 mm wide, glabrous or with patches of fine tomentum, inserted at the base of the peduncle; peduncular bract 29-46 cm x 0.7- 1.1 cm, inserted 6-11 cm below the first floral pit, with smooth, minutely tomentose outer sur- face, stiff, tough, fibrous, erect at anthesis, split- ting along one side to release the rachilla, per- sistent, or if fallen breaking irregularly, not leav- ing a clean abscission scar; rachilla 18- 31 cm x 3-7 mm, held horizontal at anthesis, bearing dispersed floral pits in a spiral, the lower lip covering the pit before anthesis, widely spreading to deflexed in fruit, the orifice ca. 2mm x 2 mm; staminate flowers ca. 6 mm, the sepals 3, imbricate, obscurely keeled, membra- nous, 3 mm, the petals fused basally, valvate api- cally; stamens 6, filaments connate basally, free apically, anthers dorsifixed basally, sagittate, in- trorse, pistillode long and narrow, almost equal to the stamens; pistillate flowers ca. 6 mm, the sepals 3, free, imbricate at base, membranous, the petals united into a membranous ovoid tube 5 mm long; stamens united into a fleshy, ovoid tube widest at middle; gynoecium trilocular, 3-lobed, style apical, stigmas 3, erect at anthesis; fruit oblong, narrowed at base, rounded apically, 11-14 mm x 7-8 mm, 1-seeded; epicarp smooth and glabrous, black, mesocarp fleshy, ca. 2 mm thick, with an inner layer of few, thick, anasta- mosing fibers adherent to the endocarp; mature fruit not seen; eophyll not seen. ADDITIONAL SPECIMENS EXAMINED. Same loc. as type, 5 March 1986, B. Hammel, G. McPherson & L. Sanders 14587 (CAS, MO, PMA); 14 km Oeste de Punta Cricamola, entrando Ensenada de Cata vel, subiendo Quebrada Nuri 8'55’N, 81'49”W, 18 March 1993, Foster et al. 14558 (SCZ). Discussion. Calyptrogyne pubescens is known only from the type locality in the Atlantic lowlands of Bocas Del Toro (Fig. 2). It is to be DE NEVERS: NOTES ON PANAMA PALMS 337 Ficure 5. Details of Calyptrogyne pubescens. A. Erect inflorescence with persistent peduncular bract and deflexed rachilla, x '4o. B. Irregularly divided leaf with the distal lobe largest, x 4 (Hammel et al. 14587). C. Peduncle, x ca. 2. D. Peduncular hairs, x ca. 15. E. Rachilla with flower pits and immature fruit, x ca. 3. A, C, D, and E were drawn from de Nevers et al. 8825. 338 sought further east in the lowlands of Veraguas province, and perhaps to the west in adjacent Costa Rica. It is unlike most species of Calyp- trogyne in that the peduncular bract is persistent, not cleanly deciduous and leaving a character- istic scar. The peduncular bract splits at anthesis, remains erect, and the spike deflexes to exit the bract and expose the flowers to pollinators. Ad- ditional distinguishing characteristics include the thick tomentum densely covering the peduncle, and the finely divided leaves. It shares the per- sistent peduncular bract only with C. anomala de Nevers & A. Henderson (de Nevers & Hen- derson 1988), which differs in its smaller stature (less than 1 m), lack of thick tomentum and branched inflorescence. Calyptrogyne ghies- breghtiana Linden ex H. Wendl. shares the acau- lescent habit of C. pubescens and is of similar size, but has a glabrous peduncle, a thinner, stri- ate peduncular bract, and fewer, wider pinnae. Calyptrogyne trichostachys, with which C. pu- bescens is sympatric at the type locality, has a similar tough, smooth peduncular bract. In C. trichostachys the peduncular bract is cleanly de- ciduous, and lacks the thick tomentum of C. pu- bescens. Calyptrogyne allenii and C. costatifrons differ from C. pubescens in their well-developed stems and lack of thick tomentum. Although no Calyptrogyne shares the remark- able pubescence of Calyptrogyne pubescens it is similar to that of Reinhardtia paiewonskiana Read, Zanoni & M. M. Mejia of Hispaniola (Read et al. 1987). Calyptrogyne kunorum de Nevers, sp. nov. Calyptrogyne ghiesbreghtiana similis sed caule erecto foliis plerumque indivisis fructibus grandibus fulvisque. Tyre: PANAMA. San Blas: Cerro Brewster, 9'18’N, 79'16’”W, 800-850 m, 20 Nov. 1985, G. de Nevers, A. Hen- derson, G. Mc Pherson, L. Brako 6261 (Holotype: CAS! iso- types: MO! NY! PMA). Fig. 6. Stem solitary, prostrate for up to 50 cm, erect for 46cm to 2 m, 6-8 cm thick, with adventi- tious roots trapping dirt and debris, covered in epiphytes; leaves 15—22, mostly simple and bifid, occasionally with gaps to one or both sides, leaf bases trapping dirt and debris, stem rooting into this debris; sheath 10-23 cm; petiole 21- 59cm x 8-10 mm, broadly channeled above, rounded beneath, brown-furfuraceous when young; rachis 35-95 cm, upraised, thin and gla- brous adaxially, upraised, thicker and tomentose abaxially; blade 22-32 cm wide at apex of the PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 rachis, lobes 42-64 cm long from apex of rachis including very long-attenuate drip-tip 6-15 cm, glabrous adaxially with prominently raised veins, glabrous abaxially with prominently raised veins these brown-tomentose when young; inflores- cence spicate, erect, arching, 86-142 cm, pen- dent in fruit; prophyll 22-51 cm x 1.2-2 cm, thin, keeled, strongly compressed, glabrous or with patches of fine tomentum, inserted at the base of the peduncle; peduncle 57-111 cm x 5- 7 mm, flattened, brown-furfuraceous; peduncu- lar bract 21-31 cm x 10-16 mm, strongly com- pressed, with winged sutures 1-2 mm wide, in- serted 1-3 cm below the first floral pit, with smooth, minutely tomentose outer surface, stiff, tough, fibrous, splitting along one side to release the spike, falling before anthesis, leaving a clean abscission scar; rachilla 21-28 cm x 6-9 mm, arched beyond the leaves at anthesis, bearing congested floral pits, the lower lip covering the pit before anthesis, widely spreading to deflexed and irregularly split in fruit, the orifice ca. 2mm x 2 mm; staminate flowers ca. 6 mm, the sepals 3, imbricate, obscurely keeled, membra- nous, 3 mm, the petals fused basally, valvate api- cally; stamens 6, filaments connate basally, free apically, anthers dorsifixed basally, sagittate, in- trorse, pistillode long and narrow, almost equal to the stamens; pistillate flowers ca. 6 mm, the sepals 3, free, imbricate at base, membranous, the petals united into a membranous ovoid tube 5 mm long, staminodial tube fleshy, ovoid, wid- est at middle; gynoecium trilocular, 3-lobed, style apical, stigmas 3, erect at anthesis; fruit oblong, narrowed at base, rounded apically, 1.6- 2.8cm x 1.1-2.8 cm, 1-seeded; epicarp smooth and glabrous, yellow to brown, mesocarp crisp, fleshy, ca. 2 mm thick, with an inner layer of few, thick, anastamosing fibers adherent to the en- docarp; seed 12-23 mm, endosperm homoge- neous; eophyll not seen. ADDITIONAL SPECIMENS EXAMINED. PANAMA. San Blas: Cerro Brewster, 9'18”N, 79'16”W, 800-850 m, 20 Nov. 1985, de Nevers et al. 6262 (PMA, NY); 21 April 1985, de Nevers et al. 5377 (MO, NY); 25 April 1985, de Nevers et al. 5548 (MO, NY); Cerro Obu, 400-500 m, 25 June 1986, de Nevers et al. 8042 (CAS, MO). Discussion. Calyptrogyne kunorum is known only from central and eastern Panama on Cerro Brewster and Cerro Obu (Fig. 2). These are two of the highest and wettest peaks in the generally low-elevation mountain chain connecting the Canal Area with the higher peaks of the Panama- Colombia border region to the east. Why this DE NEVERS: NOTES ON PANAMA PALMS 339 Ficure 6. Calyptrogyne kunorum (de Nevers et al. 6261). A. Erect habit with pendent infructescence and simple, deeply bilobed leaves, x ca. '4,. B. Infructescence showing prophyll, large fruits and peduncular bract abscission scar, x 4. C. Peduncular bract, x 4. D. Fruit with exocarp removed, partly opened to reveal seed, x ca. 2. palm has not been found on Cerro Jefe, a nearby peak of similar height and moisture regime, which shares the locally dominant palm Colpothrinax cookii Read (Read 1969) with Cerro Brewster, remains unexplained. Calyptrogyne kunorum most resembles C. costatifrons and C. allenii in its erect stem and pendent infructescence. It dif- fers from the former in its shorter, thicker stem and larger, brown (vs. black) fruits. It differs from C. allenii in its thicker, shorter stem, much larger inflorescence and larger fruits. Both C. alleniiand C. costatifrons have cleanly deciduous leaf sheaths which leave a smooth, canelike stem, while the leaf sheaths of C. kunorum adhere to the stem long after the abscission of the blade and petiole. In C. kunorum these leaf bases collect debris and develop soil like many rainforest “trashbasket plants.” In the most recent revision of Calyptrogyne five species were recognized (Wessels Boer 1968). 340 Wessels Boer did not have the advantage of fa- miliarity with the plants in the field. Field ex- perience is essential in understanding this genus. Although the species distinguished in the key below are clearly distinct in the field, herbarium material can be difficult to identify. I consider C. brachystachys and C. sarapiquensis H. Wendl. ex Burret to be synonyms of C. ghiesbreghtiana. With the addition of the two new species de- scribed above, as well as the species recovered from synonomy, I now recognize eight species. Two unresolved entities in Panama and one in Colombia may increase this number when better material is collected. Key to the Species of Calyptrogyne 1. Peduncular bract persistent, if removed not leaving a clean annular scar on the pedun- cle; inflorescence branched or spicate _..... 2 2. Inflorescence unbranched, 1.6—3.7 m; peduncle densely tomentose; leaves finely divided, 1.6-2.6 m .... C. pubescens 2. Inflorescence usually branched, 13.5- 28 cm; peduncle glabrous or finely fur- furaceous; leaves broadly and irregular- ly divided, 55-138 cm _.. C. anomala 1. Peduncular bract deciduous, leaving a clean annular scar on the peduncle; inflorescence 11 (02 | tae ee ann a AY 260, Saal eel 3 3. Stem erect, straight, 1-4 m; inflores- cence pendent, the peduncle bicarinate 4. Stem 6-8 cm diameter, the leaf bases persistent; fruit 16-28 mm; leaves Simple. ese C. kunorum 4. Stem 1.5-3cm diameter, the leaf bases cleanly deciduous; fruits 9- 20 mm; leaves trijugate to finely di- wideG: rarely simple -. Pak 2) 5. Inflorescence 67-98 cm, the spike 9-23 cm; fruit 9-11 mm, brown, dry at maturity... 2" C. allenii 5. Inflorescence 85-115 cm, the spike 19-42 cm; fruits 13- 20 mm, black, fleshy at maturity Me tein Daye sales ee: C. costatifrons 3. Stem subterranean to prostrate, 1 m or less; inflorescence erect, the peduncle te- RS Yh a LI kL SE Re thee a) 6. Inflorescence glabrous, green be- tween the pits (red in fruit); pedun- cular bract glabrous, minutely to deeply striate _..... C. ghiesbreghtiana PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 48, No. 16 6. Inflorescence densely brown, tomen- tose among the flower pits; pedun- cular bract smooth, velvety ................. 7. Flower pits densely crowded ..... Ree Me C. condensata 7; Flower pits remote. 25a phy a C. trichostachys ACKNOWLEDGMENTS This paper is dedicated to the memory of Dr. Hugo Churchill (1946-1993). Frank Almeda, Rodrigo Bernal, Michael Grayum, and Andrew Henderson provided helpful comments on an earlier version of this paper. I am indebted to Michael Grayum for the Latin diagnoses. Thanks to Kathy Bell for the superb line drawings. Field work in Panama and Costa Rica was supported by the Smithsonian Tropical Research Institute, the Missouri Botanical Garden, the National Geographic Society, and the California Academy of Sciences. RESUMEN Se presenta una clave dicdtoma para las cuatro especies de Pholidostachys. Pholidostachys kal- breyeri se reconoce como una especie distinta de P. pulchra, se describe, y se designa un neotipo. Las dos se comparan con P. synanthera y P. dac- tyloides. La ultima es un registro nuevo para Pan- ama. Se describen y se ilustran Calyptrogyne pu- bescens y C. kunorum. Se transfieran Geonoma costatifrons y G. allenii al genero Calyptrogyne. Geonoma concinna, un registro nuevo para Pan- ama, se describe, se ilustra, y se discute. Geo- noma triandra se ilustra, y se compara con las otras especies de Geonoma que tienen menos o mas de seis estambres. Geonoma chococola se nota por la primera vez en Panama. LITERATURE CITED ANpeERSON, W. R. 1979. Floral conservatism in neotropical Malpighiaceae. Biotropica 11(3):219-223. BERNAL, R. G., G. GALEANO AND A. HENDERSON. 1989. Neo- typification of Colombian palms collected by W. Kalbreyer. Taxon 38:98-107. . 1991. Notes on Oenocarpus (Palmae) in the Colom- bian Amazon. Brittonia 43:154-164. pe Nevers, G. AND A. HENDERSON. 1988. A new Calyptro- gyne (Palmae: Geonomeae) from Panama. Syst. Bot. 13: 428-431. DRANSFIELD, J. 1979. A monograph of Ceratolobus (Palmae). Kew Bull. 34:1-33. 1980. Systematic notes on some Bornean Palmae. Bot. J. Linn. Soc. 81:4—42. DE NEVERS: NOTES ON PANAMA PALMS . 1982. Pinanga cleistantha, a new species with hidden flowers. Principes 26:126-129. . 1991. Notes on Pinanga in Sarawak. Kew Bull. 46: 691-698. Evans, R. J. 1992. Cryosophila macrocarpa (Palmae), a new species from Choco Department, Colombia. Novon 2:58- 61. GALEANO, G. AND R. BERNAL. 1987. Palmas del Departa- mento de Antioquia, Region Occidental. Universidad Na- cional de Colombia, Centro Editorial, Bogota. Gou Lp, S. J. AND R. C. Lewontin. 1979. The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc. Roy. Soc. London B 205: 581-598. Grayuo, M. H. AND G. C. DE Nevers. 1988. New and rare understory palms from the Peninsula de Osa, Costa Rica, and adjacent regions. Principes 32:101-114. HENDERSON, A. 1986. Pollination studies in the Palmae. Bot. Rev. (Lancaster) 52(3):221-259. . 1988. Review of Genera Palmarum. Brittonia 40: 338-339. . 1995. The palms of the Amazon. Oxford Univ. Press. Oxford and New York. HENDERSON, A. AND G. DE NEverS. 1988. Prestoea (Palmae) in Central America. Ann. Missouri Bot. Gard. 75:203-217. Hopet, D.R. 1992. Chamaedorea palms, the species and 341 their cultivation. International Palm Soc., Lawrence, Kan- sas. Ho.pripce, L. R., W. C. GRENKE, W. H. HATHEWAyY, T. LIANG, AND J. Tost, Jk. 1971. Forest environments in tropical life zones. New York: Pergamon Press. Kress, J. W., AND J. H. BEAcH. 1994. Flowering plant repro- ductive systems. Pp. 161-182 Jn La Selva, ecology and nat- ural history of a neotropical rainforest, LS. McDade et al., eds. U. of Chicago, Chicago and London. Moore, H.E., Jk. AND N. W. Unt 1982. Major trends of evolution in Palms. Bot. Rev. (Lancaster) 48:1-70. Reap, R. 1969. Colpothrinax cookii—a new species from Central America. Principes 13:13-22. Reap, R., T. ZANONI, AND M. Mesa 1987. Reinhardtia paie- wonskiana (Palmae), a new species for the West Indies. Brit- tonia 39:20-25. Sxov, F. 1994. Geonoma polyandra (Arecaceae) a new spe- cies from Ecuador. Nordic J. Bot. 14:39-41. Skov, F. AND H. Batstev 1989. A revision of Hyospathe (Arecaceae). Nordic J. Bot. 9:189-202. Un, N. W. AND J. D. DRANSFIELD 1987. Genera Palmarum. L. H. Bailey Hortorium and the International Palm Society, Lawrence, Kansas. WESsSELs Boer, J. G. 1968. The geonomoid palms. Verh. Kon. Ned. Akad. Wetensch, Afd. Natuurk., Tweede Sect. Ser. 2, 58:1-202. © CALIFORNIA ACADEMY OF SCIENCES, 1995 Golden Gate Park San Francisco, California 94118 “Gabe cotati weed beyiit4 ¢ ee ee ee Fe ee iol Pera A aera analy ne! ares 6 hci ee ’ pet 00S if] 1 ain vp & = ail on an 7 dee = ~—