il CATBK N+

APRIL 29, 2005 * VOLUME 56 * NUMBERS 1-17

7

PROCEEDINGS

OF THE CALIFORNIA ACADEMY OF SCIENCES

G Dallas Hanna (1887-1970) Diatomist, Geologist, and Invertebrate Paleontologist

CALIFORNIA

ACADEMYorSCIENCES

FOUNDED 1853

SAN FRANCISCO, CALIFORNIA

Copyright © 2005 by the California Academy of Sciences All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without permission in writing from the publisher.

SCIENTIFIC PUBLICATIONS

Alan E. Leviton, Ph.D., Editor Gary C. Williams, Ph.D., Associate Editor Michele L. Aldrich, Ph.D., Consulting Editor

COVER IMAGE

G Dallas Hanna, Diatomist and Invertebrate Paleontologist (1887-1970); Curator of Geology, California Academy of Sciences, 1919-1970

In the early 1920s, G Dallas Hanna established the Academy’s diatom collection, reintroducing a tradi- tion for the study of fossil and Recent marine and freshwater microorganisms begun with the founding of the California Academy of Sciences in the early 1850s. Hanna’s interest in diatoms began in 1911 during the period when, as an employee of the U.S. Bureau of Fisheries, he was assigned to carry out a fur seal census in the Pribilof Islands. Thus, Hanna, when he came to the Academy in 1919, was to carry on the tradition begun by Academy co-founder John Boardman Trask and Academy member Henry G. Hanks, a tradition he reenergized when he established the first laboratory for micropaleontology on the Pacific Coast.

Today, the California Academy of Sciences houses one of the World’s premier diatom research col- lections, under the capable curatorship of Dr. J. Patrick Kociolek, and is a beehive of research activity dealing with Recent and fossil, freshwater and marine diatoms emphasizing systematics and environ- mental studies.

ISSN 0068-547X

The Proceedings of the California rbeademy of Sciences accepts manuscripts for publication in the Natural Sciences that stress systematics and evolutionary biology and paleontology, biogeography, and biodiversity, also relat- ed areas in the Earth Sciences, such as biostratigraphy, regional and global tectonics as they relate to biogeography, and paleoclimatology, and topics in astrobiology, anthropology, as well as the history of science as they relate to insti- tutions of natural history, to individuals, and to activities, such as expeditions and explorations, in the natural sciences.

Priority consideration will be given to manuscripts submitted by Academy staff and others who are affiliated with the research programs of the institution and/or manuscripts that include reference in part or in whole to Academy research collections or to Academy-sponsored expeditions. Others who do not meet the above criteria may submit manuscripts for consideration for publication on a space-available basis.

Manuscripts accepted for publication are subject to page charges; charges may be waived on a case-by-case basis.

Nore: All manuscripts submitted for publication in any of the Academy's scientific publication series (Proceedings, Occasional Papers, Memoirs) are subject to peer review. Peer review includes both internal and exter- nal review, internal review by at least one Academy scientist whose interests parallel those of the submission, and external review, ordinarily by two individuals who are recognized experts in the field.

Published by the California Academy of Sciences 875 Howard Street, San Francisco, California 94103 U.S.A.

Printed in the United States of America by Allen Press Inc., Lawrence, Kansas 66044

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 1, pp. 1-22, 6 figs., 1 table (Appendix) April 29, 2005

Redescriptions of the Nudibranch Genera Akiodoris Bergh, 1879 and Armodoris Minichev, 1972 (Suborder Doridacea), with a New Species of Akiodoris and a New Family Akiodorididae

Sandra V. Millen! and Alexander Martynov2 ! Department of Zoology, University of British Columbia, 6270 University Blvd., Vancouver, B.C., Canada V6T 124, Email: millen@ zoology.ubc.ca; 2 Zoological Museum, Moscow State University, Bolshaya Nikitskaya Str. 6, Moscow, 103009, Russia, Email: martynov @ zmmu.msu.ru

The type species of the genus Akiodoris, A. lutescens, Bergh, 1880 is redescribed from specimens found in the Sea of Okhotsk, and from the Commander Islands, Russia. A second species in the genus, Akiodoris salacia Millen sp.nov., is described from the Pacific waters off British Columbia, Canada. The type species of the genus Armodoris, A. antarctica Minichev, 1972 from the Davis Sea, is redescribed from the type and two additional specimens. The relationship of these genera with other gen- era of phanerobranch dorids is explored and a new family Akiodorididae is formed for the suctorian genera Akiodoris, Armodoris, Echinocorambe, Doridunculus, and Prodoridunculus.

Key worbDs: Opisthobranchia, Akiodoris, Armodoris, dorid phylogeny, Akiodorididae.

During the second half of 19th and early 20th centuries representatives of phanerobranch dorids of the genera Akiodoris, Prodoridunculus and Doridunculus, all with a remarkably similar radula, were discovered (G.O.Sars 1878; Bergh 1880; Odhner 1907; Thiele 1912). Bergh (1880) arranged the genus Akiodoris within the Suctoria, but Doridunculus was placed between the gen- era Goniodoris and Ancula. Following this arrangement, Fisher (1883-1887) placed the genus Doridunculus within the genus Goniodoris and the genus Akiodoris (as Aciodoris) within the genus Acanthodoris, in subgeneric ranks. Bergh (1883) created, for the first time, a phylogenetic scheme for the phanerobranch dorids where both genera Akiodoris and Doridunculus appeared at the base of Goniodorididae sensu latu as descendants of cryptobranch dorids. Later authors usually grouped the genera Akiodoris, Prodoridunculus and Doridunculus together into the base of the family Onchidorididae, probably considering them the most archaic forms (e.g., Thiele 1931; Odhner in Franc 1968), but Hgiseter (1986) arranged Doridunculus near the genus Lophodoris in the family Goniodorididae. In the second part of last century Minichev (1972) added the genus Armodoris as closely allied to Akiodoris.

Millen (1987:19), in an abstract referring to a cladistic analysis of the family Onchidorididae, isolated the four genera mentioned above as a separate clade. Martynov (1997) suggested for these genera a new family, Akiodoridae, but he did not give a formal description for this taxon. The main goal of the present study is to revise the two principal genera, Akiodoris and Armodoris, and to explore their relationship with the families Goniodorididae and Onchidorididae.

The genus Akiodoris was erected by Bergh (1879 a) for a phanerobranch nudibranch which has tripinnate gills, a sessile buccal pump, smooth lip cuticle, two cuspidate lateral teeth per side,

D PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. |

numerous rectangular marginal teeth, a penis armed with hooks and a vagina containing cilia coy- ered villi. The type species, Akiodoris lutescens Bergh, 1880 was found in the Alaskan Aleutians, and this species has not been reported since. One of the authors, Alexander Martynoyv, found a large number of specimens in several museums and research institutions in Russia, all from the Commander Islands and from the Sea of Okhotsk, mainly in the Kuril Islands. Bergh’s (1880) description was based on two specimens, one of which he considered a variety, although why he separated it is unclear. He never illustrated the external features and described them only briefly, although he described the anatomy in detail and included illustrations of some of the organs. A redescription of this species with emphasis on the morphology and additional illustrations of some of the anatomy is included in this paper. Martens (1879) emended the name to Aciodoris, replac- ing the Greek Akio (= sharp) with the Latin Acio. Because doris is a Greek name, we follow the recommendation of the Code that the components of a compound name should agree (Appendix D, II, 11) and revert to the spelling Akiodoris. Aciodoris is an unjustified emendation under the Code (art.32.5 and 33.2.3) and therefore a junor objective synonym of Akiodoris.

The other author, Sandra Millen, found 6 specimens of a small white dorid which possesses the internal characters of Akiodoris, in the waters near Vancouver, British Columbia, Canada. These animals differ in a number of external features from Akiodoris lutescens and represent a new species belonging to the genus Akiodoris. This species is described. and the relationship of the genus to other members of the family is discussed.

The genus Armodoris was created by Minichev (1972) for A. antarctica Minichev, 1972 from Tokarev Island in the Davis Sea. It has pear-shaped or conical, spiculose tubercles and a radula shape similar to that of Akiodoris, but with more teeth. There are up to six inner lateral teeth and up to eight outer lateral teeth per side. The specimen is 16 mm long and sexually mature; it is pos- sibly a mature specimen of Prodoridunculus gaussianus as only very small specimens of Prodoridunculus have been found. Without having a size range of specimens to examine, we shall retain them as separate taxa. Minichey (1972) separated Armodoris from Akiodoris because in Armodoris, the gonad surrounds the stomach, he observed a single seminal bursa, the spermathe- ca, on a long duct, and the vagina did not have a villous lining. In our observations of Akiodoris lutescens, we found that the stomach is anteriorly buried and posteriorly exposed so that the gonad- covered digestive glands surround it. The spermatheca of Akiodoris is short stalked, but has a tubu- lar shape similar to that of Armodoris antarctica. Akiodoris salacia sp. nov. lacks villi on the vagi- na. Thus, the taxonomical position of the genus Armodoris needs clarification. During a study of the collection of the Zoological Institute RAS (St. Petersburg), two addition specimens of Armodoris antarctica were found along with the dissected holotype of A. antarctica, which are used here for a redescription of this taxon.

Subclass Opisthobranchia Order Nudibranchia Suborder Doridacea

Superfamily Anadoridoidea

Family Akiodorididae Millen and Martynov, fam. nov. TYPE GENUS: Akiodoris Bergh, 1879

Akiodorididae, nom. nudum: Martynoy, 1997:233.

DIAGNOsIS. — The notum is spiculose, covered by rounded or elongate tubercles. Posterior part of the notum is round or transformed into two lobes. Branchial pocket is absent, gills are

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS — 3

arranged in semicircle with the exception of aberrant Echinocorambe where they are reduced to one simple leaf. Head has small four-corned oral veil. Anus dorsal, except for Echinocorambe where it is ventral. Buccal pump is a diverticulum of the dorsal surface of the buccal bulb. Peripheral muscle is absent and there are small, separate muscles at the anterior part of buccal pump. Lip disk thin and smooth. Radular formula is 2-14.2-6. 0-1. 2-6. 2-14. Central tooth is usu- ally present, ranging from a small plate to wide arch-shaped structure, sometimes with a central cusp. The first and following up to six lateral teeth have an irregular rectangular base and strong cusp directed downward and several strong denticles on one or both sides. Remaining laterals are in varying degrees of reduction, with a rectangular shape. Receptaculum seminis vaginal or insert- ed on the uterine duct. Stomach is entirely or partially free from the digestive gland. Prostate tubu- lar or enlarged. Penis with or without spines.

The new family includes the following five genera: Akiodoris Bergh, 1879, Armodoris Minichey, 1972, Doridunculus G.O. Sars, 1878, Echinocorambe Valdés and Bouchet, 1998, and Prodoridunculus Thiele, 1912.

Genus Akiodoris Bergh, 1879 TyPE SPECIES Akiodoris lutescens Bergh, 1880, by monotypy

DIAGNOsIS.— The notum is spiculose, covered by rounded tubercles. Posterior part of the notum is round and fully covers the foot and tail. Gills are in a semicircle. Head has a small four- corned oral veil. Anus dorsal. Buccal pump is slightly prominent anteriorly. Radular formula is 8- 13.2.1.2.13-8. The central tooth is wide and arch-shaped. The first and second laterals have only 1—2 blunt denticles on both sides. Remaining laterals are in various degrees of reduction. Cerebral and pleural ganglia are fully fused. Stomach is very large and fully free from the digestive gland. Gonad does not cover most of the digestive gland and stomach. Prostate long, tubular. Vagina may have villi. Ejaculatory duct has simple or complex spines.

The genus contains only the type species and the new species described in this paper.

Redescription of Akiodoris lutescens Bergh, 1880 (Figs. 1-2)

Akiodoris lutescens Bergh, 1879a:354—355 (nomen nudum). Akiodoris lutescens Bergh, 1879b:4-8 (plates only). Akiodoris lutescens Bergh, 1880:55—58.

MATERIAL EXAMINED?.— HOLoryPeE: The type specimen cannot be found in the University of Copenhagen Museum (Kathe Jensen, pers. commun.). OTHER MATERIAL: | specimen, 18 September 1949, Shikotan Island, Kuril Islands, Russia, 13.5 m depth, collected by E.F. Gurjanova. 1 specimen, 10 September, 1966, 44°41’N, 148°57’E, 780 m., Segsbee trawl, R/V “Vityaz,” St. 5640. Collected by V.M. Koltun. 1 specimen, 31 July, 1971, Vasilieva Pt., Paramushir Is, Kuril Islands, Russia, 20 m depth, large stones, collected by P.G. Kraynyuk. 2 specimens, 5 August, 1970, Gilyak Pt., Paramushir Is, Kuril Islands, Russia, 20 m. depth, rock, collected by A.M. Murakhveri. 1 specimen, 10 August, 1970, Anciferova Is., Kuril Islands, Russia, 15 m depth, large stones, col- lected by A.M. Murakhveri. 1 specimen, 20 September, 1971, Gromky Pt., Iturup Is., Kuril Islands, Russia, collected by V.I. Lukin. 1 specimen, 20 September 1971, Gromky Pt., Iturup Is., Kuril Islands, Russia, 15 m depth, collected by V.I. Lukin. 1 specimen, 22 September 1971, Shutka Inlet, Iturup Is., Kuril Islands, Russia, 20 m depth, rock, collected by V.I. Lukin. 1 specimen, 14 July,

° Except as indicated, no specimen numbers are given inasmuch as most are in field station collections.

4 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

1972, Korabel’naya Inlet, | km to sw of Korabel’ny Pt., Medny Is., Commander Islands, Russia, 20 m depth, collected by V.I. Lukin. 2 specimens, 5 August, 1972, Kozyrevskogo Pt., Paramushir Is., Kuril Islands, Russia, 15 m depth, stones, collected by V.N. Romanov. | specimen (juvenile), 5 September 1973, Peregrebnogo Pt., Beringa Is., Kuril Islands, Russia, 20-22 m depth, collected by B.I. Sirenko. 2 specimens, 17 September, 1973, Cherny Pt., Medny Is., Commander Islands, Russia, 10 m depth, rock, collected by B.I. Sirenko. 1 specimen, 31 July, 1978 Iona Is., Okhotsk Sea, Russia, 80 m depth, big and small stones, collected by V.A. Pavlyuchkov. 4 specimens, 11 July, 1985, Lopatka Pt., Kamchatka Peninsula, Russia, 17 m depth, rock and stones, collected by V.I. Shalukhanoy. | specimen, 10 July, 1985, Vladimira Pt., Atlasova Is., Kuril Islands, Russia, 25 m depth, rock. 1 specimen, ZMMU Lc-25738. 14 July, 1985, Sakhalinsky Bay, 54°57.5.N, 141°01.0.E, 95 m depth, R/V “G. Popov,” dredge 5, collected by Yu. I. Kantor. 1 specimen, 4 August, 1985, Okhotsk Sea, 55°26’8”N, 145°55’3”W, 160 m depth, stones, collected by A.V. Smirnov. | specimen, 20 Aug. 1986, Gladkovsky Pt., Medny Is., Commander Islands, Russia, 10-12 m depth, rock, collected by V.V. Oshurkov. | specimen, 25 August, 1986 Monati Pt., Beringa Is., Commander Islands, Russia, 25 m depth, rock, collected by V.V. Oshurkov. 1 specimen, 22 July 1991, Podutesnaya Inlet, Beringa Is., Commander Islands, Russia, 20 m depth, rock, collected by V.I. Shalukhanov. 2 specimens, 14 August 1991, Buyan Pt., Beringa Is., Commander Islands, Russia, 10 m depth, rock, collected by V.I. Shalukhanov. | specimen, 23 July 1992, Sivuchy Kamen’ Kekur Islet, Medny Is., Commander Islands, Russia, 20 m depth, rock, collected by K.E. Sanamyan. Specimens in Kamchatka’s Institute of Ecology and Environment, Petropavlovsk- Kamchatsky, Russia, the Institute of Marine Biology, Vladivostok, Russia, the Zoological Institute, Saint Petersburg, and in the Zoological Museum, Moscow State University, Russia. 1 specimen, 11 July, 1985, Lopatka Pt., Kamchatka Peninsula, Russia, 17 m depth, rock and stones, collected by V.I. Shalukhanov.

EXTERNAL MORPHOLOGY.— The body is elongate-oval and reaches a maximum preserved length of 40 mm (Fig. 1A). The dimensions of one specimen are 40 x 25 x 13 mm. The mantle edge is 4-5 mm, barely covering the sides and tail in preserved animals, although the tail probably extends beyond the mantle in live animals. The notum is densely covered with differentially sized tubercles. The small tubercles, up to 0.4 mm in width, are conical or spindle shaped, the large tuber- cles, up to 1.4 mm in width, are rounded or cylindrical with round tips (Fig. 1B). Spicules lie in the notum radially around the tubercles, most of the spicules in the tubercles have been dissolved by formalin but traces remain, radiating loosely from the centre outward. The rhinophores have 10-15 lamellae and numerous spicules. The rhinophores contract into openings which do not have raised sheaths but are edged all around with 11—12 large tubercles. There are 10—17 tri- or quadripinnate gills which contract into depressions arranged in a broad horseshoe, which is broken posteriorly by a few tubercles. The gill size decreases towards the posterior, but sometimes a few small tripinnate gills are present anteriorly between the larger gills. Within the gill circlet are many large and some small tubercles, about 80 in total. The anal opening is not raised and is slightly left of centre towards the posterior end of the circlet.

The color in life, according to Dall, is light yellow (Bergh 1880). Preserved specimens are dirty white or pinkish-tan.

The head is veliform, not high, and slightly narrower than the foot. The mouth opening is round and the rounded tentacles are thicker on the anterior edge (Fig. 1C). The foot is broad and bilabiate anteriorly, posteriorly it extends into a small leaf-like tail, which is papillate on the dor- sal surface.

ANATOMY.— The oral tube is short and muscular. The anterior half of the buccal bulb has a poorly developed, sessile, rounded buccal pump (sucking crop) with no median muscular band

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS — 5

FiGureE 1. Akiodoris lutescens, drawn with a camera lucida. A. Whole animal, dorsal view. B. Tubercles, one showing traces of spicules. C. Head and anterior of foot. D. Radula, one half row and rachidian tooth. A. adult, B. juvenile. E. Buccal bulb, A. lateral view B. dorsal view. F. Reproductive system. Key: al, albumen gland; am, ampulla; be, bursa copulatrix; hd, hermaphroditic duct: id, insemination duct; mu, mucous gland; p, penis; pr, prostate; rs, receptaculum seminis; v, vagina; vd, muscular vas deferens.

(Figure 1E). The anterior part of the buccal pump has longitudinal muscular fibres, followed by a band of transverse fibres which extends down the sides of the buccal bulb. Posteriorly there are three longitudinal bands of muscles on each side of the esophagus. Ventrally there is a small pos- teriorly projecting radular sac. The round lip disk is muscular with a thick outer layer, smooth and yellowish in color. This outer layer is fleshy with only a thin cuticle coating. The radula has 52-69 rows of reddish brown teeth with the formula (8-13.2.1.2.13-8) (Figs. 1D, 2A—B). The central tooth

6 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

FiGure 2. Akiodoris lutescens. Specimen from Lopatka Pt., tip of Kamchatka Peninsula, Russia, 11/7/85. A. SEM micro- graph of radula, scale bar = 400 um. B. SEM micrograph of central region of radula. Scale bar = 200 um. C. SEM micro- graph of entire penis. Scale bar = | mm. D. SEM micrograph of one posterior penial spine. Scale bar = 40 um.

is wide and flat in the shape of a raised arch. Sometimes a small, posteriorly directed, chitinised denticle can be seen. The two lateral teeth per side can be seen in Fig. 2A—B. The first lateral teeth have a rectangular base and a short, wide, recurved cusp with one inner and one or two outer den- ticles. The second lateral teeth have a rectangular blade with a large quadrangular cusp on the inside and 5—6 outer denticles. The marginal teeth have rectangular bases and a small projection on the inner posterior corner. They diminish in size towards the outside, the last two being very small and narrow. The remainder of the digestive tract is as described by Bergh (1880), except that the ribbon shaped salivary glands are shorter, ending at the digestive gland.

The large blood glands are pinkish, not white, in color. The central nervous system is as described by Bergh (1880).

The reproductive system is shown in Figure 1F. Bergh (1880) described this system well although he only illustrated portions of it. The vagina is lined by villi, 0.6—0.8 mm long, with a dense axis and non glandular, ciliated epithelium. The penis in Bergh’s specimen was retracted, the glans only partly everted. A fully everted penis is shown in Figure 2C and a closeup view of a pos- terior penial spine in Figure 2D. The reproductive openings are located on the right side about one third of the body’s length from the anterior mantle edge. The surface around the genital openings is covered with numerous soft villi approximately 0.3 mm in length.

NATURAL History.— The known range is expanded from its original localities of Atka Island and Kyska Island in the Aleutians, westward and northward to the Commander Islands, southward along the Kamchatka peninsula and down the chain of Kuril Islands to Shikotan near Hokkaido and into the Sea of Okhotsk to Iona Island near its centre and Sakhalinskiy Bay in the west. It has been

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS — 7

collected from low water line to depths of 780 m, usually between 10 and 25 m in the months of July, August and September. The only juvenile was collected in September. All specimens have been collected from rocky areas.

Akiodoris salacia Millen, sp. nov. (Figs. 3-4)

Akiodoris sp. 1 Behrens, 1991:51, #68

EtTyMoLoGy.— Named for the Roman sea-goddess Salacia, wife of Neptune.

MATERIAL.— HovoryPe: California Academy of Sciences, CASIZ 110807, 14 September 1985, Tyee Point, Copper Cove, British Columbia, Canada (49°22’8”N, 123°16’5”W), 20 m depth, on hydroids, rocky substrate, collected by S. Millen. PARATYpPES: Royal British Columbia Provincial Museum, RBCPM 005-000013-001, 1 specimen, collected with the holotype. Zoologi- cal Museum of Moscow University, ZMMU, LC-25737, | specimen, collected with the holotype.

EXTERNAL MORPHOLOGY.— Preserved specimens of this small dorid range in length from 4 to 6.5 mm. The largest animal measures 6.5 x 3 x 2.5 mm (I X w x h). The body shape (Fig. 3A) is elongate-oval, wider in front than behind, with a slightly trailing tail. The mantle margin is fairly wide, covering the high sides and head but not the tail. Posteriorly, above the tail, the marginal edge is indented. The notum bears elongate, slightly inflated, spear shaped tubercles with pointed tips (Fig. 4A). There is little variation in tubercle size, which ranges from 0.08 to 0.4 mm in height and 0.06 to 0.18 mm in width. The numerous tubercles are spaced evenly over the notal surface, with a few small ones in the mid-dorsal area and some medium sized ones near the mantle edge.

Spicules are found in the lower “4 of each tubercle, and their tips protrude slightly. The pro- truding spicules are rod-like, slimmer than those forming the central shaft inside each tubercle. At the bases of the tubercles, spicules extend in a radial, star-like pattern which can be seen through the notum. Within the notum there are scattered spicules forming a mesh-like network which forms a radial pattern towards the mantle margin. This pattern is only visible on the ventral surface of the mantle margin. There are scattered quadrate spicules stiffening the sides of the body.

In the foot, spicules form a mesh pattern on the sole, and the dorsal surface of the foot flange and the tail are heavily spiculated. Numerous sharp pointed spicules protrude in these areas, espe- cially on the tail. The slightly raised margins of the rhinophores each bear 4—5 tubercles, the largest being the innermost anterior one. There are no tubercles within the branchial arch.

The rhinophores are sturdy and end in a rectangular, blunt tip. The stalk is short and most of the clavus bears wide, thick lamellae supported by spicules. The 8—10 lamellae are attached along a vertical, anterior line which appears inset due to the sudden indentation of the wider lamellae. The lamellae slope ventrally and become narrower towards the posterior where they meet forming a chevron.

The 4-6 contractile gills are bipinnate and incompletely tripinnate, non-retractable, separated on the outside, but joined at their bases towards the inside. They are arranged in a semi-circle which is open posteriorly. The gills are largest anteriorly and directed towards the posterior. The gills are surrounded by tubercles, but none are within the small semicircle. The anus 1s at the posterior end of the gill arch.

Living specimens (Fig. 3A) are white or faintly yellow in color. There are sometimes a few opaque white specks in the dorsum along the sides and in front of the rhinophores. In immature animals the central area appears pinkish due to the underlying color of the reddish-brown digestive gland. In more mature animals, the digestive gland is covered by creamy-yellow gonads giving the central area a peach hue. Ventrally the digestive gland shows only faintly through the foot. The

8 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

FiGureE 3. Akiodoris salacia sp. nov. Specimen from Tyee Pt., near Vancouver, British Columbia, Canada, 14/9/85. A. Photograph of a live specimen of A. salacia, 4.6 mm in length. B. SEM micrograph of the radula. Scale bar = 20 tum. C. SEM micrograph of the lateral teeth. Scale bar = 10 um. D. SEM micrograph of the penis. Scale bar = 40 Um.

anterior edge of the foot and the head are sometimes pale yellow. The rhinophores are pale yellow and the gills are pale yellow or white. An unusual feature of this species is a series of opaque white, round, granular glands along the overhanging posterior border of the mantle. There are 8—10 of these glands, the smallest being located on the midline.

The head (Fig. 4B) is composed of two triangular lobes which end in flap-like tentacles sepa- rated from the rest of the head by a dorsal groove. The foot is narrow and elongate, wider and trun- cate anteriorly. The anterior edge is bilabiate. The triangular tail protrudes up to 0.5 mm. It is thick but does not have a dorsal keel. Noteworthy are its many, posteriorly projecting spicules.

ANATOMY.— The oral tube is a muscular ring with short labial glands surrounding it. The buc- cal bulb (Fig. 4C) has a dorsal, rounded, sucking crop which is sessile, with no median muscular band and weak peripheral muscles. The radular sac is short and protrudes from a wide, muscular odontophore. The round lip disk is thinly cuticularized and smooth.

The radula has 35—39 tooth rows. The radula formula is 3-5.2.1.2.5-3 (Figs. 3B—C, 4D). There is no well-developed rachidian tooth but the central area has a thickened, raised plate which begins at the fourth or fifth tooth row. The plate is 3—6 um long. The innermost lateral tooth has a rectan- gular base and a large, recurved, pointed cusp. It has one inner and two or three outer, well devel- oped denticles. These laterals range in width from 12.5—24 um and 13-19 um in length. The outer lateral teeth have a rectangular base with a large, recurved hook on the inside and three or four well developed outer denticles. The outer laterals are 18—19 um wide and 12—18 um long. Each mar- ginal tooth is a flat, rectangular plate with slightly rounded outer corners and a small spine on the inner posterior corner. The marginal teeth diminish in width towards the outside. The marginal

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS 9

Ficure 4. Akiodoris salacia, drawn with a camera lucida. A. Tubercles, one showing internal spicules. B. Head and ante- rior of foot. C. Buccal bulb. A. dorsal view, B lateral view. D. Radula. One half row including rachidian plate. E. Reproductive system. Key: al, albumen gland; am, ampulla; be, bursa copulatrix; hd, hermaphroditic duct; id, insemination duct; mu, mucous gland; p, penis; pr, prostate; rs, receptaculum seminis; v, vagina; vd, muscular vas deferens.

10 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

teeth range in width from 6-19 um and in length from 6-10 um.

The salivary glands are small, flattened, elongate-oval masses lying under the buccal bulb and attaching to it at the anterior end of the esophagus (Fig. 4C). The esophagus is a narrow tube which becomes slightly wider and softer before entering the digestive glands. The digestive glands appear as one elongate mass, wider in front than behind and truncated on the anterior right side. The stom- ach is sac-shaped, almost entirely separated from the digestive glands. It has a small, round cae- cum on the right border of the anterior part of the stomach, tucked in the loop of the intestine. The intestine is broad and runs to the right of the midline curving inward to the anus. The anal opening is not raised.

The pericardial sac contains a posterior, thin-walled, triangular auricle and a smaller, more muscular, diamond shaped ventricle. The flattened, compact blood glands are located just posteri- or to the central nervous system.

The fused cerebro-pleurals are oval and joined by a short commissure. The olfactory nerves have bulbous bases with small attachment points. The large, black eyes have very short stalks. The oval pedal ganglia, smaller than the cerebro-pleurals, are located directly below them and are con- nected to them by a short commissure. They are connected together by a commissure which is slightly longer than the cerebral one. The small oval, paired buccal ganglia adjoin each other.

The reproductive system is illustrated in Figure 4E. It lies on the right side with the loop of the vas deferens extending from the region of the mouth to a short distance in front of the gills. The ovotestis consists of oval, creamy-yellow lobules lying on the dorsal and lateral surfaces of the digestive glands. The branched gonoducts are broad bands which unite to form one central pre- ampullar duct. This duct widens slightly into a tubular ampulla. There is a short post-ampullar duct which bifurcates into an oviduct and a long vas deferens. A short distance from its origin, the vas deferens becomes prostatic, loops posteriorly towards the ovotestis, encircles the bursa copulatrix and continues anteriorly. It then becomes non-prostatic, loops back on itself and bends downward. The long, tubular, non-prostatic portion enlarges slightly to form a penal bulb containing the uncoiled vas deferens. The penis joins a common atrium with the vagina. It is a cylindrical papil- la. Inside, the distal portion of the vas deferens is weakly chitinised and has small spines, 5-8 um in length. The penis (Fig. 3D) is eversible, but the spines were not everted on any of the specimens.

The vagina is long and cylindrical, wider near its opening posterior to the penis in the com- mon atrium, and gradually narrowing. It loops back on itself and then gives off a short duct to the large, round bursa copulatrix. The insemination duct, after a short distance, is joined by a short duct from the tubular receptaculum seminis. The insemination duct is long and winding. It terminates at the oviduct near its junction with the post ampullary duct and entrance to the female gland mass.

The female gland mass has an oval, yellow albumen gland. The hermaphroditic ampulla runs down its anterior surface.Surrounding it, anteriorly, posteriorly, and on the inner side is a white mucous gland. The separate oviductal opening is ventral to the common genital atrium. The two reproductive openings are located high on the right side about '/ of the body’s length from the ante- rior mantle edge.

NATURAL History.— The specimens described in this paper were found in September at a depth of 20 m. on a rocky bottom swept by moderate currents. They were all found on an old hydroid colony intermingled with the ctenostome bryozoan Farrella elongata, which is presumed to be their food. The larger specimens had partially mature reproductive systems and were copu- lating, but no spawn was present or laid in the laboratory during the week subsequent to their cap- ture.

REMARKS.— The newly described species is placed in the genus Akiodoris Bergh, 1879 because it has a nearly identical reproductive system, bi- and tripinnate gills, smooth lip disk, ses-

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS 11

sile pharyngeal bulb, stomach which is free from the digestive glands, digestive caecum, and armed penis. However, in the presence of a slight posterior mantle reduction and gills arranged in a small, non-tuberculose semi-circle with a posterior anus, A. salacia is more similar to the genus Doridunculus. We considered this similarity as superficial due to the small size of A. salacia and relative frequency of appearance of a pair of lobe-like structures in the posterior part of the notum in phanerobranch dorids, apparently caused by partially delayed ontogenetic processes in notum formation.

There are a number of differences between Akiodoris lutescens and A. salacia. Externally, the tubercles differ in that A. salacia has long, similar sized, lanceolate tubercles with projecting spicules, whereas A./utescens has smaller, conical and rounded tubercles of varying sizes. In A. salacia, the gill opening is in a small semi-circle with no enclosed tubercles and the anus is at the posterior edge. In A./utescens, the gill pocket forms a large horseshoe enclosing many tubercles and the anus. The posterior mantle margin is slightly indented in the new species, but this does not appear to be the case in A. /utescens. Both species have a flat leaf like tail which is dorsally spic- ulated. The distinctive mantle glands of A. salacia can only be seen in live specimens, so their pres- ence cannot be used as a distinguishing feature. Only preserved Akidoris lutescens have been observed. Internally, the radula has fewer (4-5) marginal teeth than a similar sized (6 mm) juvenile A. lutescens, which had 7—9 marginal teeth (Fig.1D). The thickened cuticular central tooth observed in specimens of A. salacia, is not well developed, but it may develop in older specimens. Only in larger A. lutescens does the central cuticle form a chitinised spine on the rachidian tooth. The central nervous system of the two species varies in that the eyes are large and sessile in A. sala- cia and smaller and on long stalks in A. /utescens and the unusual lateral position of the pedal gan- glia found in A. lutescens was not found in A. salacia.

Genus Armodoris Minichevy, 1972 TYPE SPECIES Armodoris antarctica Minichey, 1972, by monotypy

DIAGNOsIs.— The notum is spiculose, with rounded tubercles. Posterior part of the notum is round and fully covers the tail. Gills are in a semicircle. Head has a small four-corned oral veil. Anus dorsal. Buccal pump is oval and prominent. Radular formula is (4-8.4-6.1.4-6.4-8). The cen- tral tooth is a square plate of moderate size with a slightly prominent central cusp. The lateral teeth have a rectangular base and one long cusp to the inside of center. Denticles are found on the first 6 teeth, the outer 4-8 have one reduced cusp or are reduced to rectangular bases. Cerebral and pleu- ral ganglia are fully separated. Only a small dorsal part of the stomach is free from the digestive gland. Gonad covers most of the digestive gland and stomach. Prostate massive. Interior of vagina is simple. Ejaculatory duct has simple spines.

The genus contains only the type species.

Redescription of Armodoris antarctica Minichev, 1972 (Figs. 5-6) Armodoris antarctica Minichev, 1972:366—368, fig 4; Cattaneo-Vietti et. al., 2000:175.

TYPE MATERIAL.— HOobortyPe: ZIN N 14 (dissected, most internal organs are removed and are

4 ZIN — The Zoological Institute, Russian Academy of Sciences, St. Petersburg; the numbers refer to systematic cat- alogue, without inventory numbers in this case.

5 There are no exact data in the first description of Armodoris antarctica, nor on the label that accompanies the holo- type. except for Minichev’s personal label “19-2”. The present date, depth and substrata is reconstructed by comparison with neighboring labels “19-1” and “19-3”, which both have complete or semi-complete data.

1, PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

not traceable, radula slide not found), around January 19, 19665, 11 Soviet Antarctic Expedition, Davis Sea, Tokarev Id., Between Samples 45 and 55, Depth about 30—32 m, rock, collected by A. F. Pushkin. OTHER MATERIAL: 2 specimens (ZIN N 2%, N 34), Zoological Institute, Russian Academy of Sciences, February 1, 1969, 13 Soviet Antarctic Expedition South Shetland Ids., King-George Id., La Guardia Nacional Bay, transect I, St. XXI, sample 60, 36 m depth, stones, col- lected by A. F. Pushkin.

EXTERNAL MORPHOLOGY.— Preserved specimens range in length from 13 to 16 mm. The holotype measures 16 x 9 x 5 mm (1 X w x h) (Fig. 5 A-C). The specimen, ZIN N 3, is 13 x 7.4 (1 x w). The body shape is elongate-oval, slightly wider in front than behind. The mantle margin is amply wide, fully covering the sides and tail. Posteriorly, the marginal edge is without any traces of an indent. The notum bears tubercles of different sizes, scattered small tubercles between the larger ones. In the middle of notum, longitudinally, there are some rather large, pear-shaped tuber- cles. They become more elongate or cylindrical toward to the notal margins. Tubercles sizes range from 0.07 to | mm in height and 0.06—0.7 mm in diameter. Spicules within a tubercle are rod-like, varying in size, nearly straight or slightly crooked. The spicules sometimes slightly extruded through upper part of the tubercle. The margins of the rhinophore pockets are very low and bear 6 blunt tubercles (right side of the specimen ZIN N 3) and 4 in the holotype. The rhinophores have about 10 lamellae.

Six (ZIN N 3) to nine (holotype) contractile, but non-retractile, gills are unipinnate and incom- pletely bipinnate (Fig. 5D). They are arranged in a semi-circle, which is posteriorly completed by a large tubercle. The gills are surrounded by elongate tubercles. The holotype and ZIN N 2 speci- men have three tubercles inside the gill semicircle, whereas example ZIN N 3 has none. The anus is in front of the posterior tubercle in the 13-mm length specimen. Preserved specimens are whitish-pinkish or yellowish.

The head is a small semi-oval oral veil with two short blunt lobes extending posterio-laterally (Fig. 6C) The foot is rather wide, anteriorly truncated without any traces of post oral lobes. Posteriorly, the foot narrows slightly to form a flat, blunt tail (Fig. 6C).

ANATOMY.— The oral tube is short and muscular. Almost the entire upper part of the buccal bulb has a well-developed, oval buccal pump (Fig. 6E). The buccal pump has no peripheral mus- cle (median muscular band). The anterior part of the buccal pump has longitudinal muscular fibers, followed by a band of transverse fibers, which extends down the sides of the buccal bulb. The round lip disk is smooth, covered by a colourless cuticle. Radular formula is 36—52 (4-8.4-6.1.4- 6.4-8) (Fig. 6D). The central tooth is a thin square plate of moderate size with a slightly prominent central cusp. The inner laterals are denticulate, and they gradually transform to the simple small plates of the outermost laterals. The first and at least 4-5 following laterals have a rectangular base and short or moderately long cusp (cusp of 2"4 and 3" laterals, in general, stronger than first ones) directed downward, usually straight or slightly curved. The first three or four inner laterals have 1—3 denticles on both sides of the cusp, the following 3-5 laterals have a serration (or it is reduced) on the outer side of the cusp. The remaining two to four laterals have a greatly reduced cusp, devoid of denticles, and the outermost tooth is just a simple plate. The elongate-oval salivary glands are small and attach to the buccal bulb at the anterior end of the esophagus (Fig. 6E). The latter is a rather wide, flattened tube. The stomach is sac-shaped, only a small dorsal portion is separate from the digestive gland (Fig. 6B). There is no caecum on the stomach.

The pericardial sac contains a posterior, thin walled narrow-triangular auricle and smaller, oval shaped ventricle (Fig. 6B). The blood glands are slightly visible and located posterior to the cen- tral nervous system.

The cerebral and pleural ganglia (Fig. 6B, 6F) are separated and the later are smaller. The

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIJODORIS AND ARMODORIS 13

FIGURE 5. Armodoris antarctica, holotype. A. dorsal view. B. Ventral view. C. lateral view. D. Close up of the gills and tubercles.

large, black eyes have very short stalks. The pedal ganglia, similar in size to the cerebral, lie below them and are connected to them by short commissures. The rhinophoral ganglia are roundish. The buccal ganglia are roundish-oval. Gastro-esophageal ganglia were not found. In the cerebrals, three pairs of cerebral nerves were detected, in the pleurals, four, and in the pedals, three.

The ovotestis is characterised by large female follicles (Fig. 6B). The ampulla is voluminous

14

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

(9, eT

a NY i Kv) werceprc /<* : ——-

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIJODORIS AND ARMODORIS 15

and kidney-shaped (Fig. 6G). There is a short post-ampullar duct, which bifurcates into an oviduct and a long vas deferens. The pre-prostatic duct is relatively long and wide. The prostatic part of vas deferens is a very massive tangle of a few thick, flattened loops. The prostate doesn’t encircle the bursa copulatrix. The vas deferens narrows rapidly to a non-prostatic portion, loops back on itself, and bends downward. The deferent duct is a thin, muscular, not very long, gradually enlarging duct, which widens to an elongate penial bulb. The penial bulb joins a common atrium with the vagina. The ejaculatory duct of vas deferens has inside tiny, densely placed, spines. A partially everted penis is very narrow. The vagina is rather short and wide, similar in diameter its whole length, bending in the middle and entering the large, irregularly-triangular, bursa copulatrix. Villi inside the vagina are absent. The receptaculum seminis is oval and has a long, muscular, gradually widen- ing duct, which joins the middle of the insemination duct. In the holotype, the receptaculum sem- inis is more elongate. The insemination duct is long, leaving the vagina near the middle and enter- ing the female gland mass near the oviduct.

REMARKS.— The genus Armodoris has at least four to six well developed inside lateral teeth and no clear border between inner and outer laterals, whereas the genera Akiodoris, Prodoridunculus and Doridunculus have two well-differentiated inner laterals and rest of the later- als demonstrate reduction. Echinocorambe has three inner laterals, markedly different from the outer teeth. The radulas of A. lutescens and A. salacia are similar in the shape of the two first lat- erals and in the reduction of most of the outer laterals. This is a parallel case with the genus Adalaria, where the first lateral is more or less stable, but outer laterals have tended to reduce. The stomachs of Akiodoris and Doridunculus are massive and mostly free from the digestive gland, whereas Armodoris has a rather small free part of the stomach. Armodoris has large female folli- cles, which may be due to direct embryonic development, but the stomach is not fully surrounded by the gonad, as erroneously reported by Minichev (1972). The ampulla of Armodoris is large, more like that of Doridunculus echinulatus. Both Akiodoris species have similar long, but tubular prostates, whereas the prostate of Armodoris is massive. The receptaculum seminis is very narrow and tube-shaped in both Akiodoris and Doridunculus, but in the King-George’s examples of Armodoris it is oval and in the holotype it is narrow-oval. The villi of the vaginal duct are not pres- ent in Armodoris. The buccal pump in Armodoris antarctica is more developed, although it has the same structure as found in Akiodoris, Doridunculus and Echinocorambe. Due to above differences between Armodoris and the other four genera with similar radular teeth, it is suggested that we maintain the genus Armodoris.

Compared to the King George Island examples, the dissected holotype has well preserved gills and oral veil. Minichev (1972) incorrectly interpreted the reproductive system and partially the digestive system of A. antarctica. He found only one receptaculum (“spermatheca” in his terms) and although he noted that “a “spermatocyst’ (= bursa copulatrix) is not evident in the present mate- rial”, nevertheless, he considered this case to be a “unique feature” (Minichev 1972:366). The holo- type originated from the Davis Sea, and the two additional specimens originated from another cor- ner of Antarctic, the South Shetland Islands. An additional specimen of A. antarctica is the “Italian

FiGuRE 6. Armodoris antarctica drawn with a camera lucida. A. Dorsal view, scale bar = 1 mm. B. Dissected view, scale bar = | mm. Key: bb, buccal bulb; be, bursa copulatrix; cns, central nervous system; fgm, female gland mass; g, gonad; h,

half row including rachidian tooth. E. Buccal bulb, Dorsal and lateral views. Scale bar = 0.5 mm. F. Central nervous sys- tem. Scale bar = 0.5 mm. Key: C 1-3, cerebral nerves; Cg, cerebral ganglia; P 1-3, pedal nerves; PC, pedal connective; PPC, pleuro-pedal connective; Pg, pedal ganglia; Pl 1-4, pleural nerves; Plg, pleural ganglia; R, rhinophoral nerve; Rhg, rhinophore ganglia: St, Statocyst; VL, visceral loop. G. Reproductive system. Scale bar= 1 mm. Key: am, ampulla; bc, bursa copulatrix: fgm, female gland mass; id, insemination duct; ps, penis sheath; pr, prostate; rs, receptaculum seminis; v, vagi- na: vd, muscular vas deferens.

16 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

sample” from the stomach of a fish, Trematomus bernacchii, caught in the Ross Sea, Terra Nova Bay (Cattaneo-Vietti 2000). By the courtesy of Stefano Schiaparelli, we have studied a SEM image of the radula of the Italian specimen of A. antarctica. The radulas of all four specimens of A. antarctica are similar. The rarity of Armodoris antarctica is caused by its inhabiting only shallow waters on rocky substrates, and it can only be found by diving collections, which is certainly rare in the Antarctic, especially around 30 m (both holotype and two additional samples were collected slightly below 30 m on rocky substrates and in remote localities).

PHYLOGENETIC ANALYSIS

Dorid nudibranchs are currently divided into two groups, those with gills and rhinophores which retract into a closable pocket, the Eudoridoidea (=Cryptobranchia), and those whose gills and rhinophores have no pocket or an open pocket, the Anadoridoidea (= Phanerobranchia). The phanerobrachs dorids were in turn subdivided by Bergh (1892) into two groups, those with a suc- torial buccal apparatus (suctorians) and those without (non-suctorians). These are informal groups, but the three families that Bergh placed into the suctorians, Corambidae, Goniodorididae, and Onchidorididae, are commonly arranged together. There have been numerous rearrangements with- in these three families over the years, but currently the family Onchidorididae is believed to con- tain the genera Acanthodoris, Adalaria, Akiodoris, Arctadalaria, Armodoris, Calycidoris, Diaphorodoris, Doridunculus, Onchidoris and Prodoridunculus, although Wagele and Willan (2000) consider the genera Akiodoris and Armodoris to belong to the Goniodorididae following Bergh (1883), who placed them at the base of the Goniodorididae.

Four of these genera form a distinct clade based primarily on their radular characteristics (Millen 1987). Most genera in the family Onchidoridae have on each side a single, large, strongly hooked lateral with a triangular base and narrow, hooked, marginal teeth. The rachidian, where present, has an elongate, rectangular shape. In contrast, a clade of four genera, Akiodoris, Armodoris, Doridunculus, and Prodoridunculus, have two or three (sometimes up to 6), wide, rec- tangular lateral teeth per side. The first lateral has internal and external denticles, the second and third, when present, have external denticles. The rachidian tooth, if present, is arch-shaped. The marginal teeth are wide and rectangular with a small hook. These four genera also have a relative- ly simple buccal pump.

Recent cladograms using either molecular or morphological data, which have included a vari- ety of cryptobranch and phanerobranch dorids, have suggested that the phanerobranch dorids may, in fact, be polyphyletic (Wollscheid and Wiagele 1999; Wagele and Willan 2000; Wollscheid- Lengeling et al. 2001; Vallés et al. 2001; Valdes 2002). These studies have not shown suctorian dorids to be polyphyletic, nor have they been able to use the full range of genera in the family Onchidoridae. The genera used were those with hook-shaped teeth, Acanthodoris, Adalaria, Calycidoris, Diaphorodoris, and Onchidoris, which generally clustered together. To observe the characters which distinguish Akiodoris and Armodoris from other genera and to establish their closest relationships, and to which family they should be assigned, we produced a cladogram using all of the possible Onchidorididae genera. These were compared to sister taxa consisting of three genera from the Corambidae (Corambe, Loy and Echinocorambe) and one Goniodorididae (Goniodoris). One cryptobranch dorid, family Chromodorididae (Cadlina), was used as a sister taxa to the phanerobranchs. The outgroup and basal rooting was based on the basal Anthobranchiate dorid genus Bathydoris, using B. spiralis Valdés, 2002, which is closest to the cryptobranch and phanerobranch dorids (fide Valdés 2002). In total, 15 taxa and 29 characters were analysed using MacClade and PAUP 3.1.1.

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS_ 17

The 29 characters used in this analysis are listed below: 23 characters are binary and 6 are mul- tistate. Character states are indicated by numbers; 0 is the plesiomorphic condition and 1—2 are apo- morphic. Non-applicable and unknown characters are coded with a ?. Polarities are the result of outgroup comparison. Table | (see Appendix) shows the distribution of plesiomorphic and apomor- phic states.

1. Mantle rim: The mantle rim (projecting edge) 1s narrow, partially absent in Bathydoris and some phanerobranch dorids (0) whereas it covers the sides and foot in most dorids (1).

2. Posterior mantle rim: The posterior mantle rim, where distinct, is usually entire in Bathydoris and most dorids (0). It is bilobed in some phanerobranch dorids (1).

3. Tail: The tail is extended in Bathydoris and some phanerobranch dorids (Q). It is small and covered by the mantle in most dorids (1).

4. Ridge on tail: The tail is usually smooth in Bathydoris, cryptobranch dorids and most phanerobranch dorids (0). It has a mid-dorsal crest or ridge in some phanerobranch dorids (1).

5. Notal spicules: Some Bathydoris and most dorid nudibranchs have integumentary spicules in the notum (Q). These are absent in a few dorid species (1).

6. Notal sculpture: Sculpture in the form of deciduous villi is present in Bathydoris (0); most species of dorid nudibranchs have well attached tubercles (1). Only a few species are smooth (2).

7. Head: The head in Bathydoris and most phanerobrach dorids is veliform (0). Cryptobranch dorids have a round head and separate tentacles (1).

8. Rhinophore sheath: Bathydoris and some phanerobranch dorids have rhinophores which lack sheaths and are attached directly to the dorsum (0). Other phanerobranch dorids have non-con- tractile sheaths into which the rhinophores can retract (1). Cryptobranch dorids have contractile rhinophore sheaths which close over retracted rhinophores (2).

9. Rhinophores: Bathydoris and most other dorids have lamellate rhinophores (0). A few dorids have smooth rhinophores (1) which may have vertical envelopes (2).

10. Gill protection: Bathydoris and some phanerobranch dorids have gills which lack sheaths and are inserted directly on the notum (0). Some phanerobranch dorids have sheaths into which the gills can retract but which do not close over the gills (1). Cryptobranch dorids have contractile gill sheaths which close over the retracted gills (2). For taxa with no mantle gills, this character has been treated as non-applicable.

11. Gill number: Bathydoris and most dorids have more than five gills (0). A few species of phanerobranch dorids have less than five gills which is considered a reduction (1).

12. Gill location: Bathydoris and most dorids have the gills located on the mantle some dis- tance from the edge (0). Some phanerobranch dorids have the gills posterior at the mantle edge or below it (1).

13. Notal sculpture within the gill circlet: Bathydoris and cryptobranch dorids do not have papillae or tubercles within the area encircled by the gills (0). Some phanerobranch dorids have notal sculpture in this area (1). For taxa with no mantle gills, this character has been treated as non- applicable.

14. Anus location: Bathydoris and most dorids have the anus located on the mantle some dis- tance from the edge (0). Some phanerobranch dorids have the anus posterior at the mantle edge or below it (1).

15. Lip disk: Bathydoris has thick chitinous labial armature and other dorids may have vari- ous rods or papillae to strengthen the lip disk (0); a smooth, thin cuticle is considered apomorphic (1).

16. Buccal pump: A sucking pump on the pharynx is absent in Bathydoris, non-suctorian

18 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 1

phanerobranch dorids and cryptobranch dorids (0). This pump distinguishes the suctorian phaner- obranch dorids (1).

17. Peripheral muscles: A median longitudinal band of muscles is not found on the phanynx of Bathydoris and most dorids (0). This muscle is found on some of the suctorian phanerobranch dorids (1).

18. Rachidian tooth: A rachidian tooth is present in Bathydoris and some dorids (0). Many dorids have lost this central tooth (1).

19. Number of teeth per half row: Primitive dorids and Bathydoris have many teeth, >10, per half row (0). Reduction has resulted in narrower radulas with fewer than 10 lateral teeth (1).

20. Number of inner laterals: Inner lateral teeth tend to be more highly developed than the outer lateral (marginal) teeth which are often reduced. In Bathydoris, and some dorids, only the first lateral tooth differs from the outer lateral teeth (0). In other dorids there are 2 or more teeth which can be differentiated from the outer laterals (1).

21. Shape of innermost lateral tooth: Bathydoris, some cryptobranch and some phanero- branch dorids have cuspidate innermost lateral teeth. The cusp is towards the center of a broad tooth and denticles may appear on one or both sides (0). Other, phanerobranch dorids have beak- like teeth, with a long cusp on the outer edge, and denticles, when present, only on the inner side (1).

22. Denticulation on innermost lateral tooth: Bathydoris has denticulations on both sides of the inner lateral teeth, as do some dorids (0). Other dorids, used here, may have denticulations only on the inner side (1) or they may be absent (2).

23. Marginal teeth or outer lateral teeth: In Bathydoris and many cryptobranch dorids, these are elongate, hamate teeth, with or without denticles on the outer side (0). In some phanerobranch dorids they are reduced to either an elongate-oval plate (1) or a wide, rectangular plate (2).

24. Ampulla: The hermaphroditic ampulla is narrow and undifferentiated in Bathydoris and many dorids (Q). It is wide and usually short in some dorids (1).

25. Penial spines: Spines on the eversible vas deferens of the penis are absent in Bathydoris and some dorids (0). Small spines are present in a number of dorids and their presence 1s consid- ered an apomorphy (1).

26. Seminal receptacle: This sac for sperm storage is absent in most Bathydoris species, but present in Bathydoris spiralis as a small sac with one duct (stalked) (0). In some dorids, it has two ducts (unstalked), which is considered derived (1).

27. Seminal receptacle insertion: This sac is inserted by its duct on the vagina in Bathydoris spiralis and many other dorids (0). In some dorids, it is inserted part way along the insemination (uterine) duct (1).

28. Oviduct: The oviduct carrying the eggs usually passes directly into the female gland mass (O). In a few dorids, it enters the seminal receptacle before entering the female gland mass, which is considered apomorphic (1).

29. Cerebro-pleural ganglia: These ganglia are separate in Bathydoris and a few dorids (0). They are fused in most dorids (1).

PHYLOGENETIC RESULTS

The data matrix results in one tree (Fig. 7), 64 steps long, with a consistency index of 0.562 and a retention index of 0.692. The Bremmer support index values (values to the right of the line in Fig. 7) show that most of the branches are poorly supported, with the exception of the clade con- taining those phanerobranch dorids with beak-shaped laterals (value 2) and the node containing Corambe and Loy (value 2+).

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS_ 19

Bathydoris Cadlina Goniodoris Diaphorodoris Calycidoris Acanthodoris Adataria Onchidoris Akiodoris Armodoris Doridunculus Echinocoerambe Prodoridunculus

wo loo vo fe a

_

—)

=

—

~~

a

in WN

is oe Mim _— +

2 3 18 5 18

19

24

A

26 ; 24

20 13 27 16

Figure 7. Strict consensus tree of a preliminary phylogeny. Apomorphies are indicated by character numbers included in the analysis. Characters which exhibit reversal are underlined. Bremer support values are 0 unless indicated by large numerals on the right side of the nodes.

In the resulting tree, the cryptobranch dorid Cadlina separates as a sister group to the reman- ing suctorian phanerobranch dorids, which are supported by the synapomorphies of non-contrac- tile rhinophore sheaths (#8), the presence of papillae within the gill circlet (#13) and the presence of a buccal pump (#16).

These suctorian phanerobranch dorids traditionally belong to three families, Corambidae, Goniodorididae and Onchidorididae, but some members of the three families clustered together into one clade. This clade is distinguished by the presence of a peripherial muscle in the buccal pump (#17), a reduced half row of the radula (#19), one inner lateral tooth (#22), beak-shaped inner laterals (#21), and elongate oval outer laterals (#23). The Corambidae nests within the Onchidorididae and the genus Goniodoris nests with other members of the Onchidorididae and next to Diaphorodoris. Already the Superfamily Onchidoridoidea Gray, 1827 has been used to combine family Corambidae and Onchidorididae by Valdés and Bouchet (1998). Our data suggest that the families Corambidae and Onchidorididae are closely related (see Martynov 1994) and should be combined into one family, as suggested by Martynov (1999); the family group name Onchidorididae Gray, 1827 has priority. Further study is needed to clarify the position of Goniodoris and other members of the Goniodorididae in relation to this family.

20 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. |

The second clade contains members of the ‘basal’ Onchidorididae, which have sometimes been placed into the Goniodorididae. In this clade are the genera studied in this paper, Akiodoris and Armodoris, together with the closely related Doridunculus and the poorly known Prodoridunculus. Another poorly known genus, Echinocorambe, initially was placed in the family Corambidae by Valdés and Bouchet (1998) inasmuch as it has similar paedomorphic features. It was also suggested by Martynov (2000) that it is close to Doridunculus because of its similar radu- la and buccal pump; it may, in fact, be a juvenile of the latter. The cladogram confirms that it belongs in the same clade as Doridunculus. These five genera are in a clade with similar radulas, buccal pumps, and reproductive systems. Synapomorphies which distinguish this clade are the presence of a smooth, thin, lip disk (#15), two or more inner lateral teeth (#20), and the possession of rectangular, reduced outer laterals (#23). The distant position of Goniodoris in relation to this clade suggests that these five genera should not be in the family Goniodorididae. This clade forms a sister group to the clade containing the Onchidorididae (including Corambidae) and Goniodoris. Thus, this clade is considered a distinct family, Akidorididae Millen and Martynov, fam. nov.

ACKNOWLEDGMENTS

We thank Dr. Tanya Korshunova (Moscow, IHNA and NF RAS) for drawings and photographs of Armodoris antarctica. We are also thankful to Dr. Alexi Chernyshev for figures used in plate 1 and additional sketches of Akiodoris lutescens. Ron Long, Simon Fraser University, kindly sup- plied the photograph of Akiodoris salacia. Stefano Schiaparelli provided an SEM image of the radula of A. antarctica. We also thank Sven Donaldson and John McNickol for diving assistance. This research was partly funded by the Department of Zoology, University of British Columbia to Sandra Millen. Partial funding to Alexander Martynov was from the Russian Federal Program “Study of Antarctic biota, Project N 16.” Funding for publication costs is from the National Science Foundation, PEET Program (DEB-9978155) to Dr. Terrence Gosliner.

LITERATURE CITED

BEHRENS, D.W. 1991. Pacific Coast Nudibranchs, A Guide to the Opisthobranchs. Alaska to Baja Califorina, 2nd ed. Sea Challengers: Monterey, California, USA. 107 pp.

BERGH, R. 1879a. Gattungen nordischer Doriden. Archiv. fiir Naturgeschichte 45(1):340—369, pl. 19.

BERGH, R. 1879b. On the nudibranchiate gasteropod mollusca of the north Pacific Ocean, with special refer- ence to those of Alaska. Part I. Proceedings of the Academy of Natural Sciences of Philadelphia 31:71-132, pls 1-8.

BERGH, R. 1880. On the nudibranchiate gasteropod mollusca of the north Pacific Ocean, with special refer- ence to those of Alaska. Part Il. Proceedings of the Academy of Natural Sciences of Philadelphia 32:40—127, pls. 9-16.

BERGH, R. 1883. Beitrage zu einer Monographie der Polyceraden, II. Verhandlungen der konglich-kaiserlich Zoologisch-botanischen Gesellschaft in Wein. Abhandlungen 33:135—180.

BERGH, R. 1892. Malacologische Untersuchungen, 3. System der Nudibranchiaten Gasteropoden. Pages 995—1165 in Carl Semper, Reisen im Archipel der Philippinen, Zweiter Theil. Wissenschaftliche Resultate Band 2, Theil 3, Heft 18.

CATTANEO-VIETTI, R., M. CHIANTORE, S. SCHIAPARELLI, AND G. ALVERTELLI. 2000. Shallow and deep water mollusc distribution at Terra Nova Bay (Ross Sea), Antarctica. Polar Biology 23:173-182.

FISCHER, P. 1833-1837. Manuel de Conchyliologie et de Palentologie conchyliologique ou histoire naturelle des mollusques vivants et fossiles suivi d'un appendice sur les Brachiopodes. Librairie F. Savy, Paris, France. 1369 pp.

FRANC, A. 1968. Sous-classe des opisthobranches. Pages 834-888 in P.P. Grasse, ed., Traite de Zoologie, vol. 5. Masson et Cie., Paris, France.

MILLEN AND MARTYNOV: NUDIBRANCH GENERA AKIODORIS AND ARMODORIS 21

Hols#TER, T. 1986. An annotated check-list of marine molluscs of the Norwegian coast and adjacent waters. Sarsia 71:73-145.

Martens, E. 1879. Emend. for Akiodoris Bergh, 1879. Zoological Record, Mollusca 16:61.

Martynoy, A. 1994. Materials for the revision of the nudibranch family Corambidae (Gastropoda, Opisthobranchia). Part 2. Origin of the Corambidae. Zoologicheskiy Zhurnal 73(11): 36-43 (In Russian). 1995. Hydrobiological Journal 31(7):59-67 (In English).

Martynoy, A. 1997. Opisthobranch molluscs of the coastal waters of Commander Islands with notes on their Fauna of the Far-Eastern Seas of Russia. Pages 230-241 in A. Rzhavsky, ed., Donnaya fauna i flora Komandorskikh ostrovoyv. {Benthic Fauna and Flora of the shelf of Commander Ids.] Dalnauka, Vladivostok, Russia. (In Russian)

Martynoy, A. 1999. [Abstract] Buccal pumps, gills pockets and new understanding of suctorial phanero- branchial dorids. Pages 13-14 in Systematic, Phylogeny and Biology of Opisthobranch Molluscs. 24 International Workshop of Malacology, Menfi, Italy, June 10-14, 1999.

Martynovy, A. 2000. On the taxonomic placement of the genus Echinocorambe Valdés et Bouchet, 1998 (Gastropoda: Nudibranchia). XIV (5) All-Russian Malacological Conference, St. Petersburg, pp. 50-52. (In Russian)

MILLEN, S. 1987. [Abstract] The nudibranch family Onchidorididae: a cladistic analysis. Western Society of Malacol-ogists Annual Report 20:19.

MINICHEV, Y. 1972. Opisthobranchiate molluscs of the Davis Sea. [ssledovaniya Fauny Moreyi 19:358—-382.

ODHNER, N.H. 1907. Northern and Arctic invertebrates in the collection of the Swedish State Museum (Riksmuseum). III. Opisthobranchia and Pteropoda. Kungelige Svenska Vetenskaps Akademiens Handlingar 41:1—114.

Sars, G.O. 1878. Bidrag til kundskaben om Norges Arktiske fauna. I. Mollusca Regionis Arcticae Norvegiae, Oversigt Over de I Norges Arktiske Region Forekommende. Bloddyr., xii + 466 pp., 34 pls. Universitets- program, Christiania, Norway.

THIELE, J. 1912. Die antarktischen schnecken und muscheln. Pages 183-285, pls. 11-19 in Drygalski, ed., Deutsche Sudpolar-Expedition 1901-1903. Zoologie, 13(5, 2).

THIELE, J. 1931. Handbuch der systematishen Weichtierkunde. Bd. 1. Verlag von Fischer, Jena, Germany. 778 Pp-

VaLpEs, A., AND P. BOUCHET. 1998. A blind abyssal Corambidae (Mollusca, Nudibranchia) from the Nor- wegian Sea, with a reevaluation of the systematics of the family. Sarsia 83:15—20.

Vatpes, A. 2002. A phylogenetic analysis and systematic revision of the cryptobranch dorids (Mollusca, Nudibranchia, Anthobranchia). Zoological Journal of the Linnean Society 136(4):535—636.

VALLES, Y., M. MEDINA, AND T. GOSLINER. 2001. [Abstract] Phanerobranch dorids: clade or grade? Page 368 in Abstracts of the World Congress of Malacology 2001. Vienna, Austria.

WAGELE, H., AND R. WILLAN. 2000. Phylogeny of the Nudibranchia. Zoological Journal of the Linnean Society 130:83-181.

WOLLSCHEID, E., AND H. WAGELE. 1999. Initial results on the molecular phylogeny of the Nudibranchia (Gastropoda, Opisthobranchia) based on 18A rDNA data. Molecular Phylogeny and Evolution 13: 215-226.

WOOLSCHEID-LENGELING, E., J. BOORE, W. BROWN, AND H. WAGELE. 2001. The phylogeny of Nudibranchia (Opisthobranchia, Gastropoda, Mollusca) reconstructed by three molecular markers. Organisms Diversity and Evolution 1(4):241—256.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

De,

Appendix

TABLE |. Data matrix showing the distribution of character states (see text for details).

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. |

So

N ro)

Acanthodoris Corambe Diaphorodoris Doridunculus Onchidoris Prodoridunculus Echinocorambe

Goniodoris Calycidoris

Bathydoris Adalaria Akiodoris Armodoris

Cadlina

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 2, pp. 23-30, 2 figs. April 29, 2005

First Hawaiian Record of the Grenadier Lepidorhynchus denticulatus (Macrouridae: Gadiformes: Teleostei)

Tomio Iwamoto Department of Ichthyology, California Academy of Sciences, 875 Howard St., San Francisco, CA 94103 USA; Email: tiwamoto:calacademy.org

On August 16, 2003, while night lighting for squid off the northwest coast of Oahu, Mr. Richard Cadaoas caught on hook and line an 18 to 20-inch (about 46-51 cm) squid. In its stomach was a small fish about 20 cm long in excellent condition, having suffered little deterioration from its stay in the squid’s stomach. The specimen was taken to Mr. Arnold Suzumoto, Collections Manager for the ichthyological collections of the Bernice P. Bishop Museum, Honolulu. The fish could be readily identified as a grenadier, owing to its long tapered body that ends in a point and lacks a caudal fin, the relative form and positions of the fins, the spinulated body scales, and the general form of the head. The snout in the specimen was narrow and rounded, however, unlike the pointed and often stoutly reinforced snout of most other members of the family (excluding the bathygadids), and none of the ridges on the head were strengthened by stout spiny scutelike scales. The fish did not fit the description of any previously known species from the Hawaiian Islands and was subsequently sent to me, as one of the primary workers on this group of deepsea fishes, for identification.

Even before I had taken it out of the jar, I recognized the fish as a Lepidorhynchus denticula- tus, a Species originally described by Sir John Richardson in 1846 from a dried specimen that “was thrown up on the beach of South Australia, and has lost part of its tail” (Richardson 1846:54). Despite the desiccated state of the type specimen, the original description and illustration more than adequately characterized this distinctive grenadier, which is the sole representative of the genus.

The species was previously known only from the coasts of south and southeastern Australia and off the coasts of New Zealand in depths ranging from less than 100 m to more than 1000 m. It is one of the most common grenadier in these waters and forms a significant part of the by-catch of trawlers fishing at upper-slope depths. So far as known, the species has not been recorded north of latitude 34°S off New Zealand and about 30°S off southeastern Australia. Its distribution is gen- erally confined to subtropical waters. Its occurrence off Hawaii is thus most astounding and prompts the recording of this extraordinary specimen.

METHODS

Methods for making counts and measurements follow Iwamoto and Sazonov (1988). Institutional abbreviations follow Leviton et al. (1985).

ep

24 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 2

DESCRIPTION

Lepidorhynchus denticulatus Richardson, 1846

Figs. 1-2

Macrourus vel Lepidorhynchus denticulatus Richardson, 1846:1—2, pl. 32, fig. 1-3 (holotype: BMNH 1845.11.31.6; type locality: coast of South Australia).

Coryphaenoides denticulatus: Gunther, 1862:396.

Macrurus (Optonurus) denticulatus: Giinther, 1887:147.

Optonurus denticulatus: Gilbert and Hubbs, 1916:144.

Lepidorhynchus denticulatus: Phillipps, 1927:58.

MATERIAL EXAMINED.— BPBM 39286 (200 mm total length, tail complete; sex indetermi- nate); Hawaii, Oahu, northwest side about 2 miles (3.2 km) offshore; collected by Richard Cadaoas, 16 August 2004 from stomach of 18—20-inch (46-51 cm) squid caught by hook and line at night. CAS 220920 (6 spec., 240+-456 mm total length); Australia; New South Wales, off Bermagui; 36°46’-41’S, 150°21’E; 285-310 fm [521-567 m]; collected by Ken Graham on F/V “Shelley,” by trawl; 15 February 2000.

COUNTS AND MEASUREMENTS.— (Hawaiian specimen first, followed by range of Australian specimens in parentheses.) First dorsal-fin rays II,11 (10-12); pectoral-fin rays 117 and 116 (115-118); pelvic fin rays 9 (9); branchiostegal rays 6; gill rakers first arch (outer/inner) 0+11 /3+13 (0+10—14 / 3-44+13-15 = 17-18), second arch 3+14 / 2+13 (2—3+12-15 = 14-17 / 2-3+12-14 = 14-17); scale rows below origin of first dorsal fin about 8 (about 7-8), below mid-base of first dor- sal fin about 5 (5.5), below origin of second dorsal fin about 6 (5—6.5); pyloric caeca 16 (15-25). Head length 31.5 mm (41.7—70.3 mm); the following in mm for Hawaiian specimen followed in parentheses by percent of head length for Hawaiian specimen and then Australian specimens: snout length 7.6 (22%, 20-24%): preoral length 5.1 (16%, 9—-13%); internasal width 6.8 (22%, 16-21%); interorbital width 8.8 (28%, 20-25%); orbit diameter 12.6 (40%, 35-39%); suborbital width 2.8 (9%, 8-8%):; postorbital length 12.6 (40%, 41-45%); distance orbit to angle of preopercle 11.9 (38%, 39-42%); length upper jaw 15.3 (49%, 47-50%); length barbel 3.1 (10%, 9—-15%); length

Ye — ¥ . %

a 4 t ¢ rf ¢ x : ’ ia a ,

FIGURE 1. Lepidorhynchus denticulatus, BPBM 39286 (31.5 mm head length, 200 mm total length), from off Oahu, Hawaii.

IWAMOTO: GRENADIER LEPIDORHYNCHUS DENTICULATUS IN HAWAII DES

outer gill slit 8.6 (27%, 22-27%); preanal length 47.5 (151%, 158-177%); distance outer pelvic fin to anal-fin origin 17.7 (56%, 53-69%); distance isthmus to anal-fin origin 32 (102%, 108-139%); greatest body depth 24.5 (78%, 79-90%): body depth at anal-fin origin 20.0 (63%, 64-80%); inter- space between first and second dorsal fins 23.5 (75%, 74-95%); height first dorsal fin 28.5 (90%, 78+—92); length base of first dorsal fin 11.1 (35%, 34-42%); length pectoral fin 23 (73%, 72-82%); length outer pelvic-fin ray 20.5 (65%, 57-62%).

General description of Hawaiian specimen, with condition in larger Australian specimens in parentheses, if different. Body long and slender, laterally compressed, tapering posteriorly from abdomen into a long tail that ends in a fine point. Width of body across pectoral-fin bases less than half body depth below origin of first dorsal fin; greatest depth about equal to distance from anteri- or margin of orbit to end of opercle. Head about 6.2 into total length, compressed, its greatest width slightly greater than orbit diameter. Snout rounded and narrow, its tip slightly behind vertical of tip of jaws (slightly forward in larger Australian specimens). Orbits huge, more than postorbital length (slightly less than postorbital), about 2.4 into head length; its dorsal margin entering dorsal profile of head. Mouth large, essentially terminal, unrestricted by lip folds at angle; posterior margin of premaxillary approximately below middle of orbit (below posterior margin of pupil). Suborbital region narrow, shelf and ridge well defined, although this may be partly caused by slightly desic- cated state of specimen. Preopercle broadly rounded, its ridges well defined, covering most of interopercle (except anteroventrally and distal tip) and anterior part of subopercle, which ends ven- trally in a short tab. Chin barbel short, fine, its length about equal to diameter of posterior nostril. Gular membrane narrowly con- nected at isthmus far forward under middle of orbit (posterior half of orbit), overlapping and forming a deep fold over bran- chiostegal membrane. Anus immediately in front of anal-fin origin, separated at most by one or two scale rows. Swimbladder large, elongated, with eight or nine broad, flat retia, each con- nected to a globular to elongated bean-shaped gas gland (Fig. 2).

Teeth on premaxilla in mod- erately wide band extending entire length of jaw opening; teeth small and conical or spike- like, except outer series of much larger, widely spaced canines in a single row. Dentary teeth small anteriorly and in one or two irregular series, becoming in-

creasingly larger posteriorly and Ficure 2. Lepidorhynchus denticulatus, CAS 202920, 41.7 mm _ head ina single file. Larger teeth end length, 240+ mm total length. Saggital otolith, (a) mesial and (b) dorsal views; (c) swimbladder, showing retia and gas glands; (d) mesial view of lower jaw showing comblike teeth (arrow).

abruptly and followed at posteri-

26 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 2

or angle of jaws by a series of seven (four or five) comb-like teeth, slightly reclined, closely set, and somewhat overlapping at bases. These comblike teeth unlike those seen in any other grenadier. (In much larger Australian specimens, these teeth fewer and mostly embedded in gum tissue.) As with all other macrouroid fishes, no teeth on vomer and palatines.

First dorsal fin high, triangular, with two spinous rays and 11 segmented rays, anteriormost segmented rays appear to be longest. First spinous ray short and splintlike, tightly coalesced with proximal part of long second spinous ray. Distally on second spine a few small thornlike denticles on leading edge (absent in larger Australian specimens). Interspace between first and second dor- sal fins wide, much greater than length base of first dorsal fin. Rays of second dorsal fin low and fine to end of tail, although gradually increasing in height posteriorly. Anal fin well developed throughout, its origin slightly behind vertical through posterior edge of first dorsal fin. Upper edge of pectoral fin about at mid-lateral line; fin well developed, relatively long, its length about equal to postrostral length of head, its tip extending to above 10 (4% or 5th) anal ray. Pelvic fins well developed; distal tip of outermost ray filamentous and hairlike, just reaching anal-fin origin.

Luminescent tissue on ventral surfaces of body extensive and consisting of extremely fine transverse parallel black lines, the ventral striae. Luminescent tissue extends from isthmus over abdomen up to and slightly behind pectoral-fin base (as well as lateral surface of fin base), narrow- ing posteriorly to above anus, gradually tapering above anal fin to approximately 57'* (53-57) anal- fin ray. Tissue readily observed in preserved specimens as black areas underlain by extremely fine parallel black lines; in fresh specimens, these areas of striae usually overlain by a silvery reflection and striae much less apparent. (Ventral striae are also found in several unrelated groups such as the cardinalfishes of the genus Siphamia [Apogonidae], and the berycoids Aulotrachichthys spp. and Paratrachichthys spp. |Trachichthyidae]). A small, flat, black, triangular light organ situated with- in median line of abdominal wall closely adherent to base of rectum. Light organ externally man- ifested as a round white or translucent body encircled by a narrow black ring continuous with anal surround.

Body scales of moderate size throughout, somewhat deciduous, and fully covered with fine, conical, needle-like spinules arranged in more or less parallel rows. (In larger Australian speci- mens, 30 or more rows of spinules on scales below interspace between dorsal fins.) Body scales almost entirely missing in Hawaiian specimen, but a few remain on nape and top of head. (Scales cover most of head, although lower surface of snout naked. Tiny scales incompletely cover subor- bital region.) Mandibular rami fully scaled. Gular and branchiostegal membranes naked. (Australian specimens have lowermost branchiostegal rays scaled, and in one specimen, exposed anteroventral margin of interopercle densely covered with tiny oval scales, each having a few short, upright spinules.) None of scales on head enlarged or thickened.

Color in alcohol of denuded Hawaiian specimen overall flesh colored with prominent black areas corresponding to luminescent tissue. Gill membranes, lower parts of head, and opercle pos- terior to preopercular ridge blackish. Fins clear, although first dorsal fin somewhat light dusky. (In fresh specimens, species silvery over most lateral surfaces of head and body, with dorsum gray and area Over luminescent tissue black. Pelvic fins blackish; pectoral fins light dusky.)

Comparisons

The small size of the Hawaiian specimen made comparison of the morphometry of head and body structures uncertain. Several proportions of head parts in BPBM 39286 were greater than in the larger Australian specimens: snout length, preoral length, internasal width, interorbital width, orbit diameter, and suborbital width. The postorbital and orbit-to-preopercle measurements were slightly lower, but those were probably related to the relatively larger orbit diameter, which typi-

IWAMOTO: GRENADIER LEPIDORHYNCHUS DENTICULATUS IN HAWAII pol

cally shows allometric growth in grenadiers. That the interorbital and suborbital widths were not narrower is somewhat contrary to what one would expect. The slightly lower values for the preanal length, the distance of isthmus to anal-fin origin, and the greatest body depth are not uncommon in smaller individuals of grenadier species. The weakly and sparsely denticulated distal portion of the spinous dorsal ray is also seen in the smallest Australian specimen and depicted in the original illus- tration of the holotype. In the larger Australian specimens, these denticulations are lost. All other characters appear exactly the same in Hawaiian and Australian specimens, and despite the small differences in some head proportions, there is little reason to doubt the conspecificity of the two populations at this time. The possibility of there being two species cannot be completely discount- ed, however, and a more-definitive conclusion can only be made after additional specimens from Hawaii become available.

DISCUSSION

The unusual Hawaiian record of this species leads to speculation as to how this particular indi- vidual found its way so far across the Pacific from its normal grounds some 3700 nautical miles and more than 54 degrees of latitude distant. Many species of grenadiers have broad distributions, but those can generally be separated into three categories: (1) abyssal or lower-slope species, such as Coryphaenoides armatus (Hector, 1875) and Cetonurus spp.; (2) bathypelagic species, such as Cynomacrurus piriei Dollo, 1900 and Odontomacrurus murrayi Norman, 1939; and (3) species having a long-lived pelagic juvenile stage, the prime example being Malacocephalus laevis Lowe, 1843. Lepidorhynchus denticulatus does not appear to fit into any of these categories (but its early life history is not known), so its presence in the Hawaiian Islands is a mystery. Were it to be found in intervening areas of the western Pacific, its presence in Hawaiian waters would not be so sur- prising. Yet, despite relatively extensive collecting at appropriate depths throughout the central and North Pacific, especially by Japanese vessels, L. denticulatus has yet to be recorded north of New Zealand and southeastern Australia. Extensive trawling on the Lord Howe Rise and Norfolk Ridge north of New Zealand during the cooperative (New Zealand and Australia) NORFANZ cruise of the R/V Tangaroa in 2003 captured the species only once, and that was west of Three Kings Islands off the northern tip of New Zealand’s North Island at 34°S. The extensive French survey cruises to New Caledonia and adjacent areas failed to produce any specimen of the species.

Most species of grenadiers are localized in their distributions, usually to basins, island groups, oceanic seamounts and ridges, or along continental margins that are bounded by differing oceano- graphic or geologic conditions. So far as known, besides Lepidorhynchus denticulatus, only five species of grenadiers from Hawaii are also known from Australia or New Zealand: Malaco- cephalus laevis, Kuronezumia bubonis (Iwamoto, 1974), Hymenocephalus aterrimus Gilbert, 1905, Nezumia propinqua (Gilbert and Cramer, 1897), and Trachonurus sentipellis (Gilbert and Cramer, 1897). The last three are found in Australian waters but not off the coasts of New Zealand’s main islands. Malacocephalus laevis and K. bubonis are widespread throughout most of the Pacific, Atlantic and Indian oceans, and N. propinqua is widespread in the Pacific and Indian oceans, although there is some question as to whether that name is currently being applied to more than one species. Trachonurus sentipellis is recorded from Hawaii, the New Caledonian region, and Australia, but there is again some doubt that the Hawaiian populations are the same species as the austral populations. If the identifications are correct, this species comes closest in its distribution to that of Lepidorhynchus.

It has been theorized (Marshall 1965, 1973) that most grenadiers have pelagic eggs that hatch near the bottom; the larval stage is very short before metamorphosis into a pelagic juvenile stage (sometimes called postlarva or prejuvenile) takes place. The young probably spend relatively little

28 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 2

time in the midwaters before descending to the bottom to live a benthopelagic life. A short pelag- ic early life would ensure that the settled young are over appropriate living depths and are not waft- ed too far from the narrow continental slopes on which adult populations live. The scarcity of lar- val grenadiers in collections, despite the group’s abundance throughout the world’s oceans, adds support to this idea of a short pelagic stage lived close to the bottom. Perhaps there are some species that can tolerate an extended pelagic larval or juvenile stage when wafted far beyond their normal range. The concept of “expatriate” individuals is well known for some fishes, and this may come into play with some unusual range extensions in grenadiers. Whether any of these factors may account for the extraordinary occurrence of L. dénticulatus in Hawaii 1s yet to be learned, as the early life history of the species is not known, nor is it known that L. denticulatus is actually established in the Hawaiian Islands. It seems unlikely, however, that the captured specimen repre- sents a single stray individual in these waters.

There are few examples of other demersal shelf and upper-slope fishes of southern Australia and New Zealand that have a distribution also in the Hawaiian Islands. Mr. Suzumoto informs me (in litt., Nov. 2004) that the Bishop Museum has “‘a specimen of Cheilidonichthys kumu” caught off Hawaii Island in 1926 and never seen here since.” That Bishop Museum sea robin (family Triglidae) was recorded by Pietschmann (1930) and considered of questionable validity by Springer (1982:103). It was said (Gomon et al. 1994:496) to occur off South Africa, southern Australia, and New Zealand, and also in the North Pacific off Japan, Korea, and China; however, recently published books on the Japanese fish fauna (e.g., Masuda et al. 1984; Nakabo 2002) list C. spinosus and C. ischyrus but not C. kumu. Martin F. Gomon (NMV) has informed me (in Iitt., 12 Dec. 2004) that Peter Last (of CSIRO) “has differentiated several species of Cheilidonichthys in Australia, probably none of them also in South Africa, and the Japanese species are probably something else again.” Gomon adds that “there are other Hawaiian cognates, like the Bodianus oxycephalus-like species. . .with 3 cognates in Japan, southern Pacific and SW Australia, respec- tively.” The morwong Cheilodactylus vittatus was thought to have a north-south disjunct distribu- tion: the Hawaiian Islands in the northern hemisphere, the Lord Howe Island and New Caledonia in the southern hemisphere (Springer 1982:33). Burridge (2004), however, has reported that the New Caledonia and Norfolk Island Cheilodactylus represents a different species, which he described as C. francisci. The mirror dory, Zenopsis nebulosus, is a wide-ranging shelf to upper- slope species (62—550 m) found off Australia, New Zealand and Chile in the southern hemisphere, and also in the North Pacific off Japan, Korea, Hawaii and California (Gomon et al. 1994:420-421). Of the three deep-dwelling (100-400 m) Emmelichthyidae in Hawaii, one species also occurs off Japan, Australia and the Philippines (Springer 1982:41). In light of these few exam- ples, it is apparent that many fish species formerly thought to have a broadly disjunct distribution in Hawaii and the southwest Pacific are, in fact, separable into more than one closely related species. Other examples are likely to surface after closer study using tools and knowledge not for- merly available. That possibility for Lepidorhynchus cannot at this time be discounted.

Springer (1982:117), in his valuable study on Pacific Plate biogeography, found 121 families of shorefishes that are nonmarginally represented on the Pacific Plate. Of these, 15 families (12%) are found only in the Hawaiian Islands, and 10 of the 15 occur there only in deep water. These “Hawaiian exceptions,” as so designated by Springer, “denote Indo-Pacific taxa that occur nonmar- ginally on the Pacific Plate only at the Hawaiian Islands. These taxa usually occur also in Japan and/or the Ruykyu [sic; misspelling for Ryukyu] Islands, and many are antitropically distributed in the Indo-West Pacific; some are restricted to north of the tropics and some are moderately deep dwelling.” (Springer 1982:135.). Lepidorhynchus should be added as an Hawaiian exception, but the underlying cause of its apparent disjunct distribution is unknown.

IWAMOTO: GRENADIER LEPIDORHYNCHUS DENTICULATUS IN HAWAII 29

Central to understanding the vicariant event that produced the current distribution of Lepidorhynchus, one should have a firm grasp of its closest relatives. “If one knows the cladisti- cally determined closest relatives (sister groups) of endemic taxa, and the distributions of those rel- atives, one can narrow the geographic and temporal search for the pertinent vicariant events.” (Springer 1982:6). Unfortunately, the sister-group relationships of Lepidorhynchus within the Macrouridae are uncertain and a thorough cladistic analysis of macrourid genera has yet to be made. Iwamoto and Sazonovy (1988:39), in their tentative phylogeny of macrourines with six bran- chiostegal rays, included the genera Cynomacrurus and Odontomacrurus in a clade, with Lepidorhynchus as the clade’s questionable sister group. Odontomacrurus murrayi perhaps comes closest in overall morphology to Lepidorhynchus, having a compressed head, large, terminal jaws beset with canines, and a small light organ. That bathypelagic species, however, lacks ventral stri- ae, a developed swimbladder, and a chin barbel, and it has many reductions and differences in its morphology and organ systems that reflect its bathypelagic existence and thereby distance it from L. denticulatus. The broad, almost worldwide distribution of O. murrayi in tropical to subtropical waters shows a noteworthy hiatus — so far as known, there are no records of its presence on the Pacific Plate except marginally.

Of the grenadiers with seven branchiostegal rays, the luminescence on the body, the rounded snout with little protrusion of the rostrum, and the scale spinulation are most suggestive of Hymenocephalus, but the six branchiostegal rays, the eight or nine retia and gas glands, the absence of a light organ lens on the chest, the very different dentition, and the narrowly compressed head in Lepidohynchus are among several characters that separate it from members of Hymenocephalus. The head shape, especially the narrow snout with little protrusion of the median nasal ridge, is somewhat like that of Ventrifossa atherodon (Gilbert and Cramer, 1897) and the scale morphology is quite similar, but members of Ventrifossa lack ventral striae, they have seven branchiostegal rays and two retia and gas glands, and there is a small dermal window of the light organ between the pelvic-fin bases that is absent in Lepidorhynchus. Steindachneria argentea Goode and Bean, 1886, in the monotypic family Steindachneriidae, bears some resemblance to L. denticulatus in its ven- tral striae, large terminal mouth beset with canine-like teeth, and similar retia and gas glands in the swimbladder (see Marshall 1966, fig. 3C), but differs radically in having second dorsal rays well developed and higher than most anal-fin rays, an elevated lobe in the otherwise low anal fin, vomerine teeth, no chin barbel, anus located between pelvic-fin bases and distantly separated from urogenital opening, which is immediately before the anal fin. It is apparent that a more thorough comparative study of the morphology of Lepidorhynchus denticulatus, perhaps combined with molecular investigations, will be needed to clarify the phylogenetic position and discover the sis- ter group of this perplexing species.

ACKNOWLEDGMENTS

I thank Arnold Suzumoto (BPBM) for sending me the specimen of Lepidorhynchus and for providing additional information; Ken Graham (New South Wales Fisheries, Australia) for provid- ing comparative material from Australia; Jon Fong and Mysi Hoang (CAS) assisted in the imaging of the specimen. Drs. Martin Gomon (NMV) and Clive Roberts (NMNZ) critically reviewed the manuscript and provided valuable insights and information, as well as editorial corrections.

LITERATURE CITED

BURRIDGE, CHRISTOPHER P. 2004. Cheilodactylus (Goniistius) francisi, a new species of morwong (Perciformes: Cirrhitoidea) from the southwest Pacific. Records of the Australian Museum 56:23 1-234.

30 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 2

GILBERT, CHARLES H., AND CARL L. Hupss. 1916. Report on the Japanese macrouroid fishes collected by the United States fisheries steamer “Albatross” in 1906, with a synopsis of the genera. Proceedings of the United States National Museum 51(2149):135—214, pls. 8-11.

GoMoNn, MArtIN F., JOHN C.M. GLOVER, AND RupIE H. KulTer. 1994. The Fishes of Australia’s South Coast. State Print, Adelaide, Australia. 992 pp.

GUNTHER, ALBERT. 1862. Catalogue of the Acanthopterygii, Pharyngognathi and Anacanthini in the collection of the British Museum. Catalogue of the Fishes of the British Museum, vol. 4:1—xxi, 1-534. Trustees of the British Museum, London, UK.

GUNTHER, ALBERT. 1887. Report on the deep-sea fishes collected by H.M.S. Challenger during the years 1873-76. Report on the Scientific Results of the Voyage of H.M.S. “Challenger” During the Years 1873-76, vol. 22 (part 57):1—268, pls. 1-66. London, UK.

IwAMoTO, TOMIO, AND YURI I. SAZONOV. 1988. A review of the southeastern Pacific Coryphaenoides (sensu lato) (Pisces, Gadiformes, Macrouridae). Proceedings of the California Academy of Sciences, ser. 4, 45(3):35-82.

LEVITON, ALAN E., ROBERT H. GIBBS, JR., ELIZABETH HEAL, AND CHARLES E. DAwSoNn. 1985. Standards in her- petology and ichthyology: Part I. Standard symbolic codes for institutional resource collections in her- petology and ichthyology. Copeia 1985(3):802—832.

MARSHALL, NORMAN B. 1965. Systematic and biological studies of the macrourid fishes (Anacanthini- Teleostii). Deep-Sea Research and Oceanographic Abstracts 12(3):299-322.

MARSHALL, NORMAN B. 1966. The relationships of the anacanthine fishes, Macruronus, Lyconus, and Steindachneria. Copeia 1966(2):275—280.

MARSHALL, NORMAN B. 1973. Family Macrouridae. Pages 496-665 in Daniel M. Cohen, ed., Fishes of the Western North Atlantic, Part 6. Sears Foundation for Marine Research Memoir 1.

Masuba, HAJIME, KUNIO AMAOKA, CHUICHI AARGA, TERUYA UYENO, AND TETSUO YOSHINO, EDS. 1984. The Fishes of the Japanese Archipelago. Tokai University Press, Tokyo, Japan. 437 pp.; 370 pls.

NAKABO, TETSUJI, ED. 2002. Fishes of Japan with Pictorial Keys to the Species, English edition. Tokai University Press, Tokyo. Vol. I: pp. i-xx, 1-866; Vol. I: pp. i-vii, 867-1749.

PHILLIppSs, W.J. 1927. Bibliography of New Zealand fishes. New Zealand Marine Department, Fisheries Bulletin (1):1-68.

PIETSCHMANN, VIKTOR. 1930. Remarks on Pacific fishes. Bernice P. Bishop Museum, Bulletin 73:\—24, pls. 1+.

RICHARDSON, JOHN. 1846. Fishes. In: The Zoology of the Voyage of H.M.S. Erebus and Terror, under the com- mand of Capt. Sir James Clark Ross during the years 1839 to 1843. Vol. 2. viii + 739 pp., 40 pls. London, UK.

SPRINGER, VICTOR G. 1982. Pacific Plate biogeography, with special reference to shorefishes. Smithsonian Contributions to Zoology No.67. 182 pp.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 3, pp. 31-41, 2 figs. April 29, 2005

The Phylogenetic Position of the Recently Rediscovered Philippine Forest Turtle (Bataguridae: Heosemys leytensis)

Arvin C. Diesmos!:2, James F. Parham>-4, Bryan L. Stuart5-6, Rafe M. Brown’

! Department of Biological Sciences, National University of Singapore, Block S3 14 Science Drive 4, Singapore 117543. E-mail: kaloula@i-manila.com.ph; *Herpetology Section, Zoology Division, National Museum of the Philippines, Padre Burgos Avenue, Ermita 1000, Manila, Philippines;

3Evolutionary Genomics Department, Joint Genome Institute, 2800 Mitchell Drive, Walnut Creek, CA 94598, USA. Email: parham®@ socrates.berkeley.edu; (Museum of Paleontology, University of California, Berkeley, CA 94720, USA; >The Field Museum, Department of Zoology, Division of Amphibians and Reptiles, 1400 S. Lake Shore Drive, Chicago, IL 60605-2496, USA. Email: bstuart@ fmnh.org; “Department of Biological Sciences (M/C 066), University of Mlinois at Chicago, 845 W. Taylor, Chicago, IL 60607-7060, USA; ‘Division of Herpetology, Natural History Museum, University of Kansas, Lawrence, KS 66045-2454, USA. Email: rafe@mail.utexas.edu

The Philippine forest turtle (Heosemys leytensis) is one of the least-known Asian tur- tles. Until recently it was only known from a single specimen, the neotype (CAS 60930). The rediscovery of this enigmatic turtle in the wild provides access to fresh genetic material, allowing us to rigorously test its phylogenetic position among Asian pond and wood turtles of the clade Bataguridae for the first time. We sequenced 1174 bp of mitochondrial DNA (cytb) and 1076 bp of nuclear DNA (R35 intron) from a newly acquired specimen of H. leytensis (PNM 8488) and compared our sequences to a previously published genetic survey of batagurids. Our main goal is to determine if H. leytensis is closely related to other species of Heosemys (H. spinosa [type species], H. grandis, H. depressa, and H. annandalii [comb. nov.]). Parsimony and likelihood analyses of mitochondrial and nuclear DNA data strongly support H. leytensis as the sister taxon to the Southeast Asian species Siebenrockiella crassicollis (commonly known as the ‘black marsh turtle’ or ‘smiling terrapin’). The close rela- tionship between H. leytensis and S. crassicollis is a novel hypothesis, although both share an easily recognizable diagnostic feature: vertebrals 2-4 are mushroom- shaped or ‘ginkgo leaf’-shaped. In order to eliminate two potentially monotypic gen- era, the genus Siebenrockiella is phylogenetically defined to accommodate the close relationship between S. crassicolis and H. leytensis whereas a new name, Panyaenemys, is defined for the distinct H. leytensis stem lineage.

The rediscovery of wild populations of the Philippine forest turtle (Heosemys leytensis Taylor 1920; Fig. 1B-E) in the Palawan island group of the western Philippines (Fig. 1A; Diesmos et al. 2004a), resolved a decades-long search for this species. Prior to this, our knowledge of H. leyten- sis was restricted to the original published description (the type series was lost in the destruction of the Bureau of Science, predecessor to the National Museum of the Philippines, during World War II; Brown and Alcala 1978; Buskirk 1989), one museum specimen (CAS 60930, the neotype, Fig. 1B,C) reported by Buskirk (1989), and one specimen from Palawan (UF/FSM 67515) report- ed by Timmerman and Auth (1988). Thus, for over forty years, no museum specimens of H. leyten-

3]

32 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 3

sis were known and so it remained one of the least-known Asian turtles.

The rediscovery of H. leytensis in the wild (Diesmos et al. 2004a) provides access to fresh genetic material, allowing us to rigorously test its phylogenetic position within the ‘Asian pond and wood turtles’ (Bataguridae Gray, 1870 = “Geoemydidae Theobald, 1868’; Joyce et al. 2004) for the first time. Batagurids are a species-rich group (60+ spp.) of highly threatened (van Dijk et al. 2000), but poorly studied non-marine turtles from Asia. A taxonomic reassessment of H. leytensis is nec- essary because the content and diagnosis of batagurid genera have changed dramatically since Taylor (1920) described the Philippine forest turtle as Heosemys leytensis (e.g., Ernst and Barbour 1989; Iverson 1992). The twelve characters that Taylor (1920) used to diagnose Heosemys are now known to be either plesiomorphic for batagurids or else appear in other genera as homoplasy (Joyce and Bell 2004). One of these characters, the derived absence of a temporal arch in the skull, has played a particularly important role in maintaining H. Jeytensis in the genus Heosemys. Besides being noted by Taylor (1920), this character was later used by McDowell (1964) to diagnose Heosemys (including H. leytensis) and was also described in the H. leytensis neotype (Buskirk 1989). This character is also known in Hieremys annandalii (Boulenger, 1903) as well as some species of testudinoid box turtles from China and North America (Cuora Gray, 1855 and Terrapene Merrem, 1820; Zangerl 1948; McDowell 1964; Joyce and Bell 2004). Because H. leytensis lacks the characteristic hinged plastron of a box turtle, its referral to Heosemys seems logical.

Although H. leytensis shares the absence of the temporal arch with other Heosemys, it is mor- phologically divergent from other Heosemys in other respects. For example, H. leytensis usually has mushroom-shaped or ‘ginkgo leaf’-shaped vertebral scales 2-4 (Fig. 1B, E), a feature that is otherwise restricted to the “smiling terrapin’ of southeast Asia, Siebenrockiella crassicollis (Gray 1831) (Fig. 1F—G). Additionally, the projecting gular region (anterior region of the plastron) of H. leytensis clearly distinguishes it from all other batagurid lineages, including other species referred to Heosemys. In this character, it superficially resembles some species of terrestrial tortoises (Testudinidae Gray, 1825), the sister-taxon to batagurids (Spinks et al. 2004).

Given the extensive homoplasy within batagurid morphology (Joyce and Bell 2004) and the overall distinctiveness of H. leytensis, it is difficult to confidently place H. leytensis into any exist- ing genus on morphological grounds alone. Moreover, weak nodes generated from cladistic analy- ses of morphological variation in Bataguridae (Hirayama 1985; McCord et al. 1995) are strongly rejected by molecular studies (McCord et al. 2000; Spinks et al. 2004). For this reason, we tested the phylogenetic position of H. /eyrensis with cladistic analyses of mitochondrial and nuclear DNA sequences.

Our primary goal was to determine whether H. leytensis should continue to be recognized as a member of the genus Heosemys. But in order to avoid confusion, we must establish an explicit concept of Heosemys and what species names this informal clade name should include. Spinks et al. (2004) reported DNA sequences from three species of Heosemys including one sample attrib- uted to the type species Heosemys spinosa (Gray, 1831). According to their study, Heosemys form a well-supported clade with Hieremys annandalii, and so Spinks et al. (2004) raised the possibili- ty that H. annandalii might be better placed within an expanded Heosemys. However, Spinks et al. (2004) refrained from making a taxonomic change pending future evidence for the possible place- ment of Hieremys annandalii relative to Heosemys species. We think this reticence is unwarranted inasmuch as Hieremys Smith, 1916 is a monotypic genus and H. annandalii forms a well-support- ed clade with other Heosemys. The Spinks et al. (2004) suggestion to expand the older genus name (Heosemys) to the well-supported node creates a much more stable taxonomy. Under this scheme, the widely-used genus name will remain stable regardless of the ultimate resolution of the position within the Heosemys basal polytomy. We refer Hieremys annandalii to the genus Heosemys, but

DIESMOS ET AL.: PHYLOGENETIC POSITION OF HEOSEMYS LEYTENSIS 38

FIGURE 1. A: Map showing the known distribution of Siebenrockiella crassicollis (dots) based on Iverson (1992) and Stuart and Platt (2004). The black rectangle includes the known and hypothesized range of Heosemys leytensis based on Diesmos et al. (2004b): B-C: Neotype specimen of Heosemys leytensis (CAS 60930), dorsal [note the missing “ginkgo leaf’- shaped vertebral 2] (B) and ventral view (C); D: Dorsal view of H. leytensis head showing diagnostic markings; E: Dorsal view of H. leytensis possibly from Dumaran Island (see Diesmos et al., 2004a); F: Dorsal view of S. crassicollis (FMNH 224070): G: View of S. crassicollis head showing the diagnostic markings.

refrain from defining the genus Heosemys phylogenetically until we discuss our results, especially regarding the placement of H. leytensis. In summary, our concept of Heosemys is the clade that includes the type species Heosemys spinosa, the aforementioned Heosemys [orig. Cyclemys] annandalii (Boulenger, 1903) comb. nov., Heosemys grandis (Gray, 1860), and Heosemys depres- sa (Anderson, 1875).

INSTITUTIONAL ABBREVIATIONS.— CAS, The California Academy of Sciences, San Francisco,

34 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 3

California, USA; FMNH, The Field Museum of Natural History, Chicago, Illinois, USA; PNM, National Museum of the Philippines, Manila, Philippines; UF, Florida Museum of Natural History, University of Florida, Gainesville, Florida.

MATERIALS AND METHODS

Ethanol-preserved muscle was taken from a museum specimen of Heosemys leytensis (PNM 8488, see Material Examined). Total genomic DNA was extracted from the muscle using PureGene Animal Tissue DNA Isolation Protocol (Gentra Systems, Inc.). Primers for amplifying and sequencing parts of cytb and tRNA-Thr (collectively ‘cytb’) were taken from Spinks et al. (2004) and R35 primers were taken from Fujita et al. (2004). PCR products were electrophoresed in a 1% low melt agarose TALE gel stained with ethidium bromide and visualized under ultraviolet light. The bands containing DNA were excised and agarose was digested from bands using GELase (Epicentre Technologies). PCR products were sequenced in both directions by direct double-strand cycle sequencing using Big Dye version 3.1 chemistry (Perkin Elmer). Cycle-sequencing products were precipitated with ethanol, 3 M sodium acetate, and 125 mM EDTA, and sequenced with a 3730 DNA Analyzer (ABI). Sequences were edited and protein-coding regions were translated into amino acids with Sequencher version 4.1 (Genecodes).

We compared new cytb and R35 sequences with the sequences reported by Spinks et al. (2004). Sequences were aligned by eye into the matrix used by Spinks et al. (2004) and deposited into TreeBASE (www.treebase.org, accession number $1002). Our sequences (cytb, 1174 bp; R35, 1076 and 1078 bp) were longer than those used in the Spinks et al. (2004) alignment; the extra base pairs were excluded from the analysis. For the parsimony analysis of combined data sets, we included 72 of the 79 taxa sequenced by Spinks et al. (2004), excluding seven putative hybrids. Some of these hybrids were described as new species from pet trade samples leading to taxonom- ic confusion that is not yet fully resolved (Parham et al. 2001; Wink et al. 2001; Spinks et al. 2004). We avoid these issues by removing the suspect taxa from consideration. We feel that we are justi- fied in doing this because an initial phylogenetic analysis (not shown) demonstrated that H. leyten- sis is not closely related to any of these taxa and the inclusion or exclusion of these sequences does not affect the phylogenetic position of H. leytensis. Besides H. leytensis, the only addition to the Spinks et al. (2004) data set was the sequence of one vouchered Siebenrockiella crassicollis sam- ple from Cambodia (see below).

In addition to the aforementioned combined analysis, we analyzed the mitochondrial and nuclear data sets separately in order to compare the potential congruence or conflict in the phylo- genetic signal of our markers. We also performed a maximum likelihood and Bayesian analyses on a restricted data set that includes 28 of the 79 species from Spinks et al. (2004). These 28 sequences represent all the non-hybrid taxa from the Spinks et al. (2004) study that have both cytb and R35 sequenced. The restricted data set is still appropriate for placing H. leytensis because it contains every major lineage (1.e., genus) of batagurid.

Parsimony analyses were performed using the 1000 random addition-sequence replicates of the heuristic search algorithm in PAUP*4.0b10 (Swofford 2002). Support for the parsimony analy- ses was determined using bootstrap and decay indices. For the combined analysis, we used 1000 bootstrap replicates with 10 random addition-sequence replicates each. Decay indices were calcu- lated using a PAUP command file generated by MacClade 4.0 (Maddison and Maddison 2000). For the likelihood analyses, we used hierarchical likelihood-ratio tests with ModelTest (Posada and Crandall 1998) to determine which model was appropriate. For the standard likelihood analysis the model GTR+I+G was selected, with proportion of invariable sites 0.3836, gamma distribution

DIESMOS ET AL.: PHYLOGENETIC POSITION OF HEOSEMYS LEYTENSIS 35

shape parameter 0.4762, and base frequencies as A=0.3198, C=0.3039, G=0.1336, and T=0.2427. Maximum likelihood analyses were performed with 100 random addition replicates with stepwise addition of taxa using the heuristic search algorithm and TBR branch swapping. For the Bayesian analysis, we separated the cytb and R35 data into their own partitions and estimated the best mod- els for each (GTR+I+G and HKY respectively). Using these parameters, we performed three mixed-model analyses of five million generations and four chains each using MrBayes v3.0b4 (Huelsenbeck and Ronquist 2001). We plotted the log likelihood scores against generation to deter- mine when the analysis achieved stationarity for each analysis. In each case, the seven thousand trees recovered prior to stationarity were discarded as “burn in’ and a 50% majority rule consensus was used to determine the posterior probabilities for each clade.

RESULTS

In our initial analyses, the sequence from our sample of H. leytensis is consistently placed as the sister taxon to S. crassicollis with high statistical support. None of the samples from the Spinks et al. (2004) study have vouchered specimens, so we confirmed our initial results by sequencing a vouchered specimen of S. crassicollis from Cambodia (see Material Examined). All tree statistics were calculated with the additional sample included. All analyses (parsimony, likelihood, Bayesian) place H. leytensis and S. crassicollis as sister taxa (Fig. 2). In the parsimony and Bayesian analyses, H. leytensis and S. crassicollis are united with high statistical support (100% parsimony bootstrap, 23 decay index, 100% Bayesian posterior probability). Parsimony analyses of the separate cytb and R35 data sets (not shown) show that the H. leytensis + S. crassicollis clade is independently supported by our mitochondrial and nuclear markers. Meanwhile, the other species of the genus Heosemys form a well-supported clade that is separated from H. leytensis + S. crassicollis clade by several branches (Fig. 2).

DISCUSSION

Our study demonstrates that the continued placement of H. leytensis in the genus Heosemys 1s no longer tenable. Inasmuch as the other species of Heosemys form a well-supported clade, which is readily-diagnosable with molecular and morphological data, we phylogenetically define the genus Heosemys as the crown clade arising from the last common ancestor of Heosemys [orig. Emys| spinosa (Gray, 1831), Heosemys |orig. Geoemyda] grandis (Gray, 1860), Heosemys [orig. Geoemyda| depressa (Anderson, 1875), and Heosemys [orig. Cyclemys| annandalii (Boulenger, 1903).

The hypothesis that H. leytensis and S. crassicollis are sister taxa within Bataguridae is a novel result. Both H. leytensis and S. crassicollis share two diagnostic characters that may have been inherited from a common ancestor: 1) Vertebrals 2-4 are ‘ginkgo leaf’-shaped. Orlitia borneensis Gray 1873 has a superficially similar condition, but it is restricted to the first vertebral; 2) Light spots or lines in the posterior dorsal region of the head. In S. crassicollis (Fig. 1G), this character occurs as two circles or spots that may fade with age (Lim and Das 1999) whereas in H. leytensis (1C) it is a straight line that is faint or absent in 35% of individuals (Diesmos et al. 2004a, b). A similar, but more dramatic, condition of posterior head marking occurs in parallel in the ‘eyed tur- tles’ of the genus Sacalia Gray 1870.

Additional morphological comparisons of S. crassicollis and H. leytensis, including skeletal material, may reveal more potential synapomorphies. Meanwhile, the character that has long been used to place H. leytensis in the genus Heosemys, the absence of a complete temporal arch, is more

36

__ <50 Bataguridae

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Emydidae

Rhinoclemmys rubida

Rhinoclemmys areolata

7O\ “Ament aemean CLOCIC

Hardella Ca

Melanochelys trijuga

3 S11

Volume 56, No. 3

Testudinidae

Rhinoclemmys pulcherrima Rhinoclemmys annulata

Rhinoclemmys punctularia Rhinoclemmys diademata Rhinoclemmys melanosterna Rhinoclemmys funerea

Malayvemys subtrijuga 9 Orlitia borneensis

mys hamiltonii Morenia ocellata

thurjii llagur borneoensis

Kachuga dhongoka

Batagur baska

Pangshura tecta Pangshura smithii

Pangshura tentoria

Pangshura tentoria

50 changes

|

Geoemyda spengleri

10 Sacalia bealei 94

100— Melanochelys trijuga <50, 3 Geoemyda japonica

Sacalia quadriocellata

Heosemys annandalii

Heosemys gran Heosemys depressa

| 1] Leucocephalon yuwonoi hee qi) Notochelys platynota <50|_ 14 Cyclemys atripons

—p

8 | | 89 | tole 8| 98

| |

ale

bs

Cyclemys tcheponensis Cyclemys dentata

Mauremys leprosa Mauremys reevesii Mauremys reevesii | Mauremys japonica —— Mauremys sinensis Mauremys rivulata Mauremys caspica

Mauremys annamensis Mauremys mutica

Mauremys mutica Cuora amboinensis Cuora amboinensis |= Cuora amboinensis * - Cuora amboinensis Cuora picturata Cuora mccordi Cuora flavomarginata Cuora flavomarginata Cuora mouhotii

=| Cuora zhoui Cuora pani ‘— Cuora trifasciata

— Cuora aurocapitata

* 3 "Heosemys" leytensis pei: 100 67-— Siebenrockiella crassicollis 100— Sitebenrockiella crassicollis

-—— Heosemys spinosa (type species) i

DIESMOS ET AL.: PHYLOGENETIC POSITION OF HEOSEMYS LEYTENSIS 3)//

homoplastic than previously thought. If this character is optimized onto the phylogeny (Fig. 2), it predicts that the temporal arch was lost at least three independent times within Bataguridae: once in the H. Jeyrensis stem-lineage, once in the in the Heosemys stem lineage, and at least once with- in Cuora. A polytomy at the base of Cuora (e.g., Spinks et al., 2004; Parham et al., 2004) and a poor understanding of the morphological variation in extent of the temporal arch (see Joyce and Bell 2004) preclude a definitive assessment of the polarity of this character within Cuora.

From a biogeographic perspective, the S. crassicollis + H. leytensis clade seems plausible. Heosemys leytensis is known only from the Palawan Pleistocene Aggregate Island Complex (or Palawan PAIC; Brown and Diesmos 2001). Although the Palawan region contains a significant level of endemicity in fauna and flora not found in the Sundaic region (see Widmann 1998, McGuire and Alcala 2000, Brown and Guttman 2002, Evans et al. 2004), in general the fauna of the Palawan PAIC has a much closer affinity to the Islands of the Sunda Shelf (e.g., Kalimantan/‘Borneo’, Madura, Sumatra) than do the other Philippine islands (Heaney 1986; Brown and Diesmos 2001).

Therefore, the close relationship of H. leytensis to a Sunda Shelf species (S. crassicollis) is not surprising. The combined distribution of the S. crassicollis+H. leytensis clade (Fig. 1A) is very similar to that of Cyclemys dentata (Gray 1831) and Cuora amboinensis (Daudin 1802), the two other turtle species known from Palawan PAIC (Iverson 1992). Their shared distributions may indi- cate that these three lineages may have a shared biogeographical history as well, the details of which could be compared through additional DNA sequencing of S. crassicollis, Cu. amboinensis and Cy. dentata from throughout their range.

Given the strong phylogenetic signal placing it as the sister taxon to S. crassicollis, we feel jus- tified in referring H. leytensis to a phylogenetically defined Siebenrockiella Lindholm 1929. By using the genus name Siebenrockiella to highlight this close relationship, we eliminate two mono- typic genera (Siebenrockiella and a hypothetical new genus for H. leytensis) from a literature that is lamentably crowded with monotypic genera (Spinks et al. 2004). Polytypic genera are function- ally superior to monotypic genera because they maximize the information content in each widely- used name (Parham and Feldman 2002; Feldman and Parham 2004). With this is mind, we phylo- genetically define Siebenrockiella as the crown clade arising from the last common ancestor of Siebenrockiella |orig. Emys| crassicollis (Gray 1831) and Siebenrockiella [orig. Heosemys| leyten- sis (Taylor 1920) (comb. nov.).

In addition to developing a conservative nomenclature for the widely-used genus name, we also want to name the stem-lineage that includes the morphologically and molecularly divergent S. leytensis. For example, compared to S. crassicollis, S. leytensis has a proportionately larger head and correspondent nuchal emargination, is larger in overall body size, and has an extremely pro- truding gular region of the plastron (Diesmos et al. 2004b). The morphological distinctiveness of the two known Siebenrockiella species is demonstrated by the fact that no author has ever suggest- ed a close relationship between them until now. Moreover, the genetic distinctiveness of the S. leytensis lineage is shown by a high sequence divergence from S. crassicollis (~13% cytb) that is greater than that shown by other batagurid congeners (5.0-10.7%, see Spinks et al. 2004). In order

FIGURE 2 (left). Parsimony phylogram of one of the four most parsimonious trees (4387 steps) recovered by the par- simony analysis of the combined cytb and R35 data set for batagurid turtles. For taxonomy of suprageneric clades see Joyce et al. (2004). The four equally parsimonious results differ in the placement of species within the genus Mauremys or between suprageneric genera not related to Heosemys or Siebenrockiella. Nodes that are not supported in all of the four equally parsimonious trees are indicated by ‘-’ on either side of the stem. The numbers at each of the other nodes represent support values. The top numbers are decay indices and the bottom numbers are parsimony bootstrap percentages. Support values for species clades within genera are not shown but see Spinks et al. (2004). The batagurid lineages that lose a com- plete temporal arch are shown by ‘*’.

38 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 3

to provide a name for this distinct lineage that is divorced from the vagaries of species concepts, we name a new subgenus, Panyaenemys, for S. leytensis.

SYSTEMATICS

Panyaenemys, new clade name, new subgenus TYPE SPECIES: Siebenrockiella leytensis (Taylor, 1920) comb. nov.

We designate a type species for this phylogenetically defined clade name so that it is valid under the rules of the International Code of Zoological Nomenclature (1999).

ETYMOLOGY.— Panyaen-, from the word ‘panya-en’ which means ‘enchanted’ in the lan- guage of the Pala’ wan (one of seven ethnic groups that inhabit the Palawan islands). A resident of Palawan interviewed by ACD during forest surveys in November, 2003 referred to a S. leytensis as a ‘panya-en’, because it is apparently a favorite pet of forest spirits; -emys, turtle.

PHYLOGENETIC DEFINITION.— We define Panyaenemys as the most inclusive clade containing Siebenrockiella |orig. Heosemys| leytensis (Taylor 1920) but not Siebenrockiella (orig. Emys| cras- sicollis (Gray 1831).

DIFFERENTIAL DIAGNOSIS.— The sole known member of the subgenus Panyaenemys can be distinguished from all other turtles by the combination of the following four characters: 1) Strongly projecting epiplastra; 2) “Ginkgo leaf’-shaped vertebral scales 2-4; 3) No temporal arch in the skull; 4) Light lines on the back of the head.

MATERIAL EXAMINED

HEOSEMYS LEYTENSIS.— PHILIPPINES: CAS 60930, the neotype specimen discussed in detail by Buskirk (1989) and Diesmos et al. (2004a); PHILIPPINES: PNM 8488, GenBank accession numbers = AY954911 (cytb) and AY954914 (R35), a dead specimen that was donated to the National Museum of the Philippines (PNM) by a wildlife collector in 2003 and examined by Diesmos et al. (2004b). The specimen was apparently bought from a wildlife trader from Palawan. An additional 47 live specimens of H. leytensis in captivity were examined by Diesmos et al. (2004b) comprised of 34 turtles on Palawan being held by a private collector and 13 government-confiscated turtles housed at the Widlife Rescue Center of the Protected Areas and Wildlife Bureau, Quezon City, Philippines.

SIEBENROCKIELLA CRASSICOLLIS.— CAMBODIA: FMNH 259055 (see Stuart and Platt 2004), GenBank accession numbers = AY954912 (cytb) and AY954913 (R35), Koh Kong Province, Sre Ambel District, Prek Kroch River (tributary of Sre Ambel River), 11°06’20”N 103°39’35”E, <10 m elevation, in a flooded paddy at edge of mangrove and Melaleuca forest, captured by a fisherman in bamboo fish trap set at that location, 27 August 2000, B.L. Stuart and S.G. Platt; UNCERTAIN/ MALAYSIA?: FMNH 224070, Perak, Batu Gajah, purchased in a pet shop, 19 November 1975, E.O. Moll.

ACKNOWLEDGMENTS

We thank the Protected Areas and Wildlife Bureau of the Department of Environment and Natural Resources of the Government of the Philippines for issuing a Memorandum of Agreement to conduct biological research in the Philippines and for facilitating collecting and export permits for this and related studies. We are indebted to Rogelio Sison (PNM) for providing access and a tissue sample of the S. /eytensis specimen at PNM. Field surveys of S. leytensis (between 2001 and 2003) were made possible through funding and technical support provided by the BP Conservation

DIESMOS ET AL.: PHYLOGENETIC POSITION OF HEOSEMYS LEYTENSIS 39

Programme, Royal Melbourne Zoological Gardens, North of England Zoological Society, Conservation International Philippines, Turtle Conservation Fund, Fauna & Flora International, and the National Museum of the Philippines. ACD and RMB are indebted to colleagues at “Herp Watch Palawan 2001” especially Mae Leonida-Diesmos, Genevieve Gee-Das, Carlos Infante, Nonito Antoque, Jude Dimalibot, Jason Diesmos, and Jeffrey Diesmos. We also thank Peter Widmann, Indira Widmann, Siegfred Diaz, Deborah Villafuerte, Rolito Dumalag, Snapper Poche (Philippine Cockatoo Conservation Project) and Sabine Schoppe (Western Philippines University) for their various assistance in the field.

Financial support to JEP was from a National Science Foundation Postdoctoral Fellowship. Financial support to BLS was provided by The John D. and Catherine T. MacArthur Foundation (with Harold Voris and Robert Inger). The opportunity for BLS to collect the sample of Siebenrockiella crassicollis 11 Cambodia was made possible by the Wildlife Conservation Society/Ministry of Agriculture, Forestry and Fisheries/Ministry of Environment Collaborative Program. JFP thanks Daniel G. Mulcahy (Utah State) and Ted Papenfuss (University of California, Berkeley) for technical assistance regarding the phylogenetic analyses when he was “Bayesed and confused’ and also Jens Vindum (CAS) for facilitating his examinations of the H. leytensis neotype. Sarah Drasner photographed the Field Museum carapace of S. crassicollis. Sequencing was con- ducted in The Field Museum’s Pritzker Laboratory for Molecular Systematics and Evolution oper- ated with support from the Pritzker Foundation. This work is LBNL-57255 and was performed under the auspices of the U.S. Department of Energy, Office of Biological and Environmental Research. This is University of California Museum of Paleontology Contribution # 1878.

LITERATURE CITED

ANDERSON, J. 1875. Description of some new Asiatic mammals and Chelonia. Annals and Magazine of Natural History, ser 4, 16:282—285.

BOULENGER, G.A. 1903. Report on the batrachians and reptiles. Pages 131-176 in N. Annandale and H.C. Robinson, eds., Fasciculii Malayaenseis, Anthropological and Zoological Results of an Expedition to Perak and the Siamese Malay States. 1901-1902, Undertaken by Nelson Annandale and Herbert C. Robinson. Vol 1, Zoology. University Press, Liverpool, UK.

Brown, R. M., AND A.C. Diesmos. (2001) 2002. Application of lineage-based species concepts to oceanic island frog populations: the effects of differing taxonomic philosophies on the estimation of Philippine bio- diveristy. The Silliman Journal 42:133—162.

Brown, R.M., AND S.I. GUTTMAN. 2002. Phylogenetic systematic of the Rana signata complex of Philippine and Bornean stream frogs; reconsideration of Huxley’s modification of Wallace’s Line at the Oriental- Australian faunal zone interface. Biological Journal of the Linnean Society 76:393—-461.

Brown, W.C., AND A.C. ALCALA. 1978. Philippine Lizards of the Family Gekkonidae. Silliman University Press, Dumaguete City, Philippines. 146 pp.

BuskirK, J.R. 1989. A third specimen and neotype of Heosemys leytensis (Chelonia: Emydidae). Copeia 1989:224-227.

DAUDIN, F.M. 1802. Histoire Naturelle, Générale et Particuliére des Reptiles, vol.2. Dufart, Paris, France. 342 pp-

Diesmos, A.C., G.V.A. GEE, M.L. DiesmMos, R.M. Brown, P.J. WIDMANN, AND J.C. DIMALIBoT. 2004a. Rediscovery of the Philippine forest turtle Heosemys leytensis (Chelonia: Bataguridae), from Palawan Island Philippines. Asiatic Herpetological Research 10:22-27.

Diesmos, A.C., R.V. Sison, M. DG. PEDREGOSA, AND M.J.C. CEnIzA. 2004b. The Conservation Status of Heosemys leytensis Taylor, 1920. Technical report submitted to National Museum of the Philippines, Fauna & Flora International, Royal Melbourne Zoological Gardens, North of England Zoological Society, Conservation International Philippines, and Turtle Conservation Fund. Manila, Philippines. 31 pp.

40 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 3

ERNST, C.H., AND R.W. BARBOUR. 1989. Turtles of the World. Smithsonian Institution Press, Washington, DC, USA. 290 pp.

Evans, B.J., R.M. Brown, J.A. McGuire, J. SUPRIATNA, N. ANDAYANI, A. DiEsmMos, D. ISKANDAR, D.J. MELNICK, AND D.C. CANNATELLA. 2003. Phylogenetics of fanged frogs: testing biogeographical hypothe- ses at the interface of the Asian and Australian faunal zones. Systematic Biology 52:794-819.

FELDMAN, C.R., AND J.F. PARHAM. 2004. Molecular systematics of Old World stripe-necked turtles (Testudines: Mauremys). Asiatic Herpetological Research 10:28—37.

Fusita, M.K., T.N. ENGsTROM, D.E. STARKEY, AND H.B. SHAFFER. 2004. Turtle phylogeny: insights from a novel nuclear intron. Molecular Phylogenetics and Evolution 31:1031—1040.

Gray, J.E. 1825. A synopsis of the genera of reptiles and amphibia, with a description of some new species. Annals of Philosophy 10:193—217.

Gray, J.E. 1831. Synopsis Reptilium, Part. 1, Cataphracta Crocodiles, Enaliosaurians. Treuttel, Wurtz, London, UK. 85 pp.

Gray, J.E. 1855. Catalogue of Shield Reptiles British Museum, Pt. 1, Testudinata (Tortoises). Taylor and Francis, London, UK. 79 pp.

Gray, J.E. 1860. Description of a new species of Emys lately living in the gardens of the Zoological Society. Proceedings of the Zoological Society of London 1860:232—233.

Gray, J.E. 1870. Supplement to the Catalogue Collection of the British Museum, Pt. 1, Testudinata. Taylor and Francis, London, UK. 120 pp.

Gray, J.E. 1873. On a new freshwater tortoise from Borneo (Orlitia borneensis). Annals and Magazine of Natural History, ser. 4, 11:156—157.

HEANEY, L.R. 1986. Biogeography of the mammals of Southeast Asia: Estimates of colonization, extinction, and speciation. Biological Journal of the Linnean Society 28:127-—165.

HIRAYAMA, R. 1985. Cladistic analysis of batagurine turtles. Studia Palaeocheloniologica 1:140-157.

HUELSENBECK, J.P, AND F. RoNQuist. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754—755.

INTERNATIONAL CODE OF ZOOLOGICAL NOMENCLATURE, 4TH ED. 1999. International Trust for Zoological Nomenclature, London, UK. 306 pp.

IVERSON, J.B. 1992. A Revised Checklist with Distribution Maps of the Turtles of the World. Privately pub- lished, Richmond, Indiana, USA. 363 pp.

Joyce, W.G., AND C. BELL. 2004. A review of the comparative morphology of extant testudinoid turtles (Reptilia: Testudines). Asiatic Herpetological Research 10:53—109.

Joyce, W.G., J.F. PARHAM, AND J.A. GAUTHIER. 2004. Developing a protocol for the conversion of rank-based taxon names to phylogenetically defined clade names, as exemplified by turtles. Journal of Paleontology 78(5):989-1013.

LINDHOLM, W.A. 1929. Revidiertes Verzeichnis der Gattungen der rezenten Schildkroten nebst Notizen zur Nomenklatur einiger Arten. Zoologischer Anzeiger 81:275—272.

MApDISON, D.R., AND W.P. MApDDISON. 2000. MacClade 4.0. Sinauer Associates, Inc. Sunderland, Massachusetts, USA.

McCorp, W.P., J.B. IVERSON, AND BOEApI. 1995. A new batagurid turtle from northern Sulawesi, Indonesia. Chelonian Conservation and Biology 1(4):311-316.

McCorp, W.P., J.B. IVERSON, P.Q. SPINKS, AND H.B. SHAFFER. 2000. A new genus of geoemydid turtle from Asia. Hamadryad 25:20-24.

McDoweE LL, S.B. 1964. Partition of the genus Clemmys and related problems in the taxonomy of the aquatic Testudinidae. Proceedings of the Zoological Society of London 143:239-279.

McGuire, J.A., AND A.C. ALCALA. 2000. A taxonomic revision of the flying lizards (Iguania: Agamidae: Draco) of the Philippine Islands, with a description of a new species. Herpetological Mononographs (14):81-138.

MERREM, B. 1820. Versuch eines Systems der Amphibien. Johann Christian Krieger, Marburg, Germany. 191 Pp-

PARHAM, J.F., AND C.R. FELDMAN. 2002. Generic revisions of emydine turtles. Turtle and Tortoise Newsletter 6:28-30.

DIESMOS ET AL.: PHYLOGENETIC POSITION OF HEOSEMYS LEYTENSIS 4]

PARHAM, J.F., W.B. SIMISON, K.H. KOZAK, C.R. FELDMAN, AND H. Sut. 2001. New Chinese turtles: endangered or invalid? Animal Conservation 4:357-367.

Posaba, D., AND K.A. CRANDALL. 1998. ModelTest: Testing the model of DNA substitution. Bioinformatics 14:817-818.

SmitH, M.A. 1916. A list of the crocodiles, tortoises, and turtles and lizards at present known to inhabit Siam. Journal of the Natural History Society of Siam 2:48-57.

SPINKS, P.Q., H.B. SHAFFER, W.P. MCCorD, AND J.B. IVERSON. 2004. Phylogenetic hypotheses for the turtle family Geoemydidae. Molecular Phylogenetics and Evolution 32:164—-182.

STUART, B.L., AND S.G. PLattr. 2004. Recent Records of Turtles and Tortoises from Laos, Cambodia, and Vietnam Asiatic Herpetological Research 10:129-150.

SWOFFORD, D.L. 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts, USA.

TayLor, E.H. 1920. Philippine turtles. Philippine Journal of Science 16:111—144.

THEOBALD, W. 1868. Catalogue of reptiles in the museum of the Asiatic Society of Bengal. Journal of the Asiatic Society 1868:8-88.

TIMMERMANN, W.W., AND D.L. AuTH. 1988. Geographic distribution, Heosemys leytensis. Herpetological Review 19(1):21.

WIDMANN, P. 1998. A Guide to the Ecosystems of Palawan Philippines. VisCA-GTZ and Times Edition, Singapore. 118 pp.

Wink, M., D. GUIKING, AND U. Fritz. 2001. Molecular evidence for hybrid origins of Mauremys iversoni Pritchard and McCord, 1991, and Mauremys pritchardi McCord, 1997 (Reptilia: Testudines: Bataguridae). Zoologische Abhandlungen, Staatliches Museum fiir Tierkunde Dresden 52:4\1—49.

VAN Duk, P.P., B.L. STUART, AND A.G.J. RHODIN, EDS. 2000. Asian Turtle Trade: Proceedings of a Workshop on Conservation and Trade of Freshwater Turtles and Tortoises in Asia. Chelonian Research Monographs, No. 2. Chelonian Research Foundation, Lunenburg, Massachusetts, USA. 164 pp.

ZANGERL, R. 1948. The methods of comparative anatomy and its contribution to the study of evolution. Evolution 2:351—374.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 4, pp. 42-52, 4 figs., 1 table [Appendix] April 20, 2005

A New Species of Rhacophorus (Anura: Rhacophoridae) from Myanmar (Burma)

Jeffery A. Wilkinson!3, Thin Thin2, Kyi Soe Lwin?, and Awan Khwi Shein2 'Department of Herpetology, California Academy of Sciences, 875 Howard Street, San Francisco, California 94103, Email: jwilkinson@ calacademy.org; 2Nature and Wildlife Conservation Division, Forest Department, Ministry of Forestry, Bayintnaung Road, West Gyogone, Insein, Yangon, Myanmar; 3H. T. Harvey & Associates, 3150 Almaden Expressway, Suite 215, San Jose, CA 95118

A new species of the genus Rhacophorus is described from Myanmar. The new species is most similar to R. bipunctatus but differs in the male having a larger body size, a bright green dorsal coloration, yellow in the outer portion of the iris, fainter crossbands on the limbs, a more extensive dermal fringe along the arm, more exten- sive projection on the heel, more extensive webbing on the hand, and typically two large equal-sized black spots, one in the axillary region and one on the middle of the flank.

Ten species of Rhacophorus are thought to occur in Myanmar (R. appendiculatus (Giinther), R. bipunctatus Ahl, R. bisacculus Taylor, R. dennysi Blanford, R. feae Boulenger, R. maximus Giinther, R. reinwardtii (Schlegel), R. taronensis Smith, R. turpes Smith, and R. verrucosus Boulenger). Here, we report on an additional species of Rhacophorus collected during expeditions in 2001 and 2002 to Rakhine State in western Myanmar near the Bay of Bengal and to Kachin State in northeastern Myanmar near the Chinese border (Fig. 4).

METHODS AND MATERIALS

Specimens were collected by hand, euthanized, tissue samples removed, then fixed in 10% buffered formalin before preserving in 70% ethanol. Latitude and longitude were recorded with a Garmin 12 GPS, datum WGS84. Specimens are housed in the Department of Herpetology, California Academy of Sciences (CAS) and the Division of Amphibians and Reptiles, National Museum of Natural History, Smithsonian Institution (USNM).

The preserved specimens were examined, measured, and compared with available specimens (see material examined; museum acronyms follow Leviton et al. [1985]) and published descrip- tions of currently recognized (Frost 2004) species of Rhacophorus and Polypedates from Myanmar and neighboring countries (Boulenger 1920; Smith 1924; Smith 1940; Bourret 1942; Liu and Hu 1961; Taylor 1962; Inger 1966; Berry 1975; Dring 1983; Inger et al. 1985; Sarkar and Sanyal 1985; Inger and Dutta 1986; Kiew 1987; Daniel and Sekar 1989; Yang et al. 1991; Brown and Alcala 1994; Manthey and Grossman 1997; Inger and Stuebing 1997; Iskandar 1998; Chan-ard et al. 1999; Fei 1999; Inger et al. 1999; Das 2000; Ohler et al. 2000; Vasudevan and Dutta 2000; Orlov et al. 2001; Ziegler and Kohler 2001; Chanda 2002; Harvey et al. 2002; Malkmus et al. 2002). Measurements were taken using dial calipers to the nearest 0.1 mm as follows: snout-vent length (SVL, from tip of snout to vent); head length (HL, from tip of snout to hind border of angle of jaw);

42

WILKINSON ET AL.: NEW RHACOPHORUS FROM MYANMAR 43

head width (HW, width of head at its widest point); internarial distance (IND, distance between nares); interorbital distance (IOD, minimum distance between upper eyelids); snout length (SL, from anterior border of eye to tip of snout); distance from nostril to eye (DNE, from nostril to ante- rior border of eye); forelimb length (FLL, from elbow to tip of third finger); hand length (HAL, from base of outer palmer tubercle to tip of third finger); thigh length (THL, from vent to knee); tibia length (TIL, from knee to foot); foot length (FL, from proximal end of metatarsal tubercle to tip of fourth toe); width of disk of third finger (3FDW, greatest horizontal width); and width of disk of fourth toe (4TDW, greatest horizontal width).

SPECIES DESCRIPTION

Rhacophorus htunwini Wilkinson, Thin Thin, Kyi Soe Lwin, and Awan Khwi Shein, sp. nov. Figs. 1-3. Htun Win’s treefrog

DIAGNosis.— Rhacophorus htunwini can be distinguished from all other species of Rhacophorus and Polypedates by the following combination of characters: intermediate body size in the male (SVL 37.8-50.4 mm); extensive yellow webbing between fingers; thick dermal fringe on forearm and foot, dermal projection (calcar) on heel; squared-off supracloacal fold, snout point- ed: bright green dorsal color (slate blue in alcohol); yellow in the outer portion of the iris; very faint crossbands on the limbs; reddish-orange foot webbing; two large black spots on each side of body, one in the axillary region and one at the middle of the flank.

Ho.LotyPe.— CAS 229893 (Fig. 1), an adult male, collected from Nagmung Township, Au Yin Ga Camp (27°17'36.9’N, 97°51’45.3”E), Putao District, Kachin State, Myanmar, elevation approximately 878 m, collected on 2 May 2002 by Htun Win, Young Ngai Thi Na, Ram Sar, and Hpe Ram.

DESCRIPTION OF HOLOTYPE.— An adult male with slender habitus and head slightly longer than wide, 36% of SVL; snout pointed in dorsal view, gently slopes in lateral view to nostrils, then becomes slightly spatulate extending beyond mandible (Figs. la, c); nostrils closer to eye than tip of snout and anteriorly protuberant; canthus distinct, rounded, and inwardly curved; medial rostral areas between eyes and nostrils and between nostrils and tip of snout slightly concave; lores con- cave to nostril; eye directed anterolaterally with horizontal pupil. Tympanum distinct and circular; weak supratympanic fold, curving ventrally from dorso-posterior edge of tympanum to posterior

edge of axilla. Vomerine processes with 8/6 rounded teeth respectively and approximately equal to transverse

plane, separated medially by a space equal to two times their width, and laterally in contact with anteromedial edge of choanae; choanae small, ovoid to a medial point, and wholly visible at edge of lingual shelves of maxillae when viewed ventrally; tongue deeply bifurcates posteriorly; paired vocal slits oval and lateral to tongue.

Dorsal body surface smooth; ventral abdominal surface and region below thighs areolate; pec- toral and gular regions much less areolate. Vent protrudes posteriorly, squared supracloacal fold medially notched.

Arms short and slender; hand 73% as long as foot; when adpressed, relative length of fingers is3>4>2> 1: tips of fingers rounded; digital pads on hands and feet well developed and oval, with circummarginal grooves; distal phalanges bifurcate (as seen from dorsal aspect of digital tips). Hands extensively webbed, webbing formula for digits is I2-2111-1111-11V following Myers and Duellman (1982); narrow dermal fringe extends along lateral margin of fourth finger to base of hand. Subarticular tubercles between penultimate and adjoining proximal phalange round and well developed; proximal subarticular tubercles on finger 3 smallest; right hand with one, three, and two

44 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 4

Ficure 1. (A) Dorsal and (B) ventral views of the body, (C) lateral view of the head, and ventral views of the (D) left hand and (E) left foot of the holotype of Rhacophorus htunwini sp. nov. (CAS 229893).

small supernumerary tubercles in a row between proximal tubercle and base of hand on fingers two, three, and four, respectively; left hand with less obvious supernumerary tubercles; thenar tubercle low, extends medially at base of first finger, palmar tubercle absent. Thick dermal flange extends from lateral base of fourth finger to elbow, at widest approximately 18% of width of fore- arm.

Hind limbs moderately long and slender; when adpressed to body, tibiotarsal articulation reaches beyond anterior edge of eye: webbing on foot reaches to base of pads on all toes; when adpressed, relative length of toes is 4>5 =3>2> 1: thick dermal fringe from base of pad of toe 5 extends along lateral edge of tarsus to heel where it develops into broad flange with laterally pro- jecting calcar on lateral edge of heel. A single subarticular tubercle on toes 1 and 2, two subartic- ular tubercles on toes 3, 4, and 5; proximal tubercle on toe 5 approximately same size as distal tubercle; supernumerary tubercles and an outer metatarsal tubercle absent; inner metatarsal tuber- cle flat, oval, and pointed medially.

WILKINSON ET AL.: NEW RHACOPHORUS FROM MYANMAR 45

Coloration in preservative (Fig. 1). Dorsal color of body slate blue, extending laterally and ventrally '/4 of flanks and limbs, from elbow to halfway up middorsal aspect of hindarm, and along dorsal aspect of forearm, lateral fringe, and lateral half of fourth finger to base of pad, and from vent to knee along middorsal aspect of thigh, entire dorsal aspect of tibia, and lateral half of foot, lateral fringe, and fifth toe to just short of base of pad. Lateral margins of tarsal and supracloacal fringes light cream; first to third fingers and webbing, medial side of forearm, and all but middor- sal distal half of hindarm cream yellow; similarly, all but middorsal aspect of thigh, lateral and medial sides of tibia, medial half of foot, and first through fourth toes cream yellow. Webbing between toes two through five with streaks of orange, giving an orange appearance, but cream yel- low between toes one and two; small, elongated patch of slate blue on lateral side of second pha- lange of fourth toe and within webbing at same position between fourth and fifth toes; venter and flanks cream yellow. Posterolateral fringe of lower jaw and midventral aspect of fifth toe with some slate blue pigmentation; cream yellow margin on upper lip. Two large black oval spots on sides in axillary and mid-flank; axillary spot smaller, 8.7 mm horizontal diameter, mid-flank spot larger, 10.8 mm horizontal diameter.

Color in life based on a color transparency (Fig. 2a). Dorsum bright green with sparsely scat- tered black and white pin-sized spots. Dorsal surface of fingers one through three, toes one through four, all digital pads, webbing on hand, lateral margin of fringe from foot to heel, ventrum, and sides yellow to yellowish orange. Two large jet black spots, with clusters of light blue spots on the dorsal margins, on axillary and mid-flank. Faint crossbands present on dorsal aspect of thigh and tibia.

Pupil surrounded by light grayish brown horizontally rectangular iris, with yellow above and below, more so above. Yellow color extends posterior onto interior of orbit. Black thin line sur- rounds eye at margin with eyelid.

VARIATION.— The holotype, paratypes, and referred specimens of R. htunwini are male, sex- ual dimorphism could not be determined. The paratypes and referred specimens are similar to the holotype except for the following. CAS 222065 and 222136 have smaller calcars than holotype. CAS 222065 is much smaller (SVL 37.8 mm), and specimen CAS 222136 has one large black spot in axillary region, without posterior spot on flank. Supratympanic fold of CAS 221351 strong on both sides of head and covers dorsoposterior edge of tympanum, but does so only on left side of CAS 222065, in all other specimens supratympanic fold similar to holotype. The dorsum of CAS 221351 is lighter in coloration than other specimens in alcohol, although color pattern is same, however darker bluish green in life. CAS 221351 has scattered black pin-size spotting on head, dor- sum, and dorsal aspect of thighs. Snout of this specimen rounded instead of pointed as in all other specimens, and tips of pads more squared off than other specimens.

ETYMOLOGY.— The name /itunwini is given in honor of the late U Htun Win, who devoted the last eight years of his life to the pursuit of knowledge of the diversity and natural history of the rep- tiles and amphibians within his country. As team leader of the Myanmar Herpetological Survey Team, he first recognized this frog as potentially new to science.

COMPARISONS.— Because R. htunwini is a medium size tree frog that possesses expanded discs on the fingers and toes, an intercalary element between the penultimate and terminal pha- langes, a narrow bony metasternum, a flange on the distal end of the third metacarpal, Y-shaped terminal phalanges, and extensive webbing between the fingers and toes, it has been placed with- in Rhacophorus (Wilkinson and Drewes 2000). Rhacophorus is a relatively large genus of approx- imately 60 species from Asia (Frost 2004), and members of this genus closely resemble the approx- imately 28 species of the genus Polypedates (Liem 1970). Because of this close resemblance, sev- eral species have been moved back and forth between the two genera, or Polypedates has not been

46 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

recognized as a genus separate from Rhacophorus (Dubios 1986; Fei 1999). Recently, Wil- kinson et al. (2002) provided molecular evidence to separate the genera Polypedates and Rhacophorus, and to move two species in Polypedates (P. den- nyst and P. prasinatus) back into Rhacophorus. We believe that generic level reversals are still required in order to ensure the correct taxonomic placement of many species within these two genera and therefore have includ- ed members of both genera in the comparisons below.

Following the taxonomic designations in Frost (2004), R. htunwini can be distinguished from other species of the follow- ing Rhacophorus and Poly- pedates from Bangladesh, Cam- bodia, China, India, Indonesia, Laos, Malaysia, Myanmar, Thai- land, and Vietnam as follows: from members of Polypedates in the P. leucomystax species group (P. colleti, P. cruciger, P. eques, P. leucomystax, P. macrotis, P. maculatus, P. mutus, P. otilo- phus, and P. zed), P. insularis, P. megacephalus, P. naso, P. pseu- docruciger, members of Rha- cophorus in the subgenus Rha- cophorus in the R. appendicula- tus species group (R. appendicu-

Volume 56, No. 4

A

FiGuRE 2. Photos in life of (A) a male Rhacophorus htunwini sp. nov. (CAS 229893), (B) a female R. bipunctatus (CAS 229902), and (C) a male R. bipunctatus (CAS 224676).

latus, R. bisacculus, R. verrucopus), R. baliogaster, R. baluensis, R. barisani, R. bimaculatus, R. calcaneus, R. catamitus, R. cyanopunctatus, R. exechopygus, R. margaritifer, R. modestus, R. nam- daphaensis, R. orlovi, R. poecilonotus, R. translineatus, R. tuberculatus, and R. verrucosus by a bright green dorsal color; from members of Polypedates in the P. chenfui species group (P. chen- fui, P. hungfuensis, and P. yaoshanensis), the P. dugritei species group (P. dugritei, and P. omei- montis), P. dorsoviridis, P. duboisi, P. nigropunctatus, P. pingbianensis, P. puerensis, P. zhaojuen- sis, members of Rhacophours in the subgenus Rhacophorus and the R. dennysii species group (R. dennysi and R. feae), the R. pardalis species group (R. annamensis, R. notator, R. pardalis, and R. robinsonii), R. achantharrhena, R. angulirostris, R. taronensis, and R. variabilis by the presence of a sharp dermal calcar at the heel; from members of Rhacophorus in the subgenus Rhacophorus and

WILKINSON ET AL.: NEW RHACOPHORUS FROM MYANMAR 47

FiGuRE 3. Dorsal view of representative specimens of Rhacophorus htunwini sp. nov. (top row), representative speci- mens of female R. bipunctatus (middle row), and representative specimens of male R. bipunctatus (bottom row).

the R. malabaricus species group (R. calcadensis and R. malabaricus), R. lateralis, R. pseudoma- labaricus, and R. turpes by axillary spots.

Rhacophorus htunwini closely resembles members of the R. reinwardtii species group in the subgenus Rhacophorus (R. bipunctatus, R. dulitensis, R. maximus, R. nigropalmatus, R. promi- nanus, R. reinwardtii) and R. hoangliensis, but can be distinguished from all but R. bipunctatus and R. reinwardtii by the presence of axillary spots. It can be distinguished from R. reinwardtii (SVL male 68 mm) by its smaller size and yellow hand webbing (black hand webbing in R. reinwardtii).

Rhacophorus htunwini most closely resembles R. bipunctatus, which also has axillary spots, dermal calcars at the heels, a pointed snout, and sometimes a green dorsal color (Fig. 2b). However, it can be distinguished from R. bipunctatus by size (Table 1, Fig. 3); the males of R. htunwini are larger (average SVL =45.7 mm) than the males of R. bipunctatus (average SVL = 34.9 mm). Rhacophorus htunwini also has a bright green dorsal color in life that becomes slate blue when pre- served, whereas some members of R. bipunctatus have an olive green dorsal color in life that becomes light bluish gray or brown when preserved and others are orange to tan in life with a dark- er brown pattern (blotching or an X mark) on the dorsum, which become brown when preserved (Figs. 2-3). The eye of R. htunwini contains yellow at the upper and lower portion of the iris, which is absent in R. bipunctatus (Fig. 2). Rhacophorus bipunctatus has distinct crossbands on the fore and hindlimbs (Figs. 2b-c), whereas R. htunwini has faint crossbands that disappear in alcohol (Fig. 2a). Rhacophorus htunwini has a more extensive dermal fringe on the forearm, a more exten- sive dermal calcar at the heel, and more extensive webbing on the hand than the male of R. bipunc- tatus. In all but one specimen, R. htunwini has two large black equal size spots on the sides, one in

48 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 4

the axillary region and one on the flank, whereas, the major- ity of male specimens of R. bipunctatus examined had only one spot in the axillary region, and in specimens that had two spots the posterior spot was much smaller than the ante- rior spot.

DISTRIBUTION AND NATURAL HISTORY.— At present Rhacophorus htunwini is known from Nagmung and Machanbaw Townships, Putao District, Kachin State, and from much further south in Rahkine State in the southwest- ern foothills of Rakhine Yoma, Gwa Township, and Kyauktaw Township, Sittawe District (Fig. 4). This distribu- tional pattern indicates that this species may be restricted to the Indo-Burman Mountain Range that arcs from southwest- ern Myanmar along the border with India, and the Eastern Himalayas in northern Myanmar. The absence of specimens from the Chin Hills of western Myanmar is probably due solely to a lack of surveys in the region.

The type specimens including the holotype (CAS 229913, USNM 561869) were found approximately 2 m off the ground in bamboo. Referred specimens were found in undisturbed habitat near a spring (CAS 222136) or seasonal (CAS 221351) and permanent (CAS 222065) streams. Other species of Polypedates and Rhacophorus found in the vicinity of the type locality were P. leucomystax, R. bipunctatus, and R. dennysi.

kilometers

bese Ss

FiGure 4. Distribution of Rhacophorus htunwini sp. nov. in Myanmar with type locality indicated by a star (at tip of arrow).

MATERIAL EXAMINED

Rhacophorus htunwini (paratypes): CAS 229913, USNM 561869 adult males collected at the same local- ity and date as the holotype.

Rhacophorus htunwini (referred specimens): CAS 221351, an adult male, collected between Ahtan Ga and Au Rin Ga (27°15’27.2’N, 97°50’32.4”E), Ma Chan Baw Township, Putao District, Kachin State, Myanmar, collected on 4 September 2001 by Htun Win and Ran Shaung; CAS 222136, an adult male, collect- ed from Yea Pu Camp (17°56’02.6’N, 94°38’02.9”E), Gwa Township, Rakhine State, Myanmar, collected on 8 June 2001 by Hla Tun, Kyi Soe Lwin, and Awan Khwi Shein; CAS 222065, an adult male, collected from Pin Lone Camp, Pe Chaung, near Saba Sate Village (21°00°54.9”N, 92°52’06.6”E), Kyaut Taw Township, Sittawe District, Rakhine State, Myanmar, collected on 4 July 2001 by Htun Win, Kyi Soe Lwin, and Awan Khwi Shein.

Polypedates chenfui: FMNH 232963, 232964 (China).

Polypedates colletti: FANH 234773, 235631 (Malaysia).

Polypedates cruciger: CAS 85280 (Sri Lanka).

Polypedates dugritei: CAS 64273 (China).

Polypedates eques: CAS 85281, 85282 (Sri Lanka).

Polypedates leucomystax: FMNH 239159 (Malaysia); FMNH 254649 (Lao PDR); CAS-SU 15163 (India); CAS 14943 (China); CAS 94573 (Bangladesh); CAS 103624 (Indonesia); CAS 105003 (Vietnam); CAS 105972 (Malaysia); CAS 111336 (Cambodia); CAS 172691 (Thailand); CAS 221962, 224461 (Myanmar).

Polypedates macrotis: FMANH 239107, 239119 (Malaysia): CAS 60630, 60631, 60684, 60804, 62138 (Philippines); CAS 62581 (Malaysia); CAS 64074 (Philippines); CAS 64077 (Indonesia); CAS 64089-64092 (Philippines); CAS 105974, 105975 (Malaysia).

Polypedates maculates: CAS 16922-16924 (Sri Lanka); CAS 94571, 94572, 104152, 125365—125370 (India).

WILKINSON ET AL.: NEW RHACOPHORUS FROM MYANMAR 49

Polypedates megacephalus: ROM (field numbers) 18038, 18045 (Vietnam).

Polypedates otilophus: FANH 230836, 239147 (Malaysia).

Rhacophorus angulirostris: FANH 235035 (Malaysia).

Rhacophorus annamensis: FMNH 253933, 253940 (Vietnam); ROM 29889, 29890, 29891, 29892, 29897, 29901, 29904 (Vietnam).

Rhacophorus appendiculatus: CAS 60169-60174, 62261, 64078-64086 (Philippines).

Rhacophorus bimaculatus: CAS 61840, 133178—133180, 133251 (Philippines).

Rhacophorus bipunctatus: FMNH 253122, 253124 (Vietnam); NMNS 3220 (China); CAS 224676, 228808, 229887, 229889, 229890, 229898, 229899, 229901—229907, 229910 (Myanmar).

Rhacophorus calcaneus: FANH 256456, 257933 (Lao PDR); ROM 29849, 29850, 29854, 29855, 29875, 29877, 29879, 29880 (Vietnam).

Rhacophorus dennysi: FMNH 256449, 256450 (Lao PDR); ROM 29839, 29840, 29841, 29842, 29843, 29846, 30245 (Vietnam); CAS 64224 (China); CAS 221535, 224496, 224659 (Myanmar).

Rhacophorus dulitensis: FMNH 235741 (Malaysia).

Rhacophorus feae: FMNH 257910 (Lao PDR); CAS-SU 6387, 6388 (Vietnam).

Rhacophorus gaunt: FMNH 235044, 239238 (Malaysia).

Rhacophorus maximus: CAS 221516, 221517 (Myanmar).

Rhacophorus microtympanum: CAS 85283 (Sri Lanka).

Rhacophorus nigropalmatus: FMNH 230901, 230902 (Malaysia).

Rhacophorus pardalis: FANH 235750 (Malaysia); FMNH 259530 (Philippines); CAS 60472-60476, 61386, 128725, 129267—126270 (Philippines).

Rhacophorus reinwardtit: FANH 235034 (Malaysia), FMNH 255305 (Lao PDR); NMNS 3213 (China).

Rhacophorus rufipes: FANH 231377 (Malaysia).

Rhacophorus taronensis: BMNH 1947.2.8.17 (Myanmar).

Rhacophorus turpes: BMNH 1947.2.8.69, 1947.2.8.70 (Myanmar).

Rhacophorus verrucosus: CAS 224441, 224442, 224469, 224737, 224754, (Myanmar).

ACKNOWLEDGMENTS

We thank U Shwe Kyaw, Director General, Forest Department, Ministry of Forestry, and U Khin Maung Zaw, Director, Nature and Wildlife Conservation Division, Forest Department, Ministry of Forestry for their continued support of the Myanmar Herpetological Project. We also thank Barry Clarke, Mark Wilkinson, and Nick Arnold of the British Museum, Alan Resetar and Harold Voris of the Field Museum of Natural History, Wen-Hao Chou of the National Museum of Natural Science, Taiwan, and Robert Murphy and Ross MacCulloch of the Royal Ontario Museum for loaning specimens. Fieldwork was supported by National Science Foundation Grant (DEB- 9971861) to the late Joseph B. Slowinski (Alan E. Leviton, current PI) and George Zug. Michelle Koo provided Figure 1. Dong Lin provided the photographs for Figures 2 and 4, and Hla Tun pro- vided the photographs for Figure 3. Jens Vindum and Guinevere Wogan critically read and provid- ed valuable comments on the manuscript.

LITERATURE CITED

Berry, P.Y. 1975. The Amphibian Fauna of Peninsular Malaysia. Tropical Press, Kuala Lumpur, Malaysia 130 pp.

BOULENGER, G.A. 1920. III. Reptiles and batrachians collected in Korinchi, West Sumatra, by Messrs. H.C. Robinson and C. Boden Kloss. Journal of the Federated Malay States Museums 8:285—306.

Bourret, R. 1942. Les Batraciens de |’ Indochine. Memoires de L’ Institut Oceanographique de L’ Indochine. Gouvernement General de L’ Indochine. Hanoi, Indochina (Vietnam). 547 pp.

Brown, W.C., AND A.C. ALCALA. 1994. Philippine frogs of the family Rhacophoridae. Proceedings of the California Academy of Sciences 48:185—220.

50 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 4

CHANDA S.K. 2002. Hand Book of Indian Amphibians. Zoological Survey of India, Kolkata, India. 335 pp.

CHAN-ARD, T., W. GROSSMANN, A. GUMPRECHT, AND K.-D. SCHULZ. 1999. Amphibians and Reptiles of Peninsular Malaysia and Thailand, An illustrated Checklist. Bushmaster Publications, Wuerselen. 240 pp.

DANIEL, J.C., AND A.G. SEKAR. 1989. Field guide to the amphibians of western India. Part 4. Journal of the Bombay Natural History Society 86:194—200.

Das, I. 2000. Nomenclatural history and rediscovery of Rhacophorus lateralis Boulenger, 1883 (Amphibia: Rhacophordae). Current Herpetology 19:35—40.

DRING, J. 1983[1984]. Some new frogs from Sarawak. Amphibia-Reptilia 4:103—112.

Dusois, A. 1986. Miscellanea taxinomica batrachologica (I). Alytes 5:17—95.

Fer, L. 1999. Atlas of Amphibians of China. Publishing House for Scientific and Technological Literature, Hunan, China. 432 pp.

Frost, D.R. 2004. Amphibian Species of the World: an Online Reference. Version 3.0 (22 August, 2004). Electronic Database accessible at http://research.amnh.org/herpetology/amphibia/index.html. American Museum of Natural History, New York, USA.

Harvey, M.B., A.J. PEMBERTON, AND E.N. SmitH. 2002. New and poorly known parachuting frogs (Rhacophoridae: Rhacophorus) from Sumatra and Java. Herpetological Monographs 16:46—92.

INGER, R.F. 1966. The systematics and zoogeography of the Amphibia of Borneo. Fieldiana: Zoology 52:1-402.

INGER, R.F., AND S.K. Dutta. 1986. An overview of the amphibian fauna of India. Journal of the Bombay Natural History Society 83(Suppl.):135—146.

INGER, R.F., N.L. ORLOV, AND I.S. DAREVSKY. 1999. Frogs of Vietnam: a report on new collections. Fieldiana: Zoology 92:1-46.

INGER, R.F., H.B. SHAFFER, M. KOSHY, AND R. BAKDE. 1985 [1984]. A report on a collection of amphibians and reptiles from the Ponmudi, Kerala, south India. Journal of the Bombay Natural History Society 81:551-570.

INGER, R.F., AND R.B. STUEBING. 1997. A Field Guide to the Frogs of Borneo. Natural History Publications, Kota Kinabalu. 205 pp.

ISKANDAR, D.T. 1998. Amphibians of Java and Bali. Research and Development Centre for Biology, Jakarta, Indonesia. 117 pp.

Krew, B.H. 1987. An annotated checklist of the herpetofauna of Ulu Endau, Johore, Malaysia. Malayan Nature Journal 41:413-423.

LEVITON, A.E., R.H. GipsBs, JR., E. HEAL, AND C.E. DAwson. 1985. Standards in herpetology and ichthyolo- gy: Part I. Standard symbolic codes for institutional resource collections in herpetology and ichthyology. Copeia 1985:802-832.

Liem, S.S. 1970. The morphology, systematics, and evolution of the Old World treefrogs (Rhacophoridae and Hyperoltidae). Fieldiana Zoology 57:1—145.

Liu, C.C., AND S.Q. Hu. 1961. Tailless Amphibians of China. Science Press, Peking, China. 364 pp.

Matckmus, R., U. MANTHEY, G. VOGEL, P. HOFFMANN, AND J. KosucH. 2002. Amphibians and Reptiles of Mount Kinabalu. A.R. G. Gantner Verlag Kommanditgesellschaft, Ruggel, 424 pp.

MANTHEY, U., AND W. GROSSMANN. 1997. Amphibien and Reptilien Siidostasiens. Natur und Tier Verlag, Miinster, Germany. 511 pp.

Myers, C.W., AND W.E. DUELLMAN. 1982. A new species of Hyla from Cerro Colorado, and other tree frog records and geographical notes from western Panama. American Museum Novitates (2752):1—32.

OHLER, A., O. MARQUIS, S. SWAN, AND S. GROSJEAN. 2000. Amphibian biodiversity of Hoang Lien Nature Reserve (Lao Cai Province, northern Vietnam) with description of two new species. Herpetozoa 13:71-87.

OrLov, N.L., A. LATHROP, R.W. MurPHy, AND H.T. Cuc. 2001. Frogs of the family Rhacophoridae (Anura: Amphibia) in the northern Hoang Lien Mountains. Russian Journal of Herpetology 8:17-44.

SARKAR, A.K., AND A.K. SANYAL. 1985. Amphibia. Records of the Zoological Survey of India 81:285-—295.

SMITH, M.A. 1924. New tree-frogs from Indo-China and the Malay Peninsula. Proceedings of the Zoological Society, London 1924:225-234.

SmitH, M.A. 1940. The amphibians and reptiles obtained by Mr. Ronald Kaulback in upper Burma. Records of the Indian Museum 42:465-486.

WILKINSON ET AL.: NEW RHACOPHORUS FROM MYANMAR 51

TayLor, E.H. 1962. The amphibian fauna of Thailand. University of Kansas Science Bulletin 43:265—599.

VASUDEVAN, K., AND S.K. Dutta. 2000. A new species of Rhacophorus (Anura: Rhacophoridae) from the Western Ghats, India. Hamadryad 25:21-28.

WILKINSON, J.A., AND R.C. DREwes. 2000. Character assessment, genus level boundaries, and phylogenetic analyses of the family Rhacophoridae: a review and present day status. Contemporary Herpetology 2000:2.

WILKINSON, J.A., R.C. DREWES, AND O.C. Tatum. 2002. A molecular phylogenetic analysis of the family Rhacophoridae with an emphasis on the Asian and African genera. Molecular Phylogenetics and Evolution 24:265-273.

YANG, D.T., S. Li, W. Lu, AND S. Lu. 1991. Amphibian Fauna of Yunnan, China. Forestry Publishing House, Beijing. 259 pp.

ZIEGLER, T., AND J. KOHLER. 2001. Rhacophorus orlovi sp. n., a new tree frog from Vietnam (Amphibia: Anura: Rhacophoridae). Sauria 23:37-46.

SZ

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 4

Appendix

TABLE |. Measurements of the type series and referred specimens of Rhacophorus htunwini and comparative material of R. bipunctatus. Mean (in mm) followed by range (in parenthesis) and ratio of SVL (below). See text for abbreviations.

Rhacophorus htunwini

Male N=6

Rhacophorus bipunctatus Rhacophorus bipunctatus

Female N=3

Male N= 16

SVL

HL

HW

IND

IOD

SN

DNE

BEE

HL

THL

AUG

IRIE

3FDW

4TDW

aS (SURV 9)

16.4 (13.5-17.7) 39.9 16.5 (14.9-17.5) 36 4.4 (4.3-4.9) 9.6 5.2 (4.8-5.9) 11.4 7.3 (6.1-8.1) 16 3.7 (3.1-4.2) 8 22.0 (18.6-25.1) 48.1] 13.8 (11.2-15.5) 30.3 22.5 (17.6-26.1) 49.3 23.0 (19.4-25.1) 50.3

5.4 1.9 (1.6-2.2) 4.2

Copyright © 2005 by the California Academy of Sciences

55.9 (51.2-60.2) 18.3 (17.2-19.8) S25;

17.9 (16.9-19.3) S251 4.5 (4.4-4.7) 8.1 6.6 (5.4-7.5) 11.8 8.0 (7.7-8.4) 14.2 3.6 (3.2-4.0) 6.4 27.9 (26.8-29.2) 49.8 17.6 (16.5-19.0) 31.4 27.0 (26.7-27.4) 48.2 26.2 (25.0-27.9) 46.9 25.9 (24.8-27.9) 46.3 3.2 (3.1-3.4) 5.8 2.4 (2.0-2.7) 4.3

San Francisco, California, U.S.A.

34.9 (32.0-37.7) 12.5 (11.5-14.0) 35.9 12.6 (11.4-14.1) 36 3.4 (2.7-4.2) 9.8 3.8 (2.9-4.5) 10.9 5.4 (4.6-6.0) 15.6 2.6 (2.1-2.9) 7.4 17.9 (16.1-19.7) 51.2 11.2 (10.0-12.5) 32 17.6 (14.3-19.7) 50.5 17.5 (15.6-18.7) 50.3 15.6 (13.9-16.9) 44.7 1.9 (1.5-2.4) 5.4 1.5 (1.2-1.9) 4.2

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 5, pp. 53-65, 10 figs. April 29, 2005

New Taxa of Octocorals (Anthozoa: Octocorallia) from the Northeastern Pacific Ocean

Gary C. Williams Department of Invertebrate Zoology and Geology, California Academy of Sciences, S875 Howard Street, San Francisco, California 94103, U.S.A, Email: gwilliams @ calacademy.org

The octocorallian fauna of the Alaskan region is poorly known. Recently, several research projects have been initiated to explore the faunal elements of the continen- tal shelf in the Gulf of Alaska. A wealth of material, including octocorals (mostly gor- gonians with some pennatulaceans), has been collected by trawl nets during field research conducted by agencies such as the National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration (NOAA). A new genus of gorgonian has recently been described from the Aleutian Islands as Alaskagorgia Sanchez and Cairns, 2004. The present paper reports the discovery of an additional plexaurid gorgonian from the same region, and provides a description of it as a new genus and species. Additionally, an unusual new species of sea pen, genus Cavernularia (Family Veretillidae), is described from the central Aleutian island chain. The geographically nearest veretillid taxa outside of the type locality occur in the northwestern and eastern Pacific — Japan and Panama.

According to Wing and Barnard (2004), approximately 64 species of octocorals are presently known to occur in Alaskan waters. The authors also provide a key to the known octocoral taxa and color figures for twenty species. Other kinds of coral also included are scleractinians, stylasterine hydrocorals, and antipatharians.

The gorgonian family Plexauridae Gray, 1859 has relatively recently had a plethora of genera allocated to it along with the amalgamation of several groups at the ranks of family and subfami- ly. Included here are the Muriceidae, Paramuriceidae, Plexaurinae, and Stenogorgiinae (= Paramuriceinae) (Bayer 1981). The tropical western Atlantic genera were treated by Bayer (1961). Recent works have compiled previously described taxa and have added new generic names mak- ing a total of approximately 38 genera that are presently considered to comprise the family. These works include Grasshoff (1977) for the northern Atlantic Ocean and Mediterranean Sea, Grasshoff and Barbigant (2001) and Frabricius and Alderslade (2001) for the tropical western Pacific, and Sanchez and Cairns (2004) for Alaskan waters. The present paper describes an additional new genus and species from the Alaskan Aleutian Islands, making a total of 39 worldwide genera in the family Plexauridae. The new taxon does not closely resemble other plexaurid taxa.

In addition, a new species of the veretillid pennatulacean genus Cavernularia is described, also from the Aleutian Archipelago. The only other known taxa of vertetillid pennatulaceans from the northwestern and eastern Pacific Ocean include several species from Japan, Panama, and the Galapagos Islands (L6pez-Gonzalez, Gili, and Williams 2000; Kiikenthal 1915; Hickson 1921; and Williams 1989, 1995).

53

54 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

METHODS

Material was collected on board a research vessel (the F/V “Vesteraalen” or the M/V “Sea Storm’’) using trawling gear, and was preserved in 75% ethanol. Sclerites were isolated from sur- rounding tissues by the use of sodium hypochlorite (household bleach). Micrographs were made using a Nikon Coolpix 990 digital camera, sometimes in combination with a Nikon SMZ-10 dis- secting microscope or an Olympus CH-2 compound microscope. Original drawings of sclerites were made with an Olympus CH-2 compound microscope and an Olympus drawing tube. Scanning electron micrographs were taken using a LEO 1400 series scanning electron microscope. Digital images for the making of plates were manipulated using Adobe Photoshop software. Abbreviations used in the text are CAS (California Academy of Sciences, San Francisco), NMFS (National Marine Fisheries Service), and NOAA (National Oceanic and Atmospheric Administration — of the United States Department of Commerce).

SYSTEMATIC ACCOUNT

Order Alcyonacea Lamouroux, 1816 Family Plexauridae Gray, 1859

REMARKS.— Most of the taxa allocated to this family have relatively large sclerites, over 0.3 mm long (up to a maximum length of approximately 5 mm), with tubercular sculpture rarely arranged in whorls. In addition, both Bayer (1981:924) and Fabricius and Alderslade (2001:59) use the following axial characteristics to describe the family. The plexaurid axis often appears woody and fibrous, often brown or darker. The central 5 ; core is wide, hollow, and cross-chambered. The surrounding axial cortex is soft, usually with hollow spaces called loculi that are sometimes filled with a fibrous substance or non-spicular calcite. In contrast, the family Gorgoniidae usually has small scerites, less than 0.3 mm long, with tubercles arranged in whorls. The axis is often dark brown or black. It has a nar- row core and the cortex is very dense with few or no locules.

The new taxon described here is somewhat enigmatic in that it has small scerites, less than or equal to 0.15 mm in length, which is charac- teristic of the Gorgoniidae. Additionally, the tubercles of the spindles and radiates are some- times arranged in whorls, but often they are not. Regardless of the problematic nature of these features, the new taxon is here placed in the family Plexauridae because the axis has a

FiGURE 1. Non-type specimen (CAS 164584) of

woody or fibrous appearance and wide and hol- low central core, whereas the surrounding axial cortex is soft with easily observable amounts of non-spicular calcite (Fig. 1).

Cryogorgia koolsae gen. and sp. nov. Transverse section of a branch near the apical end showing the axis and thin coenenchyme; scale bar = 3.8 mm; ca — calcitic material, em — coenenchyme, ex — axial cortex, he — hollow core.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN 55

Genus Cryogorgia Williams, gen. nov. TYPE SPECIES: Cryogorgia koolsae Williams, sp. nov., by monotypy

DrAGNosis.— Branching relatively sparse, mostly lateral but rarely dichotomous, planar, anas- tomosis absent. Terminal branches elongate, somewhat clavate. Azooxanthellate. Axis proteina- ceous, flexible with wide hollow core, brown to dark brown, cylindrical near holdfast, slightly compressed in higher branches. Polyps monomorphic, retractile into polyp mounds or completely into coenenchyme, calyces absent. Coenenchymal sclerites leaf clubs, coarsely-tuberculated spin- dles, radiates, modified and often clavate radiates (some approaching wart clubs), and crosses. Sclerites colorless. Wet-preserved coenenchyme tan-white.

ETyMOLOGyY.— The new generic name is derived from the Greek, Kryos (cold or chilly) and the commonly used suffix for a gorgonian octocoral, -gorgia; in reference to the cold nature of the type locality of this gorgonian.

Cryogorgia koolsae Williams, sp. nov. Figs. 1-8

MATERIAL EXAMINED.— HOo_ortyPe: CAS 151364, Northeastern Pacific Ocean, United States, Alaska, Aleutian Islands, 52.25°N 171.70°W, 406 m depth, Haul 79, 11 June 2000, collected by E.J. Kools aboard F/V “Vesteraalen,’ NMFS Aleutian Survey 2000, one whole colony wet preserved in 75% ethanol. PARATYPE: CAS 150650, Northeastern Pacific Ocean, United States, Alaska, Aleutian Islands, 52.07° N 177.25° W, 86 m depth, Haul 146, 26 June 2000, collected by E.J. Kools aboard F/V “Vesteraalen,’ NMFS Aleutian Survery 2000, one whole colony wet preserved in 75% ethanol. OTHER MATERIAL: CAS 164584, Northeastern Pacific Ocean, United States, Alaska, Aleutian Islands, 51.75°N 175.67°W, 83 m depth, Haul 101, 7 July 2002, collected by R.J. Van Syoc aboard M/V “Sea Storm,” three whole colonies wet preserved in 95% ethanol. CAS 168893, Northeastern Pacific Ocean, United States, Alaska, Aleutian Islands, 52.31°N 173.65°W, 402 m depth, Haul 144, collected by R.J. Van Syoc aboard M/V “Sea Storm,” one whole colony wet preserved in ethanol.

DESCRIPTION OF HOLOTYPE.— Growth form and size (Figs. 2—3): The holotype measures 370 mm long by 150 mm wide. The basal trunk is 115 mm in length before the first side branch arises. It averages approximately 10 wide and 8 mm deep, as it is slightly compressed. The branching is irregular (lateral) and sparse. The terminal branches are 30-100 mm in length.

Polyps (Figs. 3A—E): The polyps of the holotype exhibit various states of retractility. Some are preserved totally retracted and flush with the surrounding coenenchyme — especially those along the main stem and along the thicker branches, whereas others are preserved in various stages of exsertion. The length of the exserted polyps is usually less than 2 mm. They vary in shape from mound-like and hemispherical to cone-shaped or somewhat cylindrical (Fig. 3E). These mounds or cones are thickly set with coenenchymal sclerites (Fig. 5). In all cases, the anthocodiae are at least partly retracted or hidden. Examination of dissected polyps has not shown the anthocodiae to con- tain sclerites.

Sclerites (Figs. 2C, 4-7): Coenenchymal sclerites of the polyp-bearing branchs, branch tips, and polyp mounds are of several different forms: foliates or leaf clubs (0.07—0.12 mm in length); seven- and eight-radiates (0.07—0.10 mm long); clavate forms (0.08—0.15 mm in length); crosses (0.10—0.11 mm long); and modified radiates, cylinders, and spindles (0.07—0.15 mm in length). Anthocodial sclerites are apparently absent.

Color (Figs. 2—3): The axis is dark brown to black. The coenenchyme and retracted polyps are grayish white to tan-white. The sclerites are colorless.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

FIGURE 2. A-B. Holotype of Cryogorgia koolsae gen. and sp. noy.; scale bars = 65 mm. C. Three leaf club sclerites

from the surface coenenchyme of a terminal branch; scale bar = 0.05 mm.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN oy/

FIGURE 3. Holotype of Cryogorgia koolsae gen. and sp. nov.; details of branches and polyps. A—B. Details of branch-

ing patterns. CD. Details of a single branch and polyps. E. Detail of a single polyp. Scale bars: A = 25 mm, B = 25 mm,

C=5 mm, D=2 mm, E=1 mm.

58 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

Ficure 4. Cryogorgia koolsae gen. and sp. nov. Coenenchymal sclerites from a polyp-bearing branch; scale bar = 0.10 mm. A. Holotype (CAS 151364). B. Paratype (CAS 150650). C. Non-type specimen (CAS 164584). D. Non-type specimen (CAS 168893). Scale bar = 0.10 mm.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN 59

FiGuRE 5 (above) Cryogorgia koolsae gen. and sp. noy. Coenenchymal sclerites from a mound formed by a retracted polyp: scale bar = 0.10 mm. FiGurRE 6 (below). Holotype of Cryogorgia koolsae gen. and sp. nov. Coenenchymal sclerites from the base of the holdfast; scale bar = 0.10 mm.

60 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

FIGURE 7. Holotype of Cryogorgla koolsae gen. and sp. noy. Scanning electron micrographs of leaf clubs from the coenenchyme of a terminal branch. A. 0.08 mm. B. 0.11 mm. C. 0.08 mm. D. 0.08 mm. E. 0.09 mm. F. 0.08 mm. G. 0.07 mm. H. 0.09 mm. I. 0.08 mm. J. 0.10 mm. K. 0.08 mm. L. 0.06 mm. M. 0.10 mm. N. 0.12 mm. O. 0.10 mm.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN 61

ETyMOLOGyY.— This species is named for its discoverer, Elizabeth J. Kools, (Department of Invertebrate Zoology and Geology, California Academy of Sciences, San Francisco).

DISTRIBUTION.— This new taxon is presently known from four localities in the central Aleutian Archipelago of Alaska, northern Pacific; 83-406 m depth (Fig. 8).

RUSSIA Beaufort Sea

Bering Sea

CANADA

* CAS 150650 C CAS 168893

av CAS 164584 al See

CAS 151364 x

Gulf of Alaska

FiGuRE 8. Map of the Aleutian Island Chain showing collecting stations (with collection numbers) for Cryogorglia koolsae gen. and sp. nov. (); arrow shows type locality.

Order Pennatulacea Verrill, 1865 Family Veretillidae Herklots, 1858

Genus Cavernularia Valenciennes in Milne-Edwards and Haime, 1850

Cavernularia vansyoci Williams, sp. nov. Figs. 9-10

MATERIAL EXAMINED.— HOLOTYPE: CAS 168894, Northeastern Pacific Ocean, United States, Alaska, Aleutian Islands, 52.0663° to 52.07043°N; 175.30231° to 175.2824°W, 86-93 m depth, Haul Number 66, 13 June 2002, collected by Robert Van Syoc on board M/V “Sea Storm,’” NMFS Aleutian Survey 2002, one whole colony cut longitudinally into two halves and wet preserved in 75% ethanol.

DESCRIPTION OF HOLOTYPE.— Growth form and size (Figs. 9A-B, 10A): The colony is clavate, 33 mm in total length.

Polyps (Figs. 9A—D, 10A): The polyps of the holotype are all preserved completely retracted. They are flush with the surface of the rachis and nowhere do they extend past the surface. Calyces are absent. The polyps and polyp walls contain minute sclerites, similar to the sclerites of the surperficial coenenchyme and peduncular interior. As in the coenchymal tissues, these sclerites can only accurately be detected at microscope magnifications at 400x or higher.

Internal Anatomy (Figs. 9C—D): The specimen was cut longitudinally into two halves to reveal internal structures. Gastric cavities are approximately | mm wide and 4-5 mm long. Numerous ova can be concentrated in interior-most parts of some of the cavities. In the center of the colony, sev- eral vertical canals can be seen. A calcified axis is absent altogether.

Sclerites (Figs. 9E, 1OB—-C). The sclerite complement is composed entirely of very small ovals

62 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

FIGURE 9. Cavernularia vansyoci sp. nov. A—B. External views of both halves of the longitudinally sectioned holotype. C—D. Internal views of both halves of the longitudinally sectioned holotype. E. Photomicrographs of three sclerites from the superficial coenenchyme taken from the region of juncture between the rachis and peduncle of the holotype; the scle- ] }

rites are actually colorless, color shown here is due to properties of the light microscope at high power. Scale bar for A-D

= 16 mm: scale bar for E = 0.005 mm.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN 63

a. ~. uty Aleutian Archipelago

Gulf of Alaska

153°

Ficure 10. Cavernularia vansyoci, sp. nov. A. Whole wet-preserved holotype, 33 mm in length. B. Coenenchymal scle- rites from the surface of the junction between the rachis and the peduncle. C. Sclerites from the interior of the peduncle. Scale bar for B and C = 0.01 mm. D. Map showing locality of the holotype (Mf = collecting station).

64 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 5

(0.003—0.007 mm in length). The shape of these sclerites is only accurately discernable at micro- scopic magnifications of 400x or more. They vary in shape from rounded-rectilinear to oval, more- or-less round, or elliptical. Some sclerites are irregular in shape. These minute sclerites are rela- tively dense is all parts of the colony examined, including the coenenchyme of the surface of the rachis and peduncle, as well as the interior of the colony, but seem to be less densely distributed in the anthocodiae and polyp walls.

Color (Figs. 9A—D). The color of the preserved holotype is grayish white throughout. The scle- rites are colorless.

ETYMOLOGY.— This species is named for its discoverer, Robert Van Syoc, Department of Invertebrate Zoology and Geology, California Academy of Sciences, San Francisco.

DISTRIBUTION.— The new species is known only from the type locality in the Aleutian Islands of Alaska; 86—93 m depth (Fig. LOD).

DISCUSSION

The most recent compendium of Alaskan corals is that of Wing and Barnard (2004), which includes a listing of twelve species of plexaurid octocorals and twelve species of pennatulaceans. Of these, three plexaurids and three pennatulaceans are figured with color photographs. The taxa figured are identified as belonging to five genera — Muriceides, Swiftia, Protoptilum, Ptilosarcus, and Halipteris. The present paper describes a new genus and species of plexaurid gorgonian and a new species of the pennatulacean genus Cavernularia — these are newly recorded taxa to the Alaskan octocoral fauna.

ACKNOWLEDGMENTS

I gratefully appreciate the contributions and services of Elizabeth Kools and Robert Van Syoc (Department of Invertebrate Zoology and Geology, California Academy of Sciences), and the National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration (NOAA), for making this project possible.

LITERATURE CITED

BAYER, F.M. 1961. The Shallow-water Octocorallia of the West Indian Region. A Manual for Marine Biologists. Marinus Nijhoff, The Hague, Netherlands. 373 pp.

BAYER, F.M. 1981. Key to the genera of Octocorallia exclusive of Pennatulacea (Coelenterata: Anthozoa), with diagnoses of new taxa. Proceedings of the Biological Society of Washington 94(3):902-947.

FABRICIUS, K., AND P. ALDERSLADE. 2001. Soft Corals and Sea Fans — a Comprehensive Guide to the Tropical Shallow-water Genera of the Central-West Pacific, the Indian Ocean and the Red Sea. Australian Institute of Marine Science, Townsville, Queensland, Australia. 264 pp.

GrassHorF, M. 1977. Die Gorgonarien des ostlichen Nordatlantik und des Mittelmeeres Ill. Die Familie Paramuriceidae (Cnidaria, Anthozoa). “Meteor” Forschungsinstitut-Ergebnisse D 27(5—76):5—76.

GRASSHOFF, M., AND G. BARGIBANT. 2001. Coral Reef Gorgonians of New Caledonia. IRD Editions, Institut de Recherche Pour Le Développment, Collection Fauna et Flore tropicales 38. Paris, France. 335 pp.

Hickson, S.J. 1921. On some Alcyonaria in the Cambridge Museum. Proceedings of the Cambridge Philosophical Society 20:366—373.

KUKENTHAL, W. 1915. Pennatularia. Das Tierreich 43:1—132.

LOPEZ-GONZALEZ, P.J., J.-M. GILI, AND GC.WILLIAMS. 2000. On some veretillid pennatulaceas from the east- ern Atlantic and western Pacific Oceans (Anthozoa: Octocorallia), with a review of the genus Cavernularia, and descriptions of new taxa. Journal of Zoology, London 250:201-216.

WILLIAMS: NEW OCTOCORALS FROM NORTHEASTERN PACIFIC OCEAN 65

SANCHEZ, J.A., AND S. CAIRNS. 2004. An unusual new gorgonian coral (Anthozoa: Octocorallia) from the Aleutian Islands, Alaska. Zoologische Mededelingen, Leiden 78(15):265—274.

WituiaMs, GC. 1989. The pennatulacean genus Cavernularia Valenciennes (Octocorallia: Veretillidae). Zoological Journal of the Linnean Society 95:285-—310.

Wituiams, G.C. 1995. Living genera of sea pens (Coelenterata: Octocorallia: Pennatulacea): illustrated key and synopses. Zoological Journal of the Linnean Society 113:93-140.

WinG, B.L., AND D.R. BARNARD. 2004. A Field Guide to Alaskan Corals. NOAA Technical Memorandum NMEFS-AFSC-146. U.S. Department of Commerce, Washington, DC, USA. 67 pp.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 6, pp. 66-75, 9 figs., 1 table April 29, 2005

A New Species of Marionia (Gastropoda: Nudibranchia) from the Caroline Islands

Victor G. Smith and Terrence M. Gosliner Department of Invertebrate Zoology and Geology, California Academy of Sciences, 875 Howard Street, San Francisco CA 94103, U.S.A., Email: vsmith@calacademy.org, tgosliner@ calacademy.org

A new species of tritoniid nudibranch is described from the western Caroline Islands. Marionia bathycarolinensis is known only from its type location of Palau. This animal exceeds 10 cm in length, is reddish in color and is the second record from the Indo-Pacific region of a tritoniid with more than 20 pairs of branchial plumes. Internally, it is distinguished from other Indo-Pacific tritoniids by jaws with masti- catory borders composed of multiple rows of rodlets, a broad radula with elongate and distinctively shaped rachidian teeth, and a proportionally long esophagus. A muscular band carries approximately 50 chitinous plates arranged so their distal edges face the lumen of the stomach. Analysis of stomach contents shows that the holotype was feeding on octocorals of the genus Paracis. This is the first record of a nudibranch feeding on this genus. Placement of the new species in the genus Marionia is discussed in the light of taxonomic problems brought about by the treat- ment of the group by Odhner (1963).

The current state of knowledge of the tritoniid nudibranchs from the Indo-Pacific region leaves workers with a difficult set of taxonomic and nomenclatural problems to overcome as they attempt to sort out the relationships within this diverse family. Descriptions from the early 1800's through the mid 1900’s lack consistency and are based primarily on external morphology, with only scat- tered references to important anatomical details. Subsequently, many species names have been syn- onomized, and other species have not been recorded since their original description. In some cases this may be due to the relative rarity of the taxa, in others the descriptions are insufficient to con- vince modern workers of their validity. Initial taxonomy was established by and expanded upon by Odhner (1936, 1963) but other workers still find difficulties with the most recent system (Willan 1988). Odhner proposed three plate bearing genera: Marionia, Marioniopsis and Paratritonia. His division was based primarily on characters of the digestive gland (liver), jaws and radula. Marionia was defined as having a digestive gland in two masses leaving the stomach uncovered, a jaw with 3 to 6 rows of fine denticles and a radula possessing tricuspid central teeth and differentiated first lateral teeth. Marioniopsis was defined by a digestive gland in a single mass covering the upper and left side of the stomach, jaws with a single row of strong denticles and a radula as in Marionia. In his 1963 key, Odhner defined Paratritonia as having a radula with unicuspidate central teeth and undifferentiated first laterals. This does not agree with the original description (Baba 1949), which stated that the first lateral teeth differ from the outer laterals. Baba also describes the jaw as hay- ing up to 10 rows of fine denticles, and a digestive gland in a single mass.

Odhner’s classification scheme relies heavily on the morphology of the digestive gland, which he felt was of utmost importance in dividing the genera. This presents problems on more than one

66

SMITH AND GOSLINER: A NEW SPECIES OF MARIONIA (NUDIBRANCHIA) 67

level. Firstly, the progression of the nudibranch digestive gland from a compact single mass to a divided state and then to a diffuse distribution is on a continuum, is not composed of discrete steps, and appears in multiple lineages. Secondly, detailed dissections of numerous tritoniids by the sen- ior author have not always yielded results in agreement with published descriptions (e.g., Odhner 1963). In addition, there have been descriptions of new animals (e.g., Willan 1988) that do not fit well into any of the current genera, including the present study. Accordingly, the authors feel that it is most prudent for all new species of plate-bearing tritoniids to be placed in the genus Marionia, which precedes the more controversial genera Paratritonia and Marioniopsis.

Nudibranchs of the family Tritoniidae are thus far known to feed exclusively on soft corals, sea pens, and gorgonians, but exact feeding patterns are unknown for many species (McDonald and Nybakken 1999). The gut contents of Marionia bathycarolinensis sp. nov. contained skeletal ele- ments of an undescribed octocoral of the genus Paracis Ktikenthal, 1919; Family Plexauridae Gray, 1859. This is the first record of a nudibranch feeding on this genus of octocoral.

METHODS

Dissection was performed by a ventral incision through the length of the sole of the foot and around the genital and anal openings, allowing the removal of the entire visceral mass in one piece. The jaws and radula were freed from the buccal mass by partially dissolving tissues in 10% KOH solution. They were then rinsed in deionized water, air dried, mounted and coated for electron microscopy. Scanning electron micrographs (SEMs) were produced with a Hitachi S-520 or Leo 1450VP scanning electron microscope. Digestive contents were treated with undiluted household bleach to dissolve the tissue from the octocoral skeletal elements. After multiple rinses in water and 75% EtOH, sclerites and axial samples were mounted, dried and coated for electron microscopy. Slides of sclerites were also prepared for optical microscopy.

SPECIES DESCRIPTION

Suborder Dendronotacea Odhner, 1934 Family Tritoniidae Lamarck, 1809

Genus Marionia Vayssiére, 1877 TYPE SPECIES: Marionia berghi Vayssiere, 1877

Marionia bathycarolinensis Smith and Gosliner, sp. nov. (Figs. 1-9)

TYPE MATERIAL.— All material examined was collected by the Coral Reef Research Foundation (CRRF) and has been deposited at California Academy of Sciences Department of Invertebrate Zoology (CASIZ). HoLorype: CASIZ 156081, one specimen collected at 191.4 m depth, on rock, Mutremdiu 3, Palau, Caroline Islands, 07°16.27’N, 134°34.37’E, 18 March, 2001, L.J. Bell aboard the Deep Worker Submarine, dissected. PARATYPE: CASIZ 156082, buccal mass only, remainder of body not sent, Mutremdiu 3, Palau, Caroline Islands, 07°16.27'N, 134°34.37E, 25 March, 2001, P.L. Colin aboard the Deep Worker Submarine, 222.5 m depth, on rock, dissect- ed.

EXTERNAL ANATOMY.— The size of the living paratype specimen was approximately 118 mm in length. Only the buccal mass of this specimen was sent. The color of the living animal (Fig. la) is brick red, with scattered patches of lighter red and greenish-red on the notum. There are also small white flecks, patches, and short linear white markings on the body. The sheaths and shafts of

68 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 6

the rhinophores are brick red, with the apical portions a con- trasting dark greenish-red color. The preserved holotype speci- men is 155 mm in length, and 65 mm wide by 40 mm high in its largest section. The size of this animal alive was not recorded. The notum is a dull red color, with patches of a greenish tinge. The sides are of a more even red color. The foot, rhinophores and branchial plumes all have a greenish tinge. The body is elon- gate, subquadrilateral in cross section, with the largest section at about *4 of the way to the pos- terior end of the foot. The taper- ing posterior of the notum over- hangs the posterior end of the foot by about 20 mm. The broad oral veil is slightly incised at midpoint and extends about 5 mm beyond the front of the body. There are 12 velar papillae on each side of the veil. The rela- tively short velar papillae are arranged in multifid groups, with blunt rounded apices. The grooved oral tentacles usually associated with the tritoniids are present but are not distinctly

FiGuRE |. Marionia bathycarolinensis sp. nov., digital photographs. (a). demarcated, instead being incor- photo of living animal provided by CRRF. (b). Jaws of paratype, 35 mm in porated as the outer margins of _ length.

the veil. The body is finely gran-

ular all over, except for the smooth sole of the foot, with some development of low rounded tuber- cles on the notum and sides. The broad foot is rounded anteriorly. From the narrowly overhanging margin of the notum are produced the rhinophores and branchial plumes. In the holotype, the rhinophoral shafts are completely retracted into widely spaced, 6 mm-long sheaths. The photo- graph of the living animal (Fig. la) shows a typical tritoniid arrangement, with a central clavus sur- rounded by a series of pinnate projections. The branchial plumes are relatively short, the largest being slightly greater in length than the extended rhinophores in the photograph. The anterior 8 or so plumes are discrete and easy to count, whereas the more posterior plumes tend to run together, making an exact count difficult. There are approximately 22 plumes per side, with the largest in the middle third of the body. The branchial plumes are divided into three or four main branches, which then subdivide one or two more times. The gonopore is located on the right side, at about midline below and between the 4" and 5" branchial plume. It has the three-valve appearance of some tri-

SMITH AND GOSLINER: A NEW SPECIES OF MARIONIA (NUDIBRANCHIA) 69

toniids, and the tip of the penial papilla was visible. A floret of distended tissue surrounds the large opening of the anus, which is situated close to the notal margin below the 10 and 11 branchial plume on the right. The nephroproct opens between the 8" and 9‘ plume, just anterior to the anus.

DIGESTIVE SYSTEM.— The jaws are yellowish brown in color, with a darker, thickened red- dish brown masticatory margin slightly overhanging the length of the base (Fig. 1b). The mastica- tory margin appears smooth to the naked eye, but scanning electron microscopy reveals the pres- ence of 25 or more rows of jaw rodlets visible in approximately the distal 5 mm of the half jaw examined (Figs. 2-3). The jaw rodlets exhibit a complex ultrastructure, with multiple vertical ele- ments encased in an outer layer showing horizontal sculpture. The radula of the paratype is large compared with other tritoniids, with a formula of 72 (142.1.1.142) at its broadest point (Figs. 4a- d, 5a—b). The base of the rachidian is roughly rectangular, and is wider than it is long. The upper margin of the base (as oriented in the figures) is incised in the shape of a ‘V’. The central cusp aris- es as a sturdy wedge-shaped process oriented nearly 90° from the plane of the base. The top of the wedge flattens and forms a long, emarginate, blade-shaped triangular denticle. Behind the base of this denticle the top of the wedge forms two projecting ridges that follow the ‘V’ shape of the mar- gin. The rows of teeth are arranged very closely together, so that the projecting wedge of the cen- tral cusp fits neatly into the ‘V’ shaped margin and ridges of the next tooth, with the apex of the flat blade extending over the top. At the base of each wedge-shaped ridge is a small, flattened, emarginate denticle, considered here as accessory. The two outer cusps of the rachidian are short- er than the median, and are formed from the thickened outer margins of the base. Their surface and that of the accessory denticles bears a pattern of folds or ridges resembling the veins on a leaf. The

$36743 15KY

FIGURE 2. SEMs of jaw of paratype. (a). Proximal portion FiGurE 3. SEMs of jaw of paratype. (a). View of rodlets of masticatory border showing eruption of jaw rodlets. (b). near distal end of masticatory border. (b). Detail of rodlets Distal portion of masticatory border, showing part of over- from same area. (c). View of rodlets at distal end of mastica- hang broken off in preparation. (c). Higher magnification tory border. (d). Detail of rodlets seen in Figure 3c. view of proximal masticatory border. (d). Close up of rodlets in proximal masticatory border.

70 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 6

first lateral teeth are differentiated from the remaining laterals, being more heavily constructed with a short, hooked, chisel-shaped denticle fitting closely with the edges of the outer cusps of the rachidian. The remaining laterals are straight to slightly curved, and relatively broad and stout (Figs. 5 ab). A pair of floccose salivary glands extended backwards from the proximal portion of the buccal mass, with attachments to the buccal mass on either side of the esophagus. The pre- served salivary glands exhibit a greenish coloration. The esophagus extends about 8 cm from the median posterior buccal mass as a long straight tube, widening into a distended sac-like portion before turning and entering the stomach on the ventral side (Fig. 5c). A large bolus of gorgonian food material was present in the proximal esophagus, and will be described and discussed in a later section. The stomach 1s relatively small, 2 cm in diameter, with a muscular girdle visible in the proximal portion (Fig. 5c). Inside the stomach are approximately 50 stout, dark brown plates arranged so that the bases are attached to the muscular girdle and the plates point to the lumen of the stomach (Figs. 6a—b). The plates are roughly rectangular, with the exposed corners rounded, and most are about 7 mm long by 3mm high. A pair of larger, thicker and higher plates surrounds the typhlosole that continues through the stomach from the esophagus and into the intestine. There are also some thinner plates and some plates that are only 4 mm long interspersed in no discern- able pattern. The intestine (Fig. 5c) exits the stomach from the anterior left portion, curving up and around the anterior portion of the digestive gland towards the right, becoming wider for a portion before narrowing again to terminate at the anus. The compact digestive gland is divided into two lobes, each with a duct leading into the stomach. The large, spheroid posterior portion has an ante- rior hollow that fits around the posterior stomach, leaving the top exposed. The smaller anterior

FIGURE 5. SEMs from paratype, drawing from holotype. fication view of radula, older portion. (b). Rachidian teeth, (a). Middle latteral teeth from older portion of radula. (b). first and inner lateral teeth at older portion. (c). Detailed Outer laterals from newer portion of radula. (c). Visceral view of rachidian at area shown in Figure 4b. Scale bar= 300 mass (from holotype). an = anus; dga = anterior digestive tm. (d). Rachidian teeth, first and inner lateral teeth at newer gland; dgp = posterior digestive gland: esd = distal esopha- portion. gus; esp = proximal esophagus; in = intestine; st = stomach.

SMITH AND GOSLINER: A NEW SPECIES OF MARIONIA (NUDIBRANCHIA) 71

FiGuRE 6. SEMs from holotype. (a). Portion of band of stomach plates, showing typhlosole area. Scale = 1 mm. (b). Isolated plate. Scale = 400 um. (c). Portion of gut contents, showing overview of characteristic plates and 'woody' stem. Scale = 200 um. (d). Closer view of plates. Scale = 100 Lm.

lobe abuts the right anterior portion of the stomach, and is loosely connected to the posterior diges- tive gland (Fig. 5c).

REPRODUCTIVE SYSTEM.— The reproductive system is triaulic. The diffuse ovotestis covers much of the surface of the digestive glands. A narrow hermaphroditic duct connects to the narrow end of the ampulla (Fig. 7a). The ampulla is muscular and convoluted, its wide end encased with- in the compact female gland mass. The smaller albumen gland is easily discernable from the larg- er membrane and mucos glands. The proximal vas deferens emerges from the female gland mass as a thin, relatively straight tube for about 10 mm, becoming thickened and sinuous in its median portion. The distal portion then extends, thinning slightly before entering the base of the conical unarmed penis, lying near the gonopore. The muscular bursa copulatrix is an inflated oval 18 mm long. The vaginal duct is approximately equal to the bursa in length, widening at the vaginal atri- um. The oviduct exits the female gland mass and opens to the vaginal atrium.

NERVOUS SYSTEM.— The ganglia of the central nervous system sit on the dorsal esophagus, just behind the buccal mass (Fig. 7c). A thin transparent membrane not apparent in dissections of other tritoniids covers the ganglia and large nerves. The paired cerebral and pleural ganglia are dis- tinct but somewhat fused, and the pairs are joined by a short connective. The pedal ganglia are on either side of the cerebropleurals, joined to them by short connectives, and to each other by the cir- cum-esophageal nerve ring, which also contains commisures joining the cerebropleurals. A pair of buccal ganglia is present on the ventral esophagus, just anterior to the nerve ring. These are joined by a short connective (Fig 7c). The central nervous system and buccal ganglia are distinctly asym-

V2 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 6

metrical and ‘lumpy’ in appear- ance, with irregular nodules and a distinct giant nerve cells scat- tered about. The darkly pigment- ed eyes are small, connected to the central nervous system by long nerves.

Gut CONTENTS.— The con- tents of the digestive system con- tained both free sclerites and large pieces consisting of a “woody” appearing axis which is covered by distinctive large plates (Figs. 6c—d; 8a—d; 9a—b). The exposed sides of the plates have a pearly appearance, the bumpy surface giving the impression of sclerites immersed in a smooth nacreous coating. There is a distinct transition between the smooth exposed side and the rougher surface of the attached side. A crack in the membrane of the attached side shows a few sclerite-like objects imbedded in a matrix. Although it is not possible to know with any certainty the origin or exact

vdp

fgm

: : 3 S FiGuRE 7. Drawings from holotype. (a). Reproductive system. Al = albu- identity of the octocoral from min gland; am = ampulla; be = bursa copulatrix; fem = female gland mass; hd which the free sclerites were =hermaphroditic duct; od = oviduct; pn = penis; va = vaginal atrium; vd = dis- derived, the larger pieces are tal portion of vas deferens; vdp = proximal portion of vas deferens. (b). Central nervous system. ceg = cerebral ganglia; peg = pedal ganglia; plg =

consistent with the genus Paracis ae g 2 me Beis pleural ganglia. (c) Buccal ganglia. Scale bars as indicated.

Kiikenthal, 1919, Family Plex- auridae Gray, 1859 and appear to belong to an undetermined species (G.C. Williams, pers. com- mun.). Further references to octocoral information and biology can be found in Smith and Gosliner (2003).

DISTRIBUTION.— Marionia bathycarolinensis sp. nov. is known only from the type locality of Palau, in the Caroline Islands.

ETYMOLOGY.— The first part of the specific name is derived from the Greek, bathys (deep); the latter part refers to the type locality in the western-most fringe of the Caroline Islands.

DISCUSSION

In two recent papers (Avila et al. 1999; Jensen 1994), the anatomical and morphological char- acters of plate-bearing tritoniids have been tabulated. Avila et al. compared characteristics of species of the genus Marioniopsis, and included eight other species along with their newly described species. Jensen compared anatomical characters of species of Marionia and Marioniopsis, listing 17 species plus her newly described species. The combined species discussed

SMITH AND GOSLINER: A NEW SPECIES OF MARIONIA (NUDIBRANCHIA) Ws

FiGureE 8. Detailed view of detached octocoral plate. (a). Close up view of exposed (front) side. Scale bar = 20 um (b). A single detached plate, viewed from the attachment (back) side. Scale bar = 100 ttm. Small arrow indicates area detailed in 8c, large arrow indicates area detailed in 8d. (c). Detailed view of area indicated by smaller arrow in 8b. Scale bar = 100 um. (d). Detailed view of area indicated by larger arrow in 8b. Scale bar = 20 um.

FiGure 9. A Portion of the 'woody' axis of the octocoral. (a). Scale bar = 1 mm. (b). Scale bar = 100 um.

in both papers are summarized in Table 1, along with Paratritonia lutea Baba, 1949, and Marionia bathycarolinensis sp. nov.

There are two external characters that differentiate all other plate-bearing tritoniids from Marionia bathycarolinensis sp. nov.: the number of branchial plumes and the number of velar papillae. Marionia bathycarolinensis sp. noy. has 22 pairs of branchial plumes and 12 pairs of com- pound velar papillae. All other species listed have 16 or fewer branchial plumes, with the excep-

74

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

TABLE |. Comparison of Indo-Pacific species of the genus Marionia

Volume 56, No. 6

Species Branchial Velar papillae Jaws ‘h row of Stomach plumes radula teeth __ plates M. albotuberculata 9 pairs 5 pairs, some com- | row of teeth, trace 95 >100 pound of second M. babai 7 pairs 6 pairs, compound > 100 teeth, number Ds present of rows not stated M. blainvillea 10-12 pairs 7 pairs 2—4 rows of teeth 15-21 ca 40 M. chloanthes 9+ pairs 4 pairs, some com- — 5—6 rows of teeth 22 70 pound M. cucullata 12-16 pairs _ 7-11 pairs 6 rows of teeth 58 30-40 M. cyanobranchiata 9-13 pairs 4-7 pairs, simple __1 row of teeth, trace 15-50 100-120 of second M. dakini 13 pairs 6-7 pairs, compound 1 row of 45 teeth 135 present M. echinomuriceae 10-14 pairs 6-8 pairs 1 row of teeth, 65 28 indistinct M. fulvicola 7-9 pairs 3-4 pairs, most simple 4-5 rows,>100 38-42 22-32 teeth M. granularis 13-14 pairs 6 pairs unknown 50 present M. levis 9-10 pairs 3-5 pairs, compound 1-3 rows, 14-30 80-130 ca 150 teeth M. pambanensis 12 pairs 6 pairs unknown 43 present M. pellucida 13 pairs 6 pairs unknown 22 70 M. platyctena 100 pairs 5-7 pairs, simple 10 rows of teeth 71-103 30-35 M. putulasa 15 pairs 6 pairs unknown 112 25 M. rubra 10-12 pairs 6 pairs, compound | row of 100-120 50-55 present teeth M. tessellata 13 pairs 7 pairs 3 rows of teeth unknown present M. bathycarolinensis 22 pairs 12 pairs, compound 25-100 rows of 142 50 rodlets M. viridescens 10 pairs 7 pairs, some single row of teeth 90 25 compound M. olivacea 9-15 pairs 7 pairs, some 3-7 row of teeth 70-80 50-60 compound Paratritonia lutea 6—7 pairs 34 pairs 3—10 rows of teeth 110 Md)

tion of M. platyctena, which has 100. However, this animal is recorded as having a jaw with 10 rows of teeth, a velum with 5—7 pairs of simple papillae, a radular half-row of up to 103 lateral teeth and a maximum of 35 stomach plates. This is in contrast with M. bathycarolinensis sp. nov., which has a jaw armed with rodlets, a velum with 12 pairs of compound papillae, a radula with a half-row of 143 lateral teeth, and 50 stomach plates. Only M. cucullata (which may be a synonym for M. blainvillea) has a recorded number of velar papillae approaching that of the new species. However, M. cucullata does not match our new species in any of the other characters discussed, differing in the number of branchial plumes, jaw morphology, radular formula and number of stom- ach plates.

Internally, the principal apomorphy of Marionia bathycarolinensis sp. nov. is a jaw armed with

SMITH AND GOSLINER: A NEW SPECIES OF MARIONIA (NUDIBRANCHIA) TS

multiple rows of rodlets. This type of denticulation has not been recorded for any other species of tritoniid nudibranch.

The basic form of the tritoniid radula is, with few exceptions, consistent throughout the gen- era, consisting of a tricuspid central tooth flanked by a pair of first lateral teeth differentiated from the remaining curved outer laterals. The species described here does not differ from the basic form, but is distinctive in the appearance of the central tooth. The median cusp is elongate, overlapping the next row when the radula is contracted. The leaf-like shape of this flattened central cusp has not been recorded for any other species of tritoniid.

The remaining distinguishing characters are more general: large size and a relatively compact visceral mass with an elongate esophagus. Body lengths of over 10 cm are at the large end of the tritoniid size continuum, mostly occurring in temperate species without stomach plates. Both spec- imens described here are of this large size. Figure 5b illustrates the long straight esophagus and rel- atively compact visceral mass, which contrasts with published drawings of other tritoniids (Odhner 1936) and the personal observations of the present authors.

ACKNOWLEDGEMENTS

The authors would like to thank Pat Colin and Lori Bell of the Coral Reef Research Foundation (Palau) for providing the specimens and a photograph of the live animal. Dr. Gary Williams of the California Academy of Sciences provided expertise and advice on octocoral iden- tification, and edited an early draft of the manuscript. An anonymous reviewer made valuable com- ments and suggestions, which led to improvements of the final manuscript. This work was partial- ly supported by NSF PEET grant #0329054.

LITERATURE CITED

AviLa, C., D. KELMAN, Y. KASHMAN, AND Y. BENAYAHU. 1999. An association between a dendronotid nudi- branch (Mollusca, Opisthobranchia) and a soft coral (Octocorallia, Alcyonaria) from the Red Sea. Journal of Natural History 33:1433-1449.

BaBa, K. 1949. Opisthobranchs of Sagami Bay Collected by His Magisty The Emperor of Japan. wanami Shoten, Tokyo, Japan. 194 pp., 50 pls.

JENSEN, K.R. 1994. Sublittoral Notaspidea and Nudibranchia (Opisthobranchia) from Hong Kong, with a description of a new species. Pages 117—139 in B. Morton, ed., The Malacofauna of Hong Kong and Southern China III. Proceedings of the Third International Workshop on the Malacofauna of Hong Kong and Southern China, Hong Kong 13 April-1 May 1992. Hong Kong University Press, Hong Kong, China.

McDONALD, G.R., AND J.W. NYBAKKEN. 1999. A worldwide review of the food of nudibranch mollusks I. The Suborder Dendronotacea. The Veliger 42(1):62—66.

ODHNER, N. 1936. Nudibranchia Dendronotacea — A revision of the system. Mémoires du Musée Royal d Histoire Naturelle de Belgique, series 2, 3:1057—1128, pl. 1.

OpHNER, N.H. 1963. On the taxonomy of the family Tritoniidae (Mollusca: Opisthobranchia). The Veliger 6:48-62.

SMITH, V.G., AND T.M. GosLINer. 2003. A new species of Tritonia from Okinawa (Mollusca: Nudibranchia), and its association with a gorgonian octocoral. Proceedings of the California Academy of Sciences 54(16):155-278.

WILLAN, R.C. 1988. The taxonomy of two host-specific, cryptic dendronotoid nudibranch species (Mollusca: Gastropoda) from Australia including a new species description. Zoological Journal of the Linnean Society 94:39-63.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 7, pp. 76-79, 1 fig. April 29, 2005

Perulibatrachus aquilonarius, a New Toadfish Species from India (Teleostei: Batrachoididae)

David W. Greenfield! California Academy of Sciences, 875 Howard Street, San Francisco, California 94103; Email: greenfie @ hawaii.edu

The fourth known species in the genus Perulibatrachus is described. It is the north- ernmost species in the genus, and the second known from the Indian Ocean. It has a wider head than the other described species and differs from others in counts: dor- sal-fin elements II]-17, anal-fin rays 13, pectoral-fin rays 21, vertebrae 26. It has a shallow, funnel-shaped pocket in the pectoral-fin axil.

The genus Perulibatrachus is currently represented by three species. Two species are known only from the west coast of Africa: P. elminensis (Bleeker, 1863) from Ghana south to Walvis Bay, Nambia, and P. rossignoli (Roux, 1957) from Gabon south to Walvis Bay (Roux 1981; Hutchins 1986). A third species, P. kilburni Greenfield, 1996, is known only from Natal, southeastern South Africa. In 1941, A.W.C.T. Herre collected a toadfish specimen, purportedly in India, which he or someone else had identified as Chatrabus damaranus. The specimen had been deposited in the ichthyological collections of Stanford University and later transferred to the California Academy of Sciences along with the rest of Stanford’s collections. Recently, I found the specimen on the shelves at the Academy and on examination discovered that it does not belong to the genus Chatrabus but to the genus Perulibatrachus. The genus Chatrabus was described by Smith in 1949, whereas Perulibatrachus was not described by Roux and Whitley until 1972. Thus, at the time of its identification, Chatrabus would have been a logical choice in which to place this spec- imen.

The most recent key to toadfish genera is that of Smith (1952). In that key, Perulibatrachus would fall in section BIH, which includes Chatrabus, Barchatus, Tharbacus, Batrachoides, and Halobatrachus. Hutchins (1986) considers Tharbacus to be a synonym of Chatrabus. Roux (1981) presents a key to separate Perulibatrachus from Halobatrachus, which has a distinct foramen (axil- lary pore) on the upper part of the pectoral-fin axil, whereas the other genera have only a funnel- shaped pocket (Perulibatrachus) or neither a foramen nor pocket (Chatrabus and Batrachoides). In species without the pocket, the skin stretches straight across the pectoral-fin axil, whereas in Perulibatrachus there is a clear indentation, which is shallow in P. rossignoli or relatively deep in both P. elminensis and P. kilburni. The specimen from India has three dorsal-fin spines and two subopercular spines, placing it in the subfamily Batrachoidinae. The body is mostly scaled, it lacks a foramen in the pectoral-fin axil, and it has a shallow funnel-shaped pocket in the pectoral-fin axil, placing it in the genus Perulibatrachus.

| Research Associate, Department of Ichthyology, California Academy of Sciences and Emeritus Professor, University of Hawaii. Mailing address: Moss Landing Marine Laboratory, 8272 Moss Landing Road, Moss Landing, CA 95039.

76

GREENFIELD: NEW TOADFISH OF THE GENUS PERULIBRACHUS FROM INDIA Teh

MATERIALS AND METHODS

All counts and measurements follow Hubbs and Lagler (1964) except that the last two fin rays are not counted as one unless it is clear that they are joined at the base. Measurements were made to the nearest 0.1 mm using dial calipers. All measurements are expressed as percentage of stan- dard length (SL). Counts were made from a radiograph. Neither the holotype nor any additional material of P. rossignoli could be located at the Museum National d’ Histoire Naturelle, Paris, by M.L. Bauchout and J.E. Randall, so comparisons with that species are based on the literature. Institutional abbreviations are as listed in Leviton et al. (1985).

SPECIES DESCRIPTION

Perulibatrachus aquilonarius Greenfield, sp. nov. (Figs. 1A—B)

MATERIAL EXAMINED.— HOLotyPE: CAS-SU 41322, 191.4 mm SL, India, Tamil Nadu State, Ennur Fisheries Station, Madras, January, 1941, A.W.T.C. Herre. ADDITIONAL MATERIAL EXAMINED: Perulibatrachus kilburni, RUSI 28203 (1; 56.6 mm SL). Perulibatrachus elminensis, MNHN 1967-909 (1; 204 mm SL), MNHN 1970-43 (1; 150 mm SL; cleared and stained).

DIAGNOsISs.— A species of Peruli- batrachus with a wide head (43.6% SL), 17 dorsal-fin rays, 13 anal-fin rays, 21 pectoral-fin rays, 26 vertebrae, and a shal- low, funnel-shaped pocket in the pectoral- fin axil.

DESCRIPTION.— Dorsal-fin elements I-17; anal-fin rays 13; pectoral-fin rays 21; vertebrae 26; head length 45.0; head width 43.6; head depth 25.3; bony interor- bital width 9.5; orbit diameter 7.6; snout length 8.8; upper jaw length 25.9; mouth width at rictus 36.1; first predorsal-fin dis- tance 42.1; second predorsal-fin distance 61.6; preanal-fin distance 72.5; greatest body depth 25.2; caudal-peduncle depth 9.8; caudal-peduncle length 13.6; first dorsal-fin base length 11.4; second dorsal- fin base length 39.6; anal-fin base length ~ SO Par NS X 21.5; caudal-fin length 22.7, pectoral-fin a FICHE 1. Holotype of Perulibatrachus Rpailennite CAS-SU length DSO: pelvic-fin length 26.7; dis- 41322. A. lateral view; B. dorsal view. tance between pelvic-fin bases 15.3.

Head moderately depressed and wide, eyes medium in size and not raised above head profile. Pectoral-fin axil with a shallow, fun- nel-shaped pouch in upper half, glandular tissue present in rest of axil and on sides under pectoral fins. Some glandular tissue present on inner surface of pectoral fins. Body with small scales extend- ing from middle of first dorsal fin back to caudal-fin base. Scales extending forward on ventral sur- face to pelvic fins. Two lateral lines: the upper one originating above upper opercular spine, run- ning posteriorly in a straight line to below second dorsal-fin origin where it curves up to run along

78 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 7

dorsal-fin base to its end, with about 41 pores present, each with two short skin flaps; the second lateral line, with about 17 pores, runs along anal-fin base. Vomer and palatine each with single row of slightly curved conical teeth. Dentary laterally with single row of conical teeth, double row near symphysis. Premaxilla with single row of conical teeth posteriorly, double row anteriorly. Subopercle with two spines. Anterior nostrils tubular; nostril on left side with a simple major ten- tacle; nostril on right side with three terminal tentacles; a cluster of multifid cirri surrounds each tubular nostril. Cirrus at distal end of maxilla with many tips. Ventral margin of dentary with numerous cirri. Dorsal surface of neurocranium completely covered with muscles, no exposed bone.

Color in alcohol: After 64 years in preservative, general background color is cream, overlaid with rusty brown pigment pattern. Anterior portion of head from posterior margin of eyes forward rusty brown. A cream band running from posterior margins of eyes posteriorly to opercular spine forms a distinct band across the head. The area posterior to this and extending back to first dorsal- fin origin overlaid with distinct, small, rust-colored spots. Sides of body, second dorsal fin, anal fin, pectoral, and caudal fins with scattered, irregular rust-colored spots. Ventral surface of head cream, area from pelvic fins posteriorly to anal-fin origin rust.

ETYMOLOGY.— The specific epithet is an adjective from the Latin, aquilonarius, meaning northern, referring the fact that this species has the northernmost distribution of any member of the genus Perulibatrachus.

COMPARISONS.— Perulibatrachus aquilonarius differs from all other species in the genus by having a wider head: 43.6 versus 29.7 in P. kilburni, 29.1—37.9 in P elminensis, and 37.8-40.9 in P. rossignoli. \t differs from P. rossignoli by having 17 versus 19 second dorsal-fin rays, 21 versus 23 pectoral-fin rays, and 26 versus 29 vertebrae. It differs from P. elminensis by having 13 versus 14-17 anal-fin rays, 21 versus 26 pectoral-fin rays, and 26 versus 27—28 vertebrae. It differs from P. kilburni by having 17 versus 18 dorsal-fin rays, 13 versus 14 anal-fin rays, and 21 versus 19 pec- toral-fin rays. It has a shallow, funnel-shaped axillary pocket, whereas it is deep in P. elminensis and P. kilburni.

ACKNOWLEDGMENTS

C. Roux and M.L. Bauchot were very helpful in locating and providing specimens of P. elmi- nensis. M.L. Bauchot and J.E. Randall also searched for specimens of P. rossignoli. | would also like to thank Guy Duhamel and Patrice Pruvost for their hospitality while visiting the MNHN. Special thanks to the staff at CAS for lending material and providing assistance: D. Catania, W.N. Eschmeyer, J. Fong, M. Hoang, and T. Iwamoto.

LITERATURE CITED

BLEEKER, P. 1863. Mémoire sur les poissons de la céte de Guinée. Natuurkundige Verhandelingen van de Hollandsche Maatschappij der Wetenschappen te Haarlem, ser. 2, 18:1—136, 28 pls.

GREENFIELD, D.W. 1996. Perulibatrachus kilburni, a new toadfish from East Africa. Copeia 1996:901—904.

Huss, C.L., AND K.F. LAGLER. 1964. Fishes of the Great Lakes Region. University of Michigan Press, Ann Arbor, Michigan, USA. 213 pp.

HuTcuins, J.B. 1986. Family No. 100: Batrachoididae. Pages 358-361 in M.M. Smith and P.C. Heemstra, eds., Smith’s Sea Fishes. Macmillan South Africa, Johannesburg, South Africa. 1047 pp.

Leviton, A.E., R.H. Gipss, JR., E. HEAL, AND C.E. DAwson. 1985. Standards in herpetology and ichthyology. Part I. Standard symbolic codes for institutional resource collections in herpetology and ichthyology. Copeia 1985:802-832.

GREENFIELD: NEW TOADFISH OF THE GENUS PERULIBRACHUS FROM INDIA 72

Roux, C. 1957. Poissons marins. Pages 139-253 in J. Collignon, M. Rossignol, and C. Roux. Mollusques, Crustacés, Poissons marins des cétes d’A.E.F. en collection au Centre d’Océanographie de I’ Institut d’E- tudes Centrafricaines de Point-Noir. Office de la Recherche Scientifique et Technique Outre-Mer, Paris, France. 369 pp.

Roux, C. 1981. Batrchoididae. Vol. 1. Jn W. Fisher, G. Bianchi, and W.B. Scott, eds., FAO species identifica- tion sheets for fishery purposes. Eastern Central Atlantic; fishing areas 34, 47 (in part). Vols. 1-7. Canada Funds-in-Trust. Ottawa, Department of Fisheries and Oceans Canada, by arrangement with the Food and Agriculture Organization of the United Nations, Ottawa, Ontario, Canada. (No page numbers.)

Roux, C., AND G. WHITLEY. 1972. Perulibatrachus, nouveau nom de genre de Poissons Téléostéens de la famille des Batrachoididae, en remplacement de Parabatrachus Roux, 1970. Bulletin du Muséum National d Histoire Naturelle Zoologie 6(1971):349-350.

SmiTH, J.L.B. 1949. The Sea Fishes of Southern Africa. Central News Agency, Ltd., Cape Town, South Africa. 550 pp.

SmitH, J.L.B. 1952. The fishes of the family Batrachoididae from South and East Africa. Annals and Magazine of Natural History, ser. 12, 5(52):313-339.

Copyright © 2005 by the California Academy of Sciences San Francisco, California, U.S.A.

PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES

Volume 56, No. 8, pp. 80-85, 3 figs. April 29, 2005

Brotula flaviviridis, a New Species of Brotula from Fiji (Teleostei: Ophidiidae: Brotulinae)

David W. Greenfield! California Academy of Sciences, 875 Howard Street, San Francisco, Callifornia 94103

The third known species of Brotula from the Indo-Pacific, B. flaviviridis sp. nov., is described from Fiji. Brotula townsendi is an antiequitorial species, known from the Hawaiian Islands, Johnston Island, and the Marshall Islands in the northern hemi- sphere, and Chesterfield Islands, New Caledonia, Loyalty Islands, Vanuatu, and Tonga in the southern hemisphere. Brotula multibarbata is widespread in the Indo- Pacific. Brotula flaviviridis differs from both in coloration, being yellow-green. It is similar to B. townsendi in having an eye that is narrower than the fleshy interorbital width, whereas B. multibarbata has an eye that is wider. It differs from B. townsendi by having a smaller eye and narrower interorbital width.

While conducting a survey of the marine fishes of Fiji, two specimens of a yellow-green bro- tula were collected using rotenone at the barrier reef on the north side of Kanacea Island in the Northern Lau Group, Fiji. The fresh coloration of these specimens was distinctly different from the other two species in the genus Brotula known from the Indo-Pacific.

The new species is placed in the genus Brotula because: the anterior nostril is well above the upper lip, near the eye; the supramaxilla is present; the dorsal-fin rays are equal to or longer than opposing anal-fin rays; and six barbels are present on the snout and six on the chin.

Only five species in the genus Brotula are currently considered to be valid. Brotula barbata (Bloch in Bloch and Schneider, 1801) occurs in the tropical and subtropical Atlantic Ocean. Brotula clarkae Hubbs (1944) occurs in the tropical eastern Pacific from the Gulf of California to Peru, and B. ordwayi Hildebrand and Barton (1949) occurs at Peru and the Galapagos Islands. In the Indo- Pacific Ocean, B. townsendi Fowler (1900) is known from the Hawaiian Islands, Johnston Island, Marshall Islands, Chesterfield Islands, New Caledonia, Loyalty Islands, Vanuatu, and Tonga, and B. multibarbata Temminck and Schlegel (1846) is widespread. Hubbs (1944) listed ten other described species as synonyms of B. multibarbata, and these were also listed by Nielsen, et al. (1999).

MATERIALS AND METHODS

Information on the holotype is presented first, followed by that of the paratype in parentheses. The paratype died with its mouth open wide, making measurements less accurate than those for the holotype. All measurements are presented as percentage of standard length (SL), and some as per- centages of other body parts. Information on Brotula clarkae and B. ordwayi is from Allen and

! Research Associate, Department of Ichthyology, California Academy of Sciences and Emeritus Professor, University of Hawaii. Email: greenfie@hawaii.edu. Mailing address: Moss Landing Marine Laboratory, 8272 Moss Landing Road, Moss Landing, CA 95039.

GREENFIELD: A NEW BROTULA FROM FIJI 81

Robertson (1994) and Hubbs (1944). Pectoral-fin ray counts were obtained by slitting the skin on the inside of the right pectoral fin. Other fin-ray counts and vertebral counts were obtained from radiographs. Methods of counting and measuring follow Nielsen et al. (1999), and the format of the description follows Cohen and Nielsen (1982).

SPECIES DESCRIPTION

Brotula flaviviridis Greenfield, sp. nov. (Figs. 1-3)

MATERIAL EXAMINED.— HOLOTYPE: CAS 221531, 125.0 SL, Fiji, Northern Lau Group, north side of Kanacea Island at copra planta- tion, 17°14.890’S, 179°08.475’W, isolated piece of barrier reef, sand and dead coral, 7—9.5 ~ m, 5 January 2003, field number G03-15, FIGURE 1. Drawing of holotype (CAS 221531) of Brotula rotenone, collected by D.W. Greenfield, K.R. “avvirdis.

Longenecker, and R.C. Langston. PARATYPE: CAS 221532, 151.0 SL, Fiji, Northern Lau Group, north side of Kanacea Island at copra plantation, 17°14.890’S, 179°08,475’W, over- hang on barrier reef with sand at base, 12-14 bi : = m, 4 January 2003, field number GO3-I1, FIGURE 2. Fresh color of holotype (CAS 2 rotenone, collected by D.W. Greenfield, K-R. — Brotula flaviviridis. Photo D.W. Greenfield. Longenecker, R.C. Langston, and Bio Koroi

Mataitini. ADDITIONAL MATERIAL EXAMINED: Brotula townsendi: Hawaiian Islands, Oahu-CAS 221533. Johnston Island-BPBM 33975 (2), BPBM 29609 (1), BPBM 34043 (1). Chesterfield Islands-BPBM 33794 (5). New Caledonia-BPBM 22550 (1). Brotula multibarbata: Hawaiian Islands, Oahu-CAS 78987 (5), CAS 78901 (2), CAS 210003 (4). Fiji-CAS 218868 (4), CAS 219394 (5), CAS 219404 (5), CAS 219405 (4).

DiAGNosis.— A species of Brotula that is yellow-green in coloration, with an eye that is nar- rower than the interorbital width, 98-103 dorsal-fin rays, 77—79 anal-fin rays, 24 pectoral-fin rays, 43-44 vertebrae, 2—3 developed gill rakers, a fleshy interorbital width of less than 4.5 percent SL, and an orbit diameter 3.5 percent SL or less.

DESCRIPTION.— Dorsal-fin rays 97 (103). Anal-fin rays 77 (79). Pectoral-fin rays 24 (24). Vertebrae 12+32 =44 (11+32 =43). Total developed gill rakers 3 (2 one side, 3 other). Longitudinal scale series about 133. Predorsal length 26.6 (25.8). Preanal length 48.8 (48.3). Preventral length 14.2 (21.0). Body depth at vent 19.2 (21.5). Head length 22.3 (23.5). Eye diam- eter 3.2 (3.5). Snout length 4.6 (6.0). Fleshy-interorbital width 4.2 (3.8). Upper-jaw length 12.0 (13.8). Greatest maxillary width 9.7 (12.4). Pectoral-fin length 14.2 (13.8). Pectoral-fin peduncle width 6.6 (6.3). Ventral-fin length 20.3 (17.8).

Body deepest just before end of pectoral fin, tapering to tail, which is not greatly attenuate. Body scales relatively large for the genus, about 133 rows along the side of the body. Head cov- ered with scales. Scales on belly forward to ventral fins. Scales present on pectoral-fin base and also out onto the dorsal and anal fins. Head compressed. Snout bluntly rounded, longer than orbit diameter. Upper lip terminal, extending slightly beyond lower jaw. Posterior nostril with raised rim, adjacent to anterior margin of eye. Anterior nostril immediately anterior to posterior nostril, with a raised rim and barbel on posterior margin. Two barbels at front of snout, one on each side of mid-

»

21531) of

82 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 8

dle of snout. Another barbel on each side of snout, about at level of ventral margin of eye, for a total of six barbels on snout. Maxilla notably expanded posteriorly, reaching about one pupil diam- eter past rear margin of orbit. A sheath of tissue on upper third of maxilla, with a notch at end exposing top end of maxilla. Underside of lower jaw with six barbels, three on each side. Two pores at symphasis and one adjacent to posterior side of anteriormost barbel. Teeth small and granular, some with fine points, on dentary, premaxilla, palatine, and vomer. Vomerine tooth patch rounded anteriorly and V-shaped.

First gill arch with four (five) short, stubby rakers on upper arm, a longer (developed) raker at angle, lower arm with two (one) longer and ten (14) short protuberances. Longest raker about one- half eye diameter. Longest gill filament a little longer. Branchial cavity and palate pale.

Lateral line not obvious. Dorsal fin originating over end of pectoral-fin peduncle, at origin of pectoral-fin rays. Pectoral fin on a short, rounded, fleshy peduncle. Opercle with a strong spine hid- den under skin which continues as a prominent flap of skin dorsal to pectoral-fin base. Ventral fins inserted well behind symphysis of cleithra, about under opercular spine. Lining of peritoneal cav- ity pale, with scattered, small, black spots. Stomach and intestine pale.

Color in alcohol: Background color cream, overlaid with gray-brown pigment on center of scales, leaving the scales outlined. Nape, snout, and front of jaws with greater concentration of gray-brown. Barbels and posterior ends of maxilla and premaxilla white. Ventral surface of head gray. Pectoral fins clear. Pelvic fins white. Bases of dorsal and anal fins gray where scales extend out onto them. Distal margin of anterior two-thirds of dorsal fin clear. Posterior third gray with black margin. Anal fin clear distally on anterior two-thirds of fin, posterior third like posterior part of dorsal fin. Caudal fin edged in black.

Color of fresh specimen: Background color yellow with a greenish tinge. Head yellow-green. Pupil of eye black, surrounded by yellow iris. Barbels on snout orange, barbels under chin white. Snout and tip of upper and lower jaws dusky. Anterior two-thirds of body primarily yellow-green, posterior third more dusky green. Belly yellow-white. Pectoral fin yellow. Anterior half of dorsal fin yellow with orange margin. Posterior half of dorsal fin reddish grading into black posteriorly as it merges with black caudal fin. Anal fin similar to posterior half of dorsal fin. Pelvic fins white.

ETYMOLOGY.— The specific epithet is a compound adjective, combining the Latin flavus, meaning golden-yellow, plus viridis, meaning green, alluding to the yellow-green color of the species.

CoOMPARISONS.— Brotula flaviviridis differs from all known valid Pacific Ocean species by its yellow-green coloration. It differs from the two eastern Pacific species, B. clarkae and B. ordwayi, by having fewer dorsal-fin rays, 97-103 versus 108-112 (B. clarkae) and 118-125 (B. ordwayi), and fewer anal-fin rays, 77-79 versus 78-89 (B. clarkae) and 86—94 (B. ordwayi). It also differs from B. clarkae in having 24 pectoral-fin rays verses 27—28. As pointed out by Hubbs (1944), “the distinction between the relatively large scales of the Indo-Pacific species and the small scales of the American ones, particularly of B. clarkae, was obvious at sight.” Brotula flaviviridis differs from the Indo-Pacific species B. multibarbata by having an eye that is narrower than the interor- bital width, verses one that is wider. It differs from the other Indo-Pacific species, B. townsendi, by having a smaller eye (3.5 % S.L. or less) and narrower interorbital width (3.8% S.L or less) ( Fig. 3).

DiscussIONn.— Because one of the eleven species listed by Hubbs (1944) as a synonym of B. multibartata, B. townsendi, is valid, it was necessary to investigate the status of the other ten species because in many cases he worked only from the literature. Patrice Pruvost (MNHN) pro- vided me with a photograph of the dried holotype (A.8468) of Brotula burbonensis Kaup (1858). It has a large eye that is typical of B. multibarbata, not the smaller eye of B. flaviviridis. The types

GREENFIELD: A NEW BROTULA FROM FII 83

of three species of Hubbs’ synonyms are at the = ad 6 . British Museum of Natural History: B. ensi- = (ae K | formis Giinther (1862) (stuffed syntypes) from eS = | Vanuatu, B. jayakari Giinther (1909) (BMNH S Wu =} ge g @ | 1888.12.29.193) from Oman, and three syn- = 4.5 ¢ | types of B. miilleri (corrected to B. muelleri) Whos i | Giinther (1909) (BMNH 1881.10.28.8 from 2 ape 4 | Ponape. 1868.8.1.7° from Tahiti, and ; ae f ie g 1876.5.19.45 from Tahiti). In his description of

B. ensiformis, Giinther (1862) states “The ORBIT DIAMETER width of the interorbital space is less than the FiGurE 3. Interorbital diameter versus orbit diameter,

horizontal diameter of the orbit,” which clearly both as percentage SL for Brotula flaviviridis (squares), and places itn B. multirate, Oe Cr (BMNH) has kindly examined the types of joands (circles). a ;

these three species for me, and confirmed that

B. ensiformis, and B. jayakari have the large eye typical of B. multibarbata. Brotula muelleri is rep- resented by three syntypes and is a mixed type series, containing both large and small eyed species. In the original description, Giinther stated that the width of the interorbital space is much smaller than the diameter of the eyes, which clearly places this species as a synonym of B. multibarbata as concluded by Hubbs (1944). To avoid further confusion, | am designating one of the syntypes, BM(NH)1876.5.19.45 from Tahiti, which has the large eye, as the lectotype of B. muelleri, and as such, it is a synonym of B. multibarbata. The other two syntypes, now paralectotypes of B. muel- leri, have the small eye, typical of B. townsendi and B. flaviviridis.

Brotula palmietensis Smith (1935) (RUSI 299) from South Africa was listed as a synonym of B. multibarbata by Nielsen and Cohen (1986). Brotula formosae Jordan and Evermann (1902) (ZUMT 359) from Taiwan was listed as a synonym of B. multibarbata by Chen and Shao (1991). Brotula japonica Steindachner and Doderlein (1887) (NMW) from Japan was listed as a synonym of B. multibarbata by Lindberg and Krasyukova (1975) and also was not listed as a valid species in Nakabo (2002).

Two of the species were described from the Hawaiian Islands, B. marginalis Jenkins (1901) and B. multicirrata Vaillant and Sauvage (1875). Hubbs (1944) examined the type of B. marginalis at the USNM (49694) and found it to be B. multibarbata with a damaged tail. Both Fowler (1900) and Hubbs (1944) considered B. multicirrata to be a synonym of B. multibarbata. It was described as having eight rather than six barbels on the snout. Many individuals of B. multibarbata have been collected at the Hawaiian Islands and none have had eight barbels. Hubbs (1944) speculated that the nasal flap had been counted as a barbel. Considering that only two species of Brotula are known from the well-collected Hawaiian Islands, and that B. townsendi is relatively rare, treating B. mul- ticirrata as a synonym of B. multibarbata is justified.

Finally, B. ferruginosus (Tickell in Day, 1888) is not available (Eschmeyer 1998).

Brotula flaviviridis appears to be most similar to B. townsendi, which was previously known only from the Hawaiian Islands and Johnston Island but now is known to be antiequatorial. Randall (in press) lists its distribution as “Known from the Hawaiian Islands, Johnston Island, Marshall Islands, Tonga, Loyalty Islands, and Vanuatu.” It is also known from the Chesterfield Islands and New Caledonia (this paper). Brotula flaviviridis shares the relatively small eyes with B. townsen- di, and has similar counts. Brotula townsendi is an orange brown with a yellowish pectoral fin in fresh coloration. Two lots at the Bishop Museum from the South Pacific were identified as B. townsendi: BPBM 22550 from New Caledonia and BPBM 33794 from Chesterfied Islands in the

84 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Volume 56, No. 8

Coral Sea. Examination of these specimens confirmed that they appear to be conspecific with B. townsendi, clustering together in Figure 3. Brotula townsendi thus joins a number of other species from the Hawaiian Islands that demonstrate an antitropical or antiequatorial distribution (Randall, 1982). Brotula flaviviridis most likely was derived from these populations of B. townsendi to the southwest.

ACKNOWLEDGMENTS

I would like to thank K.R. Longenecker, R.C. Langston, and Bio Koroi Mataitini for assistance in collecting specimens. | also thank Captain B. Vasconcellos and the crew of the Moku Mokua Hine for assistance in our work. I am grateful to J. Seeto, GR. South, R.R. Thaman , and R.W. Tuxton of the University of the South Pacific, Fiji for facilitating our collections in Fiji. I also thank the Fijian Government and local village chiefs for permission to collect fishes. I thank Susan Monden for drawing the figure. Patrice Pruvost (MNHN) kindly photographed the holotype of B. burbonensis for me. Special thanks goes to Oliver Crimmen (BMNH) who examined the types of three species for me . W.N. Eschmeyer provided valuable guidance in nomenclatural matters. I also thank the staff at CAS for lending material and providing assistance: D. Catania, W.N. Eschmeyer, J. Fong, M. Hoang, and T. Iwamoto. This research was supported by National Science Foundation grants INT97-29666 and DEBO-1027545, and Sea Grant Project R/FM-6PD.

LITERATURE CITED

ALLEN, G.R. AND D.R. ROBERTSON. 1994. Fishes of the tropical Eastern Pacific. University of Hawaii Press, Honolulu, Hawaii, USA. 332 pp.

BiocH, M.E., AND J.G SCHNEIDER. 1801. Systema ichthyologiae iconobus ex illustratum. Sanderiano Commissum, Berlin, Germany. 584 pp.

CHEN, L.J., AND K.T. SHAO. 1991. Fishes of the families Ophidiidae and Bythitidae from Taiwan. Bulletin of the Institute of Zoology, Academia Sinica (Taipei) 30(1):9-18.

COHEN, D.M., AND J.G. NIELSEN. 1982. Spottobrotula amaculata, a new ophidiid fish from the Philippines. Copeia 1982(3):479-S00.

ESCHMEYER, W.N., ed. 1998. Catalog of Fishes. Vol. I. California Academy of Sciences, San Francisco, California, USA. 958 pp.

Fow Ler, H.W. 1900. Contributions to the ichthyology of the tropical Pacific. Proceedings of the Academy of Natural Sciences of Philadelphia (1900):493—528.

GUNTHER, A. 1862. Catalogue of the fishes in the British Museum, 4. The Trustees of the British Museum, London, England, UK. 534 pp.

GUNTHER, A. 1909. Andrew Garrett’s Fische der Siidsee Journal des Museum Godeffroy 6:261-388.

HILDEBRAND, S.F., AND O. BARTON. 1949. A collection of fishes from Talara, Peru. Smithsonian Miscellaneous Collections 111(10):1—36.

Huss, C.L. 1944. Species of the circumtropical fish genus Brotula. Copeia 1944(3):162-178.

JENKINS, O.P. 1901. Descriptions of fifteen new species of fishes from the Hawaiian Islands. Bulletin of the United States Fish Commission 19(1899):387-404.

Jordan, D.S., and B.W. Evermann. 1902. Notes on a collection of fishes from the island of Formosa. Proceedings of the United States National Museum