BULLETIN OF MARINE SCIENCE, 29(4): 530-553, 1979 A FURTHER DESCRIPTION OF GURGES/ELLA FURVESCENS WITH COMMENTS ON THE INTERRELATIONSHIPS OF GURGESIELLIDAE AND PSEUDORAJIDAE (PISCES, RAJOIDEI) John D. McEachran and Leonard J. V. Compagno ABSTRACT Additional specimens of Gurgesiella furveseens are used to supplement the original de- scription which was based solely on the holotype. The clasper, neurocranium, pectoral girdle, and pelvic girdle of this species are described and compared with those of the only known congener, G. at/antiea, and with Pseudorajajiseheri. Comparisons support Hulley's (1972b) removal of G. atlantica from Pseudoruja to Gurgesiella. Gurgesiella and Pseudo/'aja share many character states which are considered to be derived within Rajoidei, negating the hypothesis that their resemblances are due to synplesiomorphies. Gurges;eJ/a resembles Allacanlhobalis and Cruriraja in clasper morphology and AllaCtlllthobatis, Cruriruja, Raja (Rioraja) and R. (At/an/oruja) in the structure of its scapulocoracoid. The scapulocoracoid of Pseudvraja resembles those of some Psammobalis species. Both genera possess reduced rostra, that of Gurgesiella was probably derived from an ancestor with a stout or partially reduced rostrum, while the rostrum of Pseudoraja is too reduced 10 determine if it was derived from the Gurgesiella type or from a more slender Iype. However, Gurgesiella and Pseudvruja share five derived characlers, which according to our present knowledge, are unique within Rajoidei. Thus, Gurgesiella and Pseudoraja appear to be a monophyletic group and their resemblances to other taxa can be explained by secondary relationships, parallelisms and retension of primilive character states. Similarilies in shared derived char- acter states implies that separate families for Gurgesie//a and Pseudoraja arc unwarranted and Gurgesiellidae is merged with Pseudorajidae. Pseudorajidae, Pseudoraja and Gurgesie//a are redefined. The scapulocoracoid, not hitherto used in phylogenetic studies of Rajoidei, is introduced and appears to be an important taxonomic and phylogenetic character. Despite a number of regional revisional studies of the family Rajidae within the last 20 years (Ishiyama, 1958; Stehmann, 1970, 1976; Hulley, 1972a) and several revisions of related families (Hulley, ]972b, ]973) the phylogenetic interrelation- ships within the suborder Rajoidei remain unclear. Hulley (1972a, 1972b) recog- nized six families, Rajidae Bonaparte, 1831, Anacanthobatidae von Bonde and Swart, 1924, Arhynchobatidae Fowler, 1941, Pseudorajidae Bigelow and Schroe- der, 1954, Gurgesiellidae de Buen, 1959, and Crurirajidae Hulley, 1972; while Compagno (1973) recognized four families, placing Gurgesiellidae in Pseudoraji- dae and Crurirajidae in Rajidae. Discordance in the higher classification of Ra- joidei is due to the erection of more recent families before determining the range of variability within the oldest and largest family, Rajidae (containing about 80% of the 185 to 199 species in the suborder), The five more recent families have been defined by a few unique, mostly external characters and other characters which are found within Rajidae. The unique characters include presence of or number of dorsal fins, structure of the caudal fin, shape and structure of the pelvic fin, presence of an oronasal pit, structure of the rostral cartilage, and structure of the pelvic girdle. These characters are important in distinguishing lower taxa (species, subgenera and possibly genera) but may not be as important as internal structures, i.e. cranial, pectoral girdle or clasper structure, in distinguishing the higher taxa (families). The status of the families and their interrelationships will not be clear until the variability within Rajidae has been determined on a world- 530 McEACHRAN AND COMPAGNO: TAXONOMIC STUDIES OF RAJOIDEI 531 wide basis. Prior to this monumental undertaking the anatomical characters which have proved valuable in determining the interrelationships within Rajidae must be described for the remaining five families. Recently we obtained several specimens of Gurgesiella furvescens, G. atlan- tica (Gurgesiellidae) and Pseudoraja fischeri (Pseudorajidae), including mature males of the two former species, and herein describe the anatomical characters of these species which have been important in elucidating rajid interrelationships. De Buen (1959) erected the family Gurgesiellidae for his new species, Gurge- siella furvescens, which he described from a single specimen captured off Val- paraiso, Chile. He distinguished Gurgesiellidae from Pseudorajidae Bigelow and Schroeder 1954, the latter of which he considered related to Myliobatoidei, largely on the structure of the pelvic fin. Bigelow and Schroeder (1962) described another pseudorajid, Pseudoraja atlantica, from the Caribbean Sea and stated that the characters on which Gurgesiellidae was based appeared to fall within Pseudora- jidae. Hulley (1972b) examined radiographs of the pelvic girdle and cranium of Pseudoraja fischeri, P. atlantica and Gurgesiella furvescens, concurred with Bigelow and Schroeder (1954, 1962) that Pseudorajidae was a member of Rajoidei but concluded that Gurgesiellidae should be maintained as a separate family and that P. atlantica should be placed in Gurgesiella. MATERIALS AND METHODS Specimens examined in this study were borrowed from the Stanford University Fish Collection (SU) housed at the California Academy of Sciences (CAS) San Francisco; Museum of Comparative Zoology (MCZ), Cambridge, Massachusetts: National Marine Fisheries Service Systematic Labora- tory (NMFS SL) Washington, D.C.; Smithsonian Oceanographic Sorting Center (SOSe) Washington, D.C. and National Museum of Natural History (USNM) Washington, D.C. A list of material examined follows the text. Methods of Bigelow and Schroeder (1953) were followed in making external measurements. Ver- tebrae were counted from radiographs according to methods described by Krefft (1968). All anatomical descliptions were based on dissected specimens but dissections were compared with radiographs because only one specimen per species was completely dissected. Removing the cranium, pectoral and pelvic girdles resulted in considerable damage to specimens and there are few available specimens of G. furl'escens and P. fischer;. Clasper terminology follows Stehmann (1970) and Hulley (1972a). Terminology of the cranium is modified from Hulley (1972a). Several of the terms of Hulley (l972a) for cranial foramina arc changed to be consistent with the usage of earlier authors, e.g. Daniel (1934). All cranial measurements except the following are from Hubbs and Ishiyama (1968). Nasobasal length: distance from anteromedial corner of nasal capsule, at side of rostrum, to posteromedial face of occipital condyle (Fig. la). Width across otic capsules: transverse axis of cranium, across lateral walls of otic capsule above hyomandibular facets and opisthotic Iidges and below sphenopterotic ridges. Least width of basal plate: wide of basal plate at its ventrolateral junction with orbital walls. Greatest width of nasal aperture: greatest transverse or diagonal dimension across nasal aperture (ventral aperture of nasal capsule). Internasal width: least distance across interspace between nasal apertures. Hulley (1972a) suggested that the cranial measurements and proportions proposed by Hubbs and Ishiyama (1968) were of little use. However, their utility is limited by Hubbs and Ishiyama's use of cranial length (anterior tip of rostral cartilage to rear end of cranium) as the independent variable in computing cranial proportions. Cranial length includes the highly variable rostral length, which ob- scures similarities and differences in post rostral proportions. Following usage in sharks (Compagno, ms.) we use nasobasal length as an independent variable in computing proportions, but define it slightly differently in batoids due to differences in cranial structure (in sharks the nasobasallength is measured to the rear of the occipital centrum). The measurements of the lateral face of the scapulocoracoid were made as follows: Greatest length: greatest distance from procondyle to metacondyle (Fig. Ib). Greatest height: greatest distance from scapular process to base of scapulocoracoid. Premesocondyle: greatest distance from procondyle to midpoint of mesocondyle. Postmesocondyle: greatest distance from midpoint of mesocondyle to metacondyle. Anterior fenestra width: greatest horizontal distance between anterior and posterior border of fenestra. Anterior fenestra height: greatest vertical distance between dorsal border of antero- dorsal fenestra and ventral border of anteroventral fenestra (if anterior bridge is present) or greatest 532 BULLETIN OF MARINE SCIENCE, VOL. 29, NO.4, 1979 a {ftL 5 b I 7 (f\~ iT14 ~~ I OQO a 0 I --.~--9--- Figure I. a Ventral view of a rajoid neurocranium showing proposed measurements; b Lateral view of a rajoid scapulocoracoid showing proposed measurements. I = nasobasal length, 2 = width across otic capsules, 3 = least width of basal plate, 4 = greatest width of nasal aperture, 5 = internasal width, 6 = greatest length, 7 = greatest height, 8 = premesocondyle, 9 = postmesocondyle, 10 = postdorsal fenestra length, II = postdorsal fenestra height, 12 = anterior fenestra length, 13 = an- terior fenestra height, 14 = height of rear corner. McEACHRAN AND COMPAGNO: TAXONOMIC STUDIES OF RAJOIDEI 533 vertical distance between dorsal and ventral border of anterior fenestra (if anterior bridge is lacking). Postdorsal