Lowry et al.: Phytogeny of Oreosomatidae 



693 



ably higher because some are not reported and oth- 

 ers discarded (Lyle et al., 1992). The recorded ton- 

 nages of individual species are unreliable owing to 

 confusion over species identification in the catch log 

 books. As in New Zealand, the smooth and black oreos 

 dominate the Australian catch, whereas the spiky 

 oreo (Neocyttus rhomboidalis Gilchrist, 1906) and the 

 warty oreo {Allocyttus verrucosus (Gilchrist, 1906)) 

 are also important. A fifth species, the oxeye oreo 

 (Oreosoma atlanticum Cuvier, 1829), is commonly 

 caught but discarded because of its small size and 

 low commercial value. 



This paper presents the results of an allozyme sur- 

 vey of the five described Australasian species 

 (Allocyttus niger,A. verrucosus, Neocyttus rhomboi- 

 dalis, Oreosoma atlanticum, and Pseudocyttus 

 maculatus) and a new species (the rough oreo, 

 Neocyttus sp., Yearsley and Last 2 ) often captured with 

 A. niger and P. maculatus. A third Neocyttus species, 

 N. helgae (Holt and Byrne, 1908), from the North 

 Atlantic, was also examined. 



Oreosomatids not included in this study are the 

 North Pacific Allocyttus folletti Myers, 1960, the 

 southern Atlantic and Indian Ocean Allocyttus 

 guineensis Trunov and Kukuev in Trunov, 1982, and 

 the Indian Ocean Neocyttus acanthorhyncus (Regan, 

 1908). Another member of the family, the Southern 

 Ocean Pseudocyttus nemotoi (Abe, 1957), was re- 

 cently resurrected by Miller ( 1993). 



The intrarelationships of zeiforms have not been 

 discussed in the literature; thus outgroup selection 

 for this phylogenetic study is difficult. Many authors 

 consider the beryciforms to be more primitive than 

 the zeiforms but closely related to them (e.g. Green- 

 wood et al., 1966). Zehren ( 1979) found the Berycidae 

 to be more primitive than the remaining beryciform 

 families and, thus, a berycid may be a suitable 

 outgroup. However, Rosen (1984) dramatically 

 changed the placement of the zeiforms, including 

 them in the order Tetraodontiformes, with the 

 Caproidae as the sister group to all other tetra- 

 odontiforms (the caproids' placement within the 

 Zeiformes was questioned by others [Tighe and 

 Keene, 1984]). Acaproid may therefore be a suitable 

 outgroup. Furthermore, Rosen placed the zeids im- 

 mediately before the oreosomatids in his new divi- 

 sion Zeomorphi. He used "acanthurids plus chaeto- 

 dontids" to establish character polarities. Conse- 

 quently, a zeid or an acanthurid are also possible 

 outgroups. In the absence of caproid specimens, three 

 outgroups were selected for analysis: the berycid 

 Beryx splendens Lowe, 1833 (alfonsino), the zeid 



Cyttus australis (Richardson, 1843) (silver dory), and 

 the acanthurid Naso tuberosus Lacepede, 1802 

 (humphead unicornfish). 



Genetic variation present in the Australasian 

 oreosomatids and diagnostic allozyme loci for each 

 species are presented in this paper to assist in fu- 

 ture management plans for the developing deepwater 

 fishery. The phylogenetic relationships of previously 

 known species, as well as a new species from Aus- 

 tralia and a species from the North Atlantic, are dis- 

 cussed in an effort to understand more fully the sys- 

 tematics of the family Oreosomatidae. 



Materials and methods 



Samples of muscle and liver tissue were collected 

 from seven oreosomatids (Allocyttus niger, Allocyttus 

 verrucosus, Neocyttus sp. (voucher specimen: CSIRO 

 H2865.01), Neocyttus helgae, Neocyttus rhomboidalis, 

 Oreosoma atlanticum and Pseudocyttus maculatus), 

 and from the three outgroup species (Beryx splendens, 

 Cyttus australis, and Naso tuberosus). Sample de- 

 tails and species abbreviations are given in Table 1. 



Whole fish were frozen after capture and trans- 

 ported frozen to the laboratory, where tissues were 

 dissected and held at -80°C. Small pieces of tissue 

 were placed in 1.5-mL microcentrifuge tubes, homog- 

 enized manually with a few drops of distilled water, 

 and spun at 11,000 rpm in a microcentrifuge for 2 min- 

 utes. The supernatant was used for electrophoresis. 



Allozyme variation was examined with three gel 

 systems: gel system A — Helena Titan III cellulose 

 acetate plates run at 200 V with a Tris-glycine buffer 

 (0.020 M tris and 0.192 M glycine, Hebert and 

 Beaton 3 ); gel system B — Helena Titan III cellulose 

 acetate plates run at 150 V with a Tris-citrate buffer 

 (0.075 M tris and 0.025 M citric acid, pH 7.0); gel 

 system C — 8% Connaught starch gels with a histi- 

 dine/citrate buffer (gel buffer: 0.005 M histidine HC1, 

 pH 7.0; electrode buffer: 0.41 M trisodium citrate, 

 pH 7.0). Standard staining procedures were followed 

 (Richardson et al., 1986; Hebert and Beaton 3 ). 



In all, 19 enzymes, representing 27 loci, were ex- 

 amined (Table 2) and allele frequencies determined 

 (Table 3). However, the locus GPI-A* was not included 

 in the phylogenetic analyses because of poor resolu- 

 tion in three species (A. niger, N. sp., and P. macu- 

 latus). Loci and alleles are designated by the nomen- 

 clature system outlined in Shaklee et al. (1990), ex- 

 cept that peptidase loci were identified as PEP 1* and 



2 Yearsley, G.K., and PR. Last. 1995. CSIRO Division of Fish- 

 eries. Castray Esplanade, Hobart, Tasmania 7000, Australia. 



3 Hebert, P. D.N., and M.J. Beaton. 1989. Methodologies for 

 allozyme analysis using cellulose acetate electrophoresis: a prac- 

 tical handbook. Helena Laboratories, Beaumont, Texas. 



