Jordan and Bruce: Larval development of three roughy species 



77 



tralian species. Parr (1933) and Johnson (1970) de- 

 scribed 19mm and 21.5mm juvenile Korsogaster, re- 

 spectively, a genus subsequently synonymized with 

 Hoplostethus by Woods & Sonoda (1973). Crossland 

 (1981) illustrated a trachichthyid larva, possibly 

 Optivus elongatus, from northeastern New Zealand. 

 Robertson ( 1975) described an egg tentatively ascribed 

 to Paratrachichthys trailli. Kotlyar (1984) described 

 juveniles of four species of Hoplostethus (including a 

 36mm H. atlanticus), the smallest of his specimens 

 being a 15 mm H. melanopterus. Okiyama (1988) 

 figured and briefly described single specimens of an 

 unidentified Hoplostethus (10.7mm), Gephyroberyx 

 japonicus (11.0 mm ), and Paratrachichthys prosthemius 

 (27.5 mm). 



Comparatively little is known of trachichthyid lar- 

 val characters; however, common characters include 

 precocious pelvic fin development, heavy pigment, a 

 stocky body form approaching the shape of adults (in 

 larger larvae), a myomere count of 26-30, and the pres- 

 ence of minute spines over the body surface ( Keene & 

 Tighe 1984. and references therein). 



Recent commercial interest in the orange roughy 

 Hoplostethus atlanticus has emphasized the need for 

 information on the early life history of this species. As 

 yet, despite considerable effort in ichthyoplankton sam- 

 pling, no H. atlanticus larvae have been reported from 

 Australian waters. The current study describes the 

 larval development of three trachichthyid species com- 

 plexes — Paratrachichthys sp., Aulotrachichthys sp., and 

 Optivus sp. — in specimens obtained from plankton 

 samples collected primarily in Tasmanian and New 

 South Wales coastal waters from 1984 to 1986. These 

 descriptions are presented in order to further define 

 larval characters that may be of use in trachichthyid 

 systematics and future identification of other 

 trachichthyid larvae, including//, atlanticus. 



Materials and methods 



Specimens were largely obtained from ichthyoplankton 

 samples collected in 1984-86 by the CSIRO Division of 

 Fisheries, Hobart, Tasmania, as part of a study aimed 

 at documenting the distribution and abundance of lar- 

 val fishes in Tasmanian coastal and neritic waters. De- 

 tails of sampling locations and protocol are provided by 

 Thresher et al. (1989). Larvae were obtained from ob- 

 lique tows to a depth of 200 m (bottom depth permit- 

 ting) at a series of stations covering shelf and slope 

 waters, using aim diameter ring net (500u mesh). 

 Additional material was obtained from samples collected 

 with identical gear in New South Wales shelf and slope 

 waters by A. Miskiewicz (Water Board, Environ. Proj. 

 Group, P*0 Box A53, Sydney South, 2000). 



Larval samples were fixed in either 10% formalin 

 buffered with sodium tetraborate, or 959c ethanol. Mor- 

 phometric analysis and illustrations were based on for- 

 malin-fixed specimens of Paratrachichthys and Optivus. 

 However, only ethanol-fixed material was available for 

 Aulotrachichthys. No allowance was made for shrink- 

 age or distortion in preservative. 



Larvae were examined using a Wild M5 dissecting 

 microscope, and all drawings were made with the aid 

 of a camera lucida. Larvae were identified using exist- 

 ing literature (Keene & Tighe 1984, Okiyama 1988) by 

 comparison with juvenile and adult features of identi- 

 fied species, and by the establishment of developmen- 

 tal series. Comparisons with similar larvae (e.g., zeids) 

 were made with material from the CSIRO samples. 



All unspecified body lengths refer to notochord length 

 in preflexion and flexion larvae, and to standard length 

 in postflexion larvae and juveniles. We define snout 

 to anal-fin length as the horizontal distance from 

 the tip of the snout to the anterior origin of the anal 

 fin or anal-fin anlagen. Body depth at anus is the 

 vertical distance between body margins through the 

 center of the anal opening. Body depth at pectoral 

 is equivalent to 'body depth' of Leis & Rennis (1983). 

 Other definitions, such as body shape, follow Leis and 

 Trnski ( 1989 1. Nomenclature of head spination follows 

 that of Moser & Ahlstrom (1978). Larval measurements 

 were made using an ocular micrometer. Juveniles were 

 measured with vernier calipers. 



Results 



During 18 months of sampling, 119 Paratrachichthys, 

 147 Optivus, and 25 Aulotrachichthys larvae were col- 

 lected. The distribution of larvae is detailed in Figure 1. 

 No larvae that could be attributed to Hoplostethus were 

 collected. A representative series of each species was 

 deposited in the ISR Munro Fish Collection (CSIRO, 

 Hobart, Tasmania). Reference numbers: Optivus, 

 CSIRO L179-184; Paratrachichthys, CSIRO L185-190; 

 Aulotrachichthys, CSIRO L191-196. 



Identification 



In larger specimens of two of the series, the anus is 

 located between the pelvic fins. Only three trachich- 

 thyid genera have this character: Paratrachichthys, 

 Aulotrachichthys, and Sorosichthys (May & Maxwell 

 1986). Sorosichthys is separated easily on the basis of 

 a pelvic count of 1,5, compared with the 1,6 of the 

 other two genera (Table 1). The only character reported 

 in the literature to distinguish adults of Aulo- 

 trachichthys from Paratrachichthys is the presence in 



