FISHERY BULLETIN: VOL. 86, NO. 1 



ature incubation chamber set at 14.0° ± 0.2°C. In- 

 cubation jars were not aerated, and no attempt 

 was made to feed the larvae. 



All specimens used for description were fixed in 

 a IQ% formalin-seawater solution buffered with 

 sodium p-glycerophosphate and later transferred 

 to a 59^ solution. 



This description is based on a series of 74 lar- 

 vae, 2.2-34.2 mm in length, although comments 

 on pigment and meristic variability stem from 

 routine examination of several hundred speci- 

 mens. A representative series of larvae is de- 

 posited with the South Australian Museum, Ade- 

 laide, South Australia. 



Developmental terminology follows Ahlstrom 

 et al. (1976). Body measurements follow Matarese 

 et al. (1981). Length measurements are reported 

 as notochord length, NL (i.e., from the snout tip to 

 the end of the notochord) in preflexion and flexion 

 larvae, and standard length, SL (i.e., from the 

 snout tip to the posterior margin of the superior 

 hypural elements) in postflexion larvae and juve- 

 niles. Larvae were measured under a dissecting 

 microscope fitted with an ocular micrometer and 

 a camera lucida. Juveniles were measured with 

 vernier calipers. 



Meristic counts and examination of ossification 

 sequences were made on specimens cleared and 

 stained using Alizarin Red S-KOH-glycerine 

 (Hollister 1934). Caudal osteology follows Inada 

 (1981), Marshall and Cohen (1973), and Monod 

 (1968). 



Vertebral counts include the first vertebrae, 

 the neural spine of which is fused to the supraoc- 

 cipital crest (Marshall 1966), and both ural cen- 

 tra. Vertebral centra were counted as ossified 

 only when a complete band of stain connected 

 both neural and haemal spines. 



RESULTS 



Identification of M. novaezelandiae larvae was 

 based on their typical gadiform morphology 

 (large head, compact gut, tapering body form), 

 myomere count, and the development of confluent 

 dorsal-caudal-anal fins (see section on Distin- 

 guishing Features). Identification of field- 

 collected specimens was confirmed by comparison 

 with reared larvae. 



Distinguishing Features 



Prior to median fin development, myomere 

 counts are useful in separating M. novaezelandiae 



larvae (78-80) from similarly pigmented morid 

 (41-72), macrourid (10-16 + 70 > 100), gadid 

 (39-64) and other known merlucciid larvae (48- 

 58) which they superficially resemble (Marshall 

 and Iwamoto 1973; Fahay and Markle 1984; 

 present study). 



Both M. novaezelandiae and most morid larvae 

 show moderately pedunculate pectoral fins, a fea- 

 ture common in gadiform larvae with delayed 

 caudal development (Fahay and Markle 1984). 

 Macrourid larvae, in contrast, have very promi- 

 nently stalked pectorals and can further be sepa- 

 rated from M. novaezealandiae and most morids 

 by precocious development of the pelvic fin. 



Size at caudal flexion and the sequence of fin 

 development are also useful in separating M. 

 novaezelandiae from all Merluccius species. In 

 Merluccius , notochord flexion generally begins at 

 about 9 mm and the caudal fin is the first to form 

 (Dunn and Matarese 1984; Fahay and Markle 

 1984). Macruronus novaezelandiae larvae do not 

 begin caudal flexion until approximately 20 mm, 

 and the caudal fin is the second last to form. 



Macruronus novaezelandiae larvae have 1-3 

 prominent melanophores along the ventral mid- 

 line of the tail (although variable in appearance, 

 see section on Trunk and Tail Pigmentation) 

 and a double series of dorsal melanophores. When 

 expanded, melanophores in these two regions 

 coalesce to give the appearance of a broad 

 postanal band. Postanal banding patterns are 

 widespread in gadoid larvae (Fahay and Markle 

 1984); however, unlike many gadoid larvae, M. 

 novaezelandiae lacks pigment at the notochord 

 tip. 



At larger sizes M. novaezelandiae larvae de- 

 velop long-based dorsal and anal fins confluent 

 with the caudal fin. Other gadoid larvae with this 

 configuration have markedly different pigmenta- 

 tion (see Fahay and Markle 1984 for details). 

 Ophidiiform larvae have confluent dorsal, caudal, 

 and anal fins but can be separated from M. no- 

 vaezelandiae by their lack of a separate first dor- 

 sal fin and general lack of body pigment (see Gor- 

 don et al. 1984). 



Development 



Embryonic development has not been treated 

 in detail here as it is the subject of a manuscript 

 in preparation by G. Patchell (Fisheries Research 

 Centre, Wellington, New Zealand). 



The pelagic eggs of blue grenadier are spheri- 

 cal, with an unsegmented yolk and a smooth cho- 



120 



