light orange in color. The eggs were incubated at 

 irC and hatched on 10 June 1974. Larvae were 

 held in static seawater without food and pre- 

 served at 0, 3, and 7 d posthatch. From April to 

 July 1988, older larval stages were taken in 

 epibenthic plankton tows in southern British 

 Columbia, using a sled trawl with a 1.5 m x 0.8 

 m mouth and 1 mm mesh. The larvae were killed 

 in the cod end bucket with anaesthetic and fixed 

 in 5% formosaline. Several of the largest larvae 

 were removed live and transported to a 1,000 L 

 laboratory tank with through-flowing seawater 

 of 10°C, where they were fed Artemia nauplii. 

 Two individuals survived past settlement and 

 were preserved as juveniles. 



Unidentified eggs, later found to be L. cal- 

 lyodon, were collected from among barnacles 

 and rock crevices in the intertidal zone (0.6 m 

 above low water, Canadian scale) outside Sooke 

 Basin in the Strait of Juan de Fuca (48°20'N, 

 123°44'W) on 22 March 1987 and 13 March 1988. 

 In 1987, egg masses and newly hatched larvae 

 were directly preserved. A variety of egg 

 masses taken in 1988, ranging in color from 

 maroon to orange to green, were incubated in 

 inflow water of a 1,000 L through-flowing sea- 

 water tank and hatches occurred from late 

 March through mid-April from the different 

 masses. Larvae were fed Artemia nauplii that 

 had been fed supplemental omega-3 fatty acids 

 (Cooper 1988). Larvae were preserved from the 

 rearing tank at various intervals up to the 

 benthic juvenile stage. 



Positive identification of laboratory-reared 

 juveniles was based on the full spectrum of ex- 

 ternal juvenile characters from the literature 

 (e.g., Hart 1973); larval characters alone could 

 not be used to positively identify to the species 

 level. Median fin ray meristics of late larval 

 stages for these two species overlapped too 

 broadly to permit positive identification prior to 

 the juvenile stage. Juveniles used for identifica- 

 tion of both species were deposited in the British 

 Columbia Provincial Museum (BCPM 988-945, 

 BCPM 988-946, BCPM 988-947, BCPM 988-948, 

 BCPM 988-949). 



Eggs and larval morphometries were taken 

 with vernier dial calipers under a dissecting mi- 

 croscope. Upon sorting, measures of notochord 

 length (NL) or standard length (SL) were taken 

 from fixed specimens. Body depth was 

 measured near the pectoral base, where the 

 maximum body depth dimension occurred, in- 

 cluding subdermal space (i.e., dorsal epidermis 

 to ventral epidermis). Body length was 



FISHERY BULLETIN: VOL. 87, NO. 3. 1989 



measured from tip of snout to posterior margin 

 of anus, not including any portion of the ab- 

 dominal cavity posterior to the anus. Pelvic disk 

 width, not length, was measured. Too few spec- 

 imens were available to permit clearing and 

 staining, although quick-staining with ahzarin 

 red permitted viewing of external features. Il- 

 lustrations were drawn using a dissecting 

 microscope and camera lucida. Specimens of 

 newly hatched L. fucensis from 1974, together 

 with the photographs, permitted redrawing of 

 an illustration from Marliave (1975). 



Larvae of Liparis fucensis 



Pigment intensity varied between siblings 

 hatched in 1974. Similarly, larvae of hke sizes 

 and stage, which were captured from the field 

 in 1988, varied in pigment intensity. The overall 

 pattern did not vary much, although a few of 

 the more intensely pigmented individuals 

 tended to prematurely develop sets of melano- 

 phores that average larvae develop later. Also, 

 these few intensely pigmented individuals 

 developed a broader extent of particular pig- 

 ment patches and more numerous, regularly 

 spaced melanophores in rows along the anal fin 

 base and ventral fin fold. Postflexion stages, in 

 particular, tended to show a variation in extent 

 or absence of pigment. Loss of early melano- 

 phores and appearances of different sets of 

 melanophores seemed to characterize the post- 

 flexion period of development. Overall, how- 

 ever, pigment corresponded clearly to the dis- 

 tribution patterns and intensities illustrated in 

 Figures 1 and 2. 



At all stages, the epidermis had a noticeably 

 granular appearance (Fig. 1), except around the 

 outer margins of the tail fin fold. The final de- 

 velopment of the expanded subdermal space, or 

 bubble, corresponded to the extent of this gran- 

 ular appearance. On the largest preflexion larva, 

 this granular layer could be scraped away from a 

 basement membrane, on which melanophores 

 remained. 



Hatching occurred at 2.9 mm NL; the larvae 

 had a prominent yolk sac with a single anterior 

 oil droplet (Fig. 2a). At hatching, dorsal gut 

 melanophores, about 20 postanal ventral mid- 

 hne melanophores, and a hexagonal honeycomb 

 pattern of melanin on the pectoral fin bases 

 were evident. At 10°C, yolk resorption occurred 

 over a period of seven days under starvation 

 conditions. During that time, larvae grew to 3.3 

 mm NL; head length increased from 19 to 21.5% 



736 



