FISHERY BULLETIN: VOL. 80, NO. 3 



and were assumed to be about 12 h old. The time 

 to 50% hatching was determined by interpolation 

 from the linear regression of percent hatching 

 (y) against time (x). Confidence intervals on the 

 50% hatch time were calculated from the predic- 

 tion of x from y as: 



C.L. = x + 



byAyi - y) 

 D 



+ H, 



where D — b yx — t (o.o5,n-2> Sb , and 



H 



t 



(0 .05, n -2) 



~D~ 



n x 2 



Time to 50% yolk-sac absorption was determined 

 similarly. Survival of yolk-sac larvae appeared 

 to increase substantially under lighting in the 

 cold rooms, which raised temperatures in these 

 experiments to 10.5°C in the 8°C room and to 

 13.7°C in the 12°C room. The 15°C room was con- 

 sistently lighted. After hatching at temperature, 

 postyolk-sac larvae were removed, placed in new 

 jars, and observed to determine time to starva- 

 tion. 



Growth was examined by counting daily incre- 

 ments on otoliths. For verification of otolith in- 

 crements as daily marks, larvae were reared in 

 the laboratory without antibiotics in a 12-h light- 

 dark cycle. Larvae were fed Artemia salina 

 nauplii and natural zooplankton strained 

 through a 216 ^im mesh net at a concentration of 

 about 1 animal/ml. Field-caught specimens 

 were obtained from several cruises off the Cali- 

 fornia coast in 1977, 1978, and 1979. Both labora- 

 tory-reared and field-caught specimens were 

 preserved in 80% ethanol. Larvae were measured 

 (standard length) and otoliths were removed 

 under a dissecting microscope fitted with a po- 

 larizing filter. Otoliths were mounted on a glass 

 slide in protex or euparal, and rings on the oto- 

 liths were counted at 600-1000X magnifica- 

 tion. 



Larval dry weights (preserved in 80% ethanol 

 for otolith investigations and in 3% Formalin 2 for 

 respiration investigations) were determined on a 

 Cahn 25 Electrobalance after rinsing the larvae 

 in distilled water and subsequently drying them 

 for 24 h at 60°C. Weight loss due to preservation 

 was determined by comparing weights of sub- 



2 Reference to trade names does not imply endorsement by 

 the National Marine Fisheries Service, NOAA. 



samples of fresh-frozen larvae and preserved 

 larvae, all hatching from the same cohort of eggs. 

 Shrinkage in length was determined by measur- 

 ing anaesthetized larvae before preservation and 

 then again after 2-3 wk of preservation in 80% 

 ethanol or 3% Formalin. Shrinkage due to preser- 

 vation delay and death in a wet cod end during or 

 after a plankton tow was simulated by anaesthe- 

 tizing and measuring larvae, and then placing 

 them on seawater-wetted paper towels for spe- 

 cific time periods. Larvae were then preserved in 

 Formalin. These preservation effects were 

 tested only on first-feeding larvae. 



Growth, egg development, and yolk-sac ab- 

 sorption data were fitted with a Gompertz curve 

 using a least-squares nonlinear curve fitting pro- 

 gram (SPSS). The Gompertz growth function 

 was selected because it is a flexible nonlinear 

 function commonly used in studies of larval fish 

 (Zweifel and Lasker 1976). 



Metabolic Rates 



Respiration rates of larvae were measured 

 using a micro-Winkler technique (Carritt and 

 Carpenter 1966). Experiments were conducted 

 in 30 ml glass stoppered jars, at densities of 2-3 

 larvae/jar, in dim light for 11-14 h. Larvae in fil- 

 tered seawater were acclimated to temperatures 

 for 12 h. After the experiments were completed 

 and oxygen fixed, jars were kept for 2-10 d at 

 8°C in the dark before titrating. Experiments 

 were designed to include 3-5 replicates per tem- 

 perature; however, when bubbles formed during 

 the experiment (a constant problem at 15°) sam- 

 ples were discarded. 



Vertical Distribution 



Samples from vertical series of tows taken in 

 1954 and 1955 were reported by Ahlstrom 

 (1959). I sorted and measured these samples to 

 examine size-related vertical distribution. Pacif- 

 ic hake larvae from an additional three vertical 

 series taken in 1969 were sorted and measured. 

 Since there were no apparent day-night differ- 

 ences, all hauls, day and night, were combined. 

 Nonstandardized data consisting of raw num- 

 bers per haul, uncorrected for volume of water 

 filtered, were used since more detailed informa- 

 tion did not exist for many hauls. Data on num- 

 bers of larvae caught per haul were classified 

 into the depth interval where most of the tow took 

 place. 



590 



