568 



Fishery Bulletin 89(4), 1991 



higher temperatures for successful development than 

 sole eggs, probably because sole's spring spawning 

 cycle results in larval abundances later in the season 

 when sea temperatures are warmer (Irvin 1974, Fonds 

 1979). 



Our purpose was to determine how growth and sur- 

 vival of California halibut may be influenced by tem- 

 perature during development from egg to 3-month-old 

 juvenile, and how optimal temperature ranges may 

 change with ontogeny. Additionally, we were inter- 

 ested in how temperature may influence the timing of 

 an important stage of flatfish development— settle- 

 ment. We examined the possibility that month-old 

 pelagic larvae passively transported to warmer inshore 

 waters would settle significantly sooner due to a faster 

 development rate than those remaining in colder off- 

 shore areas, resulting in selective settlement of shallow 

 coastal nursery grounds. Larvae remaining pelagic due 

 to colder waters would have a continued chance of 

 transport inshore. The results of this study will aid in 

 better understanding how oceanographic conditions 

 may influence the survival of halibut early-life-history 

 stages in the field. 



Methods 



Egg development 



Eggs and sperm were stripped from ripe field-collected 

 California halibut and combined in petri dishes. Ap- 

 proximately 300 fertilized eggs were placed in each of 

 ten 3L glass jars containing filtered, ultraviolet-light 

 sterilized sea water (35%o) at an ambient temperature 

 of 16°C. To establish and maintain desired tempera- 

 tures, jars were placed in temperature-controlled water 

 baths. A low level of aeration was provided in each jar 

 to avoid the formation of temperature gradients. Soon 

 after fertilization, temperatures in eight jars were 

 raised or lowered 1°C/15 minutes from 16°C to 8, 12, 

 20, and 24°C, resulting in two replicates of each tem- 

 perature treatment. Light cycles simulated natural con- 

 ditions (12L:12D). Every 2 hours until hatching, tem- 

 peratures were recorded and at least five live eggs 

 were sampled from each jar and preserved in 4% for- 

 malin. Because of the short experimental duration, 

 water in the jars was not exchanged with fresh sea- 

 water during the experiment; when eggs in all jars had 

 hatched or died, the experiment was terminated. To 

 monitor development, egg series were examined using 

 a dissecting microscope. 



Larval growth and survival 



An experiment was initially conducted to determine the 

 influence of temperature on starvation rate of newly- 

 hatched halibut larvae. Fertilized eggs were obtained 

 from natural spawns from brood stock held in an out- 

 door 5 m diameter tank with a flow-through seawater 

 system (for a further description of brood stocks, see 

 Caddell et al. 1990). Seventy-five late-stage eggs (with 

 developed embryos) were placed in each of ten 3 L glass 

 jars of sterilized seawater at an ambient temperature 

 of 16°C. Jars were in temperature controlled water 

 baths and light aeration was provided. Temperatures 

 in eight jars were raised or lowered 1°C/15 minutes 

 from 16°C to 8, 12, 20, and 24°C; each temperature 

 treatment was replicated. Light cycles simulated 

 natural conditions (12L:12D). Twenty percent of the 

 water volume was replaced with fresh seawater twice 

 weekly; at this time, ammonia and salinity levels were 

 monitored. Dead halibut larvae were removed and 

 counted daily until total starvation had occurred. 



To determine how growth and survival of halibut lar- 

 vae are temperature-dependent, the above experiment 

 was repeated except food was added. Gadomski and 

 Petersen (1988) found that for greatest survival, first- 

 feeding halibut required food by the day of total yolk 

 absorption. Full yolk depletion occurs 6 days after 

 hatching at 16°C, but yolk absorption time varies with 

 temperature (Gadomski et al. 1990). In the current 

 study, we fed rotifers Brachionus plicatilis to halibut 

 larvae after eye pigmentation and mouth development, 

 but before total yolk absorption. Rotifers were stocked 

 at 15/mL, following the methods of Gadomski et al. 

 (1990), and supplemented thereafter as needed to main- 

 tain this density. Dead halibut larvae were removed and 

 counted daily. Seventeen days after hatching, noto- 

 chord lengths of surviving larvae were measured. 



Because final survival was the same (zero) for all 

 starvation trials, we compared survival curves of 

 starved and fed larvae using the Mantel-Haenszel test 

 (Matthews and Farewell 1985, Gadomski and Petersen 

 1988). This method tests the null hypothesis that in- 

 cremental survival rates, computed between successive 

 sampling times, are similar throughout the observation 

 period. Rapid decomposition of deceased larvae re- 

 sulted in accumulative totals of survivors and mortal- 

 ities in some trials at the end of the experiment to be 

 less than the original stocking density of 75. The 

 Mantel-Haenszel test compensated for this by adding 

 the "lost" mortalities to each experimental day pro- 

 portionate to the known number that had died that day. 

 First we tested if replicates could be combined; then 

 we tested for a significant difference between survival 

 curves of each experiment (starved and fed): 12°C vs. 

 16°C, 16°C vs. 20°C, and 20°C vs. 24°C. Because of 



