FISHERY BI'1J,ETIN VOL 77, NO 

 II8°W 



MARCH 17-20,^977 



10 km 



Figure 4. — Surface isotherms in de- 

 grees Celsius between Los Angeles and 

 Santa Catalma Island, Calif., in March 

 1977, 



Selective Mortality 



If mortality rate is a function of size, then 

 growth rates determined from size at age data will 

 provide a biased estimate of the true growth rate of 

 the survivors which are sampled iRicker 1969). 

 The effect of size selective mortality on our esti- 

 mates of growth is difficult to assess. Few sets of 

 data are extensive enough to even consider the 

 question of ontogenetic changes in mortality rate 

 of larval fish. All available estimates are based 

 upon sized but not aged specimens. They depend 

 upon an assumed growth curve (Farris 1960) and 

 are susceptible to bias by size selective avoidance 

 of the sampler. Smith ( 1973) found a difference in 



mortality rate between sardine eggs and young 

 sardine larvae, but mortality in plaice was essen- 

 tially constant through the egg and larval stages 

 (Bannister et al. 1974), Laboratory experiments 

 show that older, more active yolk-sac anchovy lar- 

 vae are less susceptible than newly hatched larvae 

 to some invertebrate predators (Lillelund and 

 Lasker 1971; Theilacker and Lasker 1974). As a 

 rough estimate of the magnitude of the maximum 

 effect of size selective mortality on our growth 

 estimates, we examined the interaction between 

 variable growth and size selective mortality and 

 determined the effect of this mortality on mean 

 size of anchovy larvae at 25 days after yolk absorp- 

 tion. 



420 



