FOLKVORD and HUNTER: VULNERABILITY OF NORTHERN ANCHOVY LARVAE 



reported that only 50% of starved, 33 mm northern 

 anchovy larvae responded to the attacks of adult 

 northern anchovy as compared with 100% for fed 

 larvae. No starved 10 mm larvae escaped attack 

 whereas 15-20% of the fed 10 mm larvae did so. The 

 numbers of observations were insufficient for a 

 statistical comparison, but recent work by Booman 

 (unpubl. data, Southwest Fisheries Center, La Jolla, 

 CA) indicates starvation can have a statistically 

 significant effect on responsiveness of 10 mm north- 

 ern anchovy larvae to adult northern anchovy 

 predators. 



The effect of the onset of larval schooling was not 

 considered in these experiments; however, escape 

 and response probabilities of individual larvae may 

 not be altered greatly by the onset of schooling. The 

 work of Major (1978) indicates that the most impor- 

 tant effect of schooling may be to reduce the rate 

 of attack by predators. He also found that the ma- 

 jority of Hawaiian anchovy captured by predators 

 were isolated individuals that had moved away from 

 the school, and predator success on schooled prey 

 was similar to that on isolated prey. The onset of 

 schooling in larval northern anchovy occurs between 

 11 and 15 mm SL, but the time spent in organized, 

 cohesive schools increases throughout the northern 

 anchovy's larval and juvenile periods (Hunter and 

 Coyne 1982). Thus attack rates of predators might 

 be expected to decline throughout later larval and 

 juvenile life as the northern anchovy spends more 

 time in cohesive schools. The onset of schooling oc- 

 curs over the size range in which we observed the 

 maximum predation rate (numbers consumed in 5 

 min) on individual northern anchovy larvae by north- 

 ern anchovy predators. Thus predation pressure 

 may be an important evolutionary factor in the tim- 

 ing of the onset of schooling during the larval stage. 



The interaction between larval growth rate and 

 size-specific vulnerability to predation may be an im- 

 portant source of interannual variation in larval 

 mortality (Shepherd and Cushing 1980; Smith 1985). 

 A simple calculation illustrates this point using the 

 size-specific vulnerability of northern anchovy lar- 

 vae (10-20 mm SL) to adult northern anchovy 

 predators. We assumed larval escape ability to be 

 an inverse measure of predator vulnerability and 

 normalized it to the average mortality rate over this 

 size interval (Table 1). Thus in our calculation, the 

 rate larval mortality decreased with increasing lar- 

 val size was inversely proportional to the rate escape 

 ability increased with size (larval escape ability in- 

 creased linearly with larval length over the 10-20 

 mm length range). Our calculation indicated that a 

 50% increase in growth rate from the average rate 



of growth in the sea resulted in a 58% increase in 

 survival in 30 d compared with average conditions. 

 Decreasing the growth rate by 50% gave a 37% 

 decrease in survival over the same interval. A longer 

 period of reduced or enhanced growth rates will, of 

 course, give a larger deviation from average survival 

 values. 



Table 1 .—Calculation of the effect of growth rate on survival of 

 10-20 mm northern anchovy larvae when mortality is inversely pro- 

 portional to length specific escape probabilities. 



Initial equations 



Z = 0.15 - (0.00625 x S) a 

 S = 10 + (G x 7") 

 dNIdt = - (Z x N) 



Final equations after substitution and integration 

 N = 0.0724 x exp (2.808 x G) 



a Mortality function generated from larval anchovy escape data with adult 

 northern anchovy as predators. Values are normalized to Z 0.05 at 16 mm 

 (Smith 1985). 



"From Smith (1985), 0.325 ± 50% also used in calculation. 



Effect of Predator Size 



We examined the existing literature on predators 

 of larval northern anchovies to determine how the 

 ability to escape a predator varied among different 

 predator species. Regardless of the predator species, 

 larval escape ability always increases with larval 

 size, but the rates vary greatly with predator size. 

 In general the smaller the predator, the faster lar- 

 val escape abilities improve with increasing larval 

 length (Fig. 6). The results of our work on E. mor- 

 dax were similar to those of Brownell (1985) on E. 

 capensis. However, capture success of the 85 mm 

 E. mordax predators used in our study was about 

 20% higher than the 34 mm E. capensis predators 

 used by Brownell. 



The extent of the predator field for a given size 

 and species of predator can be defined as the larval 

 size range in which larval escape success is <100%. 

 For adult northern anchovy predators (85 mm and 



867 



