FISHERY BULLETIN: VOL. 70, NO. 3 



equal to 0.45L2 for a cross section at X = 0.74L. 

 To estimate the volume search per unit time 

 these areas need to be multiplied by the larval 

 swimming speed. This aspect of the calculation 

 will be considered in a subsequent section. 



FEEDING RATES 



In all studies in this paper larvae were ob- 

 served and maintained in tanks containing high 

 and presumably optimum food densities {Gym- 

 nodinium 100 to 200/ml, Brachionus 10 to 20/ml, 

 and Artemia nauplii 3 to 10/ml) . It is unlikely 

 that density limited the rate of feeding under 

 these conditions. Thus, the rates recorded prob- 

 ably are near the maximum feeding rate that 

 can be sustained by larval anchovy. 



Records were kept of the frequency of feeding 

 acts of larvae ages 4 to 27 days raised in the 

 500-liter rearing tanks. A total of 325 obser- 

 vations of 5-min duration was taken at three 

 different times of day. No trend in the fre- 

 quency of feeding strikes with time of day ex- 

 isted. The larvae fed actively throughout the 

 day although their guts were filled after the first 

 half hour of feeding each morning. The aver- 

 age frequency of feeding strikes for all data 

 combined was 1.28 ± 0.144 strikes/min. Some 

 evidence existed that the rate may climb to 3 

 strikes/min for 2 to 3 days when Artemia nauplii 

 were first introduced, but no direct cause and 

 effect relationship could be established because 

 other prey were also present in the tank. 



Another estimate of feeding rate was obtained 

 from the data on feeding success described in 

 the next section. The larvae in that study had 

 empty guts at the beginning of the observation 

 period, and they were observed during their ini- 

 tial 10 min of feeding on a particular day, where- 

 as the preceding data observations were begun 

 about a half an hour after the larvae began 

 feeding and continued throughout the day. The 

 average feeding frequency for larvae studied in 

 the feeding success experiments (N = 100) was 

 1.75 ± 0.205 strikes/min and the range was from 

 0.2 to 5.8 strikes/min. Thus, during the first 

 10 min of feeding, feeding rates were on the 

 average somewhat higher than they were when 

 averaged over the entire day. Feeding rates on 

 Artemia were the same as those on Brachionus. 



FEEDING SUCCESS 



Feeding success was estimated by counting 

 the number of feeding acts completed in 10 min, 

 removing the larvae from the container, and 

 counting the number of prey in the gut. Larvae 

 were kept in the dark for 12 or more hr before 

 the test to insure that the gut was empty (an- 

 chovy larvae do not feed in the dark). I mea- 

 sured the feeding success of larvae fed Brachi- 

 onus (density, 10 to 60/ml) from the first day 

 of feeding (age 3 to 4 days) up to the 21st day 

 of larval life. A separate experiment was run 

 to determine the effect of a different prey on 

 feeding success. At age 17 days the diets of 

 some of the larvae were changed from one of 

 only Brachionus to one of Artemia nauplii (den- 

 sity, 3 to 15/ml) . Measurements of feeding suc- 

 cess were made over the first week of feeding 

 on Artemia. Gymnodinium was not used as a 

 food for first feeding larvae because it was usu- 

 ally defecated before we were able to examine 

 the contents of the gut. Two measurements of 

 feeding success of larvae fed Gymnodinium fell 

 within the scatter for Brachionus (Figure 9). 



The success of larvae fed Brachionus rapidly 

 increased over the first week of feeding (age 3 

 to 10 days), but thereafter the rate of increase 

 of feeding success was much lower. The form of 

 the relationship between age and success was 

 that of a learning curve. A semilog transfor- 

 mation of the data provided an adequate correc- 

 tion for this curvilinear trend. The regression 

 of feeding success in percent on the log of larval 

 age gave the relationship, percent success = 

 93.2 (log age) — 33.30 where si = 20.14. 



Seventeen-day-old larvae were less successful 

 in capturing A riem/a for the first time than they 

 had been in capturing Brachionus. Only 37% 

 of feeding acts were successful when larvae were 

 fed Artemia nauplii for the first time at age 17 

 days whereas ?>\[( were successful when the 

 food was Brachionus (Figure 10). The initial 

 level of feeding success for larvae fed Artemia 

 was higher than that of larvae when they first 

 began to feed at age 3 to 4 days, and less time 

 was required to obtain a high level of success on 

 Artemia than was required for first feeding 

 larvae. By the third day of feeding on Artemia 



832 



