FISHERY BULLETIN: VOL. 70, NO. 3 



Although there are some differences, most of 

 the assumptions and procedures I used were sim- 

 ilar to those used by Rosenthal and Hempel 

 (1970) to estimate food densities required by 

 herring larvae. Rosenthal and Hempel estimat- 

 ed that the minimum food ration for herring 

 larvae 10 to 11 mm was 4 to 8 nauplii/liter and 

 the maximum was 21 to 42/liter. My minimum 

 estimate based on metabolic considerations for 

 a comparable size anchovy larvae was 29/liter. 

 For the minimum density they assumed, as 

 I have, that the larva struck at all prey 

 that entered its perceptive field, but for the 

 maximum density they assumed that only a part 

 of the plankters perceived resulted in a com- 

 pleted feeding sequence. In the anchovy larvae 

 feeding on food at high density, on the average 

 only 40% of feeding sequences were completed. 

 Thus, using a similar criterion for a maximum 

 ration as Rosenthal and Hempel (1970), the 

 maximum density required for anchovy larvae 

 would be 2.5 times the minimum ration or for a 

 10- to 11-mm larva about 72 rotifer equivalents/ 

 liter. Considering specific diflferences and the 

 differences in assumptions and procedures, the 

 estimated food requirements of herring larvae 

 by Rosenthal and Hempel (1970) do not differ 

 much from the one calculated for an anchovy 

 larva of comparable size. 



In a survey of the literature given by Blaxter 

 (1965) the densities of food for larval cluepoid 

 fishes in the natural environment ranged frofn 

 1 to 68 items/liter. Arthur (1956) in a com- 

 prehensive study of the food of pelagic larvae 

 in the California Current region found 1 or more 

 nauplii/liter in 72% of his samples and 30 or 

 more nauplii/liter in about 3% of his samples. 

 Beers and Stewart (1967) measured the density 

 of copepod and naupliar and post naupliar stages 

 (35 to 103 fji) in a 600-mile transect in an area 

 of anchovy abundance and obtained densities 

 ranging from 5 to 17 copepods/liter. Thus, food 

 in the natural environment appears to be near 

 or below the minimum concentration I estimated 

 was required for older larvae but considerably 

 below that for first feeding larvae. 



Nearly every adjustment that could be made 

 in my estimates of food requirements for an- 

 chovy larvae would sharpen the differences be- 



tween early and late larval stages or elevate the 

 overall food requirement. For example, natural 

 prey would be of greater variety and could be 

 more difficult to capture, thus causing a differ- 

 ential increase in food concentration required by 

 younger larvae. In addition, older larvae are 

 capable of feeding on a greater variety of food 

 because they can feed on all food used by younger 

 stages plus larger prey as well. I assumed a di- 

 gestion efficiency of 100% but it must be less 

 than that and this would also increase the overall 

 food requirement. In short, an adjustment in 

 the estimates would probably increase the dif- 

 ference between the estimated food density re- 

 quirements taken from laboratory measure- 

 ments and natural food densities. 



Laboratory measurements indicated that lar- 

 val anchovies are more vulnerable to death from 

 starvation just after yolk absorption than at any 

 other time during larval life. This conclusion 

 agrees with those drawn from similar laboratory 

 studies on the herring and other species. Wheth- 

 er or not the increased vulnerability to a star- 

 vation death just after yolk absorption is related 

 to year class strength as contended by Hjort 

 (1914) is still a debated question (Blaxter, 1969; 

 Saville, 1971). The catch curves for larval an- 

 chovy over 10 years give no indication of an in- 

 crease in mortality just after yolk absorption 

 (Lenarz, 1972) . Three explanations for the ab- 

 sence in the anchovy of an early critical period 

 are: first, turbulence may generate random 

 movement between fish larvae and prey and 

 thereby considerably extend the search volume 

 (Murphy, 1961 ) ; second, it may be that the mor- 

 tality of larvae at all stages is so great because 

 of starvation and other causes that the increased 

 vulnerability at the youngest stages is not de- 

 tectable; and third, food may not be limiting 

 for early larvae because of patchy food distri- 

 bution. The absolute number of prey required 

 by early anchovy larvae is not great but a high 

 concentration is required; hence, the patch size 

 could be quite small. It would be of interest in 

 this regard to determine the size, density, and 

 distribution of food patches in the natural en- 

 vironment on a scale appropriate to fish larvae. 

 It would also be of interest to determine the 



836 



