ARTHUR: FOOD AND FEEDING OF LARVAL FISHES 



larva until about midway through its larval exis- 

 tence. Further research is required to determine if 

 the decline in relative physical condition indicates 

 a state of serious malnutrition and, if so, how far 

 from the well-fed state can the condition of the 

 individual vary without resulting in mortality. It 

 is also possible that laboratory-reared larvae have 

 abnormally large relative body depths. 



Food Size, Feeding Incidence, and 

 Condition of Anchovy Larvae 



The foregoing discussion points to three sig- 

 nificant trophic features of the average ocean- 

 caught anchovy larva. These features are: 



1. A lack of increase in food particle size propor- 



tional to the increase in length for larvae 

 larger than 4 mm (Figure 3b). 



2. A steep decline in feeding incidence beginning 



at 4.5 mm followed by an increase in this value 

 during the second half of the larval period 

 (Figure 8). 



3. A decline in relative morphological condition at 



lengths from at least 10 mm to 17 or 18 mm, 

 followed by an abrupt increase (Figure 9). 



Feature 1 must partly reflect the size spectrum 

 of the available plankton. Arthur (1956) and Beers 

 and Stewart (1970) have shown that there are far 

 more food particles of the size taken by the first 

 feeding larvae (50-100 jum) than there are of 

 larger particles suitable for older larvae (i.e., 200 

 /xm). Sardine and jack mackerel larvae, however, 

 are able to secure increasingly larger food parti- 

 cles (Figures 1, 5). When features 1 and 2 are 

 considered together, it would appear that the 

 average oceanic anchovy larva does not sustain its 

 original feeding intensity. 



Growth of laboratory-grown anchovy larvae 

 becomes asymptotic at 6 mm long when fed only 

 Gymnodinium and at 20 mm when fed only a 

 combination of Gymnodinium and rotifers. This 

 was noted by Hunter (in press), who concluded 

 that it is physically impossible for larvae to ingest 

 enough prey in order to grow when the prey are 

 below a certain size. Therefore, the decrease in 

 relative body depth of the ocean-caught anchovy 

 larva (feature 3) could be directly related to the 

 insufficient increase in food particle size (feature 



1). 

 Feeding intensity of clupeoid larvae decreases 



with malnutrition (Blaxter and Ehrlich 1974; 

 Hunter in press). If the decline in relative body 

 depth does denote a condition of malnutrition, 

 then the decrease in feeding incidence (feature 2) 

 is correlated with this decline, and might be the 

 causative factor. This might also result in larvae 

 spending a longer residence time at these lengths 

 which would introduce a bias in mortality 

 estimates. 



It is important to keep in mind that we are 

 considering larvae which have grown in the ocean 

 and have also been caught by plankton nets. This is 

 the reason that the expression "ocean-caught" 

 rather than "ocean-grown" has been used herein. 

 It might be reasoned that the decline in physical 

 condition is a sampling artifact produced by the 

 plankton net catching an increasing percentage of 

 sick or malnourished specimens of the larger 

 larvae as a result of the larger healthy larvae being 

 more capable of dodging the net. The same rea- 

 soning could be applied to the decline in feeding 

 incidence. An examination of the physical condi- 

 tion of over 5,000 sardine larvae (Arthur 1956) 

 revealed that there is a higher percentage of 

 larvae in poor shape (e.g., with liver deterioration) 

 taken in day hauls when healthy larvae can avoid 

 the plankton net. Such evidence led Isaacs (1964) to 

 theorize that day-caught sardine and anchovy 

 larvae represent an approximation of the percent- 

 age of the population removed by natural mor- 

 tality. Assuming this sampling bias, however, it 

 then becomes difficult to explain the increase in 

 both relative body depth and feeding incidence of 

 the older larvae taken by the same sampling 

 methods. Burdick (1969), while examining 

 Hawaiian anchovy (Stolephorus purpureus) lar- 

 vae, observed no difference of feeding incidence or 

 physical condition between samples taken concur- 

 rently with 1-m net and a plankton purse seine. 

 Assuming the plankton purse seine captures all 

 larvae, sick or well, he concluded that there is no 

 bias produced by only the healthy larvae being 

 able to avoid the 1-m net. 



The average ocean-caught anchovy is signifi- 

 cantly less robust at its midlarval lengths than its 

 laboratory counterpart, owing presumably to 

 differences in their respective rations. The first 

 feeding (4-day-old) laboratory-reared anchovy 

 larva spends 85% of the daytime in intermittent 

 swimming, 7% in feeding, and 4% at rest (Hunter 

 1972). Perhaps the undernourished average ocean- 

 caught larva, in response to the usual suboptimal 



527 



