FISHERY BULLETIN; VOL. 85, NO. 2 



Larval Rearing 



Northern anchovy eggs were spawned by admin- 

 istering hormone injections to a brood stock held at 

 the Southwest Fisheries Center (Leong 1971). Lar- 

 vae raised from the eggs were kept in 100 L circular, 

 black polypropylene tanks at constant temperature 

 (15.5°C) and photoperiod (12 h/12 h). In this study, 

 I varied the prey type and density. At the onset of 

 feeding, northern anchovy larvae have small mouths 

 that restrict their feeding to small prey (Hunter 

 1977). Thus for the initial feeding (day 3) all larvae 

 were fed Gymnodinium splendens, a naked dino- 

 flagellate having a width of about 50 ^m (Lasker et 

 al. 1970), and 2 days later the larvae were fed larger 

 prey. Because it is impossible to quantify the caloric 

 input from Gymnodinium directly (digested cells 

 cannot be counted), the experiments began on day 

 5 when the larger prey could be removed from the 

 fish stomachs, counted, measured, and subsequently 

 expressed as caloric input. 



Experimental Design 



I conducted five experiments. In the first experi- 

 ment, rotifers, Brachionus plicatilis, at a density 

 of 35/mL were fed for 20 days, and the number of 

 prey eaten was determined. The size of the rotifers 

 was not measured in this initial experiment, but in 

 subsequent experiments, widths of prey eaten were 

 measured. In the second and third experiments, 

 rotifers were offered at densities of 2 and 25/mL 

 respectively for 14 days. In the fourth experiment, 

 copepods, Tigriopus califomicus, nauplii and cope- 

 podites were fed to larvae at a density of 2/mL 

 (usually 1 nauplius and 1 copepodite/mL) for only 

 9 days instead of 14 days because of problems with 

 the copepod culture. In the fifth experiment, larvae 

 were fed copepods at 0.2/mL for 12 days without 

 the initial addition of Gymnodinium. (Mortalities 

 were high on this low density copepod diet, and data 

 are few; high mortalities are consistent with results 

 reported by O'Connell and Raymond [1970] for 

 northern anchovy raised on a similar diet. The addi- 

 tion of a Chlorella bloom to this diet improves sur- 

 vival of northern anchovy [Moffatt 1981].) 



To estimate larval growth as the increase in stan- 

 dard length, SL, and dry weight, W, larvae were 

 pipetted individually onto a slide and measured while 

 alive. Larvae were then rinsed in distilled water and, 

 to ensure dependable mean dry weight determina- 

 tions, grouped by size class onto a clean slide. Lar- 

 vae <5 mm were grouped by 0.2 mm size class, and 

 those >5 mm were grouped by 0.3 mm size class. 



Numbers of larvae per group ranged between 2 and 

 12; the larger the larval weight, the fewer larvae 

 I grouped together. After drying to a constant 

 weight at 60°C (Lovegrove 1966), larvae were re- 

 moved from the slides by using a single-edged razor 

 blade and weighed on a Cahn electrobalance to ± 2 



Feeding 



To estimate feeding rates and daily consumption, 

 each prey item removed from the larva's gut was 

 counted and its width measured. Width is the dimen- 

 sion that limits a fish larva's selection of prey (Beyer 

 1980; Hunter 1981). Prey defecated onto the slide 

 were included in the total number eaten. Gut con- 

 tents in dry weight were determined by summing 

 the width- specific weights of the prey eaten each 

 day. I used the width-specific fresh dry weights and 

 caloric values for Brachionus and Tigriopus given 

 by Theilacker and Kimball (1984) and reproduced 

 here in Table 1. 



Because northern anchovy larvae eat continuous- 

 ly, asymptotic curves (c = C^ax x (1 - e"^')) were 

 used to describe food intake, i.e., the relation be- 

 tween observed gut contents c and time t, where 

 C^ax is the asymptotic gut contents (contents in gut 

 at steady state after filling) and k is the instan- 

 taneous rate of gut filling. Using Marquardt's 

 algorithm for fitting nonlinear models, parameters 

 Cn,ax and k were estimated for 0.5 mm length 

 classes (Table 2). Fish with empty stomachs were 

 included in this analysis because all fish were used 

 in the growth estimates. Daily mean gut contents 

 c were calculated by integrating the area beneath 



Table 1 .—Width-specific dry weight and caloric value of rotifers, 

 Brachionus plicatilis, and copepods, Tigriopus californicus\ 



iProm Theilacker and Kimball 1984, table 2. 



214 



