FISHERY BULLETIN: VOL. 85, NO. 2 



Metabolism 



I determined metabolic rates for the larvae, which 

 ranged in age from first-feeding (3 days after hatch- 

 ing) to 25 days using the Winkler technique. I chose 

 the Winkler technique where large vessel volumes 

 could be used and there was no need to shake the 

 vessels during the experiment, as required for mano- 

 metric techniques. Pearcy et al. (1969) found no dif- 

 ferences between Winkler and Warburg estimates 

 of oxygen consumption. Oxygen consumption was 

 estimated at 16°C during 18-23 h experimental 

 periods with a 12 h light-dark cycle. The respiration 

 vessels were attached to a large, slowly rotating 

 wheel. Young larvae, 0.02-0.14 mg dry weight were 

 tested in 40 mL vessels in groups of 10-50, while 

 larvae older than 16 days (larger than 0.14 mg) were 

 tested individually in 60-150 mL vessels. All fish 

 tested had empty guts. Data were not used when 

 mortalities occurred during the experiment. 



To express metabolism (Q) as a function of dry 

 weight, I used a nonlinear regression to fit a power 

 equation to the data (see parameters for Model 1 

 in Table 5). The data points were weighted by their 

 sample size {n = 10-50). The Model 1 fit was un- 

 satisfactory for the whole size range, presumably 

 because each data point for the young larvae (n = 

 72) was a group mean, and the model was signifi- 

 cantly weighted toward the young larvae, causing 

 it to overestimate oxygen consumption for the few 

 large larvae {n = 17). Because the experimental 

 technique differed (i.e., respiration was measured 

 for groups of young larvae or individual older lar- 

 vae), I also fitted two separate curves. These curves 

 (Model 2) gave a good fit to the data (Table 5); the 

 Model 2 equation for younger larvae was used in the 

 present study. 



An alternate approach for estimating metabolic 

 requirements is to starve larvae of known size 

 (weight), determine the size-specific weight loss, and 

 convert the weight loss to calories. This approach 

 eliminates the need to restrict larval swimming ac- 

 tivity in a respiration vessel. Presumably the weight 



Table 5.— Parameters for equation Q = aw'' where Q is metabolic 

 rate in ^L Oj/h for northiern anchovy and w is their fresh dry 

 weight. 



'The sum of the residuals In Model 1 does not equal zero, thus the pro- 

 gram calculation of SE's Is biased. 



loss in caloric units would equal the loss due to 

 metabolic costs, excluding the metabolic cost of 

 attacking prey. Using this approach, I fed control 

 northern anchovy larvae ad libitum on Gymno- 

 dinium and Brachionus and starved the test larvae; 

 both groups were maintained in 100 L rearing tanks 

 at 15.5°C. Live standard length and dry weight of 

 groups of the same length were measured daily, as 

 described earlier in this Methods section, for 10-50 

 larvae sampled daily from each treatment. 



I calculated the caloric equivalent of northern an- 

 chovy tissue using the caloric values given by Hunter 

 and Leong (1981) for fat-free anchovy tissue, 4.129 

 cal/mg, and for anchovy lipid, 9.227 cal/mg. For ex- 

 ample, northern anchovy larvae weighing 25 ^g con- 

 tained 6 fig of lipid (unpubl. data: John Hakanson, 

 UCSD, Scripps Institution of Oceanography); using 

 the above caloric equivalents for 19 /jg of fat-free 

 tissue and 6 ng of lipid yields 5.36 cal/mg as the 

 energy equivalent of anchovy tissue. In a 20-d 

 laboratory experiment, Hakanson found that lipid 

 weight appeared to increase proportionally with an- 

 chovy weight, thus the caloric content of anchovy 

 tissue would be approximately constant for the age 

 range studied. Lipid content seems to be lower in 

 older northern anchovy larvae. The only other in- 

 formation I found was for 40-60 d-old northern an- 

 chovy where the caloric content averaged 4.9 cal/mg 

 (unpubl. data: John Hunter, Southwest Fisheries 

 Center). I used 5.4 cal/mg as the caloric value of an- 

 chovy tissue for larvae between 5 and 14 days of age. 



Evacuation 



Gut clearance times were determined for active- 

 ly feeding fish of various ages fed the rotifer and 

 copepods diets. Larvae were transferred from the 

 100 L rearing tank to a 10 L test tank. Because 

 northern anchovy larvae are sensitive to handling, 

 handling was restricted to one transfer. Transferred 

 larvae were kept in the test tank for 18 hours prior 

 to an evacuation experiment because injured larvae 

 usually die within 8-10 hours after transfer. First, 

 larvae were fed a low concentration of prey that had 

 been dyed with National Fast Blue (Laurence 1971). 

 After larvae had filled their guts, eating most of the 

 dyed prey, a known density of undyed prey was 

 added. Larvae were sampled at 5-min intervals, and 

 the time required for them to void their guts of the 

 dyed prey was determined. The number of prey in 

 the full guts was counted and converted to dry 

 weight. Evacuation rates are given as fjg prey 

 cleared through the gut per hour. Rates were re- 

 lated to fish size and to prey type. 



216 



