FISHERY BULLETIN: VOL. 70, NO. 3 



Table 4. — Estimated parameter values and 95% con- 

 fidence limits for density series separately and pooled 

 under particulate and filter feeding. 



Particulate feeding 



ArtemiaA't^er 



Intercept 



Slope (a) 



1 



2 

 Pooled 



46.8 (-316; 410) 

 181.3 (-1,004; 1,367) 

 assumed 



61.5 (53.5; 69.5) 

 58.5 (45.9; 71.1) 

 60.3 (56.3; 64.3) 



Filter feeding 



ArtemiaAi^er 



Asymptote (A'oo' 



Instantaneous feeding 

 rate (|3) 



22 20,322 (16,736; 23,908) 



66 15,384 (13,991; 16,778) 



112 22,050 (16,999; 27,102) 



Pooledi 23,788 (21,802; 25,775) 



0.0052 (0.0031; 0.0074) 



0.OO45 (0.0035; 0.0060) 



0.004O (0.00(29; O059) 



0.O036 (0.O032; 0.0042) 



1 The pooled array under filter feeding does not include the 30-, 60-, 

 and 90-min trials for 66 Artemia/\\\er. 



the data for the higher density levels with these 

 three trials removed. The estimated parameters 

 are shown, along with those for the lower density 

 levels, and also for the density levels individually, 

 in Table 4. Since the individual series under 

 particulate feeding did not have intercepts that 

 differed from zero, the parameters for the pooled 

 array were estimated with the intercept assumed 

 to be zero. Goodness of fit for the combined data 

 in both density groups was judged satisfactory; 

 in Figure 1 the calculated curves are compared 

 to the standardized trial means and standard er- 

 rors for each density series. The equations for 

 the two feeding modes can be stated as 



Np = 60.ZtD 



(3) 



and 



N, 



23,788 



I -[^ g-o.oosetVD j M) 



The asjrmptotic level, 23,788 Artemia, is in- 

 dicative of the maximum capacity of mackerels 

 at an average weight of 147 g. On the basis 

 of the wet weight of Artemia, 500 individuals/g, 

 maximum capacity would be 48 g, or 32% of fish 

 weight. This is about double the greatest weight 

 of fish food removed from the esophagus and 

 stomach, but the two kinds of estimate are not 

 necessarily inconsistent. As suggested earlier, 

 the weights of the digestive tract contents may 

 have underestimated the weights of Artemia in 

 the live state by 25 to 50% because of digestion 

 and leaching in Formalin. The estimate given 

 here is based on weight of live organisms. Since 

 digestion is going on during protracted feeding, 



112/1. (o) 

 66/1. (A) 



22/1 (•) 



2 /I. (A) 



1/I.(D) 



5 10 



60 

 MINUTES 



Figure 1. — The number of Artemia in digestive tracts 

 for different density levels in the water and for different 

 feeding durations. The lines for 1 and 2 Artemia/Mier 

 were calculated from equation (3) and the lines for 22, 

 66, and 112 Artemia/Mier were calculated from equation 

 (4). The symbols associated with each line show aver- 

 ages and standard errors for the trials in that density 

 series. Standard errors smaller than 450 are not shown. 



978 



