FISHERY BULLETIN: VOL. 79, NO 4 

 Table l. — Chemical composition (mean ± ai of Brevoortm tyrannufi, plankton food organisms, and B. tyrannus fecal pellets. 



Item 



Dry wt as 

 "o of wet wt 



Percent of dry weight 



Carbon 



Nitrogen 



Chitin 



Silicon 



Ash' 



C:N 



kcal/g dry wt 



kcal/g ash-free 

 dry wt 



B. tyrannus^ 

 Zooplankton^ mostly 



Acartia tonsa 

 Chitin 

 Phytoplankton": 



Ditylum bhghtwelli 

 Fecal pellets^; 



Zooplankton 



Phytoplankton 



33.4=1.80 

 10.86 = 0.20 



56.61=3.18 8.03 = 0.78 — 



40 05 = 2.34 10.91= .86 5.42 = 0.69 



39.37= 89 5,88= .17 — 



18.52= 19 3.04= 05 — 



10.34=1.20 1.85= 34 — 



9.29=1.34 .80= 11 — 



'May be an overestimate for plankton and fecal pellets: see text, 



^Dry weight was determined on experimental fish: other constituents estimated from measurements on Atlantic menhaden collected from Narragansett Bay R,l. 

 (Durbin e! al. unpubl, manuscr.), 

 ^Mean = ir of Experiments 1-3, 



"D. bhghtwelli from Experiment 7: mean = .r of four replicate determinations. 

 ^Mean = a from Experiments 1-3 (zooplankton) and 4-10 (phytoplankton). 



Table 2. — Food rations fed to 12 Atlantic menhaden. 



' Dry weight of zooplankton was measured in each experiment Phytoplankton dry weight was estimated from C measurements in each experiment, 

 and the conversion factor milligram C = 0.1852 (mg dry wt) (Table 1). 

 ^Basedon 12 fish = 1,212 g dry weight, 686,1 gC.97.3g N. and 7,560 kcal. 



7.8% of the Atlantic menhaden dry weight, but 

 only 0.26% to 2.56% of the estimated carbon con- 

 tent of the fish. 



Each of the 12 fish was assumed to have obtained 

 the same proportion ( 1712 ) of the plankton added to 

 the tank. This appeared reasonable since they 

 were of similar size and swam at the same average 

 speed during feeding ( Durbin et al. 1981), and thus 

 filtered similar volumes of water. 



In the zooplankton experiments the duration of 

 the feeding period was variable because of a 

 problem caused by crab zoeas present in low num- 

 bers in the plankton. It was apparent that the 

 sharp, 5 mm spines of these zoeas irritated the gill 

 rakers of the Atlantic menhaden, because in three 

 of four experiments attempted, the fish fed nor- 

 mally at first, then stopped feeding and began 

 shaking their heads and repeatedly flaring their 

 gill rakers. They stopped this unusual behavior as 

 soon as the zoeas were washed out of the tank with 

 filtered seawater. The fish were not satiated, since 

 trial experiments demonstrated that once the 

 zoeas were removed, the fish would feed readily on 

 salmon food. The number of zoeas was least in 

 Experiment 2, and in this instance the fish fed in 



their normal manner for as long as zooplankton 

 was made available, 8.1 h. Because of these prob- 

 lems, only total assimilation efficiency will be re- 

 ported from the zooplankton experiments. 



Feces Elimination 



In the phytoplankton experiments, feces began 

 to appear about 2.4 h after the beginning of feed- 

 ing (Table 3, column 2). The rate of elimination 

 continually increased during the feeding period. 

 The peak rates increased with increasing meal 

 size, and occurred during the first 1 or 2 h following 

 feeding (Figures 1, 2, 3). The elimination of feces 

 began an approximately exponential decline 2 or 3 

 h after the end of feeding (Figure 1). With the 

 exception of Experiment 6, these rates were all 

 similar (overall X = 36.6% /h, or 38.1%r/h if Exper- 

 iment 6 is excluded) (Table 3, column 8). The expo- 

 nential period lasted until about 14 h after the end 

 of feeding, after which elimination continued at a 

 low, nearly constant rate for the next 27 h (Figures 

 2,3). 



One consequence of the exponential elimination 

 of the feces was that different sized rations 



604 



