FISHERY BULLETIN: VOL. 78, NO. 4 



Table l. — Basic data on feeding experiments with a school of 12 Atlantic menhaden, total wet weight = 3,624 

 g. Oxycalorific coefficient used was 4.7 cal/ml of oxygen consumed (Kleiber 1961). Column numbers, in brackets, 

 are for text reference. 



weight of 302 g (range 248-346 g), and a mean dry 

 weight of 104 g. 



The food rations ranged from 0.79 to 7.8% of the 

 dry weight of the fish, and thus during the 7-h 

 feeding period the fish fed at rates equivalent to 

 0.11-1.11% of their dry weight per hour (Table 1). 



In a filter feeder such as the Atlantic menhaden, 

 the food ration obtained depends on the volume of 

 water filtered, corrected for the filtration effi- 

 ciency of the gill rakers. The volume filtered is 

 essentially cylindrical, with cross-sectional area 

 equal to the area of the fish's open mouth, and 

 length equal to the distance swum by the fish in a 

 unit of time. It is shown below that the fish swam 

 at about the same average speed during feeding, 



40 



u 



20 



Q 

 LlI 



^ 20 

 Q. 



in 



± 

 5 20 



Exp. 4 



J- I L 



J L 



Exp 9 



J L 



Exp. 7 



_L 



0600 



1200 



1800 



Tl M E 



2400 



hour s 



0600 



1200 



Figure l. — Meam voluntary swimming speed of a school of 12 

 Atlantic menhaden before, during, and after a 7-h period 

 (indicated by the heavy line on the x-axis) during which they 

 were fed, at a constant rate, a ration of the diatom Ditylum 

 brightwelli. Three representative experiments are shown, in 

 which total rations were: no. 4, 162.7 kcal; no. 9, 51.7 kcal; no. 

 7, 38.9 kcal. The 95% confidence limits were enclosed by the 

 symbols; horizontal bars indicate the duration of each experi- 

 ment. 



and since all fish were of similar size, each fish 

 filtered an approximately equal volume of water 

 during the 7-h feeding period. Thus we assume 

 that each fish obtained the same proportion ( 1/12) 

 of the plankton added to the tank. 



The behavior of the fish followed the same gen- 

 eral pattern in all experiments (Figures 1, 2). The 

 voluntary swimming speeds and respiration rates 

 of the fish were low during the initial measure- 

 ment and then abruptly increased severalfold over 

 the initial rates during feeding. When the input of 

 food was stopped, the fish rapidly filtered the re- 

 maining plankton from the water, decreasing 

 their swimming speed and respiration rate as the 

 plankton levels dropped. During the postfeeding 

 period there was a gradual return to prefeeding 

 activity levels and respiration rates, a transition 

 which was completed before the final two mea- 

 surements on the following morning. 



0.4- 



3 0.2 



a> 



S! 

 O 



6 02 



z 

 o 



(r 0.4 - 



oc 



a. 



CO 02 



0600 



1200 1800 2400 



TIME, hours 



0600 



1200 



Figure 2. — Mean respiration rate for Atlantic menhaden in the 

 measurements presented in Figure 1. The 95% confidence limits 

 are shown by vertical bars when they exceed the size of the 

 symbol. 



880 



