FISHERY BULLETIN: VOL. 78, NO. 4 



predicted from respiration rate-swimming speed 

 relationships during long-term swimming at con- 

 stant speed (Brett 1964; Smit 1965; Muir and 

 Niimi 1972). Explanations for this phenomenon 

 include: 1) stress, which elevates the metabolic 

 rate, is reduced when the fish are occupied by some 

 activity such as swimming against a water cur- 

 rent (Brett 1964), or feeding, if the fish are nonag- 

 gressive (this study); 2) the intermittent swim- 

 ming of spontaneously active fish, accompanied by 

 frequent accelerations and changes in direction, is 

 hydrodynamically less efficient than the smooth 

 caudal locomotion of continuous swimming and 

 thus exacts a relatively higher metabolic cost 

 (Smit 1965). 



The increased respiration rates during feeding 

 consumed a significant fraction of the energy ob- 

 tained from the ration (Table 1, column 6). Energy 

 expenditures above routine during the postfeeding 

 period averaged 4.61% of the energy contained in 

 the food ration (Table 1, column 9). The increased 

 metabolic rate during and soon after feeding ap- 

 peared to be primarily due to the increased volun- 

 tary swimming speed. Swimming speed accounted 

 for 84.3% of the variability in metabolic rate dur- 

 ing feeding and 73.3% during the postfeeding 

 period (Table 2). Other factors which may affect 

 the metabolic rate as a result of feeding include 

 changes in excitability of the fish, and the 

 calorigenic effect of the food ration (SDA, the 

 "specific dynamic affect"). 



The excitability of the fish is in practice difficult 

 to measure. Qualitative observations of the behav- 

 ior of the fish and the degree of variability in their 

 swimming behavior (Figure 3) indicated that ex- 

 citability was not a significant factor contributing 

 to the elevated respiration during feeding, but 

 could be important during the postfeeding period 

 of restlessness. The latter evidently resulted from 

 the abrupt termination of the input of food when 

 the fish were not satiated. 



The cost of digestion and transformation of the 

 food, or SDA, has generally been measured as an 

 increase in oxygen consumption following feeding 

 (Kleiber 1961; Warren and Davis 1967). In several 

 earlier studies, in which fish were fed a single 

 meal over a brief period, the increase in oxygen 

 consumption peaked several hours after the meal, 

 then gradually subsided over a prolonged period 

 (as long as 2-3 d) to the prefeeding level (Muir and 

 Niimi 1972; Pierce and Wissing 1974; Beamish 

 1974). In these species, digestion of the food also 

 occurs over an extended period. The energy loss to 



SDA was generally estimated to represent about 

 12-16% of the energy content of the ration. 



The Atlantic menhaden results differed consid- 

 erably from these earlier studies. There was no 

 peak in oxygen consumption during the postfeed- 

 ing period, but instead a rapid and continuous 

 return of the metabolic rate to the prefeeding 

 level. This rapid return is consistent with the 

 rapid digestion rates observed for menhaden."* 

 Food was assimilated within 1-2 h after ingestion, 

 and approximately 80% of the food ingested dur- 

 ing the 7-h feeding period was digested and as- 

 similated within the same period. The amount of 

 energy expended above the routine during the 

 postfeeding period was larger in the larger ration 

 experiments (Table 1, columns 7, 8), which may 

 appear to indicate some effect of SDA. However; 

 voluntary activity levels were also higher in these 

 experiments. SDA is believed to be proportional to 

 ration size, but if the oxygen consumption at- 

 tributable to swimming activity. Equation (3) (Ta- 

 ble 2) is subtracted from the total respiration rate 

 during the postfeeding period, there was no rela- 

 tionship between ration size and the amount of 

 elevated respiration which can be ascribed to 

 SDA. 



Thus, while Atlantic menhaden may be as- 

 sumed to experience some respiratory costs re- 

 lated to SDA, the major part of these will be in- 

 cluded as a part of the total respiratory increase 

 during feeding. In practice it would be very dif- 

 ficult to distinguish SDA in the total metabolism 

 because of the overwhelming effect of swimming 

 speed on the metabolic rate. 



Equations (2) and (3) (Table 2) may be extrapo- 

 lated to zero activity to obtain an estimate of stan- 

 dard metabolism. These estimates, 0.036 mg 02/g 

 per h from Equation (2) and 0.029 mg Og/g per h 

 from Equation (3) are generally lower than those 

 reported from most other fishes which have been 

 studied (Table 3). Respiration rates which 

 menhaden sustained during feeding were also 

 high relative to those which can be sustained by 

 other species, and actually exceeded the active 

 rate (the maximum which can be maintained for 1 

 h) (Brett 1 964) of a number of species, including the 

 aholehole, largemouth bass, rainbow trout, and 

 tilapia (Table 3). Since the present study mea- 

 sured only voluntary respiration rates, the active 



■'Durbin, E. G., and A. G. Durbin. Assimilation efficiency 

 and nitrogen excretion of a filter-feeding planktivore, the Atlan- 

 tic menhaden Brevoortia tyrannus. Unpubl. manuscr. 



884 



