FISHERY BULLETIN: VOL. 78, NO. 4 



the respiration rates was similarly narrow and 

 did not show any clear relationship with swim- 

 ming speed (Figure 4). 



Feeding Measurements (No. 2, 3, 4) 



As soon as food was added, the fish began "tast- 

 ing" the water by flaring their opercula and 

 swimming rapidly forward for a few seconds. After 

 1-2 min they began to feed, circuiting the tank and 

 swimming in the same direction at about the same 

 speed. Within each experiment, the mean volun- 

 tary swimming speed and the associated respira- 

 tion rate remained nearly constant throughout 

 the entire 7-h feeding period (Figures 1, 2). The 

 mean swimming speed was established during the 

 first 15 min of feeding and was related to the 

 amount of plankton in the water (Figure 4). 

 Within a range of about 1.5-4 /xg chlorophyll a/1 

 the voluntary swimming speed and respiration 

 rate of the fish were roughly proportional to the 

 chlorophyll a content of the water. Above about 4 

 ^tg chlorophyll a/1 however, the swimming speed 

 was independent of plankton concentration. In the 

 three high-ration experiments during feeding, the 

 mean speed ±95% was 41.3±1.5 cm/s and the cor- 

 responding mean respiration rate was 0.48 ±0.029 

 mg 02/g per h. The relationship between 

 chlorophyll a concentration and the voluntary 



o Initial and Final 



• Post - Feeding 



* Feeding 



0.3 — 



2 0.2 H 

 o 



I- 

 < 

 a: 



a. 

 </> 



tr 0.1 



009 

 008 — 

 0.07 — 



0.06 



10 20 30 40 



SWIMMING SPEED, cm- sec'' 



Figure 5. — Relationships between mean voluntary swimming 

 speed and mean respiration rate of a school of 12 Atlantic 

 menhaden. The x and 95% confidence limits of the initial and 

 final measurements are shown; functional regressions shown for 

 the feeding and postfeeding measurements are presented in 

 Table 2. Inset is an arithmetic plot of Equation (3); extrapola- 

 tions beyond observed data are dashed. 



swimming speed can be described by a rectangular 

 hyperbola [Equation (1), see Figure 4]. The zero 

 point for the curve was taken as the mean swim- 

 ming speed of unfed fish ( 12.2 cm/s) and the ap- 

 proximate concentration threshold (1 /u.g 

 chlorophyll a/1) (Durbin and Durbin 1975) of Z). 

 brightwelli at which Atlantic menhaden will begin 

 to feed. 



In these experiments the mean voluntary 

 swimming speed of the Atlantic menhaden during 

 feeding ranged between 29.3 and 43.4 cm/s ( 1.14- 

 1.68 BL/s). This represented a 2.2-3.3 fold increase 

 over the mean prefeeding routine swimming speed 

 (12.2 cm/s). Respiration rates during feeding 

 ranged between 0.221 and 0.538 mg 02/g per h and 

 were thus elevated 2.2-5.4 fold over the mean ini- 

 tial routine respiration rate (0.10 mg 02/g per h). 

 There was a good linear relationship between the 

 mean swimming speed during feeding and the 

 logio transformed mean respiration rate (Figure 5, 

 Table 2). 



Postfeeding Measurements (No. 6, 7, 8) 



When the input of food was stopped at the end of 

 7 h, the fish rapidly depleted the plankton remain- 

 ing in the tank. Within 5 min the swimming speed 

 of the fish decreased noticeably, and as plankton 

 levels continued to decline, the fish progressively 

 reduced their swimming speed (Figure 1, mea- 

 surement 5). Respiration rates also declined (Fig- 

 ure 2). Feeding usually became intermittent 

 within 15-20 min, and ceased entirely during the 

 next one-half hour, at which time the plankton in 

 the tank had been reduced to a negligible level. 



After the Atlantic menhaden had removed the 

 last of the plankton, they continued to "taste" the 

 water fairly frequently and were somewhat rest- 

 less, as though searching for additional food. Dur- 

 ing this postfeeding period, however, there was a 

 gradual return toward the prefeeding behavior. 

 Dusk arrived during or shortly after the second 

 postfeeding measurement. Although it was not 

 possible to photograph the fish during the third 

 postfeeding measurement at midnight, their ac- 

 tivity levels appeared to be very low as indicated 

 by their low respiration rates, and qualitative ob- 

 servations of their swimming behavior in the dim 

 light. 



The postfeeding measurements were spaced too 

 far apart to precisely define the time by which the 

 fish returned to their routine activity and 

 metabolic levels. In general the duration of this 



882 



