556 



Fishery Bulletin 89(4). 1991 



Q 

 <U 

 T3 



C 

 <D 



10 



2 



10 



20 



30 



40 



50 



60 



7.6 7.2 

 8.2 7°9 



7.6 8.2 B.3 



8°3 9°4 10°0 



9°1 10°3 11°0 



9.0 

 8°6 



8.5 7.9 8 



9°1 7°9 9 



9°4 6°2 9 



7.4 

 8°5 

 9°6 

 9°3 

 B°8 

 8°5 

 8°3 

 B°2 

 8°5 

 8?5 



7.4 

 B°0 

 9°3 

 9°4 

 9°4 

 9°7 

 9°6 

 9°0 

 9°4 

 9?1 



7.2 



7.6~* 



8.5 



9.1 



9.7 



10.8 



10.7 



9.1 



9.4 



9.0 



10  



20 



30 



40 • 



49 



B 



10 



20 



30 



40 



50 



10 20 30 40 50 60 70 80 88 



Distance across flume (cm) 



Distance across flume (cm) 



Figure 1 



Cross-sections of two flume sections showing water speed (in cm/s). Outlined areas were habitually occupied by the fish, and the en- 

 closed speed values (shown in bold) were averaged to obtain the swimming speed in the (A) slow and (B) fast sections. 



The open flume was composed of two connected 

 sections, the second section narrower and shallower 

 than the first, resulting in a more rapid flow. The first 

 section (the slow section) measured 244x88x60 cm 

 (1290 L), and the second (the fast section) measured 

 244x50x49 cm (600 L). A 20 hp pump at the down- 

 stream end of the flume recirculated about 50 L/s of 

 seawater through a feed pipe back to the upstream end 

 of the flume. Water flowed through a tapered expan- 

 sion section packed with gratings to even the flow. The 

 whole flume was inside a 15,000 L tank. Fresh filtered 

 seawater was added to the recirculating water in the 

 flume at a rate of 20-30 L/minute. Temperature was 

 maintained at 16.5-17.5°C. 



Cross-sectional current profiles in each section were 

 measured with a Marinco B-10 ducted impeller current 

 meter. Baffles upstream from each section were ad- 

 justed to make the current profiles as uniform as pos- 

 sible, and gratings at the ends of each section confined 

 the fish. Anchovy acclimated to the flume occupied 

 habitual areas within the two sections (Fig. 1). The cur- 

 rent measurements in these areas were averaged to 

 estimate swimming speed. The pump was operated 12 

 hours daily during "daylight"; the photoperiod was 

 13L:10D with half-hour "dawn" and "dusk" periods. 



Fish were anesthetized with quinaldine prior to 

 transfer to the flume. During transfer, the fish were 

 sorted and selected by size. A greater number were put 



in the larger (slow) flume section to make fish density 

 similar in both sections. Initially 1600 and 1000 fish 

 were put into the slow and fast sections, respectively. 

 The 12-hour swimming regime was built up over 8-10 

 days and then maintained for 2-3 weeks for each 

 treatment. 



Treatments 



Three levels of energy intake— high, zero, and low ra- 

 tions (in that order)— were imposed at two swimming 

 speeds, making a total of six treatment combinations 

 (Table 1). At each ration level, slow- and fast-speed 

 treatments were applied concurrently in the two flume 

 sections. The initial stock of fish was enough for the 

 high- and zero-ration treatments. The flume was 

 restocked for the low-ration treatments, but the 

 number of laboratory-acclimated fish available was low 

 (Table 1). 



Starting 8-10 days before the treatments began, food 

 type was changed to commercially obtained frozen 

 euphausiids (species not identified). Anchovy were fed 

 thawed euphausiids for 1 hour (high ration) or 0.5 hour 

 (low ration) during the end of the daily swimming 

 period. Food placed in 15 L containers of water above 

 the flume flowed into perforated tubes (2.5cm diam- 

 eter, 0.6cm perforations) stretched across the front of 

 the two sections. Screens were temporarily inserted 



