FISHERY BULLETIN: VOL. 73, NO. 4 



Table l.-Food placed in tanks, grams. 



'Tropical fish food. 



2Hatch from 0.4 g (adult) or 0.2 g (nursery) of eggs (Table 2). Silliman and Outsell (1958) found tfiat the hatch from 

 0.4 g of eggs weighed 0.125 mg. Test hatches of nauplii were not proportional in weight to the amount of eggs used 

 (Table 2), and since the total weight would be only 1/100% of the diet, no weight is indicated in the table. 



3This was combined with the Friday A.M. feeding in 35 out of 161 wk and with the Sunday feeding once. 



Table 2.-Artemia production tests. All 48-h hatches at 24°C in 

 800 ml 3% salt water. Counts from 20 samples for each test. 

 Samples were 0.3 ml, withdrawn by pipette from vigorously 

 stirred cultures killed in 0.75% formaldehyde, and replaced. 



lAII were commercial suppliers. 

 'Sample numbers times 800/0.3. 

 ^Includes shells (from hatched eggs) and unhatched eggs. 



number of fish recorded in any tank (343 guppies 

 in tank C during the last week of 3-wk period 65). 

 Dry food was placed on the surface of the water 

 and sank slowly if not eaten immediately (as oc- 

 curred with large populations). Frozen food sank 

 and was eaten as it thawed. Artemia nauplii were 

 hatched in 800-ml glass beakers (Table 2). The en- 

 tire water mass, including shells and unhatched 

 eggs, was poured through a cloth filter which was 

 rinsed with freshwater and then rinsed into the 

 fish tanks. 



Cleaning and Treatment 



Detritus including uneaten food (none in large 

 populations) was siphoned daily from the tanks 

 onto a cloth filter and the siphoned water returned 

 to the tanks. Once a week all the water was 

 removed from the tanks and one-half the volume 

 was replaced with tap water aged for 1 wk. At this 

 time the tanks and their equipment were 

 thoroughly cleaned, and the filter fiber and char- 

 coal were replaced. Also, fish in the adult tanks 

 were treated for 15 min in a 1:200 solution of a 

 commercial aquarium disinfectant "Fungistop."- 



Water Characteristics 



Water temperature in tanks A and D (Tables 3, 

 4; Figure 1) was recorded daily (Saturday excluded 

 during 3-wk periods 71-124). These end tanks were 

 chosen to reveal any temperature gradient that 

 might exist. Although there was a slight tendency 

 for tank D to vary from tank A (Figure 1), the 

 differences were mostly less than 1°C and are not 

 believed to have significantly affected population 

 growth. It will be shown in the section on oscilla- 

 tory fluctuation that deviations of population size 

 from the theoretical were not correlated with tank 

 temperature. 



-Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



874 



