42 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



sault-boat tank was stocked with 3,000 iry and 

 fed exclusively with wheat white middlings, a 

 locally available livestock feed, while the lot of 

 3,000 fry in the space-growth experiment was fed 

 trout feed. 



The weekly average lengths and weights of a 

 random subsample of 25 fish from each of the two 

 lots, for the 12-week period of the experiment, are 

 presented in table 15. Regression lines were fitted 



Table 15. — Weekly average length (mm.) and weight 

 {g.) and average absolute grmvth rates in random 

 subsamples of tilapia from 2 lots of 3,000 fry reared 

 on different diets, 1959 



Item 



Experiment began. 

 Experiment ended.. 



Initial length 



Initial weight - 



1st week: 



Length.. 



Weight.. 

 2d week: 



Length.. 



Weight.. 

 3d week: 



Length-. 



Weight.. 

 4th week: 



Length.. 



Weight.. 

 5th week: 



Length -- 



Weight-. 

 6th week: 



Length.. 



Weight. 

 7th week: 



Length-- 



Weight.. 

 8th week: 



Length.. 



Weight-. 

 9th week: 



Length-. 



Weight-. 

 10th week: 



Length.. 



Weight. 

 11th week: 



Length-. 



Weight. 

 12th week: 



Length. 



Weight. 



Average absolute growth rate 2.4 mm 



to the growth obtained in these tanks. The regres- 

 sion coeflBcient for the lot fed trout feed was 2.i 

 mm. per week while that for the lot fed wheat 

 white middlings was 1.9 mm. per week (fig. 15). 

 An analysis of variance of the results disclosed a 

 significant difference in growth rates between the 

 fish in the two tanks (7^ = 31.09, P<0.01), indicat- 

 ing that quality of the feed is highly important 

 where fast growth rates are desired. 



T — r 



1 — r 



n — I r 



> TROUT FEED 

 • WHEAT WHITE 



Y-12.96S«2.4aSX 



V- 12.686* 1.902X 



-J L 



J 1 L 



J I 



6 7 8 9 

 TIME IWEEKSI 



II 12 13 U 



Figure 15. — Regression of length on time for two lots of 

 fish reared on different diets. 



SALINITY AND GROWTH 



It has been reported by Vaas and Hofstede 

 (1952) and Brock (1954) that T. mossambica will 

 spawn and tlie young will grow in salt water. To 

 determine growth rates under varying degrees of 

 salinity, preliminary experiments were conducted 

 at the Kewalo plant laboratory using 30-gallon 

 aquariums. 



Five 30-gallon aquariums were used in the 

 experiment: two containing fresh water, two 

 brackish water (salinity of about 16Voo, and 

 one sea water. Each aquarium was stocked with 

 200 fry. Those fish to be tested in brackish water 

 and in sea water were acclimatized to sea water 

 for a period of 24 hours before being placed in 

 the aquariums. The fish in one tank in each pair 

 of the fresh-water and brackish-water aquariums 

 were fed wheat white middlings, while fish in the 

 other two tanks and in the sea-water tank were 

 fed trout feed (starter). 



From the first week, a large number of deaths 

 occurred in the brackish- and sea-water tanks and 

 the deaths continued to occur for several weeks. 

 At first, the dead fish in each of the tanks were 

 removed and counted each morning and replaced 

 with an equal number of individuals of similar 

 size from a reserve stock which was held in sea 

 water in another aquarium. Because of the high 

 mortality rate and the low growth rate in all the 



