TANK CULTURE OF TILAPIA 



29 



In view of the results that were obtained, the shelf 

 collection method was abandoned. 



FRY REARING 



As fry production increased, it became a serious 

 problem to provide sufficient spare for tlie yonn<i 

 fish. Initially, the fry were ])]aft'(l in one large 

 fry tank. After a few weeks, cannibalisin became 

 widespread: the juveniles readily devoured the 

 newly collected fry when tliey were transferi-ed 

 to the fry tank. As a remedy, removable frames 

 covered with fine-meslied Monel screen wei-e in- 

 stalled in the tank dividing it into tliree compart- 

 ments. The fry and juveniles could then be segre- 

 gated by age and size. 



With the construction of a new fi-y tank with 

 six compartments and the acquisition of four as- 

 sault-boat tanks at the pilot plant, additional 

 space for rearing the young was available. The 

 procedure became standard to place the fry in the 

 fiy tank for about 4 to 5 weeks, after which time 

 they were transferred to assault-boat tanks. 



When fry were removed from the brood tanks, 

 they usually ranged in size from 7.8 to 13.6 mm., 

 with the average about 10.8 mm. Of a total of 154 

 fry ranging from 7.8 to 10.3 mm., about 26 percent 

 carried remnants of the yolk sac. The newly col- 

 lected fry, as mentioned earlier, generally paid lit- 

 tle or no attention to the prepared feed when first 

 offered it and only took it after about 2 weeks in 

 the fry tank. 



Vaas and Hofstede (1952) observed that young 

 tilapia fed on diatoms, unicellular green algae, 

 small Crustacea, and periphyton. Varying lunn- 

 bers of these organisms were present in tlie tanks 

 and undoubtedly constituted a major portion of 

 the diet of the fry during their first few weeks 

 of life after absorption of the yolk sac. The feces 

 of the young fry were usually bright green, in- 

 dicating that algae were a major constituent of 

 their diet. 



The necessity of utilizing all fiy-rearing space 

 available during periods of heavy fry production 

 occasionally forced us to overcrowd the fry tank 

 and freqiiently resulted in an outbreak of disease. 

 A criterion for determining when an overcrowded 

 condition existed was difficult to formulate, but 

 through experience we arrived at what we con- 

 sidered an optimum stocking density. The gen- 

 eral plan was to stock the fry-tank compartments 



with appi-oximately 200 fry per square foot of sur- 

 face area and crop each compartment at fre<iuent 

 intervals, removing the larger, faster-growing in- 

 dividuals; i.e., fish that were 20 nmi. (0.75 in.) or 

 larger. In this manner, cannibalism and disease 

 were kept to a mininniin. Also, through the neces- 

 sity of occasionally overcrowding the juveniles in 

 the assault-boat tanks, we learned that they were 

 nuich more tolerant to crowding tlian were the fry 

 and also that they were afllicted less fi-eriuently 

 with ectoparasites. 



The most critical pei'iod for the fi-y appears to 

 l>e the first 4 to 5 weeks of life, for it is in this 

 period that the fiy are most susceptible to in- 

 fectious diseases and ectoparasitic infestation. 

 Proper sanitary conditions in the tanks and ad- 

 herence to an optimum stocking density for a 

 given area seem to be the two most important 

 factors detennining survival and subsequent 

 health of the f it. 



LENGTH-WEIGHT RELATION OF YOUNG 



During the operation of both plants, it was 

 frequently necessaiy to estimate the size and 

 quantity of young fish on hand. In order to have 

 a convenient means of converting length into 

 weight and vice versa, the length-weight relation 

 for tilapia 17 to 73 mm. in length was determined 

 from body measiirements of 109 fish. The loga- 

 rithmic fonnula for the expression of this rela- 

 tionship is 



log W = 3.088 log L — 4.8935 

 where IF is the weight in grams and L is the 

 length in millimeters. A cui-ve demonstrating the 

 length-weight relation is shown in figure 6. 



WATER-QUALITY DETERMINATIONS 



Some chemical determinations were made 

 routinely once each week. Only oxygen measure- 

 ments were made at the pilot plant; but at the 

 Kewalo plant, additional properties, such as free 

 carbon dioxide, bicarbonates, normal carbonates, 

 and hydrogen-ion concentration were, also meas- 

 ured weekly in selected tanks. 



Oxygen Content 



Determinations of dissolved oxygen in the brood 

 and fn- tanks were made by the modified Winkler 

 method. Analyses were made once a w-eek at 



