TANK CULTURE OF TILAPIA 



39 



tended to foul tlie tanks. On tlie other liand, pel- 

 let ized feed, sucli as iM)ndfisli and trout feed, was 

 consumed by the adult Hsh with little wasta<re. 

 Rabbit feed had a hi^h percentage of crude fil)er, 

 which seemed to pass tlirough the tisli undigested 

 and left much residue in tlie tank. Alfalfa pellets 

 were somewhat less acceptable than the other feetls 

 mentioned, probably because of their large size 

 and tlieir higli tiller content. 



An experiment to evaluate the effects of differ- 

 ent, types of feed in relation to fiy production 

 was coiulucted at the Kewalo plant in September- 

 November, 1958. For this experiment, the three 

 redwood tanks (tanks 13, 14, and 15) were each 

 stocked with 32 males and 64 females. The adults 

 in tank 13 were fed trout feed, which we believed 

 to be high in nutritional value and which was rela- 

 tively expensive. Those in tank 14 were fed rab- 

 bit feed. As stated earlier, this feed was accept- 

 able to the tilapia and was also much cheaper than 

 tront feed. PreliminaiT trials at the pilot plant 

 indicated that i-elatively good fiy production was 

 possible witli this feed. The fish in tank 15 were 

 fed millrun. whicli was locally produced and the 

 least expensive of the feeds testetl. 



Table 12 gives the production per female for 

 the 3 months that the experiment was conducted. 

 A plot of the number of hy per female per month 

 is shown in figure 13. It is evident tliat the fish 

 that were fed trout feed (tank 13) produced the 

 greatest number of fry per female, while the fish 

 that were fed rabbit feed (tank 14) and millrun 

 (tank 15) had very low fi-y production. 



Table 12. — Production per female in three brood tanks, by 

 type of feed, September- November, 1958 



TROIT FEED 



:::::=i=z-r=z= 



Figure 13. — Effect of type of feed given brood .stock on 

 production per female. 



Temperature was not considered a factor in fry 

 production in this experiment, since there was very 

 little difference in water temperatures among the 

 tanks. Assuming that other environmental factors 

 were similar among the tanks, it follows from this 

 experiment that the use of a nutritionally balanced 

 feed is highly important in obtaining good fry 

 production. We are not certain, however, of the 

 long-range effect of such high-protein feeds on 

 tilapia, which are principally herbivorous. These 

 trout feeds are manufactured primarily for trout 

 culturists who intend to market the fish rather than 

 use them as brood stock — they may possibly be 

 detrimental to spawning fish. Schaeperclaus 

 (1933: p. 98) stated that the ovaries of trout — 

 uiider intensive artificial feeding — degenerate and 

 produce few usable eggs. It is likely that spawn- 

 ing stocks of tilapia held for extended periods of 

 time should be fed a diet more in keeping with 

 their natural food or be replaced after a year or 

 two by a new stock of brood fish grown under 

 more natural conditions. 



SALINITY AND SPAWNING 



During the summer of 1959, the Hawaii Divi- 

 sion of Fish and Game .started plans foi- construc- 

 tion of a tilapia hatchery in an area with free 

 accesss to brackish water (about 10 %o). At the 

 Division's request, experiments dealing with 

 spawning and growth of young tilapia in brackish 

 water were initiated at the Kewalo plant. 



Several investigators have observed and re- 

 ported that tilapia will spawn in a saline environ- 

 ment. Vaas and Ilofstede (1952: p. 11, 16) re- 

 ported that spawning occurred in a periwl during 

 which the salinity of the water ranged from 3 to 

 4.8 percent. They pointed out, however, that "Ac- 



