TANK CULTURE OF TILAPIA 



45 



Table 19. — -Size o/ young (prey) killed and consumed 61/ 



different juvenile groups {predator) 



lA juvenile group consisted of 10 flsli of approximately equal siiej 



Otlier experiments with starved and well-fed 

 juveniles showed that the degree of cannibalism 

 increases when the fish are starved. One par- 

 ticular group of 50 juveniles, averaging 31.8 mm. 

 (1.25 in.) in length, when well fed consumed 9 

 fry averaging 11.1 nmi. (0.44 in.) in 15 minutes. 

 This same group, when starved for a period of 3 

 days, killed or consumed 17 fry averaging 14.3 

 mm. (0.56 in.) in 12 minutes. "\Mien starved, the 

 juveniles were aggressive upon introduction of the 

 fry, whereas when well fed they were not par- 

 ticularly excited by appearance of the fry and 

 generally remained near the bottom of the tank. 

 An aggressive response by well-fed juveniles was 

 noticeable only when a single fry or group of fry 

 approached closely. 



These experiments indicated the importance of 

 keeping each compartment of the fry tank stocked 

 with young fish of luiifonn size. 



Another source of attrition, although not 

 considered so important as cannibalism, was 

 predation by dragonfly nymphs. These highly 

 predacious larvae, which occurred commonly in 

 the fry and assault-boat tanks, usually preyed on 

 the smaller fry. Considerable effort was made to 

 remove these nymphs. Chemical means of control 

 (salt, jwtassium permanganate, and pyridyl- 

 mercuric acetate, commonly called PMA) proved 

 ineffective; dipnetting them iiidixidually seemed 

 the most effective method. 



SUMMARY AND CONCLUSIONS 



This study evaluates the physical and biological 

 feasibility of producing bait-size tilapia in tanks. 

 Two facilities were used during the experiments. 

 The first facility, or pilot plant, constructed on the 



grounds of the Bureau of Commercial Fisheries 

 Biological Laboratory at Honolulu, Hawaii, was 

 used from October 1956 to July 1958 ; the second, 

 at the Kewalo Basin docksite, Honolulu, was used 

 from August 1958 until September 1959. Results 

 obtained at the pilot plant were of a preliminaiy 

 and general nature, but aided in planning the 

 more detailed experiments designed to examine 

 factors associated with variations in reproduction 

 and growth at the Kewalo plant. 



Brood tanks at the pilot plant were stocked at 

 the rate of 0.9 and 1.9 square feet of floor area per 

 individual and with a sex ratio of 2 9 : 1 3 . 

 Stocking of the Kewalo tanks varied from 0.5 to 

 1.6 square feet of bottom area jjer individual, and 

 the sex ratios (2 : 3 ) were 2 : 1, 3 : 1, 4 : 1, and 

 6 :1. 



Supplementary feeding of the tilapia included 

 rice bran, chicken mash, alfalfa i^ellets. rabbit 

 feed, and pelletized pond-fish feed and trout feed. 

 Generally, for the adults pelletized feeds were 

 much more satisfactory than finely divided mash 

 and bran. Young fish were fed finely ground 

 pondfish and trout feed. 



Chemical analyses to determine concentrations 

 of oxygen, free carbon dioxide, total bicarbonate 

 and carbonate alkalinity, and hydrogen-ion were 

 made routinely at weekly intervals to follow gross 

 changes in the environment within the tanks. 



Fry production at the pilot plant started in 

 December 1956, approximately 9 weeks after the 

 initial stocking. Xest -building activity was noted 

 only after 3 to 4 inches of calcareous beach sand 

 was placed in two of the brood tanks. Evidence 

 of spawning on the bare floor, however, was noted 

 when the one tank without sand was drained. 



At both rearing plants, as tiie young emerged 

 each day they were captured with a fine-meshed 

 dipnet and counted. They wei-e then i^laced in a 

 fry tank and segregated by size in different com- 

 l^artments to prevent cannibalism. 



Mortalities among the adults were attributable 

 to factors such as handling, disease, asphyxiation, 

 and possibly hydrogen sulfide poisoning. Highest 

 mortality rates among the young were attributed 

 to outbreaks of infectious disease and infestation 

 by ectoparasites. Infestation by protozoan 

 ectoparasites such as trichodinids and Chilodon 

 spp. was rather easily controlled by chemical treat- 



