Abstract

Brood channel catfish used in this experiment were the first generation progeny of a base population originating from six different hatcheries in Georgia, Alabama, and Mississippi. Between and within family selections for body weight were performed at 40 weeks of age in the upward (W ⁺) and downward (W ^-) directions. Five-year old females and males from the two divergent lines were pair-mated in all four possible combinations to produce the first generation offspring. A randombred control line was also maintained to monitor the growth and survival in the four direct and reciprocal lines. The total of 16 spawns produced for this study were artificially hatched and the fish were reared in indoor tanks for 40 weeks and in cages for 48 weeks thereafter to determine the response to bidirectional selection for body weight. Body weight changes, measured as deviations from the control line, were about 20% in either direction at the end of the cage and tank tests. Total length difference between the upward and downward lines was 13%. Catfish produced from ♀ (W ⁺) × ♂ (W ^-) line were comparable in growth to those from ♀ (W ⁺) × ♂ (W ⁺) line but differed significantly from ♀ (W ^-) × ♂ (W ⁺) . The ♀ (W ^-) brood should not be used in fingerling production. Realized heritability estimates for body weight and total length were considerably lower than those estimated from full- and half-sib components of variance.