Selection in Small Populations of Chickens

Abstract

The results of a 10-year multi-selection experiment with White Leghorn and Fayoumi chickens are reported. The study includes 5 single-trait Leghorn lines, 3 single-trait Fayoumi lines and 2 two-trait Leghorn lines. The traits selected were high egg rate (2 lines), body size (2 up lines, 1 down line), egg size (2 up lines, 1 down line), high body weight—low egg weight (one line) and low body weight—high egg weight (one line). All lines were maintained by 8 sire matings except the high egg-rate lines maintained by 16 sires. Selection responses were measured as deviations from the Cornell Randombred Control. Even though full-sib and half-sib matings were avoided, inbreeding increased 1.6 to 1.9% per generation in the 16-sire lines and from 1.6 to 2.3% in the 8-sire lines. On the average, body weight declined 20 g, egg weight 0.6 g. and rate of egg production 2.3% for each 10% increase in inbreeding.Average selection differentials in each line were from ⅔ to one standard deviation per generation, and the response to divergent selection for body weight and egg weight was asymmetrical. Likewise, the correlated responses to selection were asymmetrical. The effective population numbers per line were large enough to give good realized heritability estimates except for the egg rate lines. The realized heritabilities were (1) for egg rate, 0.07 ± 0.04 (Leghorn), 0.03 ± 0.14 (Fayoumi), (2) for high body weight, 0.32 ± 0.04 (Leg) and 0.49 ± 0.09 (Fay), (3) high egg weight, 0.25 ± 0.03 (Leg) and 0.43 ± 0.07 (Fay) and (4) low egg weight, 0.59 ± 0.03 (Leg).The experiment with the two-trait lines was designed to examine the question of antagonistic selection, i.e., selection for two traits in opposition to the sign of the genetic correlation between them. The response to antagonistic selection was more restricted for the low body weight and high egg weight combination than vice versa. This resulted from the physical incompatibility of these two traits. Finally, the response to antagonistic selection did not seem to be rigidly controlled by the strength of the genetic correlation between the two traits.