Selection for High and Low Yolk-albumen Ratio in Chickens

Abstract

The correlated responses to the divergent selections for high and low yolk-albumen ratio in chickens on seven egg traits were studied. These seven traits, namely, egg weight, albumen weight, yolk weight, shell weight, egg length, egg width, and egg shape index were measured over seven generations. The following eight selected lines and an unselected control line were used in this study. A random-bred base population was classified into four groups according to egg weight (L, ML, MS and S). In each group, two lines were selected for the high and low yolk-albumen ratio (LL v. s. LS, MLL v. s. MLS, MSL v. s. MSS, and SL v. s. SS).The results obtained were as follows:1) The population means of egg weight regressed to the base population mean on first and second generations. It was concluded that these phenomena resulted from the effects of the classification on base population which was classified by egg size. Particularly, the calculated means of egg weight in SS and SL lines were demonstrated to increase, and decrease in LS and LL lines, respectively, on earlier generation. These lines had the smallest and largest mean egg weight, respectively, in the initial generation. Similar effects were observed in the response trends on albumen weight in first and second generations of these lines.2) The response trends in the yolk weight were observed to increase on all selected lines through seven generations, and these responses were more clearly observed in selected high lines. In the upward selection lines, the responses per generation for yolk weight were 0.50g, 0.44g, 0.28g and 0.21g in SL, MSL, MLL and LL, respectively, and were all statistically significant.3) The correlated responses of the albumen weight and egg weight wese not observed until the second generation, because of the effects of the selection caused by the egg weight in the base population. However, the both traits were demonstrated to increase in the selected low lines, except for LS line, and to stagnate or to decrease in selected high lines in the succeeding generations. The greatest responses to the selections per generation were obtained in SS and LL lines, i. e. 0.48g (SS), -0.74g (LL) for albumen weight and 0.84g (SS), -0.60g (LL) for egg weight, 4) Although the shell weight and egg width changed with the selection, the correlated responses were relatively small and showed a similar tendency as that of egg weight. The egg length and egg shape index showed differences among the selected lines within generations.5) The mean differences of the yolk weight between selected high and low lines were observed to increase slightly for seven generations. In contrast to the yolk weight, the differences in the albumen weight were demonstrated to increase markedly through generation. The differenes per generation for the albumen weight were 0.40g to 0.73g, and were all statistically significant.6) The average estimated heritabilities in egg-component traits, computed with both analyses of variance and daughter-dam regressions within lines and generations, were not parallel to the responses to selection. No differences among selected lines were found on the phenotypic correlation coefficients between egg traits within generations. On the other hand, the genetic correlations were estimated to be -0.29 to 0.61, -0.62 to 0.17 and -0.45 to 0.35 between the yolk-albumen ratio and yolk weight, albumen weight and egg weight, respectively. These coefficients had large variations within lines and generations. From the results of the analyses of varimax on these genetic correlation with matrices, it was suggested that the lines of the different genetic constitution with respect to the egg component weight were produced by the selections for the yolk-albumen ratio.