Abstract
Considerable demand exists for blue and green eggs, especially in Southeast Asia. Although some chicken breeds lay eggs with shell colors ranging from light blue to dark green (olive), genetic lines producing uniform blue eggs or green eggs are preferred for commercial production. Our previous studies demonstrated the difficulties of selection on human scoring of the visual eggshell color (AveObs; 1=light blue, 2=blue, 3=green, 4=olive), and suggested to select on 3 colorimeter measurements (L*, a*, and b*), and on Shell Color Index (SCI, =L*-a*-b*). In the present study, these 4 parameters were the criteria of selection for blue-eggshell (Blue) group (high SCI, high L*, low b*, and low a*) and for green-eggshell (Green) group (SCI≈75, L*≈80, b*≈12, and low a*). The base population was generation G5 of an experimental line called BG, of black-bone chickens with AveObs ranging from 1 to 4. Generations G6, G7, and G8, progeny of 3 cycles of selection, were evaluated in this study. In the Blue group, the first selection cycle significantly changed the G6 means of the colorimeter parameters and of AveObs, and significantly reduced the colors phenotypic variance (Vp). The following two cycles of selection in the Blue group hardly changed the means but further reduced Vp, and the heritabilities. In the Green group, in line with high heritabilities, the three selection cycles led to significant changes from G5 to G8 in the means of the eggshell colorimeter parameters, and mean AveObs increased from 2.6 to 2.9. However, the Vp in the Green group remained almost as high as in G5, probably due to the polygenic control of the combinations of the two pigments, biliverdin and protoporphyrin, that jointly determine eggshell shades from greenish blue to olive. Selection targets derived from bivariate plots suggest that one additional cycle of selection may fix uniform Blue group (AveObs=2), whereas more cycles of selection will be needed to reach uniformity of the green color (AveObs=3) in the Green group.