Systematic and Random Sampling for Estimating Egg Production in Poultry

TIHE PROBLEM OF determining the ideal period for incomplete trapnesting to estimate egg production has been investigated during the last thirty years by several workers. Thompson (1933), Olson (1939) and Hays (1946) have given good reviews of the subject. If the heritability of egg production were high, mass selection of breeding stock would be an efficient method of attaining genetic improvement (Lush, 1946) and the accuracy of the production records for individual hens would be of primary importance. Daily trapnesting would provide accurate individual records. That mass selection is not an efficient method in breeding for improved egg production has long been recognized (Gowell, 1903). Lerner and Hazel (1947) estimated that the heritability of egg production based on individual hens was about 5 per cent in the flock that they studied. Greater breeding progress is possible by selection of sire-progeny and family groups than by simple mass selection. Thus, the accuracy of the average record for the group is most important, while that for individual records is secondary. Errors in individual records are only one source of error in the average record for the group. The other important source of error is the variation caused by true differences in productive ability among the individuals of a family or sire-progeny group. In the development and testing of inbred lines and crosses among them, it is again the accuracy of the record for the group which is of primary importance. Since trapnesting is costly, there is need for an accurate assessment of the relative sizes of the errors in group records arising from various sources. It is the purpose of this paper to report the relative sizes of errors arising from incomplete trapnesting of individual hens, and from variation among the hens of a group, under several schemes of incomplete trapnesting.