Abstract:
This paper is the second in a series of studies on the genetic variation and covariation in growth and reproductive characters of the New Zealand Romney sheep. The data used were obtained over a period of 11 years (1955-1965 inclusive) from a random bred experimental flock maintained at Massey University, Palmerston North, New Zealand. Phenotypic correlations were estimated between the ewe's liveweights, oestrous performance at the hogget stage (from weaning to about 14 months), and her subsequent fertility (number of lambs born per ewe) at the first lambing (2-year-old) or over several lambings. These correlations were compared from the viewpoint of accuracy in predicting, prior to first mating, a young ewe's potential fertility over one or several lambings during her own lifetime. The results suggest that for this purpose, the predictive accuracy of the hogget characters is generally low but comparable with, and in some cases, substantially greater than, that obtainable from fertility records belonging to the dam of the young ewe. Three different methods were used to estimate the genetic correlations. In general, the pattern of genetic correlations, either within the hogget characters or between them and fertility of the ewe, is predominantly one of positive values ranging from moderate size to unity. In contrast, most of the maternal genetic correlations were negative. The relative efficiency of indirect versus direct selection for genetic gain in various characters was examined under different assumptions made with regard to the maternal effect. Indirect selection, based on the measurement of hogget (yearling) liveweight or number of hogget oestruses, for genetic gain in fertility of the Romney ewe is expected, on present evidence, to be more efficient than direct selection. Of the two characters, hogget liveweight is likely to find wider application because it can be more easily measured.