Repeatability of Dairy Cow Butterfat Records in New Zealand

Abstract

Most of the reported estimates of repeatability have been in the region of .40, and have been obtained without taking account of yearly effects on production. These have been considered in a recent analysis of some herd improvement records in New Zealand, using 55 57 records from 2436 cows in 30 Jersey herds over the dairy years 1951/2 to 1954/5. The records were age corrected by using multiplicative factors obtained from the data. Repeatability was estimated from the customary within-and between-cow analysis of variance, on a pooled within-herd basis. When year effects were neglected the resultant estimate was .49, but when a correction was made for them the estimate was .61. That the value .49 is higher than the oft-quoted .40 might be due to New Zealand''s general practice of grassland farming without any grain feeding, and to the seasonal pattern of calvings which are nearly all in early Spring. The importance of within-herd year to year variation accounts for the large increase in repeatability from .49 to .61, when yearly effects are considered.