Bioassay responses of ewes to legume swards. I. Uterine weight response: Variability, calibration, and prediction

Abstract

Statistical analyses of sheep uterine weight responses to synthetic oestrogens and oestrogenic pastures showed that wet uterine weights were relatively less variable than dry weights. Body weight and uterine score at ovariectomy could be used as independent variables to reduce the variance of wet uterine weights. The best combination of scales to reduce variability in uterine weight was log (uterine weight) (Y axis) and log (body weight) or log (score) as covariates (X axis). For calibration of stilboestrol response log (uterine weight) regressed on log (dose (µg) + l) gave the most satisfactory result. This transformation gave common slopes for all sheep and assays but did not eliminate vertical displacement between assay calibration lines. Responses of virgin Merino and virgin crossbred ewes were similar. The paper discusses the precision of comparisons within and between assays and the limitations to improvements in precision. It also sets out guides as to the allocation of sheep in future assays to obtain maximum information per sheep.