Analysis of a fertility trial with Droughtmaster cows

Abstract

Data were obtained from a grazing experiment with 96 Droughtmaster cows on a Townsville stylo-spear grass pasture. There were two replicates of 12 treatments, which were factorial combinations of stocking rate (two levels), superphosphate application (three levels), and conservation of hay (two levels). Analyses were undertaken to determine (i) any significant treatment effects on conception, and (ii) whether any such effects were solely attributable to liveweight differences or to a direct, independent effect.

It is shown that the results cannot be treated on an annual basis, since the paddocks exhibit synchronous conception failures. The method of establishing this phenomenon is outlined. The analytical methods used were the analyses of variance and covariance, manifold contingency analysis, and multiple predictive analysis, both on per cow and per paddock means.

Only superphosphate significantly affected conception, this effect being associated only with the highest level of superphosphate. Seven of the analyses strongly indicated a direct effect of superphosphate not attributable to liveweight difference, though this could be almost obscured by cows attaining abnormally high or low liveweights during mating. The exception, covariance analysis of per paddock means, failed to achieve significance, but this is probably attributable to the greater sensitivity of per paddock means to outlying single values in a continuous model.