Prediction of bull fertility

Abstract

Prediction of male fertility is an often sought-after endeavor for many species of domestic animals. This review will primarily focus on providing some examples of dependent and independent variables to stimulate thought about the approach and methodology of identifying the most appropriate of those variables to predict bull (bovine) fertility. Although the list of variables will continue to grow with advancements in science, the principles behind making predictions will likely not change significantly. The basic principle of prediction requires identifying a dependent variable that is an estimate of fertility and an independent variable or variables that may be useful in predicting the fertility estimate. Fertility estimates vary in which parts of the process leading to conception that they infer about and the amount of variation that influences the estimate and the uncertainty thereof. The list of potential independent variables can be divided into competence of sperm based on their performance in bioassays or direct measurement of sperm attributes. A good prediction will use a sample population of bulls that is representative of the population to which an inference will be made. Both dependent and independent variables should have a dynamic range in their values. Careful selection of independent variables includes reasonable measurement repeatability and minimal correlation among variables. Proper estimation and having an appreciation of the degree of uncertainty of dependent and independent variables are crucial for using predictions to make decisions regarding bull fertility.