Threonine requirements of broiler chickens: Why do published values differ?

Abstract

1. A statistical analysis of published data using restricted maximum likelihood showed that the age of the bird and the dietary crude protein concentration in test diets explained most of the variation in published threonine requirements. The sex of the bird and the main ingredients in the test diets were less important contributors of variation. The analysis resulted in a model that we tested in an accompanying paper. 2. Published data on threonine requirements are less variable than they first appear. There is no need for a plethora of studies on nutrient requirements other than those that test specific hypotheses, take a modelling approach and endeavour to explain requirements as units of nutrient per unit of production. 3. Models that use published data for predicting the responses of broilers to amino acids should first standardise the data by removing the variation due to factors such as age of bird, dietary crude protein content and the raw materials used in the ration.