Modelling the effects of thermal environment and dietary composition on pig performance: model testing and evaluation

Abstract

AbstractA deterministic, dynamic pig growth model predicting the effect of genotype, and the thermal and nutritional environments on food intake, growth and body composition of growing pigs was tested and evaluated against experimental data from the literature. Four sets of experiments meeting the necessary requirement of feeding the pigs ad libitum and reporting sufficient information on trial conditions were chosen to test the model. The parameters used in the model to describe the kind of pig were protein weight at maturity (Pm) the Gompertz rate parameter (B) and the ratio of mature lipid weight (Lm) to Pm. Values for Pm and B used to apply to the pigs in the four experiments were selected as those which gave the maximum daily gains equal to those reported at thermoneutral temperatures on diets not limiting in protein. The value of Lm was chosen as that which gave a value for food conversion ratio close to that seen in the experiment, again at a thermoneutral temperature and on a non-limiting diet. The model was run for each of the experiments from the given start weight until slaughter weight was reached. All pigs were assumed to have their desired bodily composition at the start of the experimental period, which is determined by their genetic descriptors and weight. From the conditions of the experiments, average daily gain (ADG) average daily food intake (ADFI) food conversion ratio (FCR) final body weight, body composition, average daily gains of each of the chemical body components and heat production (HP) were predicted. Generally as temperature increased or the crude protein content of the food increased, ADFI, ADG and the fatness of the pig decreased, whilst protein content increased. Quantitative differences between the model predictions and the observations, were probably due to the greater sensitivity of the model to temperature. This is likely to reflect the omission of long-term adaptation and acclimatization, or to incorrect estimation of the wetness of the pig’s skin. However, model predictions were generally in good quantitative agreement with the observed data over the wide range of treatments tested. This gives support to the value and accuracy of the model for predicting pig performance when the thermal and nutritional environments are manipulated.