Modelling the phosphorus intake, digestion, utilisation and excretion in growing and finishing pigs

Aim: 

The overall aim is to extend an existing model of pig growth and food intake in order to be able to predict of nutrient excretion (namely N and P) under different scenarios of feeding in growing and finishing pig systems.

Objectives:

1. To extend a current model of pig growth and food intake that predicts performance of pigs under different environmental and health conditions.  The extension will be towards the prediction of excretion of nutrients by these animals.
2. To describe P content and supply of the foods offered to pigs in adequate terms (available P), by taking into account the particularities of P digestion, including its relation to dietary Ca; this is equivalent to adding a P digestion extension in the above model.
3. To validate the developed model with data on pig P excretion that has not been used for model development.
4. To test for the consequences of difference scenarios of P supply (e.g. continuous vs. phase feeding, use of co products) on the excretion of nutrients by growing pigs.
5. To make recommendation to the industry about feeding strategies that would be expected to reduce the excretion of N and P from growing pigs, without penalising their performance.

Summary of findings:

The overall aim was to develop a model of Phosphorous (P) intake, digestion, utilisation and excretion in growing/finishing pigs, and use it to investigate the consequences of different P management strategies. Initially, a dynamic, deterministic model was developed (Chapter 2). It was able to predict the digestible (digP) requirements of pigs of different genotypes and stages of growth, as well as the consequences of different dietary contents of P, Calcium (Ca) and exogenous phytase. The model was also able to predict the excreted amounts of soluble and insoluble P. Subsequently (Chapter 3) the model was evaluated against independent data and a sensitivity analysis of its predictions to model parameters was undertaken. Model outputs were most sensitive to the values of the efficiency of digP utilization and the non-phytate P absorption coefficient from small intestine. The model predicted satisfactorily the quantitative pig responses, in terms of P digested, retained and excreted, to dietary variations. The model performed well with ‘conventional’, European feed ingredients and poorly with ‘less conventional’ ones, such as DDGS and canola meal. In Chapter 4 the model was converted into stochastic, by introducing variation between pig digP requirements and the consequences of two strategies were investigated (phase feeding and sorting). The former was more effective in reducing P excretion than the latter. Finally the model was extended to include uncertainty in feed composition (arising from variability in ingredient nutrient content and mixing efficiency) to investigate how this would affect the outputs of the model. Due to the assumptions made, uncertainty about feed ingredient composition contributed more to performance variation than uncertainty regarding mixing efficiency. When uncertainty about both feed composition and pig characteristics was considered, it was uncertainty about feed composition rather than pig genetic characteristics that proved to have the dominant influence on variability in pig performance.