VALIDATION OF A PORCINE GROWTH-MODEL WITH EMPHASIS ON CHEMICAL BODY-COMPOSITION

A growth model for pigs was validated. The model developed at the Massey University (New Zealand) was chosen because its assumptions and structure have been published completely. Besides this, its structure is representative for most other swine growth models. In an experiment, two strains of pigs received per strain: a protein sufficient ration at either (i) 33 MJ DE/day on average or (ii) 27 MJ DE/day on average or (iii) a protein deficient (50% of ARC recommendation) ration fed ad lib. Rations were composed by mixing two basal feeds, containing 25 and 7% crude protein respectively. At 28, 65 and 105 kg live weight, 4 pigs in every strain x treatment combination were slaughtered. Observed average daily gain, feed conversion ratio, chemical body composition at slaughter and deposition rates of protein and lipid were compared to values calculated by the model. In general, the model predicted the effects of the nutritional treatments adequately. Especially protein deposition rate and live weight gain were predicted quite accurately. However, above 6 5 kg live weight, the model underestimated lipid deposition rate in the protein-sufficient fed pigs by 20-30%. An increase in the ratio between lipid and protein deposition rate at the higher live weights was found. The model assumed no effect of live weight on this ratio. In contrast to the protein-sufficient rations, in the protein-deficient treatment, lipid deposition was overestimated considerably while protein deposition rate was predicted adequately. Results also indicated that the amount of water deposited for each unit of deposited protein is higher than predicted by the model. Main conclusion from the validation is that the parameter which accounts for the minimal ratio of lipid to protein is dependent on live weight.