| A discrete, event-driven, and stochastic swine production model for personal computer use has been constructed and validated. This model is sufficiently flexible to be used for farms over a wide range of biological variables and management policies. Simulated data are compared to actual performance of 20 farms in the PigCHAMP information system. Correlation analyses of nine tested output variables were found to have R{dollar}sp2{dollar} between 87% and 99%. Results of a parametric (regression) test and a non-parametric (Wilcoxon's signed rank) test indicated that the model has the capability to simulate the actual farms with reasonable accuracy.; This model was then used to study the production efficiency of farrow-finish operations under various biological changes and management alternatives. A fractional factorial experiment with 17 factors was chosen to produce data of 512 simulated farms over a 2-year period after a pretrial warmup period of 10 years. The farms were high investment, confinement, environmentally controlled, with 120 breeding sow capacity.; Statistical analyses were performed to study: (a) the relative importance of the seventeen factors (b) the impact of improving each factor and (c) the connection between productivity and profitability. Strategy evaluation on a sample farm was also demonstrated.; Livebirths per litter and preweaning mortality were found to be the two most important factors for production efficiency. The Feed:Gain ratios of finisher and grower were ranked as the 3rd and 4th important factors for net return per sow per year. Consideration of post-weaning factors such as pig mortality and feed efficiency were necessary to fulfill the connection between productivity and profitability. The importance of factors related to sow longevity were also discussed. |
