A comparison of alternative selection indices for non-linear profit function in closed breeding herds of swine using Monte Carlo simulation
1998
Ieiri, S. (Kumamoto-ken. Agricultural Experiment Station (Japan)) | Furukawa, T. | Hirooka, H.
Seven different selection indices for non-linear profit function were compared using stochastic simulation in closed breeding herds of swine. The indices compared were; (1): a linear index constant over generations, (2): a linear index adjusted stepwise, (3): an optimum linear index for maximization of profit in the next generation, (4): an optimum linear index for maximization of profit at the last generation, (5): an optimum linear index for maximization of accumulated profit over the planning generations, (6): a quadratic index and (7): a desired gain index. For all the indices, the simulated population consisted of 10 boars and 50 sows, each sow having 1 male and 2 female progeny as candidates of breeding animals to reproduce the next generation. Genetic and economic responses over multiple generations were simulated considering non-linear profit function with growth rate and back fat thickens combined. The alternative economic values for the linear and quadratic terms in the profit function were assumed and these changed to reflect 3 different situations; the function A based on practical situation, the function B with a one-tenth economic values for the linear and quadratic terms of back fat thickness and the function C with ten hold economic values for them. When the degree of non-linearity of profit function was high, the linear index constant over generation (1) and the quadratic index (6) yielded considerably lower economic response. Relative efficiencies of the linear index adjusted stepwise (2) and the optimum linear index for maximization of profit in the next generation (3) depended on the degree of non-linearity of the profit function and choice of generation number
显示更多 [+] 显示较少 [-]