The use of genetic algorithms for optimizing age structure in breeding populations when inbreeding depresses genetic gain through effects on reproduction

Abstract

AbstractA genetic algorithm (GA) was used to find optimal male and female age distributions in a natural mating system that maximizes cumulative response to mass selection over a 20-year time horizon for the case where inbreeding affects reproduction at 0·0 (F-0) and 0·1 (F-10) per 0·1 inbreeding coefficient. Twenty breeding female population sizes were considered ranging from 25 to 500 breeding females distributed across five age groups. Loss of response due to inbreeding effects on reproduction ranged from 19.4% and 15.5% in small breeding female populations to 2.5% and 5.2% in large breeding female populations when number of males was fixed (FX) or optimized (OP), respectively. OP resulted in an increase in response over FX ranging from 0·0 % to 69.3% for F-0 and 0·0 % to 77.6% for F-10. The potential loss of genetic gain that resulted from ignoring the inbreeding effects upon reproduction when they really existed ranged from 0·1 % to 44.6%. The potential loss of genetic gain that resulted from including inbreeding effects upon reproduction when they did not exist ranged from 0·1% to 3.9%. Optimal male and female age structures depended upon breeding female population size, the number of breeding males and inbreeding effects. Ignoring inbreeding effects upon reproduction may result in over estimation of response to selection. Use of a GA allowed accounting for complex relationships in the optimization.