Simulation of response of small self-fertilizing populations to selection for quantitative traits: I. Effect of number of loci, selection intensity and initial heritability under conditions of no dominance.

A genetic system was simulated in order to evaluate the effects of selection intensity and initial heritability on the genetic advance in a small population reproducing by selfing. A constant number of 40 individuals was measured in each generation. A quantitative trait was assumed to be controlled by 15, 30 or 60 independently segregating loci with equal additive effects, no dominance and no epistasis.It was found that the genetic advance in each generation, and the maximum possible genetic advance, expressed in actual units of measurement, were larger when fewer loci were assumed to control the trait, or when the initial heritability was higher. When the results were expressed on the basis of the initial phenotypic standard deviation, the genetic advance was smaller when 15 loci were assumed to control the trait than when more loci were assumed. An intense selection of 0.05 was most effective when selection was practised for a few generations. When selection was continued for more generations, a selection of 0.10 to 0.25 was found to be more effective. This occurred earlier when fewer loci were assumed to control the trait, or when the initial heritability was lower. The maximum possible genetic advance was attained in most cases by a selection intensity of 0.20 to 0.25.The additive genetic variance was decreased by selection at a faster rate, and its fixation occurred earlier, when fewer loci were assumed to control the trait, when selection was more intensive, or when the initial heritability was higher. The decrease of heterozygosity occurred at a faster rate when selection was more intensive or when the initial heritability was higher.