Mimulus guttatus exhibits substantial genetic variation in flower size and in other characters related to the mating system. What maintains this variation? The multitude of possible hypotheses can be classified into two broad categories based on the net effect of natural selection on variability. Selection may act primarily as a purifying force, eliminating variability that is continuously introduced by mutation (mutation-selection models). The alternative possibility is that selection actively maintains variation by some form of balancing selection. We are conducting four distinct, but complementary quantitative genetic experiments to distinguish these alternatives. The first three, the ratio test, the change in variance test, and the correlated responses test, are biometric analyses of selection experiments. The fourth analysis is a maximum likelihood estimation of genetic variance components. Results from the ratio test and variance component estimation indicate that mutation-selection models are untenable as a complete explanation for genetic variation in flower size. What fraction of genetic variation is maintained by balancing selection? A quantitative estimate awaits results from the other experiments.

Key words: flower size, Mimulus guttatus, mutation, quantitative genetics, selection