XVI International Botanical Congess

A wide range of genetic patterns, from major genes and blocks of co-adapted genes, to multiple dispersed loci, was found in numerous studies on artificial selection. Much less is known about the genetic basis of adaptation in natural populations. Two alternative concepts have been proposed. The 'gradual divergence model' assumes that adaptation is due to accumulation of many additive genes with small effects whereas the 'genetic revolution model' postulates strong epistatic interactions of fitness genes. It is not easy to distinguish between these alternatives. If controlled crosses are available, one may evaluate the contribution of epistasis using known biometrical tests. Direct estimation of epistatis using QTL mapping analysis may serve an important complementation for the biometrical approach. We found that, in contrast to the general belief, epistasis may be helpful in discriminating between alternative hypotheses about the trait genetic control, in both single- and multiple-trait analysis. The advantages of multivariate QTL analysis in testing various hypotheses concerning epistasis and quantifying corresponding effects will be presented in the talk.