The predictive value of quantitative genetic parameters in autogamous crops: Bias caused by intergenotypic competition. 2. F_∞-variance

Summary

Quantitative genetic theory for autogamous crops enables the estimation of the parameters m (F_∞-mean) and D (F_∞-variance) in the F₃-generation. With these estimated m and D a prediction can be made of the probability of finding superior inbred lines in the F_∞-offspring of the cross. The accuracy of this procedure is influenced by the correctness of the genetic model, by the magnitude of the error variance, and by the bias caused by intergenotypic competition, which is present in the environment in which the F₃ is grown and absent in the monoculture environment at which the selection is aimed, especially in a cereal crop like wheat. The influence of the latter was investigated by a special method of simulating segregating populations. In this method genotypes in the offspring of a cross are represented by varieties and segregating populations are composed by mixing the appropriate varieties according to the proper segregation frequencies. Growing the simulated population enables the estimation of m and D in the normal selection environment, while simultaneously growing the varieties in a large monoculture trial enables the calculation of ‘true’ values of m and D in the monoculture environment. Hence, a comparison is possible.

The second of this set of two papers presents the investigations on the influence of intergenotypic competition on the estimation of the parameter D. The correlation between the estimates from the selection environment and the calculated ‘true’ values from the monoculture environment was low. On the average the true values were overestimated, however in a few cases the monoculture values were significantly underestimated. Using the estimated m's and D's to rank crosses may lead to rather erroneous results.