Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/6/2008
Publication Date: 3/1/2009
Citation: Wardyn, B.M., Edwards, J.W., Lamkey, K.R. 2009. Predicted Gains from Inbred-Progeny Selection Is Inferior to Half-sib Selection for Two Maize Populations. Crop Science. 49:443-450. Interpretive Summary: Historically, genetic theory has been used to argue that corn breeders could select among inbred lines in order to improve the performance of hybrids. However, in pratice, the performance of inbred lines has been very poorly related to the performance of hybrids. We have used new results from quantitative genetics to predict that selection for inbred performance is in fact an inefficient way to improve hybrid performance. While most corn breeders do not rely the performance of inbred lines to develop improved hybrids, no genetic model has existed that could explain this observation theoretically. We have found a genetic model that much more accurately predicts the effects of selection among inbred and noninbred individuals. The impact of identifying such a model will provide new genetic insights into the relationship between inbred and hybrid performance. A better understanding of this relationship will benefit the breeders of hybrid crops, as well as the producers and consumers who benefit from improved varieties.
Technical Abstract: Historically, inbred progeny selection has been promoted as an effective means of improving both inbred and outbred performance, and in some cases superior to other methods. The predicted effectiveness of inbred-progeny selection was based on several factors, the most common being the proportion of additive variance among selection units. Advances in theory have allowed us to use a full single-locus model that correctly accounts for the effects of dominance and inbreeding in predictions of gain from selection. We developed predicted gain equations using the full model and a reduced model, which contained only the additive genetic variance in the numerator, for two maize populations, BS13(S)C0 and BSCB1(R)C13. Our results indicated that outbred progeny selection is superior to inbred progeny selection for plant height and grain yield when the response unit is the outbred population. As the response unit becomes more inbred, inbred progeny selection is superior for plant height in both populations and more effective for grain yield in BS13(S)C0. The contribution of a negative value of the covariance parameter D1 to the numerator of predicted gain reduces predicted gains for inbred progeny selection in the full model in comparison to the reduced model when D1 is ignored. In addition, previous works failed to point out that the contribution of additive variance to the numerator of the genetic gain equation is twice the contribution to the denominator for half-sib selection, whereas they are equal for inbred-progeny selection.