Strategies to improve mass selection in maize

Abstract

Mass selection is the oldest method of plant improvement and probably the most used method in plant breeding over time. Although it is relatively simple and easy to apply its efficiency of selection is often poor. The objective of this study was to assess alternative methods to increase genetic progress with use of mass selection. Use of alternative methods verified the efficiency of visual selection and strategies to improve the efficiency of mass selection. Two studies were conducted. First, we used a population of maize, in a field trial to verify the efficiency of visual selection made by ten evaluators, five in total areas and five in stratified areas, and checked the coincidence of plants visually selected with the weight of ear after the harvest. The second study was a computer simulation. That generated a population with 2025 individuals, with 1521 included in the selection. Populations were generated that considered different levels of heritability and allelic interactions. The environmental effects were simulated considering related and nonrelated errors. It was found that mass selection conducted visually within stratified areas was not efficient for the trait productivity; grain weight per plant should be determined for effective selection. The stratification of the area only improves efficiency when working with traits of lower heritability; when there is correlation of errors in the field; and the use of moving average did not increase the efficiency of mass selection.