Gene action controlling stability and adaptability in maize

Abstract

Limited knowledge about the genetic merit of maize landraces contributes to their little use in breeding, al-though some reports pointed them as a source of useful alleles. Repeated cultivation of landraces for many generations could make them a useful germplasm source to enhance yield stability. This study pretends to determine if such statement holds, and to analyze stability in terms of gene action involved. Twenty Argentine landraces were testcrossed to three US and two Argentine elite lines, and evaluated in 13 environments. Effects of general and specific combining ability for grain yield and ecovalence, as well as for regression coefficients of response to environments were also predicted. Non-additivity resulted more important than additivity for controlling ecovalence, and frequently resulted detrimental to stability. Although landraces contributed to ecovalence, their importance was less than that of lines. Testcrosses ́ adaptability to environments was variable depending on the line tester considered. Relative greater importance of lines ́ general combining ability for ecovalence than landraces’ general combining ability could be attributed to a successful selection for stability during the development of inbred lines. Sixty percent of landraces appeared more adapted to unfavorable environment, which suggests their higher level of rusticity. Argentine lines provided greater stability to their testcrosses than US lines. Correlation between grain yield and stability was mainly dependent upon non additive effects. More productive testcrosses tend to have greater ecovalence (less stability) due to specific combining ability effects. Inbred lines selected for greater general combining ability effect tend to provide their estcrosses higher stability.