COMPOSITE VARIETIES, AN OPTION FOR USING THE DIVERSITY OF MAIZE LANDRACES

Abstract

Maize (Zea mays L.) landraces have been very valuable sources of genes for breeding programs; nevertheless, efforts aimed at their improvement in the microregions where they originated have been scarce. One proposal is the formation of composite varieties; however, there are no studies documenting their potential, especially when the base material consists of outstanding landraces. Thus, the objectives of this study were to evaluate the agronomic performance and grain yield of three composite varieties compared to their parents and commercial controls and based on that, to assess the viability of this type of material. Three experiments were designed, one with white grain varieties, other with yellow grain varieties and another with blue grain varieties, each included three recombination cycles of the corresponding composite variety, the parental landraces and at least two controls. Materials of each experiment were evaluated under a randomized complete block design with three replications in three localities of the microregion of origin of the landraces. Days to silking, ear height, stalk and root lodging percentages, grain yield and several of their components were recorded. Combined analysis of variance across localities and mean comparison test were applied. Composite varieties maintained and combined favorable traits of the parent populations and surpassed them in yield, with gains between 1.8 and 10.6 % (average across recombination cycles), and showed less affectation in grain yield across environments; in addition, composite varieties equaled or exceeded the yields of the controls. It is concluded that the formation of composite varieties from outstanding landraces represents an adequate option for landrace utilization in microregional breeding programs.