Estimates of genetic variances and interrelationships of traits associated with Striga resistance in maize

The choice of the selection method to be used in the genetic improvement of a crop depends upon the type of gene action controlling the characters under study. Design 1 crosses were made in an open-pollinated maize population, TZL Composite-1 C1. As a result, 320 full-sib families were generated and tested under artificial infestation of Striga at two locations, Mokwa and Abuja in Nigeria. Study of gene action involved in the inheritance of S. hermonthica resistance was done. Additive genetic variance was a major portion of the genotypic variance for characters like ear Striga rating (sigma 2a = 0.41+/-0.15 and sigma 2d = -0.20+/- 0.23), yield of infested plants (sigma 2a = 401564+/-172649 and sigma 2d = 57127+/-315131) and for the Striga syndrome rating (sigma 2a = 0.28+/-0.12 and sigma 2d = 0.03+/-0.21). However, the number of Striga plants supported by the host was controlled by non-additive gene action (sigma 2a = 75.69+/-76.64 and sigma 2d = 211.47+/-134.48). Grain yield was positively correlated to the number of ears harvested (rg = 0.99+/-1.12), but negatively correlated to Striga syndrome rating at 8 weeks after planting (rg = -0.92+/-0.93), ear Striga rating at harvest (rg = -0.88+/-1.28), and to a lesser extent, to Striga counts at 8 weeks (rg = -0.22+/-0.46) and at 10 weeks (rg = -0.09+/-0.75) after planting.Le choix de la methode de selection a adopter pour une amelioration varietale depend du mode d'action de genes qui controlent les caracteres etudies. Des croisements de type Design I ont ete faits dans une population de mais a pollinisation libre, TZL Composite-1 C1. Ainsi, 320 familles full-sibs ont ete generes et testes sous infestation artificielle avec un parasite de mauvaises herbes, Striga [Striga hermonthica (Del.) Benth.] a deux localites, Mokwa et Abuja au Nigeria. L'etude de l'action des genes impliques dans la transmission de la resistance au S. hermonthica ete faite. La variance additive representait une importante proportion de la variance genetique pour les caracteres tels que le syndrome du Striga sur les epis (sigma 2a = 0.41+/-0.15 et sigma 2d = -0.20+/-0.23), le rendement des plantes infestees (sigma 2a = 401564+/-172649 et sigma 2d = 57127+/-315131) et pour le syndrome du Striga visible sur la plante (sigma 2a = 0.28+/-0.12 et sigma 2d = 0.03+/-0.21). Cependant, le nombre de plants de Striga supporte par la plante hote est controle par une action de genes non-additive (sigma 2a = 75.69+/-76.64 et sigma 2d = 211.47+/-134.48). Les correlations genetiques entre les caracteres associes avec la resistance au Striga ont ete estimes. Le rendement etait positivement correle au nombre d'epis recoltes (rg = 0.99+/- 1.12), mais negativement correle au syndrome du Striga 8 semaines apres semis (rg = -0.92+/-0.93), au syndrome du Striga sur l'epis a la recolte (rg = -0.88+/-1.28), et a un degre moindre au nombre de plantes de Striga 8 semaines (rg = -0.22+/-0.46) et 10 semaines (rg = -0.09+/- 0.75) apres semis