Involvement of higher order interactions addressing complex polygenetically controlled inheritance of downy mildew [Sclerospora graminicola (Sacc.) Schrot] resistance in pearl millet [Pennisetum glaucum (L.) R.Br.]

Inheritance of downy mildew [Sclerospora graminicola (Sacc.) Schrot]resistance was studied using generation mean analysis in pearl millet [Pennisetum glaucum (L.) R.Br.]. Eleven basic generations, namely, P1, P2, F1, F2, B1, B2, B1F2, B2F2, L1, L2 and L3 of three crosses involving six diverse lines for downy mildew incidence were evaluated under artificial epiphytotic conditions over two environments. The downy mildew incidence was best fitting for digenic, trigenic and tetragenic ratios when fitted into classical Mendelian ratios demonstrating involvement of two or more genes. Digenic and trigenic interaction models were adequate in the case of crosses I and III respectively, to account for the total variability in generation means. Unlike severity, comparative estimates of gene effects over two environments were mostly consistent in all crosses for prevalence. Most of the epistatic and major gene effects were found significant in all crosses for both the disease traits. Non-allelic interactions particularly at three-gene loci viz., w (additive × additive × additive) and y (additive × dominance × dominance) in cross II and all trigenic interactions in cross III were predominant. Duplicate dominance (cross I) and complementary epistasis (crosses II and III) were observed for both the traits revealing inconsistency of gene effects over crosses. The gd1 (interaction of additive gene effect with e1) and gh1(interaction of dominant gene effect with e1) were significant in crosses I and II, indicating interaction of additive and dominance gene effects with environments. Thus a breeding method that can mop up the resistant genes to form superior gene constellations interacting in a favorable manner against pathotype I would be more suitable to accelerate the pace of resistance improvement.