Sorghum Hybrids for Low‐Input Farming Systems in West Africa: Quantitative Genetic Parameters to Guide Hybrid Breeding

Abstract

ABSTRACTThe development of sorghum [Sorghum bicolor (L.) Moench] hybrids with Guinea‐race parents is a promising approach to increase yields in West Africa (WA). There is heretofore no quantitative genetic information about the genetic value of new hybrids and their parents, or about the efficiency of alternative selection methods for targeting yield performance under the predominantly low‐input, P‐deficient production conditions. This study aims to estimate the heterosis levels attainable by new Guinea‐race hybrids and the combining abilities of the current suite of parents with partial‐ to full‐Guinea‐race backgrounds of contrasting geographic origins, and to determine the effectiveness of direct selection under low‐P (LP) relative to indirect selection under high‐P (HP) conditions. Single‐cross hybrids were evaluated in 2015 and 2016 for yield under both LP and HP conditions at two locations in Mali. The hybrid yields were substantially superior to farmers’ local Guinea‐race varieties, with 20 to 80% higher means in both LP and HP environments. Average midparent and better‐parent heterosis estimates were, respectively, 78 and 48% under HP, and 75 and 42% under LP. Direct selection for yield under LP was predicted to be 20 to 60% more effective than indirect selection under HP conditions. The combining ability estimates provided initial insights into the potential benefit of germplasm from more eastern and humid regions of WA for developing a restorer pool distinct and complementary to the female pool from Mali and surrounding countries. Substantial yield superiorities of hybrids over local varieties suggested that a hybrid breeding strategy based on Guinea germplasm could help improve WA smallholder farmers’ livelihood.