Abstract

Agriseeds company produces several hybrids yearly. These hybrids need to be evaluated for yield stability before release. In this study, fifty-eight newly developed hybrids were planted at five sites and evaluated for grain yield and other traits. The objective was to assess the stability of Agriseeds hybrids in Zimbabwe and to identify strategies of minimizing evaluation cost of hybrids in multi-locations. Across site, analysis of variance indicated significant differences (p<0.001) in grain yield, days to silking, days to anthesis and anthesis-silking interval on genotypes, environments and genotype × environment interactions (GEI). Stable hybrids were 10A3WH04 (6.7 tha^-1) and 10A3WH24 (6.7 tha^-1) while hybrid 10A3WH03 (6.5 tha^-1) showed specific adaptability. Since all the evaluation sites fell into one mega-environment, a few representative sites with a few replications will be ideal to capture much of the variance due to GEI. Furthermore, Agriseeds should not establish separate breeding programmes for these environments. Rather, suitable culling and discriminating environments must be captured in few sites to be utilized and these sites are Harare, Gwebi and any one of Kadoma, Matopos and Shamwa.

Key words: Zea mays L., Genotype × environment interactions, grain yield, bi-plots, stability analysis.