**5. Perspectives: Toward breeding to for drought resistance**

Drought resistance is a complex trait controlled by many linked genes. Therefore, the probability of selecting drought-tolerant lines increases measurably as the percentage of adapted genotypes in gene combination increases. This progressively will increase the rate of genetic gains. The account below gives a summary of components (insights) of a breeding strategy for drought resistance for a practical breeding program.

#### **5.1 Identification of important secondary traits from multi-locational trials**

Analysis of genotype-by environmental interaction (GEI) was carried out in selected elite breeding material using nine characters including grain yield. The study has highlighted important secondary traits well expressed under field screening conditions because of good interaction of genotype × environment with other weather variables. Such traits; stay green, plant height, vigor (size), day to flowering, disease and pest response. The secondary traits can be used to increase progress made with primary traits (functional resistance) that is determined under growth chamber conditions. Therefore toward better integrated approaches, field screening of elite sorghum lines was carried out in hot spot locations for key sorghum production constraints; that is, drought, striga, pests and leaf spot diseases. This is because incorporated resistance to drought must have a positive effect on agronomic attributes as well as the quantity and quality of harvested products (study 1). Therefore the research captured a wide genetic variation of relevant plant characters to include in designing a breeding scheme. The GEI was approached through variance components, regression and multivariate methods (study 2 and study 3). Through these analyses, genotypes and environments were grouped, and stable genotypes were identified and ranked. The GEI for yield was attributed to various traits and genotypic correlations determined between plant height, panicle length, panicle width, days to 50% flowering, hundred grain weight, grain yield/ha, pest and disease response. The correlations ranged from negative, low, moderate, and positive and this guided selection. These analyses helped to design optimal crossing design (study 4).
