MORPHOLOGICAL AND PHYSIOLOGICAL MECHANISMS OF YIELD STABILITY IN GRAIN SORGHUM (SORGHUM BICOLOR (L.) MOENCH) ACROSS DIVERSE ENVIRONMENTS

The objectives of this research were to identify mechanisms associated with broad adaptation and yield stability in grain sorghum and to provide a focus for further studies. Yields of 3 stable and 3 unstable sorghum genotypes were evaluated across 14 environments using a modified Finlay and Wilkinson (35) stability analysis. Regression analysis was used to examine the behavior of yield components (plants/m(\u272), tillers/plant, heads/m(\u272), seeds/head, seeds/m(\u272) and 100 seed weight) of each genotype over different environments. Correlations were calculated for temperature and precipitation variables in each growth stage with yield and yield components. In three further experiments periodic sampling, bloom and maturity data were used to define the rate, length and pattern of grain filling for each genotype. Yields of stable and unstable genotypes were not greatly different in high yield environments, but the stable genotypes showed a 70% yield advantage in the poorest environments. Stable genotypes tended to maintain higher levels of all yield components in poor environments, thus maintaining greater yields. Seeds/m(\u272) was the yield component most strongly correlated with yield, though greater seed weight contributed much to the yield advantage of stable genotypes in poor environments. High temperatures between panicle initiation and bloom (GS(,2)) were associated with decreased length of GS(,2). Shorter GS(,2) periods were associated with fewer seeds/head, seeds/m(\u272) and reduced yields in unstable genotypes but not in stable genotypes. In the 14 environment study, unstable genotypes had lower mean seed weights than stable genotypes. Later experiments showed that this was not due to limited seed weight capacity, and that rate of fill (ROF) and length of fill (LOF) were independent. Though stable genotypes had greater LOF values, this could not explain the higher seed weights in the previous trials. It was concluded that: (a) Stability was more related to stress tolerance than to greater capacity for yield component compensation; (b) tolerance to high temperatures in GS(,2) was important to yield stability in this region; (c) seed weight maintenance under stress made an important contribution to yield stability and (d) further research on yield component reactions to stress is needed