Abstract
The historical and agricultural background of Barbados sugar-cane agriculture is outlined. Since the late nineteen thirties, four cycles of varieties have been grown. Regression analysis of cane yields (based upon extensive commercial data) permits the separation of varietal effects on yield from an underlying curved trend due to changing weather and husbandry. The three sets of newer varieties varied in the yield advantage achieved over their predecessors and there were evident genotype-environment (GE) effects. Collectively, the grain was large and the new canes added about 6·7 Mt to total crop over a 36 year period; they also ratooned better, a further contribution to economy of production. There is no evidence of change in sugar content. In discussion the biological and economic consequences of cane breeding, the importance of GE effects, implications for the interpretation of trials results and the need to widen the genetic base are topics touched upon.
