SummaryResponse to selection primarily for lint yield among spaced plant populations of four American and one African variety of Gossypium hirsutum L., established in 1965—6 at Kimberley Research Station, was studied in three experiments conducted between 1966 and 1970.In the first experiment, in which single-hill micro-plots of 3–5 plants per hill were used, the mean lint yield of progenies was larger than predicted from single plant heritabilities. In the second experiment, in which single-row plots were used, response to selection for lint yield was generally less than expected. In the third experiment, in which three-row plots were used, response to selection for lint yield was again less than expected and less than manifested in the second experiment. However, in all experiments responses to selection for lint percentage, lint quality properties, and height were close to expectations while some individual progenies substantially out-yielded their unselected parents.The progressive decline in response to selection for lint yield from the first to thethird experiment, and the discrepancies between the actual responses and the expected responses for lint yield in all experiments, were interpreted mainly on a basis of intergenotypic competitive effects differing in sympathy with plot size and plant disposition.It was concluded that the objective methods of selection practised were successful when judged on plant breeding criteria. It was also concluded that intra-varietal selection is an important technique by which genetic improvement can be achieved at regional breeding centres with limited resources.
Single plant selection for improving root rot disease (Phytophthora medicaginis) resistance in Chickpeas (Cicer arietinum L.)