Abstract
A systematic programme of cowpea breeding was initiated at IITA, Ibadan, Nigeria, to identify potential dual purpose varieties and parents for use in breeding dual purpose germplasm. Generally, lower total fresh leaf yields were associated with early-to-medium flowering and with high total pod and total grain yields. Total fresh leaf yields were high in late flowering/maturing lines but they fluctuated with the season. One such line, Tvx 1948-0lF, gave the best combination of a high grain plus high green leaf/fodder yield. Other lines were good in one respect but not the other. IT83S-755-l, VITA-3, IT83S-951, for instance, were the highest green leaf/fodder yielders but were the poorest grain yielders. Results appeared to suggest that an initial strategy in a crossing programme should at least include the following steps: (1) an attempt to understand and as far as possible minimize the magnitude of a genotype-environment interaction in late flowering/maturing genotypes before they are identified as parents; (2) an attempt to transfer the yielding attributes from the high grain but low leaf/fodder yielders into the high leaf/fodder but low grain yielders and; (3) effectively cull poor combinations and/or segregations in early-generation crosses made on the basis of step (2) using suitable dual purpose potential mutli-trait selection indices.
