An experiment is described in which the possible yield benefits of mixing dwarf sorghum and beans were examined under intensive farming conditions. A ‘replacement series’ of pure sorghum, two-thirds sorghum/one-third beans, one-third sorghum/twothirds beans and pure beans was used at four plant populations. A high level of nitrogen was applied to the sorghum to eliminate the effects of nitrogen transfer from the beans.

Yields of the mixtures were up to 55% higher than could be achieved by growing the crops separately. As with some earlier maize/beans experiments (Willey & Osiru, 1972), it is concluded that these yield increases must have been due to a greater utilization of environmental resources. It is suggested that the main factors which made this possible were probably the different rooting depths of the two crops and their different growth cycles.

In agreement with the earlier experiments, mixtures which consisted of two-thirds sorghum/one -third beans gave greatest yield increases at high populations and also had a higher optimum population than the pure stands. However, no such trends were apparent in the mixtures which consisted of one-third sorghum/two-thirds beans.

For both species intra-specific competition was greater than inter-specific competition. In contrast to the earlier maize/beans experiments, the relative interspecific competitive abilities of the species changed with change in the initial proportions of species in the mixtures. In the two-thirds sorghum/one-third beans mixtures the sorghum was the more competitive species, whilst in the one-third sorghum/two-thirds beans mixtures the bean was the more competitive species. Also in contrast to the maize/beans experiments, these relative competitive abilities were not affected by change in theoverall plant population pressure.