In nine breeding trials with cocoa in Ghana the experimental trees were interplanted with temporary trees of a standard variety as ‘calibrators’. The calibrators were removed after 2–2½ years before there was strong competition with the permanent trees. Trunk diameters of the calibrators were used as covariates in analyses of trunk size and yield of dry cocoa up to 7 years after planting, giving reductions in standard errors of differences between progenies of up to 25% in analyses of size and 15% in analyses of yield.

Uptake of nitrogen, phosphorus and potassium was measured for irrigated and unirrigated crops of pearl millet grown at Hyderabad, India. Irrigation slightly increased the concentration of nutrients in the crop and this, together with a higher dry weight, resulted in the nutrient content at harvest (day 82) being approximately twice that of the dry crop. Nutrient uptake by the latter ceased at about day 40 whereas the irrigated crop continued to take up nutrients for a further 7–21 days. However, both crops showed the importance of redistribution of nutrients from leaves and stems during the grain filling. The mean rate of nutrient uptake per metre of root per day (inflow) fell throughout growth in contrast with the rate of nutrient uptake per plant, which reached a maximum at about days 26 to 33. These findings are discussed in relation to possible strategies for fertilizer placement.

In trials of seedling cocoa in Ghana, biases in the estimates of yield and tree size arising on the perimeters of experiments and on the interfaces of unguarded plots were strongly correlated, characteristic of progenies and stable across environments, but were inconsistent for black pod losses. No variate showed progeny × site interactions, but those of Caribbean yield data were interpreted by principal components analysis. The results show that 25-tree plots may be used for initial tests of new varieties and, as long as the environmental variation does not exceed that of the Ghanaian trials, at single sites.

Reducing the leaf areas of different maize hybrids at the time of female flowering resulted in a reduction of up to 70–90 g/plant of grain yield, with a parallel deterioration in the health of the stalk pith (due to infection by Fusarium sp.) and an increase in lodging. However, beyond a certain limit defoliation caused a loss in fertilization and a further decline in grain yield, which was accompanied by an improvement in the health of the pith and less lodging. There were considerable differences in the reactions of six genotypes to the defoliation treatments.

Two experiments were conducted on timing the application of N to maize on Njala upland soils in Sierra Leone. Application of all the N (urea) at planting or thinning gave lower leaf N values and relatively poor yields compared to split applications. Lower grain yields in the minor season might have been due to poor and inadequate rainfall. Basal application of SCU does not seem to release enough N for maximum grain yield of maize. Total N (urea) in the major and minor seasons may be phased in two splits, viz. ½ N 30 DAP + ½ N 50 DAP or ⅓ N at planting + ⅔ N 25 DAP, respectively.

Sweet potato (Ipomoea batatas L.) virus disease (SPVD) causes severe disease symptoms of various combinations of leaf strapping, vein-clearing, puckering and stunting. The disease is widespread, especially in Africa, and yield is adversely affected, though there are no figures to quantify this effect nor is it known how yield is affected. Loss of tuberous root yield due to SPVD was studied in sweet potato showing symptoms of the disease. Slips were obtained from plants with and without symptoms and two yield trials in 2 years were conducted by planting them alternately in rows. Yield of fresh tubers was reduced by 78% in plants showing SPVD symptoms but the reduced chlorophyll level of diseased leaves did not appear to affect yield.

In a long-term fertility study it was observed that judicious use of chemical fertilizers maintained the levels of soil pH, total N, organic C, CEC and exchangeable Ca. Highest accumulation of exchangeable Ca and Mg was found in the 40–60 cm layer. Absence of Mg and K from the fertilizer schedule resulted in a decrease in the amounts of those elements in the soil. Profile CEC and exchangeable Ca were not affected much by the application of K or trace elements. Compared with adjacent grass fallow or virgin bush land, only the values of C/N ratio, CEC and exchangeable Mg decreased on cultivation. All the crops responded well to fertility levels but not to K; trace element response was confined to cotton and groundnut.

No natural population of Rhizobium japonicum specific for soyabeans (Glycine max L.) could be found in any soil type in Israel. A crop of 5450 kg seed/ha was obtained in a field of sandy soil, inoculated with commercial granular inoculant, by maintaining soil at a moisture level near to field capacity during the growing season. In another field, on heavy soil inoculated with different strains of R. japonicum, the highest yield was obtained with strain 31-SM, a mutant derived from strain 61A75. Nitrogen fixation, as measured by the acetylene reduction method, was highest with strain 31-SM. Under local conditions soyabean plants snowed 100% nodulation when inoculated by the slurry method, by granular inoculant or by spraying bacteria into the furrow.

Growth characteristics of dry bean genotypes of determinate (Type I), indeterminate small vine (Type II), and indeterminate large vine (Type III) types were evaluated over three spacings and two growing seasons. Important observations include: (1) Type I did not give increased yields in response to planting rows closer together in Trial II because there was a large decrease in harvest index, which did not occur for Types II and III; (2) accumulation of reproductive nodes prior to anthesis is important for high yields, as are biological yield, harvest index and leaf area; and (3) within-habit spacing combinations, factors such as flowering duration, branch-, leaf- and raceme-numbers, and stem length could not be used to predict yield.

Ten tomato varieties, tested at three locations for 3 years in Western Nigeria, gave variety × environment interactions, indicating that some were only suitable for specific areas and none possessed wide adaptation. It is suggested that Western Nigeria could usefully be divided into sub-zones in which specific varieties could be developed.

With the exception of leafy and daylength-sensitive cultivars, cowpea (Vigna unguiculata (L.) Walp), produced larger yields in the first season (April-July) than in the second season (September-November) at Ibadan, Western Nigeria, whether or not the crop was protected by insecticide. The populations of major cowpea pests were low at the start of the rainy season and reached their peaks in September or October. The implications of these results in insect control strategy are discussed.

Three cultivars of mustard (B. juncea) were grown on field plots with five levels of fertilizers and with or without irrigation. Irrigation gave 59% more seed yield than no irrigation, accompanied by improved performance of ancillary characters. Significant increases in seed yield and most of the other characters were noted with increased levels of fertilizers, with a significant cultivar × fertilizer interaction due to the negative response of one cultivar to increased levels of fertilizers. The importance of irrigation and fertilizer in mustard seed production was thus demonstrated.