The implementation of a ‘Green Revolution’ programme in Nigeria, known as the National Accelerated Food Production Project, was linked with earlier programmes in other countries derived from CIMMYT and IRRI and with a previous FAO Fertilizer Programme in Nigeria. The methodology was based on ‘minikit’ demonstration trials on farmers' fields. Details are given of the the forms used in the field and plot size, number and choice of treatments and subsequent package demonstrations, and production kits are discussed.

Photosynthetic rates of coffee plants grown under the full sun and shade were compared. The saturating irradiance of 300 μE−1 m s−1 of shade plants shifted to near 600 μE in sungrown plants, but shade plants possessed substantially higher photosynthetic rates. Photosynthetic rates decreased above 25°C, which was apparently due to a decline in the mesophyll conductance, as stomatal conductance remained more or less unchanged between 25° and 35°C. Most of these measurements were made on seedlings in plastic bags. It is concluded that coffee is more suited to high density plantings where mutual shading, resulting in low light intensities and lower leaf temperatures, are likely to produce a favourable environment for efficient photosynthesis and growth.

Experiments were carried out to relate soil moisture to leaf water potential (Ψ1), and to determine the effects of varying Ψ1, on leaf conductances and photosynthesis in coffee. Stomatal conductance was maximum at 0900 h, but plants growing in drier soil showed marked mid-day stomatal closure. After 1500 h, stomata began closing although plant water status improved. Photosynthesis in relation to changing Ψ1 appeared to exhibit roughly three different rates. At the fixed experimental temperature (25°C) low Ψ1 reduced photosynthesis throughits influence on stomata, but under field conditions low Ψ1 and an accompanying rise in temperature could lower the rate by lowering both mesophyll and stomatal conductances.

Various intercrops were examined in an alternate row pattern with pigeonpea or sorghum on both Alfisol (red soil) and Vertisol (black soil). The slow-establishing and later-maturing pigeonpea combined well with earlier cereals and legumes to give very large yield advantages as measured by the Land Equivalent Ratio. In the pigeonpea/cereal combinations, the earlier the cereal the bigger the yield advantage tended to be, attributed to improved use of resources over time as the difference in maturity periods of the component crops increased. Sorghum was generally more competitive than pigeonpea and intercropping advantages tended to be less. But even where there was little difference in maturity periods of the component crops, both sorghum/legume and sorghum/cereal combinations gave substantial and statistically significant advantages, suggesting that improved ‘spatial’ use of resources was also important.

In a 4 × 4 × 4 NPK field experiment with coconut on a sandy soil, N increased girth, height and leaf production during the pre-bearing period, but phosphorus (up to 6½ years) and K (up to 3½ years) had no effect on any vegetative growth parameters. The rates of fertilizers for optimum copra and nut yield (13 to 16 years) were 0.35, 0.73 and 0.64 kg N, PO and KO/palm/ year respectively. The high P requirement was due to the low solubility of the saphos phosphate in these sandy soils. A positive N × P interaction was also observed.

The effects of rootstock on tree development, leaf nutrient levels and yields of two old-clone lemon cultivars, Eureka and Villafranca, were studied. The largest canopy development and yields were obtained on Alemow and Volkameriana; red rough lemon failed due to foot rot; Rangpur lime trees were severely stunted; and Citremon induced high chloride and relatively low boron contents in the leaves.

A range of legumes was evaluated in North-east Thailand for dry season forage production on stored moisture following harvest of the paddy rice crop. Crotalaria juncea (sunnhemp) and Lablab purpureus (lablab) produced the best dry matter yields of 1950 and 2020 kg−1 ha respectively on a cultivated paddy field without further fertilizer. Stored moisture was adequate for these two but growth of other species was poor, especially on plots that received no prior cultivation. In pot experiments, sunnhemp did not respond to either nitrogen application or Rhizobium inoculation, indicating that native strains were effective for nitrogen fixation. Lablab responded to both treatments and gave a 55% dry matter response to inoculation with Rhizobium in the field.

Results clearly demonstrate that Crotalaria juncea (sunnhemp) is well adapted to the dry season paddy environment; peak dry matter production was 3170 kg ha−1 in a field experiment with rows 30 cm apart and 100 kg−1 ha of seed. A pot experiment gave yield responses of 179 and 16% to P and S application respectively but there was no response to fertilizer in the field. Cultivation increased DM production in one experiment from 86 to 1590 kg−1 ha, attributed to a reduction in bulk density from 1·65 to 1·19 g cm−3. This effect was not fully substantiated in a second experiment on a coarser soil, where good DM production of sunnhemp was achieved even on broadcast uncultivated plots. The toxicological implications of using sunnhemp as a forage legume are briefly discussed.

N, P and K fertilizers improved mulberry leaf, stem and fruit yields on peat, the optimum rates being 100 kg/ha/yr N, 25 kg P and 62·5 kg K, with further leaf responses to N and K if applied together in the ratio of 3:2. Harvesting every 2 months at 60 cm height gave better yearly leaf yields than cuttings 3–12 months apart, but annual stem yields were not affected by harvesting frequency and fruit yields were optimum when plants were cut at intervals of 5–6 months. Harvesting quarterly at 60 cm produced better leaf and stem yields than cutting lower (15 and 30 cm) or higher (100 and 180 cm). Two Thai cvs (Noi I and Mii) were female-sterile but the Local Malaysian produced fruits.

In Ghanaian cocoa progeny trials plot values were adjusted by covariance on neighbouring plots. Standard errors of differences between progenies were usually reduced by 10% and often by over 20% when covariates were formed from the residuals of neighbouring plots and analyses were iterated using successive improved estimates of treatment effects to derive fresh residuals. The method gave larger reductions in error and was effective over a wider range of variates than adjusting on early growth data from interplanted temporary trees of a standard variety.

Before an innovation can be judged as superior to existing technology, its relevance should be jointly evaluated by researchers and farmers by testing the proposed practice in a farm environment. Using a proposed maize technology as an example, it is demonstrated how site-related and management variables can be used to quantify the factors influencing actual yields on farmers' fields. Similarly, economic and management analysis of the practice provides an estimate of the probability of the farmer being better off by using the technology. The integrated analysis of the practice provide insights for the design of problem-oriented research projects.