Gas exchange, leaf carbon isotope discrimination, growth, yield and beverage quality were evaluated for two Coffea arabica cultivars (Catuai and Mundo Novo), grafted on to C. canephora and C. congensis progenies growing in open fields. During the years 1994 to 1997, grafting resulted in an average increase in bean yield of 151 and 89% for Catuai and Mundo Novo respectively. As analysed by sensory analyses and by the ratio between the mono-isomers and di-isomers of caffeoylquinic acid, beverage quality of the C. arabica was not altered by grafting. Shoot growth was significantly greater in grafted plants, showing an increase of 52% in total leaf area compared with the non-grafted plants. Under conditions of water excess in the soil there was little difference in the transpiration and stomatal conductance rates between the grafted and non-grafted plants, but the net photosynthesis was higher in grafted plants. With an accentuated water deficit in the soil in the dry period, the grafted plants showed significantly higher transpiration and stomatal conductance rates than the non-grafted plants, and similar values to those of C. canephora. Carbon isotope discrimination was greater in the grafted plants, suggesting greater root hydraulic conductance. The results suggest that the better performance of the grafted plants during the dry period was due to the greater capacity of the root system of C. canephora to provide water to the shoot thereby maintaining greater gas exchange in the leaves and consequently a greater carbon gain.

In the semi-arid Mwea Division of Kenya's Eastern Province, the modern pigeonpea (Cajanus cajan) variety Nairobi Pigeonpea 670 (NPP 670) had become known to all farmers and was being grown by 68% of them within a period of 12 years. The only injection of this seed to the area was from a single on-farm demonstration. Three-quarters of farmers found out about the variety from observing it growing in the field, and obtained seed primarily from other farmers in the village. Factors favouring the diffusion of the variety included its attractiveness as a cash crop, the ease with which it could be distinguished from other varieties, the low seed rate, and the relative ease with which growers were able to maintain seed purity. Farmers expressed a willingness to pay for fresh seed, which suggests that more effort needs to be made to involve the formal seed sector. The present regulatory system does not favour the development of a formal supply system. NPP 670 was released more than 10 years after it was first tested.

On-farm experiments were conducted between 1990–91 and 1996–97 in the Indian districts of Jalandhar (Punjab) and Ghazipur (Uttar Pradesh), on rice-wheat cropping systems. The aim was to compare yields and net returns from three treatments: (i) farmers' nutrient management practices; (ii) integrated use of organic manures and fertilizer; and (iii) NPK fertilizer inputs alone. Productivity of the rice-wheat systems, in terms of grain yield per unit area, was greater in Jalandhar district than in Ghazipur district. In Jalandhar, a high productivity zone, greatest yields were achieved with integrated use of green manure and fertilizer NPK inputs, while in Ghazipur, a low productivity zone, yields were highest with inputs of fertilizer NPK alone. Increases in yield due to these improved practices, over farmers' practice, were 6.98% in Jalandhar and 46.6% in Ghazipur. At both locations, net returns were greater with the use of fertilizer NPK alone. However, benefits from NPK alone over farmers' practice were 6.96% in Jalandhar and 79.6% in Ghazipur. After 11 years of a long-term experiment, at Ludhiana (Punjab) and Faizabad (Uttar Pradesh) soil organic-C and available phosphorus contents have increased, and available potassium content decreased compared with the initial levels.

In a genotype × environment experiment at four sites with four tea (Camellia sinensis) clones in Kericho, Kenya, differences in ground cover and total dry matter production and partitioning were found between clones and between sites. The major contributor to these differences was the daily intercepted solar radiation that differed by as much as 30% between sites. Differences in radiation use efficiency (RUE) between the sites were small, but varied from 0.3 to 0.45 g MJ−1 between the clones. The rate of dry matter production was affected by the prevailing weather conditions. During the dry seasons the rate of dry matter production differed between the sites and clones by a maximum of 2 g m−2 d−1. Harvest indices were found to vary between the sites and increased with time. Partitioning to harvested shoots ranged from 10% for clone TN14-3 to a maximum of 19% in clone S15/10 at site 4. Partitioning to roots showed distinct differences between clones, being least in clone S15/10 (10–20%) compared with 15–32% in other clones. Rooting depths were greatly influenced by the prevailing temperatures.

Instructions are given for the presentation of the results of statistical analyses in Experimental Agriculture. Statistical codes accepted by the journal are defined. Appropriate styles of presentation are illustrated for estimated parameters and variances for a selection of standard statistical analyses. The value of presentations of results of significance tests is discussed. Formats for acceptable tables and graphs are shown. References to further guiding texts are included for non-standard situations.

In an experiment on genotype × environment interactions in tea (Camellia sinensis), yield differed between the genotypes at all sites. The highest yield in the third year after planting was 3760 kg ha−1 for clone S15/10 at site 4 (1800 m altitude), while the lowest was 1610 kg ha−1 for clone 6/8 at site 1 (2200 m). The dry matter contents of harvested shoots from each clone ranged from 0.24 for clone TN14-3 down to 0.19 for clone S15/10. Yield response to soil water deficits was more pronounced at site 4, where larger deficits were observed. Hail damage affected the yield of two sites and a temperature influence on recovery was evident.

Under rainforest vegetation, the central Amazonian Ferralsols are characterized by relatively high availability of N in relation to other nutrients. After forest clearing, several tree crops also have not shown yield responses to N fertilizer. To elucidate the mechanisms of this apparent N sufficiency, the mineralization of soil N was measured under three tree crops and a leguminous cover crop (Pueraria phaseoloides) in a multi-strata agroforestry system at two fertilizer input levels on a Xanthic Ferralsol in central Amazonia. In situ incubations of topsoil (0–10 cm) were carried out using the buried-bag method on five occasions over ten months. The highest mineralization rates were found under the cover crop, intermediate rates under rubber trees (Hevea brasiliensis) where the soil was also covered by the cover crop, and lowest rates under peach palm (Bactris gasipaes) and cupuaçu (Theobroma grandiflorum) with no cover crop. The increased N mineralization under the cover crop was due to more total N in the soil, higher soil moisture and, presumably, a larger pool of readily mineralizable soil N compared with the soil under the tree crops. Other fertility parameters also differed significantly between sampling positions within the plots, but this had no major influence on net N mineralization. Also, the input of NPK fertilizer and dolomite had no significant influence on N mineralization, indicating that N mineralization was not nutrient-limited. High total N mineralization rates in the soil (approximately 350 kg ha−1 a−1 at 0–10 cm depth) explained earlier observations of nitrate leaching into the subsoil under multi-strata agroforestry at this site. Considering the spatial patterns of N mineralization with maximum values under the cover crop, the exploration of the soil volume by crop roots should be maximized to increase the uptake of mineralized soil N by the crops and reduce nitrate leaching. Appropriate measures are narrow tree spacing, use of annual and semi-perennial intercrops and encouragement of the lateral root development of the trees. In addition, the mineralization of soil N close to the tree crops can be influenced through the management of the cover crop. In view of the high total N mineralization rates in the system and unclear yield responses of tree crops to N fertilizer, the application of N fertilizer to tree crops with well-developed root systems and a well-managed cover crop may often be unnecessary on this soil type. This may facilitate the further development of tree crop agriculture in the region.

Growth variability in pearl millet (Pennisetum glaucum) over short distances is a severe constraint on the interpretation of agricultural experiments in the West African Sahel. The purpose of this study, therefore, was to compare different non-destructive methods to estimate, spatially, millet growth and final yields. Aerial photography, georeferenced radiometric measurements and a chlorophyll meter were tested during three rainy seasons in a nitrogen rate×density×genotype experiment in western Niger. For the radiometric measurements, normalized difference vegetation indices (NDVI) obtained and calibrated for individual millet hills spaced 1.5 m apart were aggregated for the entire experiment with 6000 samples per hectare. A simple calibration procedure was used to correct for variation in soil background reflectance and incident light. For NDVI measurements of individual planting hills, the correlation between plant total dry matter (TDM), leaf weight, leaf area and NDVI was high (r2 = 0.89–0.91) and regression parameters were genotype-specific. Aggregated georeferenced NDVI measurements at the plot level correlated with grain and TDM at harvest (r2 = 0.40–0.87). The analysis of true-colour and infrared aerial photographs permitted the monitoring of millet growth and the quantitative evaluation of treatment responses throughout the growing season. The infrared images were the most efficient in the detection of vegetation followed by the normalized green band of true-colour images. The red band was the least effective because of the influence of soil albedo and image vignetting. Although chlorophyll meter measurements reflected relative differences in plant nitrogen status between treatments, their interpretation required destructive sampling and proved unsuitable to predict millet yields. The results demonstrate the potential of georeferenced radiometric data and aerial photographs to improve soil sampling strategies, sequential plant growth monitoring and the statistical design and analysis of experiments. By providing intermediate data sets, the tested tools can also help in the upscaling of ground truth to satellite data in yield prediction studies.

To assist commercial producers with optimizing the use of irrigation water, the responses to drought of mature and young tea (Camellia sinensis) crops (22 and 5 years after field planting respectively) were compared using data from two adjacent long-term irrigation experiments in southern Tanzania. Providing the maximum potential soil water deficit was below about 400–500 mm for mature, and 200–250 mm for young plants (clone 6/8), annual yields of dry tea from rainfed or partially irrigated crops were similar to those from the corresponding well-watered crops. At deficits greater than this, annual yields declined rapidly in young tea (up to 22 kg ha−1 mm−1) but relatively slowly in mature tea (up to 6.5 kg ha−1 mm−1). This apparent insensitivity of the mature crop to drought was principally due to compensation during the rains for yield lost in the dry season. Differences in dry matter distribution and shoot:root ratios contributed to these contrasting responses. Thus, the total above-ground dry mass of well-irrigated, mature plants was about twice that for young plants. Similarly, the total mass of structural roots (>1 mm diameter) to 3 m depth was four times greater in the mature crop than in the young crop and, for fine roots (<1 mm diameter), eight times greater. The corresponding shoot:root ratios (dry mass) were about 1:1 and 2:1 respectively. In addition, each unit area of leaf in the canopy of a mature plant had six times (by weight) more fine roots available to extract and supply water than did a young plant. These results show that young tea should be irrigated in preference to mature tea, especially where the maximum soil water deficit is likely to exceed 250 mm.

The mean time for 50% germination at 20 °C of 12 Indian wheat (Triticum aestivum) cultivars was nearly halved, from 51 h to 27 h, by soaking seed in water for 8 h prior to sowing. A delay of 24 h without further soaking, intended to simulate postponement of sowing, reduced the time saved by priming to 16%. Priming had no effect on final germination percentage. These results were used in the design of 275 on-farm, farmer-managed, participatory trials of seed priming in wheat during the 1997–98 and 1998–99 rabi (post-monsoon) seasons. In forty-one trials in tribal areas of Bihar and West Bengal states of India, seed priming gave a 13% grain yield advantage for farmers growing wheat in marginal areas with low levels of agricultural inputs. Mean benefits from seed priming of wheat in nine trials in Chitwan, Nepal were 17%. In high potential areas of Gujarat, India, 205 trials had higher rates of input use. Yield benefit from priming in these trials averaged only 5% but constituted an extra 200 kg ha−1 grain at little or no cost. In 20 trials on marginal land, with slightly saline irrigation water in Ahmadwala, Pakistan, an average yield increase of 36% was obtained using seed primed with a 0.2% gypsum solution. Collaborating farmers reported that priming wheat seed overnight resulted in faster, more complete emergence, more vigorous early growth, better tillering, earlier flowering, larger ears, earlier maturity and higher yields. In addition, many farmers also reported that foliage in primed plots was a darker shade of green than that in non-primed plots, suggesting that primed plants may have been using nitrogen more efficiently. Seed priming was popular with farmers, most of whom reported that they would prime wheat seed the following year. A survey in Gujarat in 1998–99 of 63 farmers who had tested priming in 1997–98 showed that, while 65% had primed some of their own seed, none had primed more than 50 kg, suggesting that there were practical difficulties in priming larger volumes of seed.