A field experiment will sometimes give data so unexpected that their correctness is doubted. It would be wrong to discard them because they go against the experimenter's preconceptions but sometimes enquiry will reveal why they are aberrant. Sometimes no reason can be found but they still raise doubts. The problem is approached using a published body of data.

The calculation of residuals and total residuals is commended as a means of enquiry. Exclusion of data is to be avoided if possible. It can have serious consequences and methods are given for assessing them. It is important to seek alternative approaches and one possibility is the use of covariance adjustments on fertility trends shown by study of the residuals. It is shown also that extreme values sometimes arise from unsuitable data and can be eliminated by a suitable transformation, but it must be meaningful and not just a statistical expedient to avoid difficulties.

A procedure was used to calibrate the DSSATv3 CERES–Maize (Zea mays) model and to evaluate its performance in simulating growth and development of maize using input data collected from campesino farmers instead of using data obtained from on-station experiments or from the literature. The problems encountered in the calibration process are illustrated, particularly the failure of the model to simulate the growth and development of local highland maize cultivars (HMC). It is argued that the low ambient temperatures to which HMCs are exposed in the Toluca Valley are responsible for this failure, because HMCs respond differently to temperature and have a different optimum temperature range from temperate and tropical maize cultivars. It was concluded that the model needs to be adjusted to allow for consideration of the effects of constant low temperatures on the prediction of plant phenology and production, and partition of biomass in HMCs, since grain yield is not the main criterion used by smallholders when selecting maize cultivars.

In southern and central India, farmers crop Vertisols only in the post-rainy season, to avoid land management problems in the rainy season. In 1983 ICRISAT established a long-term trial seeking to intensify cropping. The trial included intercrops, sequential crops and appropriate Vertisol management technology to allow consecutive rainy-season and post-rainy season crops to be grown. Benefits provided by legumes to succeeding rainy-season sorghum (Sorghum bicolor) were analysed in relation to a non-legume system of sorghum + safflower (Carthamus tinctorius). Rainy-season sorghum grain yield production was sustained at about 2.7 t ha−1 over 12 years within a continuous sorghum–pigeonpea (Cajanus cajan) intercrop system. With a cowpea–pigeonpea intercrop system, succeeding sorghum benefitted each year by about 40 kg N ha−1 (fertilizer nitrogen (N) equivalent). Without N fertilizer application the sorghum grain yield was around 3.3 t ha−1. Legume benefits were less marked in the chickpea (Cicer arietinum)-based rotation than in the pigeonpea system, in which a 12-year build up of soil total N (about 125 μg g−1) was observed. Although sorghum benefitted from this system, pigeonpea yields declined over time due to soil-borne fungi and nematodes. Wider rotations of crops with pigeonpea may help to overcome these problems, while sustaining sorghum production.

The effects were evaluated of six methods and six durations of seed treatment on the germination of wild sunflower (Tithonia diversifolia) seeds. The treatments were as follows: (1) hot water at 80 °C, (2) hot water at 100 °C, (3) oven-drying at 80 °C, (4) oven-drying at 100 °C, (5) immersion in concentrated sulphuric acid and (6) immersion in 10% hydrogen peroxide. The durations were 0, 2.5, 5, 10, 15 and 20 min. Treatment with hot water at 80 or 100 °C for 11–14 min, and oven-drying at 100 °C for 20 min all resulted in more than 65% germination 10 d after treatment. Oven-drying at 80 °C, immersion in concentrated sulphuric acid and in 10% hydrogen peroxide resulted in less than 50% germination. Based on the higher cumulative germination, lower cost, and environmental concerns, hot water treatment at 80 or 100 °C for 11–15 min is recommended.

Relative palatability and some agronomic traits were studied for nine legumes including Aeschynomene histrix, Centrosema brasilianum (two accessions), Centrosema pascuorum, Chamaecrista rotundifolia, Stylosanthes guianensis (two accessions) and Stylosanthes hamata (two accessions). All species were consumed by cattle but the relative palatability varied according to season. Only the two S. guianensis accessions were positively selected throughout the year. A positive relationship was established between the ability to retain green leaves in the dry season and palatability. Once established, the two C. brasilianum accessions out-performed the other legumes in their ability to stay green and retain leaves in the dry season; related to the ability to retain green leaves under drought conditions, C. brasilianum ILRI 155 was among the most palatable legumes in the dry season. S. guianensis accessions had the highest dry matter (DM) yields and one of them (ILRI 15557) had the highest nutritive value. In contrast with earlier results in the same environment, C. rotundifolia and A. histrix performed poorly in terms of DM productivity. The implications of the agronomic characteristics of the legumes for the management of legume-based pastures in crop/livestock systems are discussed and the concept of legume–legume mixtures, composed of several compatible species, is advocated.

The effect of tillage and crop rotations can only be seen over many crop years. Crop yield and soil results are evaluated from two long-term trials, established in 1978–79 and 1985–86 to investigate various forms of tillage and the timing of such operations in various wheat (Triticum aestivum)-based rotations on a Calcixerollic Xerochrept in northern Syria. In a tillage systems trial involving two wheat–legume–watermelon (Citrullus lanatus) rotations, deep tillage showed no advantage over a shallow sweep-tillage system, either for soil moisture storage or yield increase of any crop. The zero-till system suited legume crops but gave lower productivity in wheat due to a build up of grassy weeds, and was not suitable for watermelon. Minimum tillage, with its higher energy-use efficiency and yield levels equal to or even slightly above those of deep-tillage systems, appears promising for the lowland areas of West Asia and North Africa. In a tillage timing trial, wheat in a wheat–lentil (Lens culinaris) rotation yielded best after conventional deep disc-plough tillage, but lentil yields were higher in a zero-till system.

The effect of seven planting densities on yield and witches' broom disease incidence in hybrid cacao (Theobroma cacao) trees in the Brazilian Amazon was investigated. Data collected over a three-year period showed that it was possible to optimize regional cacao yield by high planting densities (2500 and 1666 trees ha−1) and that there was no density × year interaction. However, high planting density also favoured witches' broom incidence. A growing regional interest in planting cacao clones at high planting density should be pursued with caution, since high planting densities of clones in environments under strong pressure of selection for the witches' broom pathogen, such as in the Amazon region, are still a risky strategy. Alternatives to the high planting density system are presented and discussed.

Weeds are major constraints to wider adoption of wet seeded rice. Two split-plot experiments on water management during crop establishment of wet seeded rice were conducted in the dry and wet seasons of 1994 to quantify crop stand establishment and weed suppression by herbicides and rice genotypes. The latter consisted of five previously identified hypoxic-tolerant lines and three standard cultivars, sown at approximately 300 seeds m−2. The tested hypoxic-tolerant genotypes had superior seedling growth, but not higher crop stand establishment than the standard ones. Genotypes that had superior crop stand establishment or faster seedling growth did not necessarily give better weed suppression. Genotypes with high tillering ability were more competitive against weeds. Echinochloa glabrescens dominated the weed flora, especially in farmers' practice and anaerobic seeding (seeding into soft mud), followed by flooding at 7 d after seeding (DAS). The relative proportion of Monochoria vaginalis increased in anaerobic seeding with flooding 3 DAS and seeding into standing water. Pretilachlor + fenclorim effectively controlled weeds in farmers' practice and anaerobic seedings, and 2,4-D in water seeding. Without herbicide, water seeding and anaerobic seeding flooded 3 DAS reduced dry weed weight by 73–88% compared with farmers' practice. Anaerobic seeding with flooding 7 DAS controlled weeds effectively when the weed pressure was low (dry season, about 136 g m−2 of weed biomass in plots without herbicide) but not when weed pressure was high (wet season, 513 g m−2). Water seeding could not sustain high rice yields due to low crop stand population caused by flotation of seedlings. Anaerobic seeding with flooding 3 DAS allowed genotypes to sustain high yield and increased water productivity (rice production per unit volume of water used in the field) without having to use herbicide or with only half of the recommended herbicide rate. The correct water management during the crop establishment stage may effectively lower the economic and possible runoff costs of herbicides without reduction in yield or water productivity.

Seven maize cultivars and seven soyabean cultivars were grown in monocultures, and in all possible 49 mixtures of maize and soyabean, at two locations to assess the intercropping advantages of mixtures. On average, intercrops produced 6% more dry matter and 17.5% more protein than the monocultures of maize, the most productive component. Both the dry matter and protein yields of intercrops were closely correlated with the yield of the maize component, but not with that of the soyabean component, indicating that there was no need for a specific breeding programme for maize to be used in intercropping systems. Although the average intercropping advantage, the monoculture yields of maize cultivars and, to a lesser extent, the monoculture yields of soyabean cultivars differed between locations, there was no evidence that the yield of various intercrop combinations differed between locations.

The millet (Pennisetum glaucum)-based cropping systems that dominate the Sudano–Sahelian Zone of West Africa cannot, as they are currently practised, meet the growing food needs of the region. They must therefore be intensified in a sustainable manner. The present study was initiated in 1986 and continued until 1996 to evaluate the effects of phosphorus (P) fertilization, tillage and rotation with sole cowpea (Vigna unguiculata) on an operational scale with two cropping systems, namely, sole millet and millet–cowpea intercropping. A randomized complete block design with four replications was used. The effects of P fertilization, ridging with animal traction and planting on ridges (AT), and rotation with sole cowpea increased the productivity of millet substantially in 10 of the 11 years. Based on the 11-year average, P fertilization alone improved grain yield by 52%, and AT with P fertilization improved grain yield by nearly 135%. Combining AT, P fertilization and the sole cowpea rotation resulted in a 200% increase in grain yield compared with the traditional system of production. Millet productivity did not show a significant decline when intercropped with cowpea. Stability and relative stability analysis showed that the traditional system was more stable than the various agronomic packages, but had the least yield. Conversely, the agronomic package with the highest yield advantage over the traditional system was the least stable. A major portion of the annual variation in the environmental index for grain yield and total dry matter was attributed to the seasonal variation in rainfall and organic matter depletion. Organic matter levels declined linearly with years of cultivation. Significant differences were found in the rate of depletion between the various agronomic treatments tested. After 11 years, nearly 60% of the organic matter was depleted irrespective of the agronomic treatments.

Since shading often occurs under normal growing conditions in various cropping systems (for example, intercropping, alley cropping), field experiments were carried out to determine the effects of shading on stomatal density, leaf size, leaf dry matter, and leaf lamina thickness in the major tropical root and tuber crops, tannia (Xanthosoma sagittifolium), sweet potato (Ipomoea batatas), yam (Dioscorea esculenta), cassava (Manihot esculenta), and taro (Colocasia esculenta). Shading decreased stomatal density in the lower epidermis of tannia, sweet potato, yam and cassava, and in the upper epidermis also in tannia and sweet potato; the upper epidermis of yam and cassava were devoid of stomata. In contrast to the other species, taro under shade had an increased stomatal density in both the upper and lower epidermis, a finding which was confirmed in subsequent pot experiments. This response of taro was postulated as a possible manifestation of greater shade adaptation by the species.

For all the species, shading generally resulted in the production of larger (in terms of surface area) but thinner leaves, with a decreased dry matter concentration. For the two species (yam and cassava) that had stomata only on the lower epidermis, normal sun-grown plants had about twice as many stomata per unit area of lower epidermis than did corresponding plants of the other three species. The agronomic and physiological significance of the findings are discussed.

This study, quantifying variation in flower and pod production in lentil (Lens culinaris), aimed to answer the question: Will selection for more pods per inflorescence increase seed yield? In Season 1 (1992–93) all open flowers were tracked to maturity in a field experiment with two lentil genotypes sown on two dates. Genotype Talia 2 had a higher rate of flower abortion than pod abortion, in contrast to genotype ILL 2581 which showed the reverse. Flower abortion accounted for 15% of flowers opened in early sowing and increased to 22% in the late sowing. Pod abortion was 19% (of flowers opened) in early sowing and 23% in the late sowing. These are the first quantitative estimates of flower and pod abortion in lentil. From the data, a rapid sampling method was developed to estimate the average number of pods per inflorescence at maturity. In Season 2 (1993–94) an experiment was conducted at two locations to estimate the average number of pods per inflorescence of 81 genotypes and to relate this to seed yield. Although the broad-sense heritability (h2) of the number of pods per inflorescence was 0.68 and its phenotypic correlation with seed yield was r = 0.71, the highest-yielding genotypes were not those with the most pods per inflorescence. Selection for the number of pods per inflorescence cannot be recommended for increasing seed yield in lentil.

Sustainability of intensive irrigated agriculture in Egypt has become a critical issue, as land and water resources are limited on the one hand and population is increasing rapidly on the other. Salinization, heavy input use, nutrient export and pollution all threaten the health of soils that have been feeding Egypt for centuries. At the same time, the build-up of newly reclaimed desert soils to economically sustainable productive capacity is a major challenge. In a collaborative effort between the Agricultural Research Center (ARC) of the Ministry of Agriculture and Land Reclamation in Egypt and the International Center for Agricultural Research in the Dry Areas (ICARDA), a long-term resource management programme, funded by the European Union, has been developed to address the issue of resource management in a multidisciplinary way. Long-term agronomic trials (with such variables as water quantity, water quality, nutrient inputs and crop rotations) have been set up at sites representing the old lands, the newly reclaimed areas and the rainfed areas. These trials are complemented by extensive long-term monitoring in villages close to the experimental sites. This covers farmers' perspectives, farming practices and the condition of farmers' soils and crops and is aimed at identifying over time the sustainable and non-sustainable production practices and the social and economic factors that underline them. The project activities began with a Preparatory Phase which comprised inventory studies, rapid rural appraisal and multi-disciplinary surveys. This knowledge was used in the planning of the two closely related activities of long-term trials (LTT) and long-term monitoring (LTM). As the programme was implemented, the complementarity of the LTT and LTM approaches became the most important feature of this programme.

Land resources are scarce in Mauritius and to fulfil the export quota sugarcane (Saccharum hybrids) has to be grown in marginal areas which are not conducive to optimal growth. In order to increase overall productivity there is a need to breed and select sugarcane varieties specifically adapted to the extremely dry and extremely wet areas. In this study, 154 genotypes representing four families were planted in a randomized complete block design at two sites. The genotypes were replicated at the two sites and evaluated in the plant cane and first ratoon crops. Family × environment and genotype × environment interactions were determined using the mixed model analysis of variance. Significant differences between families, genotypes and environments were found in stalk height, stalk diameter, industrial recoverable sucrose % cane (IRSC) and cane yield per hectare in tonnes (TCH). The genotype × environment interaction was significant for stalk height, stalk number, stalk diameter, sucrose content, TCH, tonnes sugar per hectare (TSH) and kilobrix (parameter used for selection in the preliminary phase of the selection programme at the Mauritius Sugar Industry Research Institute) in both plant cane and first ratoon crops indicating that the relative performance of the genotypes is not consistent across the environments for these characters. The family × environment interaction was found to be significant for cane and sugar yields whereas it was not significant for field Brix in either the plant cane or the ratoon crops. This showed that, for the parameter Brix, the relative performance of the families is similar in both environments and there is no need to replicate in more than one environment. The environmental sensitivities measured with respect to sugar yield showed that families and genotypes with nearly equal means over all environments displayed inconsistent performances. Differences in the sensitivities of the different families and genotypes were observed and some of the family × environment and genotype × environment interactions could be ascribed to them. The results also suggest that mass selection could be more effective than family selection and selection for each specific environment is advisable.

Three alternate-row intercrop experiments of sorghum (Sorghum bicolor)–cowpea (Vigna unguiculata) and/or millet (Pennisetum glaucum)–cowpea were grown at plant densities of 20 000 to 80 000 plants ha−1 at two locations in northern Nigeria between 1990 and 1992 to examine relations between yield and plant density. These relations were examined using the reciprocals of yield per plant to determine effects of environment, species and genotype on the theoretical maximum yield and optimum plant density. The intercrops were dominated by cereals, and cowpea biomass (BY) and seed yield (SY) were <10% of cereal BY and SY. Cowpea yields decreased as cereal plant density (D) increased, whereas cereal and total intercrop yields increased asymptotically with increasing D. Biomass yield of all intercrops responded to increasing D in a similar manner and the theoretical maximum intercrop BY was 12 290 kg ha−1 in all experiments. The response of SY to D varied among intercrops and was greatest and least with early and late maturing cereals respectively. The optimum D required to produce 90% maximum intercrop BY and SY varied between 15 600 and 30 000 plants ha−1, and 0 (no response to D) and 120 000 plants ha−1 respectively, and was higher for sorghum than for millet intercrops. The implications of these responses for agronomic management and germplasm improvement of cereal–cowpea intercrops are discussed.

In many highly weathered soils of the humid tropics, crop exploitation of the subsoil environment is limited through acid soil infertility. Since the use of mechanical profile modification is often prohibitive, surface incorporation of soil amendments is often the only means available to rectify this problem. A field trial was established with sugarcane on a strongly Acidic Dystrophic Brown Dermosol (Oxic Humitropept) in 1978 to evaluate the effects of surface incorporated lime additions on yield and performance of sugarcane. Eighteen years after the establishment of this trial, significant responses in cane yield were still evident following a single application of 5 t lime ha−1 made in 1978, as well as repeated applications of 5 t ha−1 on three occasions over the past 18 years. Progressive reductions in exchangeable acidity were accompanied by increases in subsoil Ca2+ and Mg2+. Soil pH increased significantly to a depth of 100 cm, this being attributed to the formation of ion pairs with NO3 in the surface soil, the subsequent leaching of these complexes and the differential uptake of NO3 at depth by roots. The results from this long-term study indicate that surface incorporation of lime is an economically viable approach to the remediation of subsoil acidity on soils such as those studied, namely, those with a low inherent cation exchange capacity and anion exchange capacity.

Long-term field experiments were conducted at six different sites (representing dryland and irrigated environments) with wheat (Triticum aestivum) on sandy loam soils to generate a database relating available water supply and fertilizer nitrogen (N) rate to grain yield. Step-wise multiple regression analysis showed that water supply and fertilizer N at more than 53 cm a−1 and 103 kg ha−1 a−1 were unproductive. A method, based on the principle of equi-marginal productivity of water in different periods of crop growth, has been used to estimate the optimum allocation of the variable amounts of water supply over the growing season. The optimized quantities of water in different periods of crop growth were then employed to compute fertilizer N requirement to achieve a given yield target. For medium grain yield targets (4–5 t ha−1), the range of substitution between fertilizer N and water for efficient resource use was quite wide.

Green revolution technologies of the 1960s and 1970s have led to the emergence of rice (Oryza sativa)–wheat (Triticum aestivum) (R–W) as a major cropping system in the Indo-Gangetic Plains (IGP). Expansion of irrigation through a network of canals and privately owned tube wells was one of the key factors in the success of the new technologies. Since canal water supplies are not demand-driven, farmers increasingly relied on the use of ground water by installing shallow tube wells. The strategy of ground water development and its use in conjunction with canal supplies paid rich dividends in control of waterlogging and secondary alkalization and in increasing the pace of alkali soils reclamation programmes in areas underlaid with fresh quality aquifers. Rice–wheat systems were the preferred choice of farmers. Adoption of the R–W system, however, changed the salt and water balance in north-western parts of the IGP. In the north-east of Punjab and Haryana states it led to receding water tables and waterlogging, and in south-west parts to associated problems of secondary salination. In eastern parts of the IGP in West Bengal, development of ground water aquifers to meet irrigation water requirements for the intensification and diversification of agriculture has led to serious problems of ground water contamination due to dissolution of arsenic-bearing minerals under altered soil moisture regimes, conducive to oxidation of pyritic sediments. These experiences suggest that there is an urgent need for a change in the way food is produced in the IGP for sustainability of the natural resource base. This will depend on the ability to predict long-term consequences of intensification and diversification of agriculture on the farm and regional scale. Whereas issues of favourable regional salt and water balances need attention in the north-west of the IGP, rain-water management and alleviating drainage congestion can facilitate the adoption of agricultural technologies and improve productivity of the R–W systems.