New initiatives in agroforestry are seeking to integrate indigenous trees, whose products have traditionally been gathered from natural forests, into tropical farming systems such as cacao farms. This is being done to provide from farms, marketable timber and non-timber forest products that will enhance rural livelihoods by generating cash for resource-poor rural and peri-urban households. There are many potential candidate species for domestication that have commercial potential in local, regional or even international markets. Little or no formal research has been carried out on many of these hitherto wild species to assess potential for genetic improvement, reproductive biology or suitability for cultivation. With the participation of subsistence farmers a number of projects to bring candidate species into cultivation are in progress, however. This paper describes some tree domestication activities being carried out in southern Cameroon, especially with Irvingia gabonensis (bush mango; dika nut) and Dacryodes edulis (African plum; safoutier). As part of this, fruits and kernels from 300 D. edulis and 150 I. gabonensis trees in six villages of Cameroon and Nigeria have been quantitatively characterized for 11 traits to determine combinations defining multi-trait ideotypes for a genetic selection programme. I. gabonensis fruits are rich in vitamin A (67 mg 100 ml−1), while the kernels are rich in fat (51.3%) and contain a polysaccharide that is a food thickener. The fruits of D. edulis are also rich in oil (31.9%) and protein (25.9%). This poverty-reducing agroforestry strategy is at the same time linked to one in which perennial, biologically diverse and complex mature-stage agroecosystems are developed as sustainable alternatives to slash-and-burn agriculture. To meet the objective of poverty reduction, however, it is crucial that market expansion and creation are possible. Hence, for example, it is important to determine which marketable traits are amenable to genetic improvement. While some traits (such as fruit and kernel mass) that benefit the farmer are relatively easy to identify, there are undoubtedly others that are important to the food, pharmaceutical or other industries which require more sophisticated chemical evaluation. There is a need, therefore, for better linkages between agroforesters and the private sector. The domestication activities described are relevant to the enrichment of smallholder cacao farms and agroforests. This diversification is seen as being important for the support of the cacao industry.

The tubers of yam, principally those of Dioscorea rotundata (white Guinea yam) and D. alata (water or greater yam), are important staple foods and sources of carbohydrate in West Africa. Yams are grown in diverse environments – from the high-rainfall forest zone on the coast to the seasonally arid savannas of West Africa, that is in situations in which the duration and the timing of the onset of the growing season vary appreciably. Dormancy in both underground and aerial tubers of the Dioscoreaceae is an important adaptive mechanism that helps to maintain organoleptic quality during storage and also ensures that tubers germinate at the start of the growing season. Plant breeders are especially keen to manipulate the duration of the dormant period in order to synchronize growth periods and, therefore, to produce more than one generation per year. The control of tuber dormancy, however, is poorly understood. This review examines critically those factors that affect tuber initiation, dormancy and sprouting, and makes recommendations for future priorities in research.

Using experience with bambara groundnut (Vigna subterranea), this paper examines how local knowledge, genetic evaluation, research in fields, glasshouses and laboratories, and crop simulation modelling might be linked within a methodological framework to assess rapidly the potential of any underutilized crop. The approach described is retrospective in that each activity was not clearly defined and structured at the outset. However, the experience gained may help to establish a methodology by which growers, researchers and international agencies can integrate their knowledge and understanding of any particular underutilized crop and apply similar principles to accelerate the acquisition of knowledge on other underutilized species. The use of a methodological framework provides a basis for activities that maximize knowledge, minimize duplication of effort, identify priority areas for further research and dissemination, and derive general principles for application across underutilized crops in general. It also allows policy makers and planners to make comparative decisions on the nutritional, economic and research importance of different underutilized and more-favoured species. In particular, the incorporation of a generic crop simulation model within the methodological framework may assist growers, extension agencies and scientists to refine general recommendations for any particular crop to local conditions. Also, the incorporation of information gathered from the field, laboratory or market can be used to update rapidly the predictive capacity of the model for each crop.

The role of water in the development and yield of the coffee crop (Coffea arabica L.) is reviewed. A period of water stress, induced either by dry soil or dry air, is needed to prepare flower buds for blossoming that is then stimulated by rain or irrigation. Although attempts have been made to quantify the intensity and duration of stress required, these have not yet been specified in ways that are commercially useful. Water must be freely available during the period of rapid fruit expansion to ensure large, high-quality seed yields. Depending on the time and uniformity of flowering this can occur at times when rainfall is unreliable, particularly in equatorial areas.

Although there are differences in their responses to drought, commercial cultivars have retained many of the characteristics adapted to the shady environment of the forests in the Ethiopian highlands in which C. arabica is believed to have originated. These include partial closure of the stomata when evaporation rates are high as a result of large leaf-to-air saturation deficits (>1.6 kPa), even if the soil is at field capacity. This is thought to be an adaptive mechanism that minimizes transpiration at high irradiances when the leaves are light-saturated.

Our understanding of the actual water use of coffee crops grown in diverse ways is imperfect. For mature crops, well supplied with water, the crop coefficient (Kc) appears to have a value in the range equivalent to 0.7–0.8 times the evaporation from a US Weather Bureau Class A pan. There is some evidence that Kc values are less than this on days when evaporation rates are high (>7 mm d−1). For immature crops allowance has to be made for the proportion of the ground area shaded by the leaf canopy, but this alone may underestimate rates of water use. Present methods of calculating crop water requirements for the purposes of irrigation scheme design and management are imprecise and, probably, subject to large errors depending on local circumstances.

The need for irrigation, and its role in controlling the timing of flowering, varies depending on the rainfall distribution, the severity of the dry season, and soil type and depth. Two geographic areas need to be distinguished in particular; those close to the equator with a bi-modal rainfall pattern and those at higher latitudes with a single rainy season and an extended dry season. Despite the international importance of irrigation in coffee crop production, the benefits to be derived from irrigation, in yield and in financial terms, have not been adequately quantified in either location. Allowable soil-water deficits have been specified for deep-rooting crops (2–3 m) on water retentive soils, usually linked to conventional over-tree sprinkler irrigation systems. Other, potentially more efficient, methods of irrigation are now available for coffee grower use, in particular, micro-jet- and drip-irrigation systems. However, there appears to be little advice, based on sound experimental work, on how to design and operate these to best advantage.

There is a need to interpret and apply the scientific understanding of the role that water plays in the growth and development of the coffee plant into practical advice that can assist the grower to plan and to use water efficiently, whether rainfall or irrigation, for the production of reliable, high-quality crops. Future research opportunities are identified.

In view of soil fertility problems and financial constraints identified in small-scale farming, on-station investigations on Low-External-Input and Sustainable Agriculture (LEISA) were started by the Farming Systems Research Team Luapula Province (FSRT-LP) in order to assess the potential of three annual and six perennial pasture legume species for green manuring purposes. Rhodes grass (Chloris gayana cv NIRS Boma) was used as a control. Over two seasons, a pasture legume-maize rotation trial was established. Different mineral fertilizer amounts were applied to the maize. Of all the species and cultivars tested, the perennial glycine (Neonotonia wightii cv Cooper) had the strongest green manure effect when no mineral fertilizer was applied. Of the annual pasture legumes, sunn hemp (Crotalaria juncea cv. NIRS 3) and velvet bean (Mucuna pruriens cv. NIRS 16) showed the greatest green manure potential at zero mineral fertilizer level. There were strong indications that the carbon:nitrogen ratio influenced the green manure potential of a species more than its ability to produce biomass. The need to assess the economics and sustainability of current mineral fertilizer practices on maize has been identified.

The concept of an integrated pest management (IPM) strategy is explored using case-study evidence from smallholder agriculture in southern Malawi. The conventional concept of strategy as a ‘plan’ is contrasted with the use of strategy as a game, as a performance, and as an accident. This pluralistic view is echoed in recent literature on business strategy. A typology is developed that relates farmers' choice of IPM strategy to the predictability of crop losses, the stability of the crop environment, and farmers' knowledge of pests. Some implications for IPM are explored. Where conditions favour ‘adaptive’ rather than planned IPM strategies, formal experimentation to verify farmers' strategies may be an inefficient use of resources. Where the crop environment is volatile and several pests attack the crop simultaneously, there may be limited scope to increase the adoption of IPM strategies by improving farmer knowledge of pest biology. Addressing the interactions between pest and crop management is critical in making IPM relevant for resource-poor farmers. A deeper understanding of farmers' management strategies is required to frame meaningful recommendations.

A computer simulation model was developed to investigate strategies for control of the parasitic weed species of Orobanche. The model makes use of data from published literature and predicts infestation levels in a dynamic and deterministic way. It is predicted that sustainable control of the parasite can only be achieved by reducing the soil seed bank to levels of 1000–2000 seeds m−2 and maintaining it at that level in subsequent years. When cultural control methods such as hand weeding, trap/catch cropping, delayed planting, resistant cultivars and solarization were considered individually, a relatively high level of effectiveness was required to contain the soil seed bank. An integrated approach with a selection of appropriate cultural methods is therefore recommended for further testing and validation in the field. The simulations demonstrate the importance of preventing new seeds entering the soil seed bank and that although reducing the soil seed bank may not increase yield for the first few years, it will ultimately increase production.

The results of 86 replicated experiments on soybean (Glycine max) cultivation practices in eight provinces – planned, conducted and analysed by farmers – were examined to identify local differences in the effects of cultural practices on yield, and to explore the potential for farmers to improve their practices. Plant spacing consistently improved yield relative to broadcast seeding. The comparison between a moderate and a high dose of N, and between local spraying practices and integrated pest management, also had a relatively consistent effect on yield. Conversely, the effects of straw mulch, of moderate doses of N compared to zero N, and of weeding, were highly location-specific. A change in cultural practices influenced yield and economic benefits more strongly in low- than in high-yield situations. Modifications of current cultivation practices often resulted in a significant yield increase. This suggests that farmers need training in how to improve their cultural practices and that it should utilize locally conducted field experiments. Training in experimental skills made farmers less dependent on external measures and advisers, and enabled them to become ‘experts’ who utilize science.

Phosphorus at the recommended dose of 10 kg P2O5 ha−1 divided between three equal applications, one on each of 0, 7 and 14 d after Azolla inoculation (DAI), increased biomass production but sporulation was invariably inhibited. Of the different options tested, lowering the phosphorus application rate to 4–8 kg P2O5 ha−1 significantly improved sporulation in A. microphylla (strain 202) but there was a substantial reduction in biomass production. On the other hand, changing the schedule of phosphorus application from 0, 7 and 14 DAI to 0, 3 and 6 DAI did not hamper biomass production and improved sporulation frequency and sporocarp number in A. microphylla (strain 202), A. caroliniana and A. pinnata. It was comparable to the no-P treatment for the number of sporocarps, with slightly lower sporulation frequency. In these species, the sporulation frequency and sporocarp number of Azolla enriched with 30 or 60 kg P2O5 ha−1 and then grown without any further added phosphorus were higher than those of unenriched Azolla grown with 10 kg P2O5 ha−1 and mostly comparable to those of Azolla grown without phosphorus. Foliar spray of 2.5 μg ml−1 gibberellic acid (GA) solution (7 DAI) along with the application of 10 kg P2O5 ha−1 (split between 0, 7 and 14 DAI) to unenriched Azolla increased the sporulation frequency and number of sporocarps in A. microphylla (strains 202 and 203), A. caroliniana and A. pinnata, not only over that of phosphorus application alone but also over the untreated control. Combining the use of P-enriched Azolla (A. microphylla strains 202 and 203) with application of GA was more effective for increasing sporulation than was the use of P-enriched Azolla without GA, or application of phosphorus plus GA to unenriched Azolla.

In the process of developing an improved fallow system based on forage legumes, the potential interest of crop-livestock farmers in enhancing soil fertility restoration and dry-season feed supply was assessed during a socio-economic study on 11 farms of settled agro-pastoralists in the subhumid zone of West Africa. Simultaneously, eight farmers in the group participated in simple and largely farmer-managed on-farm experiments testing the establishment of selected forage legumes on fallow land. In addition to agronomic parameters, the participatory approach included an evaluation of the technology by farmers both during farm visits and field days.

With cropping and dry-season feeding strategies increasingly being limited by land availability, the agronomic performance of some of the tested legume species, in particular Stylosanthes guianensis, promised substantial productivity gains once they could be integrated into the traditional fallow system. The major concerns of the farmers were animal health and labour supply for cropping activities, rather than soil fertility and feed constraints. The need for initial weed control within most of the legume species, therefore, limited their acceptability by farmers at this early stage of innovation testing. Nevertheless, farmers' interest grew remarkably in the course of the study. Options for the selection of appropriate species and management practices, which aimed at facilitating the establishment of legumes on fallow land, are discussed as a means of enhancing further adoption of the innovation.

The participatory approach of Farmer Research Groups (FRGs) was used in Farmer Adaptation of Starter Technology (FAST) with small-scale farmers in Luapula Province, Zambia. The introduction of a starter technology proved to be a suitable method to induce self-help initiatives with farmers. Their first move in FAST related to the improvement of food security and income generation by the cultivation and marketing of new maize (Zea mays) varieties in wetlands during the dry and early rainy seasons. By informal on-farm seed multiplication the expenditure on external agricultural inputs was reduced. The technical aspects of the starter technology began with the integration of pasture legumes as a pioneer crop for green manuring purposes in maize production. In this respect, FRGs developed individual risk-aversion strategies to ensure early planting of the maize with the onset of the rains. The groups made adaptations such as biomass transfer, intercropping and crop rotations in order to integrate pasture legumes into current cropping systems for green manuring purposes.

The study of interactions between cacao (Theobroma cacao) and coconut (Cocos nucifera) in Lampung, Indonesia, examined different combinations of age, plant lay-out, planting chronology and choice of planting material under changing environmental conditions. Four coconut-cacao intercropping trials were used to assess the performance of each intercrop under limiting or non-limiting environmental conditions. In intercropping experiments with young cacao trees and young coconut palms, delayed cacao tree development and reduced yields were observed. When coconut palms were aged five years or over, coconut and cacao growth were satisfactory under virtually normal environmental conditions; death rates remained reasonable and yield percentages differed little from those of the monocultures for each crop. The performance of both plants, however, changed when water became a limiting factor.

Seedling growth in the nursery and the magnitude of shock experienced by the seedlings after transplanting can affect subsequent growth and development of transplanted rice (Oryza sativa). These two processes have not been adequately represented by rice growth simulation models. In this study, seedling growth in the nursery was described with respect to endosperm mobilization, dry matter (DM) distribution and seed size. In addition, the transplanting shock process was represented describing the crop development and growth in relation to seedling age at transplanting. These relationships, as two sub-models, were incorporated into CERES-Rice, a rice growth model, and their performances compared with independent field data. The nursery growth sub-model was also evaluated with another model, RIBHAB. Results of the comparison showed that the nursery growth sub-model, either with CERES-Rice or RIBHAB, closely predicted observed seedling growth. On the other hand, CERES-Rice overestimated root and leaf DM and underestimated culm DM, while RIBHAB underestimated all the growth traits considered in the evaluation. Transplanting shock period predicted by the sub-model closely mimicked observation, whereas CERES-Rice highly overestimated it. It is concluded that the developed nursery growth and transplanting shock sub-models can be incorporated into CERES-Rice and other rice growth models for better simulation of those growth processes.

Farmers' Participatory Varietal Evaluation (PVE) was conducted on sorghum (Sorghum bicolor) crops in the Kile-Bisidimo plains of eastern Ethiopia for three consecutive years, 1995–1997. The study aimed at providing farmers with alternatives to their landrace to enable them to overcome crop losses and to identify farmers' varietal selection criteria for inclusion in future breeding work. In 1995 constraints and opportunities in sorghum growing and farmers' varietal matching characteristics were identified through an informal survey. This was followed by a search for varieties and acquisitions of seed. Subsequently, eight varieties were evaluated of which five varieties were released and three were at the pre-release stage.

In 1996 and 1997 farmer-managed on-farm trials were conducted and farmers evaluated the performance of the experimental varieties against a locally grown cultivar. Through pairwise and matrix ranking, farmers' selection criteria were listed, preferred varieties identified and initiatives taken to diffuse the varieties through the local seed system. Out of the eight varieties that research workers considered the best, farmers selected only three. The study negated the generally accepted view that farmers in lowland areas of eastern Ethiopia are reluctant to grow short-duration varieties. The selection and introduction of three new varieties into a farming system where farmers were growing only one local variety reconfirmed PVE to be a means for enhancing adoption and increasing genetic diversity. The study also confirmed that increasing farmers' access to their preferred varieties would result in a faster rate of diffusion through farmer-to-farmer seed exchange.

Manure is a key input to smallholder farming systems, especially in the semi-arid environment of West Africa where cost and availability limit the use of inorganic fertilizers. This paper considers manure management by farmers in an intensive integrated farming system in the Nigerian savanna. The paper reports farmers' indigenous knowledge concerning manure production, quality and application, chemical analysis of manure nutrient content and application rates of manure. The potential manure supply of the livestock population of the Kano close-settled zone is calculated and compared with application rates. Recommendations are made concerning methods for improving manure quality through changes in management practices.

Chickpea (Cicer arietinum) is a promising post-wet-season crop to follow rainfed rice (Oryza sativa) in the High Barind Tract of northwestern Bangladesh. Yields in farmers' fields, however, remain low (<1 t ha−1) primarily due to such factors as poor crop establishment, late sowing, and terminal drought and heat stress. Having been shown to improve plant stand and yield of chickpea in other environments, seed priming – soaking the seed overnight before surface drying and sowing the next day – was tested for its efficacy under Barind conditions.

In the 1998–99 season, 30 trials grown entirely on residual soil moisture were conducted in farmers' fields. A statistically significant mean yield response to priming of 47% was obtained. In 1999–2000, 105 on-farm trials and 15 demonstrations comparing presence and absence of priming were conducted. The crop received rain during this growing season. In a randomly chosen subset of 35 trials, scientists recorded a mean yield increase due to priming of 20%; it was 22% (from 1.02 to 1.25 t ha−1) in the remaining 64 trials (6 of the 105 trials were abandoned) where farmers recorded yields. Using a different variety, the mean yield response to priming in 15 demonstrations was 17% (from 1.25 to 1.46 t ha−1).

The priming response was attributed mainly to rapid seedling establishment, with higher plant stand and earlier crop maturity allowing escape from end-of-season stresses. Priming also reduced the incidence of stem and root diseases, and increased nodulation by native rhizobia. This simple technology can substantially increase chickpea yields to remunerative levels for the resource-poor farmers in this difficult environment. Further, it is suggested that this technology can act as a catalyst for the introduction of further technologies that will permit reliable and profitable cultivation of post-rainy-season crops and thus improve the livelihoods of the rural population.

In Mozambique the sweet potato (Ipomoea batatas) is often grown as both a leafy vegetable, the terminal shoots or vines being progressively harvested during the season, and as a root crop. This paper reports the results of experiments designed to evaluate the effects of drought and vine harvesting frequency on the productivity of both yield components (cv. TIS 2534). Experiments were conducted during the rainy and dry seasons, with supplementary irrigation treatments superimposed. As the frequency of vine harvesting (equivalent to the number of harvests) increased, the total fresh weight of vines increased. There was a corresponding reduction in the yield of storage roots, however, particularly under well-watered conditions. As a result, the total harvested yield (vines plus roots) was remarkably stable in both wet and dry seasons (43–45 t ha−1 fresh weight). The cumulative dry weight of harvested vines increased with the number of harvests at constant rates, depending on water availability (from +150 to +250 when rain-fed, up to +340 to +440 kg ha−1 harvest−1 when irrigated). The corresponding rates of reduction in storage root yields varied from −3 to −130 (rain-fed) down to 310 to 400 kg ha−1 harvest−1 (irrigated). Total dry weight yields under well-watered conditions were constant at about 11 to 13 t ha−1, or double this on an annual basis. Irrigation water-use efficiencies (by dry weight of harvested crop) were generally higher in the dry season than in the rains. For vine production they increased with the number of harvests from 1.6 to 3.5 (rains) up to 0.9 to 6.7 kg ha−1 mm−1 (dry season). The corresponding values for root production were 7.5 to 13.1 (rains) and 12.2 to 19.1 kg ha−1 mm−1 (dry season). For the combined dry weight yields the water-use efficiencies were, with one exception, independent of harvesting frequency at 11.2 (rains) and 19.0 kg ha−1 mm−1 (dry season). Irrigated plants harvested at weekly intervals yielded vine dry weights of about 0.5–0.6 t ha−1 week−1. Compensatory vine growth (reported elsewhere) was observed in previously droughted plants following a rainfall event. The practical implications of these results are discussed. Future papers describe in more detail the physiological aspects of the observed yield responses.

Intercropping of pearl millet (Pennisetum glaucum) with clusterbean (Cyamopsis tetragonoloba) is a common practice in the arid zone of northwestern India. Field experiments were conducted for two years to examine the performance of two contrasting genotypes of each of these crops grown both as sole crops and as intercrops using all possible genotypic combinations of the pearl millet with the clusterbean. The seed yields of both crops were always lower in intercropping than in sole cropping, though the degree of reduction in the seed yield of each intercrop component was influenced greatly by the genotype of its companion crop. The seed yield of both clusterbean genotypes was reduced more when intercropped with the tall and late-maturing pearl millet, MH 179, than with the medium-statured and early-maturing HHB 67. On the other hand, the two clusterbean cultivars had similar effects on intercropped pearl millet. As a result, the highest land equivalent ratios (LERs) were obtained with intercrops containing pearl millet HHB 67 rather than MH 179, while the genotype of clusterbean had little overall effect on LER. The results obtained are discussed in the context of developing cultivars for intercropping.

Experiments to assess the yield and economic performance of a maize-groundnut rotation compared with continuous maize (both when inorganic fertilizer was applied to maize and when not), were conducted under management by smallholder farmers in Zimbabwe over six years. The experiment was planted on-station near Harare and at six smallholder sites in northeast Zimbabwe, predominantly on sandy soils. Fertilizer rates and practices were those used by farmers, as described in surveys. On-farm grain yields from continuous maize without fertilizer were generally in the range 0.5–0.8 t ha−1 over five years. Maize yield responses to inorganic fertilizer on smallholder farms were highly variable, but moderate (up to 29 kg grain per kg N) with adequate rainfall. With no inorganic fertilizer applied to maize, the on-station groundnut crop (producing 0.260–0.355 t ha−1 shelled grain) almost doubled the grain yield of the following maize crop (in 1995–96), increasing output from 2.46 t ha−1 to 4.61 t ha−1. Where inorganic fertilizer was applied to maize, the rotation produced even more additional maize grain (an increase of 2.93 t ha−1). Up to 0.50 t ha−1 extra grain was obtained in the second year of maize following groundnut (1996–97). With inorganic fertilizer, groundnut improved the grain yield of following maize crops at only two of five on-farm sites. Without fertilizer, the groundnut rotation increased maize grain yields at five on-farm sites by an average of only 0.28 t ha−1. For the on-station groundnut and two subsequent years with maize, discounted net benefits (DNBs) over cash costs (seed and fertilizer) were greater for the rotation than for continuous maize, irrespective of whether or not inorganic fertilizer was applied. When labour costs were added, continuous maize plus fertilizer showed better returns than did the rotation, while the returns for the rotation and continuous maize without fertilizer were almost the same. On-farm the rotation was far less profitable. At only two sites, DNBs over cash costs were higher for the rotation whereas DNBs over all costs (including labour valued at a local casual-worker wage) were always negative or close to zero. At three sites, it was far more profitable to grow continuous maize, especially with fertilizer. These findings of low groundnut yield, marginal to zero profitability, and high labour cost of groundnut-maize rotations, support and explain the general trend by smallholder farmers to reduce groundnut area in Zimbabwe.

Dry matter production and yield of tea were recorded in a genotype × environment interaction experiment with four tea clones planted at four sites at a range of altitudes in Kericho, Kenya. Large responses to environment in dry matter production and yield were found. After 34 months from planting, the total plant dry matter production ranged from 18 to 22 t ha−1 across the sites. One clone, TN14-3, produced the most dry matter, reaching 29 t ha−1 at the lowest altitude, while clone S15/10 produced the least (15 t ha−1). Analysis of genotype × environment inter actions showed that clone TN14-3 was above average stability in dry matter production across the four sites, but below average in yield. Clone S15/10 showed above average stability for tea yield, but was below average in dry matter production.