Rainfed agriculture in sub-Saharan Africa (SSA) is the mainstay of the continent's food and feed production. Nearly 90% of staple food and feed production comes from, and will continue to come from, rainfed agriculture (Rosegrant et al., 2002). In spite of this, investment in this vital production system, and hence its productivity, has stagnated. There are many complex and interrelated issues that contribute to this state of affairs. The outcomes of lack of investment and low production of rainfed agriculture reinforce each other leading to poverty traps and increased vulnerability of livelihoods to climatic and other shocks (World Bank, 2000). This has become well recognized and an emerging political will, both within and outside SSA, to support increased investment in rainfed agriculture appears to be gaining momentum (Sanchez et al, 2009).

The effect of seed priming and micro-dosing (the application of small amounts of mineral fertilizers) was studied in sorghum and pearl millet in on-station and on-farm experiments for three seasons under rainfed conditions in the North Kordofan State, western Sudan. Seed priming consists of soaking the seeds for eight hours in water prior to sowing. Seed priming increased sorghum grain yield in the on-station experiments across three seasons from 482 kg ha−1 to 807 kg ha−1. Micro-dosing of 0.3 g, 0.6 g and 0.9 g NPK fertilizer (17-17-17) per pocket increased sorghum grain yield by 50.4, 68.8 and 109.7% respectively compared to the control. Seed priming did not significantly increase pearl millet yield while the micro-doses of 0.3, 0.6 and 0.9 g fertilizer increased millet yield by 31.3, 30.7 and 47% respectively. On-farm seed priming increased sorghum yields by 32.6% while seed priming plus 0.3 g fertilizer increased yields by 69.5%. For millet, the corresponding yields increased by 29.8% and 71% respectively. Fertilizer use efficiency for both crops increased remarkably with seed priming, although this effect was more apparent in sorghum than in millet. In sorghum, seed priming and the application of 0.9 g fertilizer per pocket increased the gross margin from 49.5 to 206.5 US$/ha. For millet, the gross margin increased from 44.9 in the control to 90.0 US$/ha with the combination of seed priming and 0.3 g fertilizer per pocket. These technologies are simple to apply; they offer low financial cost and low risk and are affordable for resource-poor farmers.

Maize (Zea mays) seed orientation at planting can influence emerging leaf angle. If leaf angle were consistent among plants without leaf overlap, large numbers of these bilaterally symmetrical plants could be arranged to optimize light interception and possibly increase grain yields or maintain grain yield with a lower population. The objectives were to evaluate the effect of seed orientation in soil on the angle of maize leaves relative to the planted row and on emergence rate. Seeds were planted 2.5 cm deep in diverse combinations of flat, cotyledon down, cotyledon up, on their side, radicle up and radicle down. Each seed orientation was repeated 10 times. Data on days to emergence and leaf angle were collected. In three experiments, maize seeds planted flat with the cotyledon up resulted in homogenous and faster emergence, and maize leaves aligned perpendicular to the direction of the maize row. Similar results were achieved with maize seeds planted parallel to the row with the radicle down. Random placement of maize seeds resulted in random orientation of maize leaves and lower emergence rates. The effects of controlled leaf geometry could facilitate planting higher populations with the potential for increasing grain yield and/or allow the maintenance of grain yields while reducing seed rates.

The results of research on the water relations and irrigation needs of sugar cane are collated and summarized in an attempt to link fundamental studies on crop physiology to irrigation practices. Background information on the centres of production of sugar cane is followed by reviews of (1) crop development, including roots; (2) plant water relations; (3) crop water requirements; (4) water productivity; (5) irrigation systems and (6) irrigation scheduling. The majority of the recent research published in the international literature has been conducted in Australia and southern Africa. Leaf/stem extension is a more sensitive indicator of the onset of water stress than stomatal conductance or photosynthesis. Possible mechanisms by which cultivars differ in their responses to drought have been described. Roots extend in depth at rates of 5–18 mm d−1 reaching maximum depths of > 4 m in ca. 300 d providing there are no physical restrictions. The Penman-Monteith equation and the USWB Class A pan both give good estimates of reference crop evapotranspiration (ETo). The corresponding values for the crop coefficient (Kc) are 0.4 (initial stage), 1.25 (peak season) and 0.75 (drying off phase). On an annual basis, the total water-use (ETc) is in the range 1100–1800 mm, with peak daily rates of 6–15 mm d−1. There is a linear relationship between cane/sucrose yields and actual evapotranspiration (ETc) over the season, with slopes of about 100 (cane) and 13 (sugar) kg (ha mm)−1 (but variable). Water stress during tillering need not result in a loss in yield because of compensatory growth on re-watering. Water can be withheld prior to harvest for periods of time up to the equivalent of twice the depth of available water in the root zone. As alternatives to traditional furrow irrigation, drag-line sprinklers and centre pivots have several advantages, such as allowing the application of small quantities of water at frequent intervals. Drip irrigation should only be contemplated when there are well-organized management systems in place. Methods for scheduling irrigation are summarized and the reasons for their limited uptake considered. In conclusion, the ‘drivers for change’, including the need for improved environmental protection, influencing technology choice if irrigated sugar cane production is to be sustainable are summarized.

Crop-livestock systems in sub-Saharan Africa (SSA) are mostly rainfall-dependent and based on fragmented marginal lands that are vulnerable to soil erosion, drought and variable weather conditions. The threat of water scarcity in these systems is real, due to expanding demand for food and feed, climate variability and inappropriate land use (Amede et al., 2009). According to recent estimates, farming, industrial and urban needs in developing countries will increase water demand by 40% by 2030 (FAO, 2009). Water shortage is expected to be severe in areas where the amount of rainfall will decrease due to climate change. The lack of capacity of communities living in drought-prone regions to respond to market opportunities, climatic variability and associated water scarcity also results from very low water storage facilities, poverty and limited institutional capacities to efficiently manage the available water resources at local, national and basin scales. The spiral of watershed degradation causes decline in water budgets (Awlachew and Ayana, 2011), decreases soil fertility and reduces farm incomes in SSA (Amede and Taboge, 2007) and reduces crop and livestock water productivity (Descheemaeker et al., 2011). In areas where irrigated agriculture is feasible, there is an increasing demand for water and competition among different users and uses.

The effect of seed priming and micro-dosing in groundnut, cowpea and sesame was studied for three years in on-farm and on station experiments under rainfed agriculture in North Kordofan, Sudan. The on-station trials showed that seed priming increased groundnut pod and hay yields by 18% and 20% respectively. Micro-dosing of 0.3, 0.6 and 0.9 g fertilizer per pocket increased groundnut pod yield across the three years by 36.7, 67.6 and 50.8% respectively compared to the control. The highest yield increases were consistently obtained when micro-dosing was combined with seed priming. A combination of seed priming and micro-dosing of 0.6 g increased groundnut yield by 106%. Priming alone did not significantly affect sesame seed or hay yield, but micro-dosing of 0.6 g per pocket increased the grain yield by 38% over the control. Cowpea grain yield in the on-station experiments was not significantly affected by seed priming or micro-dosing. However, both seed priming and micro-dosing increased cowpea hay yield. In the on-farm trials, seed priming increased groundnut and cowpea yields by 18.2 and 25.5% respectively, and seed priming combined with 0.3 g fertilizer increased their yields by 42.2 and 54.5% respectively compared to the control. For sesame the yield increase after 0.3 g fertilizer per pocket was 46.3%. The economic analyses of the on-station experiments showed that the highest gross margin was obtained when combining seed priming with 0.6 g micro-dosing for all the crops. These results show that the combination of micro-dosing and seed priming has the potential to increase productivity and improve net return in the crops tested.

In the East African highlands, crop yields tend to increase with proximity of the farm plots to homesteads. Farmers identified soil erosion as the most detrimental cause of low crop yield in the outfields followed by soil compaction due to livestock trampling. The main objective of this study was to determine whether zai pits (i.e. small water harvesting pits) developed for dryland regions of the Sahel could increase crop yield and water productivity of degraded outfields in high rainfall areas, where mean annual rainfall exceeds 1300 mm but soil water infiltration is reduced by slope, low soil organic matter and hardpans. The pits were enlarged to resist strong runoff flows. The research was conducted over three years from 2004 to 2006. Potatoes and beans were used as test crops. Overall, compared to control plots, the zai pits, in combination with nitrogen (N) inputs, increased potato yields from 500% to 2000% (p ≤ 0.001). The pits contributed more to increased crop yield than N inputs. Similarly, bean yields from the zai pits were up to 250% higher. Crop water productivity was 300–700% higher with zai pits than with control plots. The income of farmers who used zai pits was up to 20-fold higher than the labour costs required to prepare them. Contrary to conventional wisdom, this study reveals that the major constraint of the outfields is not nutrient deficiency per se rather low soil water holding capacity, which hinders crop growth and efficient utilization of available nutrients.

Global food security is under threat by climate change, and the impacts fall disproportionately on resource-poor small producers. With the goal of making agricultural and food systems more climate-resilient, this paper presents an adaptation and mitigation framework. A road map for further agricultural research is proposed, based on the CGIAR Research Program on Climate Change, Agriculture and Food Security. We propose a holistic, integrated approach that takes into account trade-offs and feedbacks between interventions. We divide the agenda into four research areas, three tackling risk management, accelerated adaptation and emissions mitigation, and the fourth facilitating adoption of research outputs. After reviewing specific technical, agronomic and policy options for reducing climate change vulnerability, we acknowledge that science and good-faith recommendations do not necessarily translate into effective and timely actions. We therefore outline impediments to behavioural change and propose that future research overcomes these obstacles by linking the right institutions, instruments and scientific outputs. Food security research must go beyond its focus on production to also examine food access and utilization issues. Finally, we conclude that urgent action is needed despite the uncertainties, trade-offs and challenges.

Two long-term field experiments, each consisting of three treatments (organic compost treatment, chemical fertilizer treatment and an untreated control) were established in 1993 and 1997, respectively. Soil samples were collected from each plot in June 2004 and 2005 after crop harvest and were used to determine soil physical-chemical properties, biological and biochemical activity, and the nematode community. Soil physicochemical parameters, microbial biomass, biological activities and nematode communities were significantly influenced by long-term application of organic compost. In general, soil total organic carbon, dissolved organic carbon, total nitrogen, alkaline-hydrolysable nitrogen, available phosphorus, and available potassium, microbial biomass, basal respiration, urease activities, total number of nematodes and bacterial-feeding nematodes were significantly higher in the compost plots than in the chemical fertilizer and control plots at two experimental sites and two sampling dates. Soil bulk density and pH values were significantly lower in the compost plots. We conclude that soil physical-chemical properties, size and activity of soil microbial biomass, metabolic quotient (qCO2), urease activity, total number of nematodes and bacteria-feeding nematodes could be used as indicators of soil quality.

The results of research on the water relations and irrigation needs of coconut are collated and summarized in an attempt to link fundamental studies on crop physiology to drought mitigation and irrigation practices. Background information on the centres of origin and production of coconut and on crop development processes is followed by reviews of plant water relations, crop water use and water productivity, including drought mitigation. The majority of the recent research published in the international literature has been conducted in Brazil, Kerala (South India) and Sri Lanka, and by CIRAD (France) in association with local research organizations in a number of countries, including the Ivory Coast. The unique vegetative structure of the palm (stem and leaves) together with the long interval between flower initiation and the harvesting of the mature fruit (44 months) mean that causal links between environmental factors (especially water) are difficult to establish. The stomata play an important role in controlling water loss, whilst the leaf water potential is a sensitive indicator of plant water status. Both stomatal conductance and leaf water potential are negatively correlated with the saturation deficit of the air. Although roots extend to depths >2 m and laterally >3 m, the density of roots is greatest in the top 0–1.0 m soil, and laterally within 1.0–1.5 m of the trunk. In general, dwarf cultivars are more susceptible to drought than tall ones. Methods of screening for drought tolerance based on physiological traits have been proposed. The best estimates of the actual water use (ETc) of mature palms indicate representative rates of about 3 mm d−1. Reported values for the crop coefficient (Kc) are variable but suggest that 0.7 is a reasonable estimate. Although the sensitivity of coconut to drought is well recognized, there is a limited amount of reliable data on actual yield responses to irrigation although annual yield increases (50%) of 20–40 nuts palm−1 (4–12 kg copra, cultivar dependent) have been reported. These are only realized in the third and subsequent years after the introduction of irrigation applied at a rate equivalent to about 2 mm d−1 (or 100 l palm−1 d−1) at intervals of up to one week. Irrigation increases female flower production and reduces premature nut fall. Basin irrigation, micro-sprinklers and drip irrigation are all suitable methods of applying water. Recommended methods of drought mitigation include the burial of husks in trenches adjacent to the plant, mulching and the application of common salt (chloride ions). An international approach to addressing the need for more information on water productivity is recommended.

Low crop and livestock productivities in the mixed farming systems of Ethiopia hamper efforts to meet the increasing food demands from a stressed natural resource base. Important reasons for the low agricultural productivity are water scarcity and poor spatial and temporal rainfall distribution. Although improving agricultural water productivity would safeguard people's livelihoods and the environment, the lack of information on best bet interventions and strategies to achieve this impedes targeted decision making. Therefore, the aim of this study was to conduct an ex-ante evaluation of the potential effect of selected interventions on livestock water productivity (LWP) in mixed crop-livestock systems. Baseline data were collected from a water scarce area in the Ethiopian highlands. An analysis of productivity gaps and stakeholder interviews helped to identify promising interventions, which were categorized in three groups related to feed, water and animal management. A spreadsheet model was developed that was composed of the various production components of the farming system, their interactions and influencing factors. By linking water use for feed production with livestock products through the energy supplied by the feeds, the potential effect of interventions on LWP could be simulated. The evaluation showed that the various interventions targeting feed, water and animal management could result in LWP improvements ranging from 4 to 94%. Feed and energy water productivity increased particularly with interventions like fertilizer application, and the introduction of fodder trees, concentrates, improved food-feed crops, and soil and water conservation measures. Combining the different interventions led to a stronger improvement than any of the single interventions. The results of the evaluation can inform policy-makers and development actors on which best bets to promote and invest in.

The present study concerns identification of the most profitable and water and nitrogen use efficient best management practice (BMP) in a rice–wheat system using a combined approach of field experimentation and simulation. In the field study, two independent experiments, (1) effect of three transplanting/sowing dates, two cultivars and two irrigation regimes and (2) effect of four nitrogen (N) levels with four irrigation regimes, were conducted for two seasons of 2008–09 and 2009–10 at Punjab Agricultural University, Ludhiana, India. Integrating the treatments of the two independent field experiments, simulations were run with the CropSyst model. The BMP demonstrated was transplanting of rice on 20 June and sowing of wheat on 5 November, irrigation to rice at 4-day drainage period and to wheat at irrigation water depth/Pan–E (open pan evaporation) ratio of 0.9, and fertilizer N of 150 kg ha−1 to each crop for medium-duration varieties. This practice gave higher profit (35%), equivalent rice yield (16%), crop water productivity (15%), irrigation water productivity (51%), economic water productivity (34%) and economic N productivity (94%) than the existing practice by the farmers. The improvement in crop water productivity by shifting the transplanting/sowing date was due to reduction in soil water evaporation and increased transpiration and fertilizer N productivity through increased N uptake.

Early vigour in wheat (Triticum aestivum) is an important physiological trait to improve water-use efficiency and grain yield, especially on light soils in Mediterranean-type climates. Potential interactions for plant stand and seedling vigour between seed density and various seed quality treatments were examined for wheat grown in two experiments, conducted under controlled and field environments in Western Australia. Seed lots were graded into seed size classes and seed density fractions using saturated solutions of ammonium sulphate or sodium polytungstate. Dense seed improved plant stands or produced seedlings with greater early seedling vigour than their low-density counterparts in all three field environments. Artificial ageing reduced germination and emergence in the controlled environment. When grown in the field at Merredin, Western Australia, on the sandy soil, plant development was delayed with aged seed, and total leaf area and dry weight of plants were reduced. Fungicide application diminished total plant dry weight in sandy soils, but had a much larger detrimental effect when applied to aged and low-density seeds than normal seeds, retarding development, total leaf area and total plant dry weight. Our results indicate that an increase in seed density, particularly in small seed, can potentially improve plant stand and seedling vigour independently of seed size, and may be especially important for wheat grown on sandy soils of poor fertility and low water-holding capacity. The results also suggest consistency in seedling vigour may benefit from combined screening against small seed size and low seed density, which may also reduce the likelihood of adverse reactions to seed-applied fungicides. More attention should be paid to seed density as a valuable trait for improved reliability in plant stand and seedling vigour.

Leaf gas exchange and leaf water potential (Ψleaf) were measured seasonally on non-grafted and grafted Coffea arabica on Coffea canephora in the field to investigate whether grafting would be able to protect the carbon balance against the rise of in vapour pressure deficit (VPD) and air temperature (Tair) under future climate change. The net maximum photosynthetic rate obtained from the net photosynthesis (PN) curve as a function of photosynthetic photon flux density (PPFD) in wet and dry periods was used to estimate the integrated potential diurnal net CO2 assimilation (IPPN) around midday. The difference between IPPN and the integrated values of PN during diurnal courses (IPN) was measured to test grafting as suitable practice for minimizing midday depression of PN. Higher values of PN in grafted plants around midday showed that grafting was important even when environmental conditions were favourable in field conditions. Reduced susceptibility of grafted plants to midday depression was revealed by lower values of Ψleaf associated with higher values of PN and leaf transpiration (E) on sunny days in summer and spring, and by higher values of stomatal conductance (gs) around midday in autumn, winter and spring. The differences of E, gs, PN and Ψleaf between non-grafted and grafted plants were higher in dry periods in winter and spring. In addition, the ratio IPN/IPPN in grafted was double that in non-grafted plants around midday in sunny summer and in spring. Indeed, PN and gs of non-grafted plants showed higher dependence on VPD than grafted ones. The lower susceptibility of grafted plants to water stress demonstrated the graft efficiency for increasing positive components of leaf carbon balance of C. arabica in the field, especially under high VPD in projected future climate conditions.

We review the use and value of seasonal climate forecasting for agriculture in sub-Saharan Africa (SSA), with a view to understanding and exploiting opportunities to realize more of its potential benefits. Interaction between the atmosphere and underlying oceans provides the basis for probabilistic forecasts of climate conditions at a seasonal lead-time, including during cropping seasons in parts of SSA. Regional climate outlook forums (RCOF) and national meteorological services (NMS) have been at the forefront of efforts to provide forecast information for agriculture. A survey showed that African NMS often go well beyond the RCOF process to improve seasonal forecast information and disseminate it to the agricultural sector. Evidence from a combination of understanding of how climatic uncertainty impacts agriculture, model-based ex-ante analyses, subjective expressions of demand or value, and the few well-documented evaluations of actual use and resulting benefit suggests that seasonal forecasts may have considerable potential to improve agricultural management and rural livelihoods. However, constraints related to legitimacy, salience, access, understanding, capacity to respond and data scarcity have so far limited the widespread use and benefit from seasonal prediction among smallholder farmers. Those constraints that reflect inadequate information products, policies or institutional process can potentially be overcome. Additional opportunities to benefit rural communities come from expanding the use of seasonal forecast information for coordinating input and credit supply, food crisis management, trade and agricultural insurance. The surge of activity surrounding seasonal forecasting in SSA following the 1997/98 El Niño has waned in recent years, but emerging initiatives, such as the Global Framework for Climate Services and ClimDev-Africa, are poised to reinvigorate support for seasonal forecast information services for agriculture. We conclude with a discussion of institutional and policy changes that we believe will greatly enhance the benefits of seasonal forecasting to agriculture in SSA.

The results of research on the water relations and irrigation need of oil palm are collated and summarized in an attempt to link fundamental studies on crop physiology to drought mitigation and irrigation practices. Background information is given on the centres of origin (West Africa) and of production of oil palm (Malaysia and Indonesia), but the crop is now moving into drier regions. The effects of water stress on the development processes of the crop are summarized followed by reviews of its water relations, water use and water productivity. The majority of the recent research published in the international literature has been conducted in Malaysia and in Francophone West Africa. The unique vegetative structure of the palm (stem and leaves) together with the long interval between flower initiation and the harvesting of the mature fruit (ca. three years) means that causal links between environmental factors (especially water) and yield are difficult to establish. The majority of roots are found in the 0–0.6 m soil horizons, but roots can reach depths greater than 5 m and spread laterally up to 25 m from the trunk. The stomata are a sensitive indicator of plant water status and play an important role in controlling water loss. Stomatal conductance and photosynthesis are negatively correlated with the saturation deficit of the air. It is not easy to measure the actual water use of oil palm, the best estimates for mature palms suggesting crop evapotranspiration (ETc) rates of 4–5 mm d−1 in the monsoon months (equivalent to 280–350 l palm−1 d−1). For well-watered mature palms, crop coefficient (Kc) values are in the range 0.8–1.0. Although the susceptibility of oil palm to drought is well recognized, there is a limited amount of reliable data on actual yield responses to irrigation. The best estimates are 20–25 kg fresh fruit bunches ha−1 mm−1 (or a yield loss of about 10% for every 100 mm increase in the soil water deficit). These increases are only realized in the third and subsequent years after the introduction of irrigation and follow an increase in the number of fruit bunches as a result of an improvement in the sex ratio (female/total inflorescence production) and a reduction in the abortion of immature inflorescences. There is no agreement on the allowable depletion of the available soil water, or on the associated optimum irrigation interval. Drip irrigation has been used successfully on oil palm.

In Ethiopia, irrigation is mainly implemented in small-scale irrigation schemes, which are often characterized by low water productivity. This study reports on the efficiency and productivity of a typical small-scale irrigation scheme in the highlands of the Blue Nile, Ethiopia. Canal water flows and the volume of irrigation water applied were measured at field level. Grain and crop residue biomass and grass biomass production along the canals were also measured. To triangulate the measurements, the irrigation farm management, effects of water logging around irrigation canals, farm water distribution mechanisms, effects of night irrigation and water losses due to soil cracking created by prolonged irrigation were closely observed. The average canal water loss from the main, the secondary and the field canals was 2.58, 1.59 and 0.39 l s−1 100 m−1, representing 4.5, 4.0 and 26% of the total water flow respectively. About 0.05% of the loss was attributed to grass production for livestock, while the rest was lost through evaporation and canal seepage. Grass production for livestock feed had a land productivity of 6190.5 kg ha−1 and a water productivity of 0.82 kg m−3. Land productivity for straw and grain was 2048 and 770 kg ha−1, respectively, for teff, and 1864 kg ha−1 and 758 kg ha−1, respectively, for wheat. Water productivities of the crops varied from 0.2 to 1.63 kg m−3. A significant volume of water was lost from small-scale irrigation systems mainly because farmers' water application did not match crop needs. The high price incurred by pumped irrigation positively affected water management by minimizing water losses and forced farmers to use deficit irrigation. Improving water productivity of small-scale irrigation requires integrated interventions including night storage mechanisms, optimal irrigation scheduling, empowerment of farmers to maintain canals and proper irrigation schedules.

Rainfed wheat is frequently exposed to periods of water stress that generate low and variable grain yields. Field experiments (with studies in Tunisia and Morocco) carried out in the context of a European research project of co-operation with Mediterranean partner countries (WatNitMED) showed that nitrogen (N) fertilization may be a tool to increase productivity of rainfed wheat in Mediterranean environments. However, most farmers in Northern Africa do not fertilize their rainfed cereals. In the present study, we aimed to analyse whether the generally accepted positive yield response to N fertilization in rainfed Mediterranean conditions corresponds to actual advantages achieved in the fields of working farmers, attempting a further up-scaling of knowledge from field experiments to real fields. We attempted to apply research results to Tunisian working farmers’ fields by conducting a farm pilot experiment. The pilot experiment was conducted in two different regions (a low-yielding region and a relatively high-yielding region) of cereal production in Tunisia, where wheat production represents typical rainfed Mediterranean agro-ecosystems in North Africa. First, we compared the yield response to N fertilization against unfertilized conditions (a common situation for many of the farmers in North Africa), and secondly we compared what the farmers suggested as an optimal N fertilization practice in their fields against the WatNitMED's recommendation which was based on an N-fertilization scheme derived from field experiments from the European research project in Mediterranean conditions. The WatNitMED fertilization scheme suggested higher rates of fertilization than those considered optimal by farmers (on average 40 kg N ha−1 higher). Unfertilized grain yield across both locations ranged from about 1 to 3.5 Mg ha−1 (typical of farmers’ yields in the region), and fertilizing increased grain yields in most situations. Within the two alternative fertilization treatments, WatNitMED fertilization produced higher yields than the fertilization rate considered optimal by farmers. This trend was observed at the low-yielding location as well as at the high-yielding location. These responses demonstrated that fertilization in working farmers’ field conditions may be a reliable means of improving dryland wheat grain and straw yields. They also showed that rates of fertilization regarded as optimal by real farmers were below the optimum for these regions.

Rainfall variability, both within and between seasons, is reflected in highly variable crop growth and yields in rainfed agriculture in sub-Saharan Africa and results in varying degrees of weather-induced risk associated with a wide range of crop, soil and water management innovations. In addition there is both growing evidence and concern that changes in rainfall patterns associated with global warming may substantively affect the nature of such risk. Eighty-nine years of daily rainfall data from a site in southern Zambia are analysed. The analyses illustrate approaches to assessing the extent of possible trends in rainfall patterns and the calculation of weather-induced risk associated with the inter- and intra-seasonal variability of the rainfall amounts. Trend analyses use monthly rainfall totals and the number of rain days in each month. No simple trends were found. The daily data were then processed to examine important rain dependent aspects of crop production such as the date of the start of the rains and the risk of a long dry spell, both following planting and around flowering. The same approach is used to assess the risk of examples of crop disease in instances when a ‘weather trigger’ for the disease can be specified. A crop water satisfaction index is also used to compare risks from choices of crops with different maturity lengths and cropping strategies. Finally a different approach to the calculations of these risks fits a Markov chain model to the occurrence of rain, with results then derived from this model. The analyses shows the relevance of this latter approach when relatively short daily rainfall records are available and is illustrated through a comparison of the effects of El Niño, La Niña and Ordinary years on rainfall distribution patterns.

To assess genotype adaptability to variable environments, we evaluated five irrigated rice genotypes, three new varieties, WAS161, a NERICA, IR32307 and ITA344, and two controls: Sahel 108, the most popular short-duration variety in the region, and IR64. In a field experiment conducted at two locations, Ndiaye and Fanaye, along the Senegal River, rice was sown on 15 consecutive dates at one month intervals starting in February 2006. Yield (0–12.2 t ha−1) and crop cycle duration (117–190 days) varied with sowing date, genotype and site. Rice yield was very sensitive to sowing date and the associated temperature regimes. Spikelet sterility due to cold stress (T < 20 °C) was observed when the crops were sown in August (Ndiaye), September (Ndiaye and Fanaye) and October (Ndiaye and Fanaye), and heat stress (T > 35 °C) resulted in spikelet sterility when sowing took place in April (Ndiaye and Fanaye) and May (Fanaye). For all experiments the source and sink balance was quantified and showed that yield was most limited by sink size when sowing between July and October. Variety WAS 161 was least affected by genotype × environment interactions, resulting in lower interactive principal component values. An increase in minimum temperature of 3 °C could decrease spikelet sterility from 100 to 45%. These changes in temperature are likely to force rice farmers in the Senegal River to adjust the cropping calendar, e.g. to delay planting or to use heat-tolerant genotypes.