Rainfed agriculture plays and will continue to play a dominant role in providing food and livelihoods for an increasing world population. We describe the world's semi-arid and dry sub-humid savannah and steppe regions as global hotspots, in terms of water related constraints to food production, high prevalence of malnourishment and poverty, and rapidly increasing food demands. We argue that major water investments in agriculture are required. In these regions yield gaps are large, not due to lack of water per se, but rather due to inefficient management of water, soils, and crops. An assessment of management options indicates that knowledge exists regarding technologies, management systems, and planning methods. A key strategy is to minimise risk for dry spell induced crop failures, which requires an emphasis on water harvesting systems for supplemental irrigation. Large-scale adoption of water harvesting systems will require a paradigm shift in Integrated Water Resource Management (IWRM), in which rainfall is regarded as the entry point for the governance of freshwater, thus incorporating green water resources (sustaining rainfed agriculture and terrestrial ecosystems) and blue water resources (local runoff). The divide between rainfed and irrigated agriculture needs to be reconsidered in favor of a governance, investment, and management paradigm, which considers all water options in agricultural systems. A new focus is needed on the meso-catchment scale, as opposed to the current focus of IWRM on the basin level and the primary focus of agricultural improvements on the farmer's field. We argue that the catchment scale offers the best opportunities for water investments to build resilience in smallscale agricultural systems and to address trade-offs between water for food and other ecosystem functions and services.

No abstract

No abstract

Considering the persistently growing pressure on finite freshwater and soil resources, it becomes increasingly clear that the challenge of feeding tomorrow's world population is, to a large extent, about improved water productivity within present land use. Rainfed agriculture plays a critical role in this respect. Eighty percent of the agricultural land worldwide is under rainfed agriculture, with generally low yield levels and high on-farm water losses. Ninety-five percent of current population growth occurs in developing countries and a significant proportion of these people still depend on a predominantly rainfed-based rural economy. This chapter presents the agrohydrological rationale for focusing on water productivity in rainfed agriculture, identifies key management challenges in attempts to upgrade rainfed agriculture and presents a set of field experiences on system options for increased water productivity in smallholder farming in drought-prone environments. Implications for watershed management are discussed, and the links between water productivity for food and securing an adequate flow of water to sustain ecosystem services are briefly analysed. The focus is on sub-Saharan Africa, which faces the largest food-deficit and water-scarcity challenges. The chapter shows that there are no agrohydrological limitations to doubling or even quadrupling on-farm staple-food yields, even in drought-prone environments, by producing more 'crop per drop' of rain. Field evidence is presented suggesting that meteorological dry spells are an important cause of low yield levels. It is hypothesized that these dry spells constitute a core driving force behind farmers' risk-aversion strategies. Risk aversion also contributes to the urgent soil-fertility deficits resulting from insignificant investments in fertilizers. For many smallholder farmers in the semiarid tropics, it is simply not worth investing in fertilizers (and other external inputs) so long as the risk for crop failure remains a reality every fifth year and the risk of yield reductions every second year. These high risks are associated with periodic water scarcity during the growing season (i.e. not necessarily cumulative water scarcity). Results are presented from on-farm agrohydrological field research with innovations in water harvesting and conservation tillage among smallholder farmers in semiarid rainfed farming systems in Burkina Faso, Kenya and Tanzania. These results indicate that upgrading rainfed production systems through supplemental irrigation during short dry spells can lead to large increases in water productivity. Downstream implications of increased upstream withdrawals of water for upgrading of rainfed food production are discussed. Finally, it is argued that some of the most exciting opportunities for water-productivity enhancements in rainfed agriculture are found in the realm of integrating components of irrigation management within the context of rainfed farming, e.g. supplemental or microirrigation for mitigating the effects of dry spells. Combining such practices with management strategies that enhance soil infiltration and improving water-holding capacity and the potential of water uptake of plants can have a strong impact on agricultural water productivity. This suggests that it is probably time to abandon the largely obsolete distinction between irrigated and rainfed agriculture, and instead focus on integrated rainwater management.

No abstract

No abstract

The paper describes a hydrological model for agricultural water intervention in a community watershed at Kothapally in India, developed through integrated management and a consortium approach. The impacts of various soil and water management interventions in the watershed are compared to no-intervention during a 30-year simulation period by application of the calibrated and validated ARCSWAT 2005 (Version 2.1.4a) modelling tool. Kothapally receives on average 800 mm rainfall in the monsoon period. 72 per cent of total rainfall is converted as evaporation and transpiration (ET), 20 per cent stored by groundwater aquifer and eight per cent exported as outflow from the watershed boundary in current water interventions. ET, groundwater recharge and outflow under no intervention conditions are found to be 64 per cent, nine per cent and 19 per cent, respectively. Check-dams helped in storing water for groundwater recharge, which can be used for irrigation, as well minimizing soil loss. In-situ water management practices improved the infiltration capacity and water holding capacity of the soil, which resulted in increased water availability by 10-30 per cent and better crop yields compared to no intervention. Water outflows from the developed watershed were more than halved compared to no intervention, indicating potentially large negative down-stream impacts if these systems were to be implemented on a larger scale. On the other hand, in the watershed development program sediment loads to the streams were less than one tenth. It can be concluded that the hydrological impacts of large scale 2 implementation of agricultural water interventions are significant. They result in improved rain-fed agriculture and improved productivity and livelihood of farmers in upland areas while also addressing the issues of poverty, equity and gender in watersheds. There is a need for case specific studies of such hydrological impacts along with other impacts in terms of equity, gender, sustainability and development at the meso-scale.

No abstract