Irrigated agriculture counts for some 70% of total water withdrawals. It will therefore be of importance to continue with efforts to increase productivity of irrigation water use, while taking into account multiple other factors, including environment, related to irrigation. For example, much of the production in arid and semi-arid zones is based on the mining of groundwater resources (Plusquellec, 2002). However, even with the most effective measures in the field of water saving it will still be required to increase the withdrawals for irrigation combined with increase in storages. Related to the required increase in food production an increase of 10 - 15% would have to be considered (Van Hofwegen and Svendsen, 2000). Pollution of water resources and environmental concerns with respect to application of agro-chemicals may reduce the potential for their use in agriculture. This would have to be achieved by the stakeholders that are shown in principle in Figure 1.

To maintain food security or food self-sufficiency, many countries in the arid and semi arid zones have reached, or are already beyond their water carrying capacity: they use more than the renewable amount. Importing food as an alternative to producing it themselves will alleviate stress on their water resources and make water available for priority and more productive purposes. Financing food imports requires access to world markets, which for most of the least developed countries is difficult due to the subsidy and tariff systems of most developed and emerging countries, but also related to quality and food safety standards which are difficult to attain at the stage of development these countries are in.

Another aspect that plays a role here is the livelihood of the small-holder of landless farmers, especially in the emerging and least developed countries under the increasing pressures of scarcity. This aspect will have to be included when we discuss sustainable development of the rural areas.

Figure 1 Indicative schematisation of actors in agricultural water management (Schultz, 2001)

Large variability of land productivity (ton ha) exists between countries. Statistical records for different crops exist over a long period of years. This would represent Actual Yield at field level (Ya). An example of this variability over the years is reported in Figure 1 for the case of the maize crop. Each country has its main crop(s) that can be taken as reference to monitor rainfed land productivity. One criterion for global land productivity can be the selection of major staple crops (e.g., rice, wheat and maize). For each of these crops, a statistical record can be derived over time. A time baseline needs to be identified: we recommend in principle the average of the years 2005-07. The ponderate aggregation of the different countries yields would then represent the global yield. This can be monitored over time, departing from the base-line. The slope of the interpolating line of the yields over the years would indicate the % change of land productivity (departing from the base line).

Figure 2.Land productivity of maize for different countries and regions over the 1866-2005 period (from ‘A Comprehensive Assessment of Water Management in Agriculture’).

Each country (or sub-national though significant systems) then needs to assess the land-productivity gap (or yield-gap) between (Ya) and Potential Yield (Yp), i.e., the yield of an adapted cultivar when grown with the best management and without natural hazards (such as hail, frost or lodging) and without water, nutrient or biotic stress limitations. This yield gaps would provide the scope for increasing land productivity (e.g., where the yield-gap is largest there the progress for expected yield increase is likely to be faster). However, the different causes for the yield-gap need to be identified in order to derive the exploitable yield-gap and the typology and affordability of intervention measures that would allow an effective reduction of the overall yield-gap. In other words, causes can be infrastructural, economic/market induced, management or non-management related, etc. And for each of them there are different measures to be taken in order to reduce the yield-gap. Standard methodologies have been developed to assess yield-gaps at different scales (field, country, global).

The resulting yield increase as aggregation of the individual country will generate the expected land productivity increase at global level.