It's likely there will be a major paradigm shift, probably within the next 25 years in many areas around the world. Both urban and rural users will need to lower their expectations of how much water is available for them and how it is employed. In the US, the front range of Colorado is going through this now - urban demands for limited water have led to several thousands of acres of irrigated lands being idled. And southern California and central Arizona are not far behind. While the water crisis is most acute in the western US, it is also a big concern in the southeast, where aquifers are being mined, and it will only get worse.

My take home message is that the greatest potential for water conservation/savings will be achieved by deficit irrigation strategies - management techniques which limit the total water that a crop receives with great attention to the timing as well as the amount of the applications. This will be supported by better crop selection and plant breeding for drought tolerance.

Biotechnology is needed, but I believe that improved management of water offers a much greater possibility of reducing water consumption than biotechnology and breeding, at least in the short term. That said, it will take advanced irrigation technologies to fully implement deficit irrigation strategies. The resulting water savings can range from 10-50%, whereas the potential from genetic approaches alone appears to be of the order of 5-10%. The greatest yield increases from biotechnology will probably be in dryland crops where water is limited and not controllable.

Deficit irrigation strategies are yet to be developed for most crops, and will in any case change as biotechnology affects plant responses. These techniques include ways to match crop/plant needs with inputs over time, adjusting plant populations for various drought levels, varying inputs such as fertilizers, figuring out when stress is more easily tolerated with least yield loss, and means to minimize pests below economic thresholds. Excellent weed control is essential to reduce water losses.

Different species, and even varieties within species, can respond differently, and it may not be economically possible to deficit irrigate crops such as potatoes due to quality issues. However some fruits such as apples and wine grapes can do very well under specific tailored deficit programmes. Deficit irrigation strategies can be very effective in managing supplemental irrigations in humid as well as semi-arid areas as irrigation, by definition, is supplemental to rainfall - when limited water is applied at critical growth stages to minimize yield reductions due to drought.

Under deficit irrigation, the concept of application efficiency - the water available to the plant divided by the water delivered to the plant - has little value or meaning. That's because almost all the water applied is used by the plant. Water is only lost by evaporation from the soil or plant during its application, depending on the irrigation method. As a result we need to think in terms of water use efficiencies instead -maximizing yield per unit of water rather than the current concept of maximizing yield per unit of area. That's a major change.

Another way to save water is the spatial optimization of crops by changing where they are grown around the world. That means moving growing locations to where the plants are best adapted and grow most efficiently rather than cultivating them in many areas just because they are close to markets.

Advanced technologies - such as precision irrigation, site-specific management (where water and/or water-soluble nutrients are applied differentially across a field to spatially match plant needs), remote sensing, and within-field real-time sensor systems tied to decision support systems - collectively have great potential to reduce water quantity and quality problems in irrigated agriculture.

The use of real-time sensor-based irrigation scheduling techniques and precision applications of water through centre-pivot machines is the next step in the evolution of this technology. This should both result in substantial labour, water, and energy savings and minimize losses to groundwater and surface runoff. These savings and potential environmental benefits accrue both to the irrigation manager and, ultimately, the general populace. In the future, it's likely that both groups will perceive these benefits as increasingly critical.