Discussion 126 

Norms and recommendations for water resources use were developed in Central Asian countries during the era of collective farming systems. Despite reforms initiated since independence, these former practices still dominate the mind-set of the present practitioners. At the same time, the share of areas with medium and heavy salinization and shallow GW levels is continuously growing, indicating the inefficiency of such practices (Ibrakhimov et al., 2007). Shallow GW levels are dynamic in nature and play an important role in root-zone water balancing. Therefore, to update the norms for the

existing conditions is a prerequisite for sustainable water resources management in the area.

Water balance models are used to develop an optimum irrigation schedule (Brown et al., 1978; Odhiambo and Murty, 1996; Mishra, 1999). Pereira (1989) developed the IRRICEP and ISAREG water balance model and tested its performance against field data. Paulo et al. (1995) validated the IRRICEP model for selected sectors of the Sorraia irrigation system in Brazil. Khepar et al. (2000) developed a water balance model to predict deep percolation loss during wet and dry periods. However, these models are applied in the conditions where GW is deep, and this deep GW does not affect the root-zone water balance. George et al. (2004) reported that the problem of irrigation scheduling is complicated by a number of factors (weather, response of crop to irrigation, spatial and temporal variability of infiltration characteristics, soil water availability, etc.), and therefore a user-friendly irrigation-scheduling model is required.

Due to shallow GW conditions in the WUA Shomakhulum, the capillary rise contribution cannot be neglected (Forkutsa et al., 2009). None of the developed irrigation scheduling models can estimate this contribution, which is an important parameter in the context of the Khorezm conditions. Smith (1992) developed a simulation model CROPWAT, which has a user-friendly interface and therefore was selected for this study. Capillary rise contribution estimated against four different groundwater levels (see Chapter 5) by the HYDRUS-1D model (see Chapter 6) was successfully introduced in the CROPWAT model using the ‘User Defined Irrigation’ option in the model.

Results of the irrigation schedule for cotton under the business-as-usual scenario show that 9 to 13 % of the irrigation water can be saved as compared to following the norms in optimum irrigation schedule. The water saving in the optimum irrigation schedule is due to the introduction of the capillary rise contribution in the CROPWAT model. The capillary rise contribution (see Chapter 6) is within the range of the studies conducted in the region by Forkutsa et al. (2009) and the studies conducted in regions with the same climatic conditions, e.g., in Pakistan (Kahlown et al., 2005). Therefore, capillary rise has a significant impact on the surface water

The results of the scenarios show that by improving the irrigation efficiency to the target value (from scenario S-A to scenario S-D), the irrigation quota for cotton would increase by 8 to 15 % compared to the business-as-usual scenario. This increase in surface water demand is due to the decline of the GW levels. 19 % of the average surface water demand for all the crops can be met through the capillary rise contribution (Chapter 6).

Instead of improving the irrigation scheduling, overall irrigation efficiency in the WUA could be improved. As the major losses are during the field application, these could be reduced if the farmers were to implement water saving measures such as double-side irrigation, surge irrigation, siphon irrigation, or drip irrigation. The improvement in application efficiency would also become feasible through the use of laser leveling especially on the large fields. These approaches are successfully practiced in many part of the world, and water saving linked to these approaches has been successfully tested in Khorezm. For example, benefits of double-sided irrigation are mentioned for Khorezm conditions by Paluasheva (2005), while Masharipova (2009) suggested the use of laser leveling among the solutions to increase the field application ratio in Khorezm. However, these modern techniques are expensive, e.g., farmers cannot implement drip irrigation on their own and therefore need help from the land and water resources management institutions. Intensified maintenance of canals would lead to higher conveyance efficiency. However, lining is not recommended due to the shallow GW, as lining requires plentiful resources and to maintain the lining with fluctuating GW levels needs additional resources. An increase in irrigation network efficiency can also be realized by reducing operational losses. Better coordination of operational activities would avoid or at least reduce the overflow of water from the canals into the drains.

By improving the irrigation efficiency to the target value, 36 % surface water can be saved. However, this would lower the GW levels, and consequently cause an 11 % reduction in the capillary rise contribution. The decline in GW level is not a feasible option for the farmers or for the water management institutes, because it not only provides the necessary soil moisture to the roots at the existing level, but is also needed as a safety net against the unreliable water supplies. When farmers are not supplied with the surface water, they block the drains and thus raise the GW levels. Forkutsa (2006)

reported that farmers grow cotton under shallow GW levels even when they have no surface water supplies. For the sustainability of agriculture in the Khorezm region under the current situation, a decline in GW levels is not a feasible option (Chapter 6).

Therefore, the impact of efficiency improvement on the current irrigation and drainage system assessed in this study clearly shows that compensation for the safety net function of the groundwater needs to be considered by institutional strengthening to make water supply at the farm level more reliable.