Effect of irrigation efficiency enhancement on water demand of date palms in a Tunisian oasis under climate change

Consideration of future change in water salinity is important for estimating irrigation demand in salinity-prone arid regions. Further, it is important to evaluate the contribution of irrigation efficiency enhancement to climate change resilience. Based on field measurements in 2019, a simulation approach from 2019 to 2050 was carried out in this study to investigate the impact of climate change and its consequences (i.e., change in water salinity) on the future gross irrigation demand of date palms and possible applied dose of water in a Tunisian oasis considering different irrigation efficiency enhancements. The estimation was done under very high (RCP 8.5), medium (RCP 6.0), and low (RCP 4.5) emission scenarios using the CROPWAT model. Results first showed an increase in gross irrigation requirement under inefficient surface irrigation (37% efficiency) from 3,340 mm year 1 in 2019 to 3,588–3,642 mm year 1 in 2050 for different climate change scenarios. This significant increase is mainly attributed to a significant change in climate variables and a high increase in water salinity. Second, considerable water savings (up to 1,980 mm) can be achieved if surface irrigation efficiency increases from the current value of 37–70%. Finally, much water can be saved only by reducing the overdose amount of water.


INTRODUCTION
Annual evapotranspiration (ET) greatly exceeds the annual rainfall (R), and the R/ET ratio is less than 0.65 in dry regions (Huang et al. ). During recent decades, the rapid growth of population (≈ 2.1 billion people, growth rate ≈ 18% over 2008-2018; IPCC ) coupled with on-going socio-economic development has led to increased dependence on irrigated agriculture and agricultural intensification to satisfy growing food demand. Consequently, a significant pressure on water resources has become evident over recent years, especially in water-stressed arid regions (Xie et al. ). Because of increasing air temperature and changing drought and rainfall patterns, climate change is intensifying this pressure and presents foreseeable water governance challenges (Al-Faraj et al. Much research has shown that water losses from irrigated fields (e.g., conveyance loss, distribution loss, evaporation, seepage, percolation, and run-off) can contribute to an important increase in water demand (e.g., Nikkami ; Ali & Klein ; Nassah et al. ). For example, Nassah et al.
() concluded that a water loss of 37% from deep percolation contributed to increasing the irrigation demand of citrus crops by more than 11%. Accordingly, the reduction of water losses is likely to decrease irrigation water demand, such as by enhancement of irrigation efficiency (i.e., optimal crop productivity with less irrigation water), improvement of the knowledge-based irrigation demand management, and development of proper irrigation management practices (irrigation method, timing, water quantity, and irrigation duration). As reported by IPCC (), such improved practices will need to be supported in many arid regions due to their outdated irrigation networks that cause inefficient irrigation water management, largely because farmers lack the necessary information and skills and the absence of water conservation irrigation technologies.
Based on climate data projections and their inclusion in crop water use simulation models (e.g., CROPWAT), it was recently noted that many studies have been conducted to evaluate the effects of climate change on irrigation water demand (e.g., Khattak et al. ; Zhang et al. ). However, as suggested by IPCC (), under expected high risks of climate change (such as drought and rainfall decrease), continued research is needed to identify the implications of irrigation efficiency enhancement on agricultural water demand, and to improve on water resources management in arid regions. Also, irrigation water salinity plays a key role in improving the knowledge-based irrigation demand management (Haj-Amor et al. ). Accordingly, future changes in irrigation water salinity as caused by pumping groundwater of increasing salinity must be considered as well, to evaluate its effect on agricultural water demand (Letey et al. ).
In the Tunisian oases, date palm (Phoenix dactylifera) is an important horticultural crop and is at the heart of oasis cultivation systems. For those oases, it is expected that date palm water demand will increase over the next few decades due to rising temperatures and evaporative demand (Haj-Amor et al. ).
Climate change may increase this water demand because of a projected decrease in effective rainfall and an increase in crop ET. Therefore, the effects of climate change and its consequences (i.e., irrigation water salinity) on irrigation water demand should be quantified to improve on irrigation efficiency, water conservation, and for long-term water resources development and planning. In this context, the objectives of this research were (1) to estimate the irrigation water demand of date palms during 2019-2050 in Metouia Oasis (i.e., one of the important oases in Tunisia) utilizing downscaled climate scenarios, (2) to identify the implications of different levels of irrigation water salinity on water demand, and (3) to assess implications of different levels of irrigation efficiency on irrigation water demand. The 2019-2050 period was selected so that necessary information is becoming available toward the implementation of irrigation water management options in the Tunisian oases for the next decades. The study also provides recommendations to enhance water availability for irrigation under climate change.

Study area
This research work was based on 310 ha of date palm fields located in the Tunisian coastal oasis called Metouia Oasis ( Figure 1). The oasis area is divided into 155 equally sized plots of farmland. Each plot is rectangular in shape, typically 100 m long and 50 m wide. The date palm (Phoenix dactylifera) is the main crop cultivated and plays a key role in agricultural production, as in several Mediterranean countries. Even though the date palms are adapted to the harsh climate of the oasis (i.e., average yearly temperature ≈ 25.2 C, average yearly rainfall ≈ 162.2 mm year À1 ), enormous amounts of water (1,900-2,500 mm year À1 ) are required to produce commercial yields (Haj-Amor et al.

METHODOLOGY
The methodology involves nine successive steps ( Figure 2) as follows: Step 1: data collection during field visit  Company, UK) and found to be 4.2 dS m -1 . The main cultivating crop was date palm (Phoenix dactylifera), and the plant density was 100 plants ha -1 .
Step 2: collection of climate data The daily climate data including rainfall, temperature, wind speed, relative humidity (RH), and solar radiation were obtained from the Tunisia Meteorological Department.
These data are summarized in Table 1. The 2019 climatic data were used for estimating the current irrigation demand in Metouia Oasis (details in step 4), whereas the data of the 1964-2013 period were mainly used as a reference period for validating future climate projections (details in step 5).
Step 3: assessment of the current irrigation efficiency The irrigation efficiency was determined in 2019 for one research plot after each irrigation event. In the Metouia Oasis region, irrigation water is conveyed from the well to the irrigated plots through a distribution network of concrete canals. Because water conveyance losses from the well to the plot were low, the irrigation efficiency (E a in %) was calculated by the ratio of the volume of irrigation water stored in the root zone (V s in mm) and the total volume of irrigation water delivered to the plot (V f in mm) (Burt et al. ): (1)  Table 2.
Step 4: estimation of the current irrigation demand where ET 0 is the reference evapotranspiration (mm day À1 ); Δ is the saturated vapor pressure slope (kPa C À1 ); G is the heat flux density of soil (MJ m À2 day À1 ); R n is the net radiation (MJ m À2 day À1 ); T is the mean temperature ( C); u 2 is the average daily wind speed (m s À1 ); e s is the saturation vapor pressure (kPa); e a is the actual vapor pressure (kPa); (e s -e a ) is the deficit of vapor pressure (kPa); and γ is a psychrometric constant (kPa C À1 ). GIR was calculated based on the following four successive steps: (1)    Step 5: projection of the future climate variables The climate data (i.e., temperature, rainfall, wind speed, and NorESM1-M) were evaluated to identify the optimal model which can provide the best climate data modeling.
The performances of these models in simulating climatic data (e.g., temperature and rainfall) for the 1964-2013 reference period were assessed against measurements.
Using various statistical indicators (e.g., relative root mean square error and Kling-Gupta efficiency), results showed that HadCM3 was the desired optimal model to be applied for the climatic forecasting. Three   Step 6  Step 7: analysis of trends of changes in water loss and overdose amount Water loss was defined as the difference between gross (GIR) and net irrigation requirement (NIR), i.e., GIR-NIR.
Moreover, we define overdose as the amount of water where x i and x j are time series elements with i < j. Furthermore, the ratio of simulated leaching requirement (LR) and total loss of water (i.e., the difference between applied dose and NIR) was estimated to illustrate the importance of salinity management in relation to irrigation water management.
Step 8: estimation of the future irrigation demand with improved irrigation efficiencies In this step, GIR was re-estimated by inputting three variations of irrigation efficiency: 50, 60, and 70%. These percentages were adapted from FAO () for efficient surface irrigation.
50 and 60% refer to efficient surface irrigation, whereas 70% refers to very efficient surface irrigation which could be ensured through perfect drainage and irrigation conditions.
Here, the objective is to evaluate the effect of surface irrigation efficiency enhancement on future GIR.
Step 9: summarize and analyze water availability for irrigation In this final step, the possible and effective ways to reduce irrigation water loss and increase irrigation efficiency were identified, using technical reports provided by the ADO, Tunisia.

Current irrigation efficiency
Twelve irrigation events were applied in the investigated irrigated plot of the Metouia Oasis, following one irrigation per month. Accordingly, 12 values of irrigation efficiency (E a ) were identified. These values are shown in Figure 6. Based As suggested by FAO (), good drainage conditions and proper maintenance of the irrigation system may contribute to a significant increase in the surface irrigation efficiency (up to 70%). In addition to these four factors, based on field measurements, it was observed that during each irrigation event, the applied water dose was much more than the required dose (see data in Figure 7). These overdoses could be the main factor behind the inefficient

Current irrigation demand
In 2019, the total irrigation demand of date palms (GIR) in Metouia Oasis was calculated based on the observed climatic variables (data shown in Table 1), measured irrigation water salinity (data shown in Figure 4), and irrigation efficiency (data shown in Figure 6). Annual and monthly values of GIR are presented in Figure 8. From this latter figure, it was observed that the annual GIR was 3,347 mm. Also, the monthly GIR varied from 259 mm in January (coldest month) to 321 mm in August (hottest month). The important difference between these two values (i.e., 259 and 321 mm) demonstrated the contribution of climate conditions to GIR variation. Differences in monthly irrigation demand are common in many date palm-growing countries such as Algeria (Mihoub et al. ) and Syria (Brunel et al. ). In addition to the climate conditions, the high value of annual GIR (3,347 mm) could also be related to the high LR (¼14%), as a consequence of the high salinity of the applied water (data shown in Table 4) and the low value of irrigation efficiency (data shown in Figure 6). However, as mentioned in Figure 8, the big difference between the annual net irrigation demand (1,389 mm) and the annual GIR (3,347 mm) confirmed that irrigation efficiency (hence water losses within the irrigated fields) may have a larger impact (i.e., increasing trend) on the GIR than climate and LR factors. Accordingly, water losses from the irrigated fields (e.g., overdoses, percolation, and run-off) must be reduced in order to decrease the water demand of date palms. For that purpose, farmers must be provided with the water conservation skills, as currently farmers over-irrigate (see data in Figure 8).

Future irrigation demand
Under various climate change scenarios (RCP 8.5, RCP 6.0, and RCP 4.5), Figure 9 shows the variation trends of future GIR in the study area during 2019-2050. Under these scenarios, a continuous increase in future GIR is predicted from 2019 to 2050. GIR will increase from 3,347 mm year À1 in 2019 to 3,588-3,640 mm year À1 in 2050, depending on the simulated climate change scenario. The data in Figure 9 revealed also that the highest increase in future GIR is noted for the highest greenhouse gas emissions scenario (RCP 8.5), whereas the lowest increase is noted for the lowest scenario (RCP 4.5). This confirms the important contribution of climate change, especially increasing temperature and decreasing rainfall (data shown in Table 5)

Trends of water loss and overdose with the current level of practice
The difference between the gross irrigation requirement and net irrigation requirement (i.e., GIR-NIR) reflects the amount of water loss which can only be managed by improving the irrigation system efficiency. The difference between applied dose and gross irrigation requirement (i.e., AD-GIR) is an additional loss of water which can easily be managed by measured and controlled application of irrigation of  mm year -1 ) in overdose, if farmers continue to apply same overdose/NIR ratio as present for the future periods for all scenarios (see Figure 10 for RCP 8.5). Hence, it is extremely important to ensure controlled application of irrigation with proper measurement techniques to reduce the overdose amount. With possible increase in NIR in the future, it is therefore essential to ensure more careful application of This finding can more strongly be observed by increasing trends of LR/(AD-NIR) in Figure 10. While increased overdose amount (AD-GIR) indicates that improper irrigation application (i.e., over-irrigation) may contribute to further future water shortage.

Irrigation efficiency enhancement and irrigation demand
The analysis of enhancing surface irrigation efficiency at various levels (50, 60, and 70%) showed that considerable water saving can be ensured over 2019-2050 under the simulated three climate change scenarios. Figure 11 illustrates only the results of the RCP 6.0 scenario; however, similar results were obtained for the other climate change scenarios (RCP 8.5 and RCP 4.5). The data presented in Figure 11 revealed that the future annual GIR is projected to decrease Institute. The results of these experiences (unpublished technical reports) are available in the ADO, Tunisia. In Table 6, the major ways and techniques that can help us to reduce irrigation water loss and increase irrigation efficiency are illustrated. However, these recommended practices of Table 6 are not yet adopted in the Tunisian oases due to the difficulty of integrating smart irrigation techniques with traditional irrigation networks. Moreover, the amount of investment required is often too high for it to be profitable for small-holder farmers.

Limitations and scope for future research
The irrigation demand may vary as a result of land-use change (e.g., an increase in the date palm area) which was low irrigation efficiency value (E a ¼ 37%). Accordingly, the results of this study showed that inefficient irrigation in combination with climate change and increasing groundwater salinity may significantly affect water resource availability.
Moreover, irrigation overdose by local farmers may hamper water resource conservation. Furthermore, this study revealed that despite increasing water demands by future climatic changes, considerable water savings can be achieved if surface irrigation efficiency increases from a current value of 37 to 70%. Therefore, the implementation of improved irrigation efficiency, the continuous control of the salinity of applied water, and the controlled application of irrigation water by farmers are required for decreasing GIR and solving the encroaching problem of water scarcity.