The HYDRUS-2D model was experimentally verified for water and salinity distribution during the profile establishment stage (33 days) of almond under pulsed and continuous drip irrigation. The model simulated values of water content obtained at different lateral distances (0, 20, 40, 60, 100 cm) from a dripper at 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 120, 140 and 160 cm soil depths at different times (5, 12, 19, 26 and 33 days of profile establishment) were compared with neutron probe measured values under both irrigation scenarios. The model closely predicted water content distribution at all distances, times and soil depths as RMSE values ranged between 0.017 and 0.049. The measured mean soil water salinity (ECsw) at 25 cm from the dripper at 30, 60, 90 and 150 cm soil depth also matched well with the predicted values. A correlation of 0.97 in pulsed and 0.98 in continuous drip systems with measured values indicated the model closely predicted total salts in the root zone. Thus, HYDRUS-2D successfully simulated the change in soil water content and soil water salinity in both the wetting pattern and in the flow domain. The initial mean ECsw below the dripper in pulsed (5.25 dSm -1 ) and continuous (6.07 dSm -1 ) irrigations decreased to 1.31 and 1.36 dSm -1 , respectively, showing a respective 75.1 and 77.6% decrease in the initial salinity. The power function [y = ax -b ] best described the mathematical relationship between salt removal from the soil profile as a function of irrigation time under both irrigation scenarios. Contrary to other studies, higher leaching fraction (6.4-43.1%) was recorded in pulsed than continuous (1.1-35.1%) irrigation with the same amount of applied water which was brought about by the variation in initial soil water content and time of irrigation application. It was pertinent to note that a small (0.012) increase in mean antecedent water content (h i ) brought about 8.25-9.06% increase in the leaching fraction during the profile establishment irrespective of the emitter geometry, discharge rate, and irrigation scenario. Under similar h i , water applied at a higher discharge rate (3.876 Lh -1 ) has resulted in slightly higher leaching fraction than at a low discharge rate (1.91 Lh -1 ) under pulsing only owing to the variation in time of irrigation application. The influence of pulsing on soil water content, salinity distribution, and drainage flux vanished completely when irrigation was applied daily on the basis of crop evapotranspiration (ETc) with a suitable leaching fraction. Therefore, antecedent soil water content and scheduling or duration of water application play a significant role in the design of drip irrigation systems for light textured soils. These factors are the major driving force to move water and solutes within the soil profile and may influence the off-site impacts such as drainage flux and pollution of the groundwater.

Keywords:Modelling HYDRUS Mandarin Fertigation Nitrate leaching Soil salinity s u m m a r y Estimation of all water fluxes temporally and spatially within and out of the crop root zone, and evaluation of issues like salinity and nutrient leaching, are necessary to fully appraise the efficiency of irrigation systems. Simulation models can be used to investigate these issues over several seasons when the cost of long term monitoring is prohibitive. Model results can be used to advise growers if improvements are required to various aspects of irrigation system operations. In this study, HYDRUS-2D was used to evaluate data measured during one season in a young mandarin (Citrus reticulata) orchard, irrigated with an intensive surface drip fertigation system. Water contents, salinities, and nitrate concentrations measured weekly in the field were compared with model predictions.The temporal mean absolute error (MAE) values between weekly measured and simulated water contents ranged from 0.01 to 0.04 cm 3 cm À3. However, modelling error (MAE) was slightly larger at 10 cm depth (0.04 cm 3 cm À3), as compared to greater depths (0.02-0.03 cm 3 cm À3). Similarly, the errors were larger in the surface soil layer (25 cm depth) for nitrate-nitrogen, NO 3), as compared to greater depths. The spatial and temporal soil solution salinity (EC sw ) and NO 3 À -N data showed accumulation of salts and nitrate within the soil up until day 150 of the simulation (December, 2006), followed by leaching due to high precipitation and over irrigation at later times. Only 49% of applied water was used by the mandarin trees, while 33.5% was leached. On the other hand, the simulation revealed that a significant amount of applied nitrogen (85%) was taken up by the mandarin trees, and the remaining 15% was leached. The results indicate that the irrigation and fertigation schedule needs modifying as there was overwatering from December onwards.Different permutations and combinations of irrigation and fertigation scheduling were evaluated to optimise the water and nitrogen uptake and to reduce their leaching out of the crop root zone. Slightly higher nitrogen uptake (1.73 kg ha À1) was recorded when fertigation was applied second to last hour in an irrigation event, as compared to applying it earlier during an irrigation event. Similarly, a 20% reduction in irrigation and N application produced a pronounced reduction in drainage (28%) and N leaching (46.4%), but it also decreased plant N uptake by 15.8% and water uptake by 4.8%, and increased salinity by 25.8%, as compared to the normal practice. This management would adversely impact the sustainability of this expensive irrigation system. However, reducing only irrigation by 30% during the 2nd half of the crop season (January to August) reduced drainage and N leaching by 37.2% and 50.5%, respectively, and increased N uptake by 6.9%. Such management of irrigation would be quite promising for the sustainability of the entire system. It is concluded that judicious manipulations of irrigation and fer...

No abstract

a b s t r a c tAccurate estimation of evapotranspiration (ET) and its partitioning into transpiration and evaporation is fundamental for improving water management practices in water-limited environments and under deficit irrigation conditions. This investigation was conducted to estimate the water balance and ET components of subsurface drip (SDI) irrigated Chardonnay wine grapes for two seasons (2010-2011 and 2011-2012) using a numerical model (HYDRUS-2D). Treatments involved the application of different volumes [51% (I 1 ), 64% (I 2 ), 77% (I 3 ), and 92% (I 4 ) of normal application] of water for irrigation. A modified version of the FAO-56 dual crop coefficient approach was used to generate daily transpiration and evaporation as inputs to the HYDRUS-2D model. The calibrated and validated model produced estimates of actual evapotranspiration (ET Cact ), actual transpiration (T pact ), and actual evaporation (E sact ), and deep percolation under varied irrigation applications. The model-simulated values were then used to estimate actual crop coefficients (K cact and K cbact ), and water productivity of wine grape under different deficit irrigation conditions. Seasonal ET Cact simulated by HYDRUS-2D for different treatments varied between 239 and 382 mm. However, seasonal evaporation accounted for 44-59% of seasonal ET Cact losses in different treatments. The modelled daily transpiration rate in I 4 treatment (T p4act ) varied from 0.11-2.74 mm/day. Deep percolation accounted for 35-40% of the total water applied by rainfall and irrigation. The mean value of actual crop coefficient (K cact ) estimated by HYDRUS-2D simulated ET C act over the two seasons was 0.27, which matched with other investigations. Similarly, values of K cbact for initial, mid and end stages were 0.13, 0.27 and 0.14, respectively. Monthly values of evaporation coefficient (K e ) ranged from 0.1 to 0.32, with a mean value of 0.18. Water productivity with respect to ET losses (WPET C ) ranged from 5.9 to 6.2 kg/m 3 of water use. However, water productivity for transpiration (WPT C ) almost doubled as compared to WPET C in all treatments. The impact of deficit irrigation on berry juice composition (Brix, pH and titratable acidity) was lower than the inter-seasonal variability. These results can help develop better irrigation management strategies for SDI irrigated wine grapes under water scarce conditions.

No abstract