Deficit irrigation under water stress and salinity conditions: The MOPECO-Salt Model

Agricultural Water Management 98 (2011) 1451–1461

Contents lists available at ScienceDirect

Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat

Deficit irrigation under water stress and salinity conditions: The MOPECO-Salt Model A. Domínguez a,∗ , J.M. Tarjuelo a , J.A. de Juan a , E. López-Mata a , J. Breidy b , F. Karam c a b c

Centro Regional de Estudios del Agua (CREA), Universidad de Castilla-La Mancha, Ctra. de las Pe˜ nas, km 3.2, 02071 Albacete, Spain Department of Plant Breeding, Lebanese Agricultural Research Institute, P.O. Box 287, Zahleh, Lebanon International Center for Agricultural Research in the Dry Areas, P.O. Box 2897, Aleppo, Syria

a r t i c l e

i n f o

Article history: Received 15 October 2010 Accepted 26 April 2011 Available online 17 May 2011 Keywords: Salinization Leaching fraction Water stress AquaCrop Solanum tuberosum L. Allium cepa L.

a b s t r a c t In both arid and semi-arid areas the use of saline water for irrigation is a common practice, even though it may cause a drop in crop yield and progressive soil salinization. In order to determine the most suitable irrigation strategy for higher yield, profitability, and soil salinity management of certain crops, the MOPECO-Salt Model has been developed. This model was first validated in the Eastern Mancha Agricultural System in Albacete (Spain) through a test carried out on onion crop in April–September 2009, where the simulated yield was 2% lower than the observed one. The model was then tested at Tal Amara Research Station in the Central Bekaa Valley Agricultural System (Lebanon) using data from a 5-year experiment on the effects of deficit irrigation on two cultivars of potato (Spunta: July–October 2001, and June–September 2002; and Agria: March–August 2004, 2005, and 2007). Furthermore, these results were compared with those obtained through AquaCrop, which does not currently assess crop response to salinity. Differences between observed and simulated yields were lower than 3% for MOPECO-Salt and up to 12% for AquaCrop. According to findings from simulations, the irrigation strategies without leaching fraction employed in both areas are remediable since the off-season rainfall is sufficient to wash out soluble salts supplied with irrigation water. Results showed that as much as 14.4% water could be saved when this strategy was adopted for onion crops. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Salinity has long been identified as a major threat to agriculture, leading to policies aimed at improving irrigation and drainage practices in many parts of the world (FAO, 2007). In arid and semiarid areas, where 25% of the irrigated land is currently affected by salts (Arshi et al., 2010), salinity has become a great hindrance for agricultural production. So, a wide range of processes such as seed germination, seedling growth and vigor, vegetative growth, flowering and fruit set are affected by high salt concentration in the soil, ultimately causing poor crop quality and diminished economic yield (Hasegawa et al., 2000; Sairam and Tyagi, 2004; Qadir et al., 2008). With this in mind, it is of utmost importance to address this serious environmental issue. Many authors have modelled the behaviour of crops when undergoing water or salt stresses (Childs and Hanks, 1975; Letey et al., 1985; Bresler, 1986; Majeed et al., 1994; Castrignanò et al., 1998; Allen et al., 1998; Ferrer-Alegre and Stockle, 1999; García et al., 2006; Pereira et al., 2007). Three of these models were used

∗ Corresponding author. Tel.: +34 967599200; fax: +34 967599269. E-mail address: [email protected] (A. Domínguez). 0378-3774/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.agwat.2011.04.015

in this study: Allen et al. (1998) (Model 1) relate actual crop evapotranspiration (ETa ) to soil water and soluble salts content in the root zone. Inputs required in this model are the Readily Available soil Water in the root zone (RAW) and the depletion factor (p), which is a fraction of Total Available soil Water in the root zone (TAW) that a crop can extract from this zone without suffering water stress. According to Allen et al. (1998), the main limitation of this model is the assumption that RAW and p are kept constant for any soluble salts content in the root zone. In order to solve this handicap Pereira et al. (2007) developed a model (Model 2), based on Model 1, which allows balancing the values of RAW and p depending on the variable soil salinity conditions. Finally, García et al. (2006) (Model 3) relate potential evapotranspiration of the crop (ETm ) to osmotic potential and matric pressure head in the root zone. The MOPECO model (Ortega et al., 2004; López-Mata et al., 2010) is a tool for identifying optimal production plans, and water irrigation management strategies. The model estimates crop yield, and gross margin (GM) as a function of the irrigation depth. Finally, these GM functions are used to determine an optimum cropping pattern and irrigation strategy to maximize the GM on a farm in a specific scenario. A major limitation of the MOPECO model is the fact that it does not contain a salinity module; consequently it should not be used to simulate cases where saline water is used

1452

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

Fig. 1. Location of Eastern Mancha “EMAS” (Spain) and Bekaa Valley “BVAS” (Lebanon) agricultural systems.

for irrigation. This is the case of the Eastern Mancha (EMAS) (Spain) and Bekaa Valley (BVAS) (Lebanon) agricultural systems, where the salinity of the irrigation water may affect crop yields, and where farmers do not apply a leaching fraction (LF) at each irrigation supply, as recommended elsewhere by Doorenbos and Pruitt (1977) and Rhoades (1982), mainly because of the limited available water resources for irrigation in both areas. Therefore, this irrigation management practice may lead to a progressive salinization of the arable lands, which in turn may have negative impacts on the environment and socio-economic conditions of the farmers (Minhas, 1996; De Nys et al., 2005; Darwish et al., 2005). Results obtained in the present study are based on field tests carried out on onion and potato. Many irrigation experiments have shown that these crops are sensitive to water stress (Lynch et al., 1995; Shock et al., 1998; Bekele and Tilahun, 2007; Jimenez et al., 2010) and soluble salts (Wannamaker and Pike, 1987; Patel et al., 2001; Chauhan et al., 2007; Levy and Veilleux, 2007), factors which produce negative effects such as a drop in yield. The objectives of this paper are (i) to select the most suitable model among the three proposed models for developing the new MOPECO-Salt Model; (ii) to validate the new model under the EMAS conditions, and (iii) to test the new MOPECO-Salt Model under the semi-arid conditions of BVAS in Lebanon. The specific objective is to assess the different irrigation strategies already in use in both areas, where no LF is applied by farmers, using the new proposed MOPECO-Salt Model. The suitability of the irrigation strategies will be evaluated in the light of the results on soil management sustainability and water productivity under saline conditions. 2. Materials and methods 2.1. Description of Eastern Mancha and Bekaa Valley agricultural systems EMAS belongs to the Júcar River Basin in the Castilla-La Mancha Region (Spain) (Fig. 1). This is a semi-arid climate area with an average annual reference evapotranspiration of 1320 mm and a mean annual rainfall of 377 mm. The irrigated area is about 105,000 ha. Barley and maize are among the most cultivated annual crops. Sprinkler irrigation is commonly used in the area, mainly with center pivot and solid-set systems. About 90% of the irrigation water used originates from the Eastern Mancha aquifer, with an average electrical conductivity (EC) of 0.85 dS m−1 (CHJ, 2004) and an average annual irrigation availability of 4400 m3 ha−1 (Domínguez and de Juan, 2008). Results of a trial carried out at the “Aguas Nuevas” experimental farm located in Albacete (Spain) (38◦ 56 53 N, 1◦ 53 51 W, 690 m above sea level) on onion crop in 2009 (Domínguez et al., 2010a,b) were used for developing and validating the new MOPECO-Salt Model. Climatic series were obtained

from a weather station situated within the experimental plots (http://crea.uclm.es/siar/datmeteo/datos hist.php). BVAS extends from the western foothills of the Mount Lebanon Chain to the Eastern foothills of Anti-Lebanon Mountains. BVAS has a well defined semi-arid climate with an average annual reference evapotranspiration of 1500 mm and a mean annual rainfall of 592 mm (Karam et al., 2003). Main crops are winter wheat and potato. The latter is cropped during two different growing seasons: March–August and July–October. In the last two decades, the extensive use of groundwater for irrigation purposes has led to a significant depletion in ground resources along with a salinization of agricultural lands, especially where localized irrigation systems are in use without applying a leaching fraction (Darwish et al., 2005). Data and results of a five-year experiment (2001, 2002, 2004, 2005 and 2007) on potato (Karam et al., 2005, 2009) were used for testing the MOPECO-Salt Model. Climatic series were obtained from a weather station (AURIA 12E, DEGREANE, France) situated within the experimental plots at Tal Amara Research Station in the Central Bekaa Valley (33◦ 51 44 N, 35◦ 59 32 E, 905 m above sea level). 2.2. Model presentation 2.2.1. Model of Stewart et al. (1977) MOPECO uses the model proposed by Stewart et al. (1977) for estimating crop yield as a function of the ETa /ETm ratio in the different growth stages. When ETa < ETm , the plant suffers from any stress that may cause a drop in yield (actual yield (Ya ) < potential yield (Ym )).

 Ya = Ym n=4



 1 − kyj

j=1

1−

ETaj



ETmj

(1)

where Ya and Ym are actual and potential crop yields (kg ha−1 ); n is the number of growing stages (Allen et al., 1998); j is the actual growing stage; ky is the crop yield response factor; while ETa and ETm are seasonal crop evapotranspiration (mm) yielding Ya and Ym , respectively. In this study, low water content and high salt concentration in the root zone were considered stress-causing factors, corresponding to actual evapotranspiration under water stress conditions (ETaw ), salinity conditions (ETas ), or both stress conditions (ETa ws ). The first part of the paper consists in selecting a model from the literature that will estimate actual evapotranspiration (ETa ) under water stress and/or saline conditions, which would be replaced in Eq. (1). To achieve this aim, three models were selected. 2.2.2. Model 1 The evapotranspiration capacity of a crop is directly related to the soil water content in the root zone. If soil water content is higher than p (defined as the fraction of TAW that a crop can extract without suffering water stress), then the crop is not subject to water stress conditions. According to Allen et al. (1998), ETaw may be estimated as follows: If TAW − Dr ≥ (1 − p)TAW = TAW − RAW; then ETaw = ETm Otherwise : ETaw =

TAW − Dr ETm (1 − p)TAW

(2)

where Dr is the root zone depletion at a given time (mm). Allen et al. (1998) stipulated that crops may decrease their evapotranspiration capacity if soluble salts existing in the root zone, expressed as electrical conductivity of the soil saturation extract (ECe ), exceed a threshold value of (ECet ). The following equation was proposed by Allen et al. (1998) to evaluate the combined effect

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

of water and salinity stresses on ETa :



(1)

Table 1 States of the variables used in the simulations.



ETa ws b TAW − Dr (1) = Ksc = Kss Ksw = 1 − (ECe − ECet ) ETm (1 − p)TAW ky 100 (3) (1)

where ETa ws is ETa under water stress and saline conditions (mm); Ksc is a dimensionless transpiration reduction factor dependent on Ksw and Kss ; Kss is a dimensionless transpiration reduction factor dependent on electrical conductivity of soil saturation extract [0–1], where Kss = 1 if ECe ≤ ECet ; Ksw is a dimensionless transpiration reduction factor dependent on available soil water [0–1], where Ksw = 1 if Dr ≤ RAW; ECe is the actual electrical conductivity of the soil saturation extract, as average value of the root zone (dS m−1 ); ECet is the threshold of electrical conductivity of the soil saturation extract above which the crop yield is affected by salinity (dS m−1 ); while b is a crop specific parameter, which describes the rate of yield decrease per unit of excess salts (% dS−1 m). General values of ECet are included in Doorenbos and Pruitt (1977) and Allen et al. (1998). 2.2.3. Model 2 Salinity increases the soil water content at the wilting point because roots have to overcome the combined elevated matric potential and increased osmotic potential (Beltrão and Ben Asher, 1997). For this reason, TAW and p (Eq. (3)) vary with the soluble salt content in the root zone. Pereira et al. (2007) propose three (2) (2) equations for estimating TAW, p, Ksw and ETa ws under saline conditions: pcor

b =p− (ECe − ECet ) p 100

(4)

TAWsalt = (ϑFC − ϑWPsalt )Zr (2)

(5)





ETa ws b TAWsalt − Dr (2) = Ksc = Kss Ksw = 1 − (ECe − ECet ) ETm (1 − p)TAWsalt ky 100 (6) where pcor is p modified due to saline conditions (mm); TAWsalt is TAW modified due to saline conditions (mm); ϑFC is the soil water content at field capacity (mm mm−1 ); Zr is the root depth (mm); while ϑWPsalt is the soil water content at wilting point for saline conditions (mm mm−1 ), which may be estimated as follows: ϑWPsalt = ϑWP + b

 EC − EC  e et 100

(ϑFC − ϑWP )

(7)

where ϑWP is the soil water content at wilting point under nonsaline conditions (mm mm−1 ). 2.2.4. Model 3 For calculating ETa ws García et al. (2006) developed a model that relates ETm to the soluble salts content and the matric pressure head in the root zone:

Variable

State 1

State 2

State 3

Model Texture Maize: ECiw (dS m−1 ) Barley: ECiw (dS m−1 ) Onion: ECiw (dS m−1 ) ETa ws /ETm objective

1 Sandy-loam 0.85 0.85 0.85 0.90

2 Sandy-clay-loam 2.00 5.00 1.50 0.75

3 Sandy-clay 3.00 9.00 2.00 0.60

mated through the equations proposed by Adiku et al. (2001) and Campbell (1974), respectively: = −400ECsw = −400

h = hb

SCr WCr640

 WCr −d ϑs Zr

(9)

(10)

where ECsw is the electrical conductivity of the soil water (dS m−1 ); SCr is the soluble salts content in the root zone (mg m−2 ); WCr is the equivalent water depth to the soil water content in the root area (mm); hb is the air entry pressure (cm); ϑs is the saturated water content (mm mm−1 ); while d is Campbell’s parameter (Campbell, 1974). In the calculation, it was assumed that 1 dS m−1 corresponds to 640 mg l−1 of total dissolved salts (Abrol et al., 1988; ASCE, 1996).

2.3. Water and salt balances in the soil during crop growth stages The estimation of daily ETa requires daily calculations of water and salt balances. To perform these calculations, the soil profile is divided into three zones: the first is the zone occupied by roots; the second is the zone that is not currently occupied by the roots but will be so after their complete development, and the third is the zone that is occupied by roots in any moment of crop growth. Water and salt balances are calculated in the first and second zones. Water balance WCri = WCri−1 + Ri + Iwi + Gwi − ETa1 − Pwri If WCri−1 + Ri + Iwi + Gwi − ETa1 − Pwri ≥ ϑFC Zri ; then WCri = ϑFC Zri and Pwri = ϑFC Zri − (WCri−1 + Ri + Iwi + Gwi − ETa1 ) Otherwise : Pwri = 0 Gwi =

WCgi−1 (Zri − Zri−1 ) Zt − Zri−1

(11)

WCgi = WCgi−1 + Pwri − Gwi − Pwgi If WCgi−1 + Pwri − Gwi − Pwgi ≥ ϑFC (Zt − Zri ); then WCgi = ϑFC (Zt − Zri ) and Pwgi = ϑFC (Zt − Zri ) − (WCgi−1 + Pwri − Gwi )

(3)

ETa ws 1 = ETm 1 + ((ah + )/2

1453

50 )

3

(8)

where is the salt concentration in the soil water (units of equivalent pressure head) (cm); 50 is the salt concentration in the soil (units of equivalent pressure head) that produces a 50% reduction in uptake by a certain crop (cm) (Doorenbos and Pruitt, 1977); a = 50 /h50 where h50 is the analog of 50 but for matric pressure head (cm) (i.e., the matric pressure head resulting in a 50% loss in yield for a certain crop); while h is the matric pressure head, which is a function of water content. Both and h may be esti-

Otherwise : Pwgi = 0 where WCr and WCg are water content in the root zone (zone 1) and in the zone that will be occupied by roots at the end of their development (zone 2), respectively (mm); R is the effective rainfall (mm); Iw is the water depth supplied by the irrigation system (mm); Gw is the water depth supplied to the root area from deeper areas due to the root growth (mm); Pwr and Pwg are the water depths that leave the current root area and the future root area due to percolation, respectively (mm); while Zt is the maximum depth

1454

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

reached by roots at the end of their development (mm).

Raes et al., 2009).

Salts balance

Evaporation from bare soil

SCri = SCri−1 + Isi + Gsi − Psri

Ei = Kei 1 − fm

Isi = Iwi ECiw 640

Kei = Kri (Kcmax − Kcb)

Gsi = Psri =

SCgi−1 (Zri − Zri−1 ) Zt − Zri−1





EToi

Kcmax = max (1.2, Kcb + 0.05) (12)

(SCgi−1 + Isi + Gsi )Pwri ϑFC Zri + (Pwri /2)

SCgi = SCgi−1 + Psri − Gsi − Psgi Psgi =

M 100

(SCgi−1 + Psri − Gsi )Pwgi ϑFC (Zt − Zri−1 ) + (Pwgi /2)

where SCr and SCg are the soluble salts content in the root zone (zone 1) and in the zone that will be occupied by roots at the end of their development, respectively (mg m−2 ); Is is the quantity of salts provided by irrigation water (mg m−2 ); ECiw is the electrical conductivity of the irrigation water (dS m−1 ); Gs is the soluble salts content supplied to the root zone from deeper layers due to the root growth (mg m−2 ); while Psr and Psg are the quantity of soluble salts that leave the current root area and the future root zone due to percolation, respectively (mg m−2 ). For calculating soluble salts dissolved by the percolated water, it was assumed that just before the onset of percolation, the total amount of soluble salts is dissolved in a volume of water equal to field capacity, corresponding to the maximum salt concentration. During the percolation, the salt concentration decreases up to the end of this process. Therefore, the quantity of soluble salts dissolved per unit of percolated water volume is maximum at the beginning and minimum at the end of the percolation. Assuming that soluble salts content decreases linearly in the percolated water, the amount of salts dissolved per unit of percolated water volume is calculated by dividing the total amount of salts at the beginning by the volume of water at field capacity plus one-half of the percolated volume (Eq. (12)). 2.4. Water and salt balance in the soil between irrigation seasons (bare soil) In order to reduce the number of input variables for MOPECO, the basal crop coefficient (Kcb) (Allen et al., 1998) was not used to estimate ET. However, in order to reach more precise water and salt balances in the soil during the off-season, the model has adopted this methodology in order to estimate evaporation from soil (E). As a result, Eqs. (11) and (12) can be used during the off-season by replacing ETa with E in Eq. (11). For the calculation of E, the soil profile is divided into two layers: the topsoil (Ze) layer, where E may occur and which usually varies in depth between 100 and 150 mm, and a deeper layer where water can be stored or percolated to deeper layers (Allen et al., 1998). The total amount of water that can be evaporated when the topsoil has been completely wetted initially is named Total Evaporation Water (TEW) and is conditioned by ϑFC and ϑWP . Moreover, a portion of TEW has a higher capacity for being evaporated and is named Readily Evaporable Water (REW). Finally, depending on the soil moisture content of Ze, these authors define the E coefficient (Kr) that indicates the higher or lower capacity of the soil for evaporating water. Furthermore, it is possible to insert other coefficients that may affect E (i.e. remains of previous crops on the surface of the soil;

(13)

TEW = (ϑFC − 0.5ϑWP ) Ze If WCZei−1 + Ri ≥ ϑFC Ze − REW; then Kri = 1 If WCZei−1 + Ri ≤ ϑFC Ze − TEW; then Kri = 0 Otherwise : Kri =

WCi−1 + Ri − 0.5ϑWP Ze TEW − REW

where E is evaporation from soil (mm); Ke: soil evaporation coefficient (dimensionless); fm (≤1) is the adjustment factor for the effect of mulches on soil evaporation, which varies between 0.5 for mulches of plant material and 1 for plastic mulches (Allen et al., 1998); M is the percent of soil covered by mulch (%); ETo is the reference evapotranspiration (mm); Kr is the evaporation reduction coefficient dependent on the cumulative depth of water depleted (evaporated) from the topsoil (dimensionless); Kcb is the basal crop coefficient, which is 0.15 for bare soil (Allen et al., 1998); Kcmax is the maximum value of Kc following rain or irrigation (dimensionless). Due to Kcb = 0.15 for bare soil, then Kcmax = 1.2; WCZe is the actual soil moisture content in the topsoil (mm); TEW is the total evaporable water (mm); while REW is the readily evaporable water (mm). Although this methodology is widely used, Torres and Calera (2010) carried out an experiment on a farm close to our experimental farm, concluding that the model proposed by Allen et al. (1998) overestimates E. Therefore, they propose to decrease REW from 8 to 4.6 mm for Ze = 100 mm and sandy-clay-loam texture, and insert a reduction coefficient of Kr named fc (equal to 0.15 for sandy-clay-loam soils): Kri = fcKri

(14)

2.5. Methodology for selecting the model to be integrated into MOPECO-Salt Selection of the model to be incorporated into MOPECO-Salt was achieved by comparing the results of 243 simulations obtained with the three selected models (81 per model). Simulations were executed for three typical crops of EMAS with varying salt sensitivity: barley, maize and onion. In order to increase the amount of data for the analysis, three irrigation depths were supplied through three ETa ws /ETm objectives. Model 1 was used for calculating the irrigation schedule that reaches each ETa ws /ETm objective depending on both the soil texture and the crop but without considering the saline effect. Later, three different levels of EC were assigned to the irrigation water of each irrigation schedule (Table 1). Climatic data belong to year 2005 and the software used for the statistical analysis was STATGRAPHICS Plus version 5.1 (Statistical Graphics Corporation, 2001). Because Model 1 is simpler than the other two models, the former was used as a reference for comparing yields of Model 1 (Oi ) with those obtained with Models 2 (P2i ) and 3 (P3i ). Error was calculated as follows:



RMSE = n−1



(Pji − Oi )2

0.5

(15)

where RMSE is the root mean square error; n is the number of simulations; while Pji and Oi are the compared yields.

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

The mean square error, expressed as a percentage of the average yield, was used as a measurement of the Relative Error (RE): RE =

RMSE 100 Oave

(16)

where Oave is the average yield according to Allen et al. (1998). The Index of Similitude (IS) was also used as a relative measure of the difference between yields:



IS = 1 −



(Pji − Oi )2

((Pji − Oave ) + (Oi − Oave ))2

(17)

The main data of the selected crops in EMAS are shown in Table 2a and b. Economic data for barley and maize are not shown since the economic study was only conducted for onion. 2.6. Validation of the equations used in the soil water balance Validation of the equations used in the soil water balance was made through a comparison of the daily data of soil water calculated by the model and the readings of the soil water sensors installed in the soil during the onion trial in the 2009 growing year (Domínguez et al., 2010a). In total, 16 Enviroscan® probes that measure volumetric soil water content were installed at four depths (10, 20, 30 and 40 cm) of the soil profile. From the soil parameters determined through the soil analysis, the soil water retention curves were calculated using Richard plates (Richards, 1948). Soil water sensors were installed after crop establishment (May 28th) and were removed ten days prior to harvest (September 4th). Therefore, total ETa calculated by the sensors is lower than the ETa for the entire crop duration. In addition, an irrigation schedule based on 48 irrigation events was monitored throughout the growing season, and 33 irrigation evaluations were carried out in order to measure the real water supplied by the irrigation system to the crop. 2.7. Validation of the equations used in the soil soluble salt balance Four soil samples were taken and analyzed from the onion plot at four different periods: before the irrigation season in 2009; at the end of that irrigation season; after a period of huge rainfall between December 2009 and January 2010; and before the beginning of the next irrigation season, in this case for an oat crop. Each sample was obtained through 25 subsamples of the top 40 cm of soil layer (maximum root depth of onion in the plot). Parameters analyzed were: texture; saturated soil moisture content, field capacity and wilting point; pH; EC in the saturation extract; and content of bicarbonate, sulphate, chloride, potassium, sodium, magnesium and calcium. To determine the soluble salts added to the soil by the irrigation water, the ECiw was measured in three different periods of the irrigation season. The average was 0.85 dS m−1 with standard deviation (SD) of 0.02 dS m−1 , which coincides with CHJ (2004). 2.8. Application of MOPECO-Salt for simulating a potato crop in the Bekaa Valley To validate the model, the results of five potato tests were compared with those calculated by MOPECO-Salt when taking saline water into account (ECiw > 0 dS m−1 ). The main characteristics of this crop in the area are shown in Table 2a. Potato (Solanum tuberosum L.) was planted under field conditions in late March–early April during five growing years, using cultivars Spunta (2001 and 2002) and Agria (2004, 2005 and 2007). The soil at the experimental site has a high clay content (40%) and relatively low organic matter (1.9%). Laboratory-based measure-

1455

ments of the soil water content at field capacity and permanent wilting point indicate ϑFC = 0.420 and ϑWP = 0.220 mm mm−1 . The plant available water holding capacity within the top 90 cm is 180 mm. As for salinity, the average ECiw is 1.7 dS m−1 and the minimum ECe of the soil is 1.1 dS m−1 (Darwish et al., 2005). In order to emphasize the effects of salinity on yields when simulating crops under deficit irrigation, the test yields were compared with those calculated by AquaCrop (Raes et al., 2009), which presently does not have an available saline module. Finally, for the comparison between MOPECO-Salt and AquaCrop, a third simulation was carried out using irrigation water with ECiw = 0 dS m−1 . In other words, the ECiw is so low that ECe < ECet at any moment of the crop growing period. This comparison enables to check the goodness of fit between both models under similar simulation conditions. Consequently, if the goodness of fit is high, the parameters inserted in AquaCrop for the simulation of these two potato cultivars would be presented as results. 2.9. Evaluation in terms of soil sustainability of the irrigation strategies used in the Eastern Mancha and Bekaa Valley agricultural systems To determine the sustainability of the strategy used in both areas in terms of soil management, the soluble salts content progression in the root zone was calculated from the end of the actual irrigation season to the beginning of the following one. In this manner, if the water supplied by rainfall between the two irrigation seasons is able to decrease the ECe of the soil in the root zone up to the minimum value fixed for each area, then it would be possible to conclude that this irrigation strategy is suitable. 2.10. Evaluation in terms of yield and gross margin of the irrigation strategy used in the Eastern Mancha agricultural system for onion The strategy used by the farmers in EMAS for irrigating onion in 2009 was compared with two theoretical ones. The characteristics of each irrigation schedule were: (1) the ECiw is very low (it was assumed ECiw = 0 dS m−1 ), thus the crop was not affected by salt water. In this situation, five irrigation objectives were proposed through five ETa ws /ETm rates (1.0, 0.9, 0.8, 0.7, and 0.6), which were kept constant for the four growing stages of the crop. Yields obtained with this schedule were considered maximum as per “Yield vs. Irrigation depth” rate. Of course, if regulated deficit irrigation (RDI) (English, 1990) were put into practice, yield could be higher for the same irrigation depth; (2) the same schedules used in the first one were applied to the crop but assuming ECiw = 0.85 dS m−1 , which is the average in the area. This option can be considered as the real irrigation strategy of the farmers; (3) the model calculated five irrigation schedules for the same ETa ws /ETm rates of option 1, but assuming ECiw = 0.85 dS m−1 . Therefore, the application of an LF was required. By simulating these three irrigation scheduling scenarios, three “Yield vs. Irrigation” functions were generated and analyzed for determining the most suitable irrigation strategy. These functions were translated into irrigation productivity in terms of yield (IPY) by dividing yield by the irrigation depth. Therefore, this parameter shows the amount of yield reached per unit of irrigation water applied to the crop. Finally, the “Yield vs. Irrigation” and “IPY vs. Irrigation” functions were compared for their validation with the results of an RDI test carried out in 2001 in the area using water from the aquifer (Martín de Santa Olalla et al., 2004) (Table 3). Results from 2001 to 2009 can be compared because the climatic conditions of both years were similar: the ETo was 939.4 mm and 891.1 mm, and the effective rainfall was 17.7 mm and 63.3 mm, respectively.

1456

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

Table 2 (a) Area, recommended irrigated depth, average yield and parameters related to the sensitivity of the selected crops in Eastern Mancha and Bekaa Valley; (b) economic data of the onion crop in Eastern Mancha. Eastern Mancha

(a) Area (ha) Irrigation (mm) Yield (kg ha−1 ) Ym (kg ha−1 ) ECet d (dS m−1 )

Bekaa Valley

Barley

Maize

Onion

Potato “Spunta”

Potato “Agria”

16,456a 250b 5750c 7500 8.0

7890a 650b 12,200c 17,000 1.7

6730a 560b 70,450c 100,000 1.2

8000 560 40,000 50,000 1.7

3200 700 60,000 80,000 1.7

Economic data of the onion crop in Eastern Manchae Sale price (D kg−1 )

Water cost (D m−3 )

Yield (kg ha−1 )

Variable costs (D ha−1 )

(b) 0.12

0.20

100,000 80,000 60,000 ≤40,000

4927.8 4495.0 3845.8 2980.3

a b c d e

ITAP (2009). JCRMO (2004). MAPA (2009). Allen et al. (1998). Results of local farmer surveys.

Table 3 Actual scheduling of regulated deficit irrigation, and irrigation and effective rainfall. Stage

Treatments (ETa /ETm ) T1

Settling Growth Tuber bulking Ripening Rainfall (mm) Irrigation (mm) Yield (kg ha−1 )

1.00 0.80 1.20 0.25 17.7 695.1 67,290

T2 1.00 0.80 0.90 0.25 17.7 585.4 65,310

T3 1.00 1.00 1.20 0.25 17.7 727.0 73,850

To study the effect of the irrigation strategy on crop profitability, the “Yields vs. Irrigation” functions were translated into “GM vs. Irrigation” functions through the economic data shown in Table 2b. As well, the irrigation productivity in terms of GM (IPGM) was obtained by dividing GM by the irrigation depth. 3. Results and discussion 3.1. Model selection for integration in MOPECO-Salt Comparisons made between the three models (Table 4) showed a high goodness of fit between simulated yields by Model 1 and yields simulated by the other two models. Therefore, the three models offer similar results when simulating the three selected crops under the proposed conditions of water stress and salinity. Consequently, because Model 3 is the most complicated one and because Model 2 does not significantly improve the results of Model 1, this last model was integrated into MOPECO-Salt. Probably, calibration and adjustment of the variables used for the simulation of maize using Model 3 would improve the similitude with results of Model 1. However, to avoid a possible slant and to stress the differences among the three models, it was decided to insert into the models the value of the variables assigned in the references consulted. However, since some of the parameters used in the simulations were only calibrated through experimental field tests in the area for onion crops, these results are only suitable for comparing the selected models.

T4

T5

1.00 1.00 1.20 0.50 17.7 754.3 70,730

1.00 0.80 0.90 0.50 17.7 612.7 73,130

T6 1.00 0.80 1.20 0.50 17.7 722.4 68,230

T7 1.00 1.00 0.90 0.25 17.7 617.3 67,710

T8 1.00 1.00 0.90 0.50 17.7 644.6 74,630

3.2. Validation of the equations used in the soil water balance Soil analysis demonstrated that soil texture in the experimental plot is sandy-clay-loam, with ϑFC = 0.228 mm mm−1 and ϑWP = 0.126 mm mm−1 . The soil water balance calculated during the period of measurements of the 16 soil moisture probes indicated ETa = 448.84 mm (SD = 8.32 mm), while MOPECO-Salt estimated ETa = 438.06 mm. Thus, it can be asserted that the model estimates the ETa (Fig. 2a) in a suitable way. As an example, time course evolution of the soil water volumetric content in the top 40 cm of the soil profile calculated by MOPECO-Salt is compared with the proportional average of the readings of one probe, showing a good fit between them (Fig. 2b). 3.3. Validation of the equations used in the soil soluble salts balance The four ECe measurements obtained in the laboratory were represented for validating the ECe progression calculated by MOPECO-Salt in the top 40 cm of the soil profile in the onion plot (Fig. 3). Hence, the maximum ECe estimated by the model (2.91 dS m−1 ) was slightly lower than the one obtained with soil sampling (3.16 dS m−1 ). On the other hand, a fast wash of the soluble salts occurred after an unusually abundant rainfall during the period of December 2009–January 2010. This fact was favoured by a soil water content at the beginning of December 2009 that was close to field capacity. The three samples analyzed before the irrigation season (one in early 2009 and two in 2010) proved that

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

1457

Table 4 Comparison among the three models. Crop

Model

N

Oave

Pave −1

(kg ha Barley

2 3 2 3 2 3

Maize Onion

27 27 27 27 27 27

Pave /Oave (%)

a

)

R2

RMSE (kg ha−1 )

RE (%)

IS

0.999** 0.999** 0.998** 0.981** 0.999** 0.998**

41.6 182.9 491.9 1328.8 990.2 3490.7

0.80 3.54 7.28 19.67 2.02 7.11

0.999 0.989 0.991 0.939 0.998 0.972

YO = aYP

5165.8 5165.8 6755.6 6755.6 49115.0 49115.0

5144.7 5200.3 6358.9 5641.2 48195.5 46278.5

1.004* 0.995* 1.053* 1.139* 1.018* 1.059*

99.6 100.7 94.1 83.5 98.1 94.2

Where N is the number of simulations; Oave is the average yield according to Model 1; Pave is the average yield according to Model 2 or 3; YO is the yield according to Model 1; YP is the yield according to Model 2 or 3; “a” is the slope of the equation (regression coefficient); while R2 is the determination coefficient. * p-Value < 0.05. ** p-Value < 0.01.

Volumetric content (%).

b

1,10 1,08 1,06 1,04 1,02 1,00 0,98

Progression of volumetric water content 30 28 26 24 22 20 18 16 14 12

0,96

28 -0 5 04 - 09 -0 6 11 -09 -0 6 18 -09 -0 6 25 -09 -0 6 02 - 09 -0 7 09 - 09 -0 7 16 - 09 -0 7 23 -09 -0 730 09 -0 7 06 -09 -0 8 13 - 09 -0 8 20 - 09 -0 809

ETa measured / ET a calculated

a

Comparison of measured ET a with calculated ET a

0,94 0,92

Date

0,90

Calculated by MOPECO-Salt Sensor at 10 cm depth Sensor at 30 cm depth

1 2 3 4 5 6 7 8 9 10111213141516 Enviroscan probe

Probe 0-40 cm depth Sensor at 20 cm depth Sensor at 40 cm depth

27/03/2010

0

25/02/2010

0.0 26/01/2010

4

27/12/2009

8

0.5 27/11/2009

1.0

28/10/2009

12

28/09/2009

1.5

29/08/2009

16

30/07/2009

2.0

30/06/2009

20

31/05/2009

24

2.5

01/05/2009

3.0

01/04/2009

-1

ECe (dS m )

28

Simulated yield / Observed yield

EC e progression in onion plot

3.5

Rainfall and irrigation (mm)

Fig. 2. (a) Comparison of measured ETa by EnviroScan probes and ETa calculated by MOPECO-Salt; (b) time course progression of the volumetric soil water content in the onion plot according to an EnviroScan probe during the 2009 growing season and calculated by MOPECO-Salt.

Observed yields vs. Simulations 1.20 1.15 1.10 1.05 1.00 0.95 0.90 Spunta

0.85 2000

2001

2002

-------------Agria-------------

2003

2004

2006

2007

2008

2009

2010

Year

Date ECe Calculated by MOPECO-Salt Min ECe Rainfall

2005

Onion

ECe Measurements ECet Onion Irrigation

Fig. 3. Comparison of calculated and measured ECe progression in the plot cropped with onion (Eastern Mancha) (2009–2010).

minimum ECe of the soil was 0.78 dS m−1 (SD = 0.07 dS m−1 ), which is somewhat lower than ECiw . In terms of management, onion was under saline stress for about two thirds of the growing period, thus affecting both crop development and yield. 3.4. Application of MOPECO-Salt in Eastern Mancha and Bekaa Valley agricultural systems Real yields observed in onion (Domínguez et al., 2010b) and potato (Karam et al., 2005, 2009) trials were compared with yields simulated by MOPECO-Salt (Fig. 4). In general, the model achieved

MOPECO-Salt (ECiw > 0 dS/m)

MOPECO-Salt (ECiw = 0 dS/m)

AQUACROP (ECiw = 0 dS/m)

Simulated yield = Observed yield

Fig. 4. Comparison between observed and simulated yields.

results close to the ones observed (differences between yields < 3%). On the other hand, yield overestimation by AquaCrop was probably due to the fact that the evapotranspiration rate of potato was not decreased because of the ECe raise, but in 2007 the simulated yield was similar to the one observed. This result may be conditioned by the weather conditions of that year. Anyway, according to several authors (Farahani et al., 2009; Heng et al., 2009), differences between observed and simulated yields lower than 10% are acceptable. Furthermore, the same authors consider that one model is calibrated when the percentage of matches is around 70% or higher. Finally, when comparing AquaCrop with MOPECO-Salt under no salt irrigation water conditions (ECiw = 0 dS m−1 ), the dif-

1458

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

Table 5 Comparison between the evapotranspiration rates estimated by MOPECO-Salt and AquaCrop for the potato simulations. Year

2001 2002 2004 2005 2007

Real data (mm)

MOPECO-Salt ECiw = 0 dS m−1 (mm)

AQUACROP (mm)

ETo

Rain

Irrig.

ETa ws

ETm

ETa ws /ETm

ETa ws

ETm

ETa ws /ETm

651.3 801.9 864.1 839.0 840.1

0.0 0.0 14.2 18.0 67.0

553.0 563.0 747.0 670.0 658.0

445.6 558.3 651.9 593.6 582.5

484.1 627.5 706.3 679.0 696.4

0.92 0.89 0.92 0.87 0.84

447.0 501.5 664.2 654.1 664.2

609.0 656.3 747.9 741.7 723.7

0.73 0.76 0.89 0.88 0.92

Table 6 AquaCrop parameters for simulating two potato cultivars (Spunta and Agria). Parameter Development Initial canopy cover Plants per haa Initial C cover Planting method Canopy development C expansion Max. C cover (%) C declinea (d) Tuber formation Building up HIa , b (d) Start tuberizationa , b (d.a.s) Root deepening Max. EDa (m) Shape factor Max. deptha , b (d.a.s) Temperatures ET Ke and Kcb coeff. C coverc (%)

Variety “Spunta”

Variety “Agria”

70,000 High Sowing

70,000 High Sowing

Fast 92 91

Fast 92 29

41–49 56–61

63–77 63–70

0.60 2.0 56–61 n.c

0.60 2.0 77–91 n.c

90

95

Parameter WEP Max. RE (mm d−1 ) Upper ¼ (%) Second ¼ (%) Third ¼ (%) Bottom ¼ (%) Production Crop wp WP (g m−2 mm−1 ) Harvest index HIoa (%) Water stress C expansion Stomatal closure Early C senescence Aeration stress FERT. stress TEMP. stress

Variety “Spunta”

Variety “Agria”

8.1 40 30 20 10

10.2 40 30 20 10

15.5

20.0

75

85

S S S MT n.c n.c

S S S MT n.c n.c

C is the green canopy that transpires water; HI is the harvest index, which is obtained dividing the yield by the biomass (dimensionless); d is days; d.a.s is days after sowing; ED is the effective rooting depth, which indicates the profile of the soil where roots are taking up water from; n.c means not considered; ET is evapotranspiration; WEP is the water extraction pattern. Since root density is generally highest near the soil surface and declines with depth, a 40, 30, 20, and 10% water extraction pattern was used for the upper, second, third, and bottom quarter of the root zone, respectively; RE is the water root extraction over the root zone being 0.60 m; WP is the water productivity, which is defined as the weight of the biomass per m2 and per mm of cumulated water transpired over the time period in which the biomass is produced); HIo is the reference harvest index under no stress conditions; S means sensitive to water stress; MT means moderately tolerant to water logging; FERT. means fertilization; while TEMP. means temperature. a Real data. b The range of real results achieved during the five years of field tests are shown. c Effect of canopy shelter in the late season.

ference between yields of four out of five years in this study was lower than 2%. In addition, the ETa ws estimated by both models were practically the same for 2001 and 2004, but showed slightly higher than 10% disagreements in years 2002 and 2007 (around 12% of difference) (Table 5). Nevertheless, when comparing ETm , the goodness of fit between both models is high (around 5.5% of difference) except for 2001 (20.5% of difference). Since AquaCrop uses the Kcb method, in that year the model estimated higher E losses from the top layer of the soil profile at the beginning of the growing period. This fact may be attributed to the late sowing time (July 4th) and the slower canopy development of this cultivar (Spunta) observed in 2001. Therefore, MOPECO-Salt could underestimate the E losses from the topsoil during the early crop growth stage although, according to Farahani et al. (2009) and Heng et al. (2009), these differences are acceptable. Therefore, under the experimental conditions of this study, the compared models are valid for the simulation of potato even when the salt effect on yields is not taken into account. Based on these results, it becomes evident that the parameters inserted in AquaCrop for the simulation of the two potato cultivars can be considered suitable (Table 6). Results of the onion simulations are analyzed in depth in Section 3.6, while yields of potato obtained in this study are in line with the findings of Fabeiro et al. (2001). These authors observed that treatments with deficit irrigation during the last part of the crop cycle have the lowest tuber production when compared to

those with deficit irrigation during the early growth stages since water stress late in-the-season has negative impacts on tuber size rather than tuber number. Moreover, since Spunta is characterized by both medium and large size tuber production, this cultivar has been shown to be more sensitive to water stress than Agria, which is characterized by medium sized tubers. On the other hand, the lower harvest index obtained in Spunta in comparison to Agria is mainly because the former has less dry matter content than the latter. The sensitivity of potato vis-à-vis water stress in terms of stomatal closure obtained in this experiment has also been reported by Karam et al. (1997) on potato irrigated with saline water. A comparison made between Spunta and Agria showed that root extraction was higher in the latter than the former. This finding may be due to the fact that Spunta showed more sensitivity to water stress than Agria. 3.5. Model evaluation in terms of soil sustainability of the irrigation strategies Even though irrigators in EMAS do not supply an LF to onion or to any other crop, no salinization problems have been encountered in this area. However, results revealed in Fig. 3 are not conclusive when determining if this irrigation strategy is sustainable in terms of soil management since only one year was studied and the rain-

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

1459

Fig. 5. (a) Simulated ECe progression in a maize plot (2000–2009) (Eastern Mancha); (b) ECe progression in potato farms (Bekaa Valley).

during the growing season (Fig. 5b). Thus, this irrigation strategy enabled farmers to save a considerable amount of water without affecting the sustainability of the soil. In that sense, Darwish et al. (2005) pointed out that improper cropping practice, combined with poor and inappropriate irrigation policy on the part of farmers, contributed to the building up of salt in the top layers of the soil profile, especially in areas where localized irrigation is in use. Results obtained by the MOPECO-Salt Model showed that this problem can be remediable whenever off-season rain is enough to leach out accumulated salts during the in-season irrigations, and before the onset of the next growing season in spring. Indeed, results of the different simulations showed that the soil salinity level started to decrease at the onset of the rainy season. On the other hand, our results confirmed the hypothesis that potato growers in the Bekaa Valley of Lebanon are not applying an LF during the different irrigations because of the increasing costs of energy, and also because of the limited ability of drip irrigation to provide efficient leaching requirements to crops in fine-textured soils. This point has obliged some farmers to use sprinkler irrigation instead of drip irrigation, an indication that the introduction of modern irrigation systems is insufficient when improvements in agricultural practices, as well as an upgrading of farmers’ skills, are also required (Darwish et al., 2005). The proposed model allows the irrigation decider to define specific production options and resource constraints under different biophysical conditions.

fall pattern was unevenly abundant at the end of 2009 and the beginning of 2010. Furthermore, onion is used in the local cropping pattern within the farms simply to avoid disease attacks. If water requirements of the successor crop are lower and the crop is tolerant to salts (e.g. spring barley), this situation may compensate the excess of salts added to the soil and disguise the consequences of this irrigation strategy. For this reason, the sustainability study was carried out over a ten-year period for maize, largely because water requirements for this crop are higher than onion (Table 2a), and maize is usually grown on the same plot for successive years (Fig. 5a). In this example, in three out of the ten years (2001, 2005 and 2006) rainfall was not able to wash out the soluble salts added by the irrigation water before the beginning of the next irrigation season. In spite of the fact that maximum ECe values were reached in the following irrigation seasons (around 4.5 dS m−1 ), the subsequent rainfall period was able to wash soluble salts accumulated during two seasons. Consequently, this irrigation strategy may cause negative effects on maize yield grown after periods of low autumn–winter rainfall, but it will not cause a progressive salinization of the agricultural soils in the area. On the contrary, results of the potato experiment in the Bekaa Valley in Lebanon showed that in all years the off-season rainfall was able to wash out the salts added with irrigation water

c

d

-1

6

Irrigation (mm)

Irrigation (mm) Option 1: ECiw = 0 dS/m Option 3: ECiw = 0,85 dS/m (with LF) Martín de Santa Olalla et al., 2004

Irrigation (mm)

1500

1500

1250

1000

500

750

250

1500

1250

1000

750

0

500

0

1500

60

1250

20000

750

2

1000

1500

250

80

500

40000

1250

4

1000

3000

500

100

750

60000

IPGM vs. Irrig. 8

-1

-1

-1

GM (€ ha )

-1

IPY (kg ha mm )

GM vs. Irrig. 6000

4500

120

-1

Yield (kg ha )

80000

IPY vs. Irrig. 140

250

b

IPGM (€ ha mm )

Yield vs. Irrig. 100000

250

a

Irrigation (mm)

Option 2: ECiw = 0,85 dS/m (no LF) Recommended irrigation depth (mm)

Note: Each function is composed of six ETa/ETm rates (1, 0.9, 0.8, 0.7, 0.6 and 0.5). Fig. 6. Determination of the optimal irrigation schedule for onion (2009) in terms of: (a) yield; (b) irrigation productivity in terms of yield (IPY); (c) gross margin (GM); and (d) irrigation productivity in terms of gross margin (IPGM) under saline stress conditions. Note: each function is composed of six ETa /ETm rates (1, 0.9, 0.8, 0.7, 0.6 and 0.5).

1460

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461

3.6. Model evaluation in terms of yield and gross margin of the irrigation strategy used in the Eastern Mancha agricultural system for onion Results illustrated in Fig. 3 demonstrate that supplying a suitable LF may decrease salinity stress, which would cause a higher onion yield. However, the higher amount of water applied to the crop may compromise its profitability. The irrigation schedules proposed in Section 2.10 were used to obtain “Yield vs. Irrigation”, as well as “IPY vs. Irrigation” functions for onion in 2009 (Fig. 6a and b). As expected, only scenarios 1 and 3 were able to achieve Ym , and a 27% yield drop occurred for the irrigation schedule ETa /ETm = 1 in scenario 2. Nonetheless, reaching Ym implies supplying a large amount of water compared to the recommended irrigation depth in the area. Under the prevailing 2009 weather conditions and recommended irrigation depth, scenario 2 seems to be the best strategy. It is also important to mention that for irrigation depths lower than 490 mm there are negligible differences among irrigation strategies since ECe < ECet for the entire growing period (Fig. 6a). On the other hand, the recommended irrigation depth in the area practically coincides with the maximum IPY (104.4 kg ha−1 mm−1 ), decreasing IPY from 690 mm (Fig. 6b). Yields and IPYs achieved by Martín de Santa Olalla et al. (2004) (Table 3) were compared with the simulations, obtaining similar values (Fig. 6a). The use of RDI justifies reaching yields similar to the “Yield vs. Irrigation without salt water” function. Therefore, the use of this technique in the area would partially neutralize the effect of the no LF irrigation strategy. The maximum GM and IPGM values reached by scenario 2 were for an irrigation depth higher than the one recommended (Fig. 6c and d). According to these simulations, it is feasible to reach higher GM for higher irrigation depths. However, due to water scarcity in the area, increasing irrigation depth for onion would imply decreasing the availability of water for other crops on the farm. Consequently, the total GM of the farm may decrease. In-depth economic analyses such as those made by MOPECO are therefore required before advising an increase in irrigation depth for onion in the area.

4. Conclusions The model proposed by Allen et al. (1998) allows for estimating the yield of herbaceous crops and vegetables under water stress and/or salt conditions through a daily balance calculation of water and soluble salts content in the root zone. The results of this study are in agreement with those obtained by Pereira et al. (2007) for modified TAW and p, and García et al. (2006) for daily calculations of osmotic potential and matric pressure in the root zone. However, in this case a more affordable methodology has been employed. Simulation of the behaviour of crops under water stress should include the effect of salt stress, especially in arid and semi-arid areas, where regulated deficit irrigation shows a high potential. The use of salt water that may affect yield and crop development is a common practice. Thus, for potato crop in the Bekaa Valley (Lebanon) and depending on the cultivar, the use of models without a salt water module may overestimate the simulated yield up to 14%. In areas where water is scarce, if electrical conductivity of irrigation water is not too high and yield is not greatly affected, it would become a common practice not to supply an LF at each irrigation event, a factor which may cause a progressive salinization of the agricultural soils. The MOPECO-Salt Model may help farmers and irrigators to determine whether that irrigation strategy is sustainable in terms of soil management, because rainfall is able to wash out the soluble salts accumulated in the root area. In this way, it

is possible to save considerable volumes of water and to increase the gross margin on farms. Hence, in both Eastern Mancha (Spain) and Bekaa Valley (Lebanon) agricultural systems and for the crops studied, it is not necessary to supply an LF, which may save up to 14.4% of irrigation water required by an onion crop. However, the natural washing of salts hugely depends on the evaporation from soil when the plot is under bare conditions. Therefore, due to the importance of this process in arid and semi-arid agricultural areas, further research on this matter is required, mainly when using FAO56 methodology because of the disagreements found by several authors. Finally, the MOPECO-Salt Model may indicate the irrigation strategy that optimizes the yield and the gross margin of a crop or a pattern of crops depending on irrigation water availability. Acknowledgements This paper has been developed within the framework of two European projects funded by EC: FLOW-AID “Farm Level Optimal Water Management: Assistant for Irrigation under Deficit” N◦ 036958 (GOCE), and DeSURVEY “A Surveillance System for Assessing and Monitoring of Desertification” (SUSTDEV-CT-2004003950-2). The authors also wish to thank the reviewers for their efforts in improving the quality of this paper. References Abrol, I.P., Yadav, J.S.P., Massoud, P.I., 1988. Salt Affected Soils and Their Management. FAO Soils Bulletin Paper No. 39. Food and Agriculture Organization of the United Nations (FAO), Rome, Italy. Adiku, S.G.K., Renger, M., Wessolek, G., Facklam, M., Hecht-Bucholtz, C., 2001. Simulation of the dry matter and seed yield of common beans under varying soil water and salinity conditions. Agric. Water Manage. 47, 55–68. Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration: Guidelines for Computing Crop Water Requirements. Irrigation and Drainage Paper No. 56. Food and Agriculture Organization of the United Nations (FAO), Rome, Italy. Arshi, A., Ahmad, A., Aref, I.M., Iqbal, M., 2010. Effect of calcium against salinityinduced inhibition in growth, ion accumulation and proline contents in Cichorium intybus L. J. Environ. Biol. 31 (6), 939–944. ASCE, 1996. Agricultural Salinity Assessment and Management. Water Quality Technical Committee of the Irrigation and Drainage Division of ASCE, New York, USA. Bekele, S., Tilahun, K., 2007. Regulated deficit irrigation scheduling of onion in a semiarid region of Ethiopia. Agric. Water Manage. 89 (1–2), 148–152. Beltrão, J., Ben Asher, J., 1997. The effect of salinity on corn yield using the CERESmaize model. g. Drain. Syst. 11, 15–28. Bresler, E., 1986. Application of a conceptual model to irrigation water requirement and salt tolerance of crops. Soil Sci. Soc. Am. J. 51, 788–793. Campbell, G.S., 1974. A simple method for determining the unsaturated conductivity from moisture retention data. Soil Sci. 117, 311–314. Castrignanò, A.M., Katerji, N., Hamdy, A., Karam, F., Mastrorilli, M., 1998. A modified version of CERES-Maize model for predicting crop response to salinity stress. Ecol. Model. 111, 107–120. Chauhan, C.P.S., Shisodia, P.K., Minhas, P.S., Chauhan, R.S., 2007. Response of onion (Allium cepa) and garlic (Allium sativum) to irrigation with different salinity waters with or without mitigating salinity stress at seedling establishment stage. Indian J. Agr. Sci. 77 (8), 483–485. Childs, S.W., Hanks, R.J., 1975. Model of soil salinity effects on crop growth. Soil Sci. Soc. Am. J. 39, 617–622. CHJ, 2004. Júcar Pilot River Basin. Provisional Article 5 Report. Pursuant to the Water Framework Directive. Confederación Hidrográfica del Júcar, Ministerio de Medio Ambiente, Madrid, Spain. Darwish, T., Atallah, T., El Moujabber, M., Khatib, N., 2005. Salinity evolution and crop response to secondary soil salinity in two agro-climatic zones in Lebanon. Agric. Water Manage. 78, 152–164. De Nys, E., Raes, D., Le Gal, P., Cordeiro, G., Speelman, S., Vandersypen, K., 2005. Predicting soil salinity under various strategies in irrigation systems. J. Irrig. Drain. Eng. 131 (4), 351–356. Domínguez, A., de Juan, J.A., 2008. Agricultural water management in Castilla-La Mancha (Spain). In: Sorensen Magnus, L. (Ed.), Agricultural Water Management Research Trends. Nova Science Publishers, Inc., New York, USA. Domínguez, A., de Juan, J.A., López-Mata, E., Tarjuelo, J.M., 2010a. Calibration of irrigation schedule module of MOPECO model through soil moisture sensors. In: Paltineanu, I.C., Vera, J. (Eds.), Transactions of “The Third International Symposium on Soil Water Measurement Using Capacitance, Impedance and TDT”. April 7–9, Murcia, Spain.

A. Domínguez et al. / Agricultural Water Management 98 (2011) 1451–1461 Domínguez, A., Medina, D., García-Pardo, G., de Juan, J.A., Tarjuelo, J.M., 2010b. Efecto de la altura y del tipo de emisor en los sistemas pivote sobre el cultivo de la cebolla (Allium cepa L.). Transactions of “XXVIII Congreso Nacional de Riegos”, (Ed.) Consejería de Agricultura y Ganadería de la Junta de Castilla y León, Aso˜ ˜ de riegos y drenajes, ciación espanola de riegos y drenajes, y Comité espanol June 15–17, León, Spain. Doorenbos, J., Pruitt, W.O., 1977. Crop Water Requirements. Irrigation and Drainage Paper No. 24. Food and Agriculture Organization of the United Nations (FAO), Rome, Italy. English, M.J., 1990. Deficit irrigation. In: Analytical framework. J. Irrig. Drain. Eng. 116, 399–412. Fabeiro, C., Martin de Santa Olalla, F., de Juan, J.A., 2001. Yield and size of deficit irrigated potatoes. Agric. Water Manage. 48, 255–266. FAO, 2007. Advances in the Assessment and Monitoring of Salinization and Status of Biosaline Agriculture. World Soil Resources Reports, No. 104. Food and Agriculture Organization of the United Nations, Rome, 72 pp. Farahani, H.J., Izzi, G., Oweis, T.Y., 2009. Parameterization and evaluation of the AquaCrop model for full and deficit irrigated cotton. Agron. J. 101 (3), 469–476. Ferrer-Alegre, F., Stockle, C.O., 1999. A model for assessing crop response to salinity. Irrig. Sci. 19, 15–23. García, L.A., Foged, N., Cardon, G.E., 2006. Development of GIS-BASED model to estimate relative reductions in crop yield due to salinity and water logging. J. Irrig. Drain. Eng. 132 (6), 553–563. Hasegawa, P.M., Bressan, R.A., Zhu, J.K., Bohnert, H.J., 2000. Plant cellular and molecular responses to high salinity. Annu. Rev. Plant Physiol. Plant Mol. Biol. 51, 463–499. Heng, L.K., Hsiao, T., Evett, S., Howell, T., Steduto, P., 2009. Validating the FAO AquaCrop model for irrigated and water deficient field maize. Agron. J. 101 (3), 488–498. ITAP, 2009. Anuario Técnico del ITAP 2008 ,. Instituto Técnico Agronómico Provincial de Albacete. Diputación de Albacete, Albacete, Spain. Jimenez, M., de Juan, J.A., Tarjuelo, J.M., Ortega, J.F., 2010. Effect of irrigation uniformity on evapotranspiration and onion yield. J. Agric. Sci. 148, 139–157. JCRMO, 2004. Normas del plan de explotación de la Junta Central de Regantes de la ˜ 2005. [On line] www.jcrmo.org. Junta Central de Mancha Oriental para el ano Regantes de la Mancha Oriental. Date: January 3rd 2008. Karam, F., Katerji, N., Hamdy, A., Mastrorilli, M., 1997. Solanum tuberosum water status and yield as affected by salinity and soil texture. Acta Horticulturae 449 (2), 635–640. Karam, F., Breidy, J., Stephan, C., Rouphael, J., 2003. Evapotranspiration, yield and water use efficiency of drip irrigated corn in the Bekaa Valley of Lebanon. Agric. Water Manage. 63, 125–137. Karam, F., Lahoud, R., Masaad, R., Stephan, C., Rouphael, Y., Colla, G., 2005. Yield and tuber quality of potassium treated potato under optimum irrigation conditions. Acta Horticulturae 684, 103–108. Karam, F., Rouphael, Y., Lahoud, R., Breidi, J., Colla, G., 2009. Influence of genotypes and potassium application rates on yield and potassium use efficiency of potato. J. Agron. 8 (1), 27–32. Letey, J., Dinar, A., Knapp, C., 1985. Crop-water production function model for saline irrigation waters. Soil Sci. Soc. Am. J. 49, 1005–1009.

1461

Levy, D., Veilleux, R.E., 2007. Adaptation of potato to high temperatures and salinity—a review. Am. J. Potato Res. 84 (6), 487–506. López-Mata, E., Tarjuelo, J.M., de Juan, J.A., Ballesteros, R., Domínguez, A., 2010. Effect of irrigation uniformity on the profitability of crops. Agric. Water Manage. 98, 190–198. Lynch, D.R., Foroud, N., Kozub, G.C., Farries, B.C., 1995. The effects of moisture stress at three growth stages on the yield, components of yield and processing quality of eight potato cultivars. Am. Potato J. 72, 375–386. Majeed, A., Stockle, C.O., King, L.G., 1994. Computer model for managing saline water for irrigation and crop growth: preliminary testing with lysimeter data. Agric. Water Manage. 26, 239–251. MAPA, 2009. Anuario de Estadística Agroalimentaria 2008. Ministerio de Agricultura, Pesca y Alimentación, Madrid, Spain. Martín de Santa Olalla, F.J., Domínguez, A., López-Urrea, R., 2004. Production and quality of the onion crop (Allium cepa L.) cultivated under controlled deficit irrigation conditions in a semi-arid climate. Agric. Water Manage. 68, 77–89. Minhas, P.S., 1996. Saline water management for irrigation in India. Agric. Water Manage. 30, 1–24. Ortega, J.F., de Juan, J.A., Martín-Benito, J.M., López-Mata, E., 2004. MOPECO: an economic optimization model for irrigation water management. Irrig. Sci. 23 (2), 61–75. Patel, R.M., Prasher, S.O., Donnelly, D., Bonnell, R.B., 2001. Effect of initial soil salinity and subirrigation water salinity on potato tuber yield and size. Agric. Water Manage. 46 (3), 231–239. Pereira, L.S., Gonc¸alves, J.M., Dong, B., Mao, Z., Fang, S.X., 2007. Assessing basin irrigation and scheduling strategies for saving irrigation water and controlling salinity in the upper Yellow river basin, China. Agric. Water Manage. 93, 109–122. Qadir, M., Tubeileh, A., Akhtar, J., Larbi, A., Minhas, P.S., Khan, M.A., 2008. Productivity enhancement of salt-affected environments through crop diversification. Land Degrad. Dev. 19, 429–453. Raes, D., Steduto, P., Hsiao, T.C., Fereres, E., 2009. AQUACROP v. 3.0. Reference Manual. FAO, Land and Water Division, Rome, Italy. Rhoades, J.D., 1982. Reclamation and management of salt-affected soils after drainage. In: Proc. First Annual Western Prov. Conf. Rationalization of Water and Soil Resources and Management , Alberta, USA. Richards, L.A., 1948. Porous plate apparatus for measuring moisture retention and transmission by soil. Soil Sci. 66, 105–110, NA. Sairam, R.K., Tyagi, A., 2004. Physiology and molecular biology of salinity tolerance in plants. Curr. Sci. 86 (3), 407–421. Shock, C.C., Feibert, E.B.G., Saunders, L.D., 1998. Potato yield and quality response to deficit irrigation. HortScience 33 (4), 655–659. Statistical Graphics Corporation, 2001. STATGRAPHICS Plus Versión 5.1. Statistical Graphics System. User’s Guide. STSC, Inc. Stewart, J.I., Hagan, R.M., Pruitt, W.O., Kanks, R.J., Riley, J.P., Danilson, R.E., Franklin, W.T., Jackson, E.B., 1977. Optimizing crop production through control of water and salinity levels. Utah Water Res. Lab. PWRG 151-1. Utah, USA. Torres, E.A., Calera, A., 2010. Bare soil evaporation under high evaporation demand: a proposed modification to the FAO-56 model. Hydrol. Sci. J. 55 (3), 303–315. Wannamaker, M.J., Pike, L.M., 1987. Onion response to various salinity levels. J. Am. Soc. Hort. Sci. 112 (1), 49–52.