Soil Science Society of America Journal 63:1397-1403 (1999)
© 1999 Soil Science Society of America
DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION
A New Drip-Injection Irrigation System for Crop Salt Tolerance Evaluation
R. Aragüésa,
E. Playánb,
R. Ortiza and
A. Royoa
a Unidad de Suelos y Riegos, Servicio de Investigación Agroalimentaria (Diputación General de Aragón) and Laboratorio Asociado de Agronomía y Medio Ambiente, Apartado 727, 50080 Zaragoza, Spain
b Departamento de Genética y Producción Vegetal, Estación Experimental de Aula Dei (Consejo Superior de Investigaciones Científicas), and Laboratorio Asociado de Agronomía y Medio Ambiente, Apartado 202, 50080 Zaragoza, Spain
aragues{at}mizar.csic.es
 |
ABSTRACT
|
|---|
An irrigation system was developed for the establishment of salinity gradients in field experiments that are aimed at obtaining salinity-yield response functions of crops. The drip-injection irrigation system (DIS) consists of a parallel pump system (a centrifugal pump for fresh water and an injection pump for saline water) and a conventional drip irrigation system composed of various irrigation sectors. The number of emitters installed in each irrigation sector determines the discharge of the centrifugal pump that blends with the fixed discharge of the injection pump. So, for the rest of fixed variables, the number of emitters (N) installed in a given irrigation sector determines the salinity of the irrigation water (ECiw):
, with a measured coefficient of determination of 
99%. The DIS was validated in an experiment where the salinity-yield response functions of ten barley cultivars were obtained using nine ECiw salinity treatments, with two replications per treatment. The DIS proved to be accurate and robust in that: (i) the measured ECiw gradient was similar to the target ECiw gradient (r positively correlated at P < 0.0001); (ii) the soil salinity (ECe) horizontal and vertical (0–50 cm depth) uniformities within each salinity treatment were low (average coefficient of variation [CV] of the pooled salinity treatments equal to 16–22% for the horizontal soil salinity and equal to 10% for the vertical soil salinity) and the temporal variability of soil salinity was low to moderate (average CV of the pooled salinity treatments equal to 18% during the studied period); and (iii) ECiw and ECe were positively correlated (P < 0.001). We concluded that the DIS is an excellent, low cost irrigation system for conducting field crop salt tolerance evaluations.
Abbreviations: CV, coefficient of variation • DIS, drip-injection [irrigation] system • DU, uniformity distribution coefficient • EC, electrical conductivity • ECa, four-electrode probe EC • ECe, saturation extract EC • ECfw, fresh water EC • ECiw, irrigation water EC • ECsw, saline water EC • EC1:5, 1:5 soil:water extract EC • FEP, four-electrode probe • H, pressure • LF, leaching fraction • N, number of emitters • Q, discharge • qu, unit emitter discharge • Viw, volume of irrigation water • 
g, gravimetric water content
|
INTRODUCTION
|
|---|
The salinization of soils and waters is becoming an increasingly serious constraint for the production of crops in the arid and semiarid regions of the world. Increasing the salt tolerance of crops might improve their productivity in these regions and provide stability of yield in subsistence agriculture. Salt-tolerant crops could also reduce the salinization of irrigated fields with poor drainage since, for a given irrigation water salinity, an increase in salt tolerance could lead to decreased leaching requirements and, therefore, the risk of rising water tables. These lower leaching requirements could also contribute to decreased salt load in the irrigation return flows and, therefore, to a decrease in salt loading's potential deleterious effects on the receiving systems (i.e., external effects). On the other hand, if an increase in salt tolerance promotes the use of more saline irrigation waters and/or the application of lower leaching fractions, then soil salinity or sodicity in the crop's root zone (i.e., internal effects) could dramatically increase. An appropriate balance between these external and internal effects is needed for the sustainability of agriculture. Addressing such salinity problems should therefore be made in a much more integrated manner, combining the management approach (i.e., limiting the build-up of soil and water salinity) and the biological approach (i.e., genetic improvement in salt tolerance). The work presented here is related to the biological approach.
Many salinity tolerance studies have been performed in controlled environments (hydroponics, sand cultures, greenhouse or growth chamber environments, etc.). However, genotypic differences observed in these situations may not correspond to those observed in the field (Shannon, 1997), so evaluation of the salt tolerance of crops under natural field conditions would be preferred. However, the high variability in soil salinity often observed under field conditions makes it difficult to obtain reliable and reproducible data needed for agronomic and plant breeding studies. The lack of rapid and precise field-controlled evaluation methods has hindered the development of accurate field-based salt tolerance research (Shannon and Qualset, 1984; Blum, 1988; Subbarao and Johansen, 1994; Flowers and Yeo, 1995).
Various automated systems for producing a range of irrigation water salinities for screening purposes and crop tolerance studies have been developed in the last decade. They may be classified into three groups: (i) systems in which fresh and saline waters are mixed in the irrigation line: mixing junction, brine injection, and mixing manifold systems (Pasternak et al., 1986); (ii) systems where fresh and saline waters are mixed in the air using sprinklers: triple-line source (Royo et al., 1987; Aragüés et al., 1992) and double-line source systems (Magnusson et al., 1988; Frenkel et al., 1990); and (iii) systems where fresh and saline waters are mixed in the soil: double drip-line system (Pasternak and De Malach, 1994) or double-emitter source (De Malach et al., 1996). A discussion of the advantages and limitations of these systems is out of the scope of this paper, but such matters are considered in the cited literature.
We present here the design and validation of a DIS, which may be classified in the above group (i) (i.e., mixing of saline and fresh water in the irrigation line). Relative to other systems used in the past, the DIS is different in that it establishes a salinity gradient in the irrigation water by simply installing a different number of emitters in each irrigation sector. This paper (i) describes the theory upon which the system is based, the system components, and the hydraulic circuitry, (ii) presents data on the salinity and volume of the irrigation water delivered at each target salinity treatment, (iii) evaluates the spatial and temporal variability of soil salinity using such a system, and (iv) determines the relationships between irrigation water salinity and soil salinity.
|
Theory
|
|---|
Pumping systems are often analyzed using performance curves. These curves establish relationships between discharge (Q) and pressure (H). When a drip irrigation system is analyzed, the performance curve relates the target system discharge to the required pressure at the inlet. The shape of the system performance curve depends on the head losses imposed by the pipes and fittings. A typical system performance curve is represented by curves S1 and S2 in Fig. 1
, where the head losses increase exponentially with discharge. The systems S1 and S2 differ only in the number of emitters. System S1 is characterized by a low number of emitters and requires a high pressure to attain a certain system discharge, whereas system S2 is characterized by a large number of emitters and requires a lower pressure to supply the same discharge.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1 Pump and system performance curves for the drip-injection system (DIS); S1, S2 = performance curves of a drip irrigation system with low and large number of emitters, respectively; c = performance curve of a centrifugal pump; i = performance curve of an injection pump; T = performance curve of the centrifugal and injection pumps working in parallel; P1, P2 = system–pump operating points for S1 and S2, respectively; Qi = fixed discharge of the injection pump; Qc1, Qc2 = variable discharge of the centrifugal pump obtained at P1 and P2, respectively; Q1, Q2 = total operating discharges obtained at P1 and P2, respectively
|
|
When the characteristics of a pump are analyzed, the curve provides information regarding the trade-off between service discharge and service pressure. In our work, two types of pumps were analyzed: centrifugal and injection. The performance of a typical centrifugal pump is represented by curve c (Fig. 1). When the pump is required to deliver a large discharge, it will do it at the expense of supplying a low pressure. On the other hand, an injection pump, also known as positive displacement pump, supplies a constant discharge independently of pressure. Injection pumps are based on a cylinder whose content is forced to flow inside the supply pipeline by a piston. The invariance of discharge with respect to pressure has made this type of pump very adequate for fertigation in drip and sprinkler irrigation. A pressure exists at which an injection pump can no longer push the piston. If this pressure is not attained, the performance curve (i) of an injection pump can be represented by a straight line (Fig. 1).
The theory behind the DIS is based on a system with two circuits, one for fresh and one for saline water. The fresh water is pumped into the system using a centrifugal pump, whereas the saline water is pumped using an injection pump. The injection point is located on the fresh water pipeline, downstream from the centrifugal pump. Thus, the DIS is hydraulically composed of two pumps in parallel and a drip irrigation system. When two pumps work in parallel, the total discharge is the sum of the discharge pumped by each pump. When this criterion is applied to the performance curves, the total pump performance curve is obtained. This curve is represented as T (Fig. 1). The last step in this analysis is to determine the system–pump operating point (i.e., the intersection of the system and the total pump curves). The points P1 and P2 (Fig. 1) are obtained for the systems S1 and S2, respectively. The corresponding operating discharges are Q1 and Q2, Q2 being larger than Q1. Both discharges can be expressed as the sum of the constant discharge of the injection pump (Qi) plus the variable discharge obtained from the centrifugal pump (Qc1 at P1 and Qc2 at P2). As shown in the figure, Qc1 will be smaller than Qc2. In each case, the water flowing through the system will be a blend of a fixed discharge of saline water (Qi) plus a variable discharge of fresh water. As a result, the salinity of the irrigation water will be a function of the number of the constant-flow emitters installed in the irrigation system.
If we define the variable x as the ratio of Qi (discharge of the injection pump) to Qt (total discharge or discharge of the injection plus the centrifugal pumps), then the total discharge can be obtained from the system performance curve. If pressure-compensating emitters are used in the system, the variable x can be expressed as
 | (1) |
where N is the total number of emitters installed in a given irrigation line, qu is the unit discharge of the emitter, and K is a constant. Equation [1] can be used to compute the electrical conductivity of the irrigation water (ECiw). This computation is based on the electrical conductivity of the fresh water (ECfw) and the saline water (ECsw):
 | (2) |
Substitution of Eq. [1] in Eq. [2] yields
 | (3) |
Equation [3] indicates that, for a given choice of pumps and water sources (i.e., fixed ECsw and ECfw), the salinity of the irrigation water (ECiw) is a unique function of the number of emitters (N). The qualitative relationship between N and ECiw is presented in Fig. 2
. The resulting curve is a potential relationship that asymptotically approaches ECfw as N 
. Equation [3] is only valid for Qt 
Qi. When Qt < Qi, backflow occurs at the centrifugal pump, and ECiw remains equal to ECsw independently of N. This situation is not desirable, since it implies contamination of the fresh water source and serious risk of equipment damage. Therefore, for each DIS set-up, a minimum number of emitters (Nmin) will be required to ensure proper functioning of the system.
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 2 Relationship between the electrical conductivity of the irrigation water (ECiw) and the number of emitters (N) installed in the irrigation line. ECsw and ECfw are the electrical conductivities of the saline and the fresh water sources, respectively
|
|
Once the system is in operation, a number of problems can occur. If a leak develops in the pipe fittings, the result will be the same as if new emitters were added to the irrigation system (the system performance curve will shift to the right in Fig. 1). Consequently, leaks are easy to identify, since the ECiw will be abnormally low. If one of the polyethylene pipes is plugged or kinked, then the result will be that total head losses will increase due to the new flow constriction. As a result, the system performance curve (Fig. 1) will shift to the left, and the resulting ECiw will be abnormally high.
Description of the System
The drip-injection system (DIS) was developed to supply a range of salt solutions, primarily to a drip irrigation set-up. The DIS is fully automated, controlled by an electronic logic unit, and all the components are commercially available. A schematic diagram of the DIS is presented in Fig. 3
. Polyethylene pipelines with diameters of 19.1 and 12.7 mm (3
4 and 12 inches) were used for the mainlines and the irrigation laterals, respectively.
Water from two sources (fresh and saline) enter the system via two independent inlets. The filtered fresh water is driven into the main irrigation line by an electric 1.86-kW centrifugal pump. A pressure meter and a flow meter are installed downstream from this pump to monitor water pressure, the instantaneous discharge, and the flow direction (to prevent backflow). After passing through various filters, the saline water is injected into the fresh water line by an electric injection pump with a discharge independent of the water pressure in the main irrigation line. The discharge of the pump can be regulated by modifying the piston displacement. This adjustment was performed during the set-up phase of the system to obtain an ECiw–N relationship that matches the target ECiw interval. The discharge of the injection pump was not modified, either during each irrigation event, nor during the irrigation season.
The fresh and saline waters are mixed in the main irrigation line in a proportion that depends on the discharges of the centrifugal and injection pumps. For a fixed discharge of the injection pump, the discharge of the centrifugal pump depends on the total number of constant-discharge emitters installed in each irrigation sector. By varying the number of emitters in each irrigation sector we can produce various dilutions of the saline water in a series of different sectors operated sequentially by means of the solenoid valves (Fig. 3). A feature of the DIS system is that the solutions of different ECiw are produced at different injection points and delivered to each irrigation sector separately; the pipes used to irrigate one sector are not used afterwards to irrigate the next sector. Hence, irrigation waters of different ECiw do not mix. This property results in an accurate establishment of the desired salinity levels. A commercial irrigation programming computer controls the solenoid valves (their open–close sequence), the on–off sequences of the two pumps, and the stirrer installed in the saline water tank (dashed lines in Fig. 3). The computer is equipped with an error detection routine that protects the system from energy surges and cuts and from programming errors.
Since the number of emitters in each irrigation sector varies as a function of the target ECiw, the irrigation time of each sector was programmed so that the applied volume of irrigation water (Viw) was the same in each sector. However, if desirable, the possibility of having Viw as an experimental variable is an important advantage over the triple-line source system, the design of which imposes a constant Viw throughout the salinity treatments (Aragüés et al., 1992). In the present system the emitters were characterized by a manufacturer pressure-flow equation,
. Within the range of pressures used, the unit discharge (qu) of the emitters was basically constant. The current cost of the automated DIS described here is around U.S. $5000, although this cost could decrease to $3000 if the computer and the solenoid valves were substituted by a manual, nonautomated system.
The DIS was validated in an experiment where the salinity-yield response functions for 10 barley cultivars were obtained using nine ECiw salinity treatments, with two replications per treatment (Ortiz, 1997). Thus, the total number of the individual 1.5-m2 barley plots was 180. The saline treatments were applied by means of 19 irrigation sectors [details below] so that the minimum number of emitters installed in each sector was greater than Nmin (Fig. 2) and sufficient in all cases to uniformly wet the area of each individual barley plot.
|
Results and discussion
|
|---|
Salinity and Volume of the Irrigation Water
Various preliminary trials were performed to obtain a satisfactory relationship between the number of emitters (N) and the ECiw. Figure 4
shows the ECiw–N relationships for two levels of salinity of the saline water
when the flow rate of the injection pump was set equal to 122 L h-1 and the salinity of the fresh water (ECfw) was 1.8 ± 0.2 dS m-1. Since the saline water of
(made up of Na and CaCl2 in a 2:1 weight ratio) was ultimately used in the experiments, the equation
was used to determine the number of emitters to install in each sector to obtain their target ECiw values. In practice, N was varied slightly from the above theoretical numbers in order to compensate for other minor head losses. These results indicate that, when the other variables are fixed (see Theory), the gradient in ECiw is easily and accurately determined from the number of emitters installed in each irrigation sector.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4 Relationship between the electrical conductivity of the irrigation water (ECiw) and the number of emitters (N) installed in the irrigation line, obtained for two salinities of the saline water
. The discharge of the injection pump was set equal to 122 L h-1. The regression equation of best fit and coefficient of determination (R2) are given
|
|
There were nine target salinity treatments
. A total of 19 irrigation sectors were established (one sector for each of the 2, 5, and 8 dS m-1 treatments, two sectors for each of the 11 and 14 dS m-1 treatments, and three sectors for each of the 17, 20, 23, and 26 dS m-1 treatments). The division of the treatments into various irrigation sectors was done to satisfy two criteria: salinity target and irrigation uniformity. In treatments with a high target ECiw, a low number of emitters is required to obtain the desired salinity (Fig. 4). At the same time, irrigation uniformity requirements forced us to consider a minimum emitter density (i.e., a minimum of 5 emitters m-2). In order to satisfy both criteria some treatments were split into two or three sectors that were irrigated sequentially. In this way, the required low number of emitters was distributed over a smaller irrigated area.
We used the treatments with more than one irrigation sector to ascertain the variability of the measured ECiw values among sectors of a given salinity treatment. Thus, for a total of 73 irrigations, the average CV of the ECiw values measured in two or three irrigation sectors of the same salinity treatment varied between a minimum of 1.6% (treatment with
) and a maximum of 3.5% (treatment with
), indicating that the irrigation sectors pertaining to the same salinity treatment gave similar ECiw values.
The ECiw was measured throughout the barley irrigation season (Feb.–May 1995) at one emitter from each of the 19 irrigation sectors. On Irrigations 9, 25, and 47 the number of emitters installed in six, nine, and four sectors, respectively, was slightly modified (generally the changes were within ±1 emitter m-2) so that the measured ECiw would more closely approach the target ECiw values. On Irrigations 57, 66, and 70, a solenoid valve was damaged and remained open after its sequence of irrigation. Since the number of emitters diverting water increased (emitters of a given sector plus those of the sector with the open solenoid valve), the result of this malfunctioning was an abrupt decrease in ECiw. It should be noted that any operational problem of the DIS was easily detected in the field by measuring the ECiw with a portable EC meter, since increases or decreases in head losses imply increases or decreases in ECiw. Apart from these problems, the ECiw gradient imposed by the DIS was close to that intended.
The CV of the mean ECiw for the first 40 irrigations varied among sectors between 17.1% (sector of 2 dS m-1) and 4.5% (sector of 23 dS m-1), with an average CV of 8.4% for all sectors. The relatively high variability of the 2 dS m-1 sector was due to changes in the salinity of the canal water (average EC of 1.8 dS m-1), since this sector did not receive saline water from the injection pump. The average CV for all sectors increased to 11.5% when all the irrigations were considered because of the system failures already mentioned after Irrigation 56. This higher variability may then be attributed to operational problems, and not to design problems of the DIS. In any case, the low CV of the mean ECiw values indicates that the system worked exceptionally well.
The relationship between the target ECiw and the mean irrigation season ECiw measured in each of the 19 irrigation sectors was highly consistent. The regression equation was
. The correlation coefficient
was significant at P < 0.0001, the intercept was not significantly different from zero (P > 0.05), and the regression coefficient was not significantly different from unity (P > 0.05), indicating that the measured ECiw values were not significantly different from the target ECiw values. The standard error of the ECiw estimate was 0.44 dS m-1. We concluded from these results that the DIS established a salinity gradient that was highly reliable (i.e., with a low spatial and temporal variability and with salinities similar to the target ECiw).
One potential problem in drip irrigation using saline waters is clogging of the emitters by precipitates. This is an important problem for the DIS, since any emitter's clogging will increase the head losses and, therefore, the ECiw. We examined this potential problem by measuring the flow rates delivered by an emitter in each irrigation sector during 50 irrigations in the period between 10 March and 29 May. The mean flow rates varied between a maximum of 6.3 L h-1 in Sector 26b and a minimum of 4.0 L h-1 in Sector 5. The maximum CV of the mean flow rates was 27.1% (in Sector 23c, where its solenoid valve did not work properly at the end of the irrigation season), and the minimum CV was 2.0% (in Sector 8). The average flow rate for all sectors and irrigations was 5.8 L h-1, and the composite CV was 6.3%. We also evaluated the uniformity of water application for each irrigation event by means of the uniformity distribution (DU) coefficient proposed by Merriam and Keller (1978). The DU were high (between 80–95%) and the average flow rates of the 19 emitters measured in each irrigation varied between 5.3 and 6.2 L h-1. We concluded from the results in this study that clogging of emitters was not important and that the spatial (i.e., among irrigation sectors) and temporal (i.e., among irrigation numbers) variabilities of the emitter's flow rates were quite low. However, other waters with higher tendencies for salt precipitation might contribute to higher rates of emitter clogging.
Soil Salinity and Soil Water Content: Spatial and Temporal Variability
An analysis of the uniformity of the salinity and the water content imposed in the soil by the DIS is needed for validation purposes. The horizontal spatial variability was determined by taking at two dates (6 April and 11 May 1995) soil samples (0–50 cm depth) in various 1.5-m2 barley plots randomly selected within each salinity treatment. A total of 146 samples were brought to the laboratory and the EC1:5 (soil:water) extract and the gravimetric water content (g) were measured by standard methods (USSL, 1954). The uniformity within each salinity treatment was assessed by calculating the CV of the average EC1:5 and g values measured in n soil samples (Table 1)
. The CV of the average EC1:5 varied between a maximum of 35.2% (treatment of
sampled on 11 May) and a minimum of 9.3% (treatment of ECiw = 2 dS m-1 sampled on 11 May). The average CV of the pooled salinity treatments was 16.4% (6 April) and 22.5% (11 May). The CV of the average g values were low (6.8% and 7.6% for 6 April and 11 May samplings, respectively) and presumably reflected the high irrigation application rates and frequencies imposed by the DIS. We concluded that the DIS established soil horizontal (0–50 cm soil depth) uniformities that are acceptable (average CV < 23%) in terms of soil salinity and high (average CV < 8%) in terms of soil water content.
View this table:
[in this window]
[in a new window]
|
Table 1 Soil horizontal spatial variability of salinity (EC1:5) and gravimetric water content (g): coefficients of variation (CV) of the mean EC1:5 and mean g of n soil samples (0–50 cm depth) taken in each salinity treatment (target ECiw) on 6 April and 11 May 1995
|
|
The uniformity with depth of soil salinity and soil water content was determined by taking on two dates (6 April and 11 May 1995) soil samples at 0- to 25-, 25- to 50-, and 50- to 100-cm depths in various plots (n in Table 1) that were randomly selected within each salinity treatment. The average CV of the pooled salinity treatments for the 0- to 25- plus 25- to 50-cm soil depths were 10.4% (EC1:5) and 6.9% (g), whereas for the 0- to 25- plus 25- to 50- plus 50- to 100-cm soil depths they were 30.4% (EC1:5) and 13% (g). These results indicate that the 0- to 50-cm salinity and water content profiles were fairly uniform, whereas the 0- to 100-cm soil salinity profiles were more variable, due to the lower EC1:5 values measured at the 50- to 100-cm soil depths (reflecting the 4-mo experimental period).
Since the 0- to 50-cm soil profile is most important from the point of view of crop's responses to salinity in high-frequency irrigation, we compared the EC1:5 and g values measured at 0- to 25- and 25- to 50-cm soil depths. Figure 5
shows the mean and standard deviation values of EC1:5 and g measured in 6 to 10 random plots of each salinity treatment, as well as the percent differences (0–25 cm - 25–50 cm) of the means for both depths. Although the EC1:5 and g values were generally higher in the 0- to 25- cm than in the 25- to 50-cm depth, their differences were in all cases <17% and not significantly different (P > 0.05). We concluded that the 0- to 50-cm soil profile is quite uniform both in terms of salinity and water content, which is very advantageous for establishing the response of crops to soil salinity. However, it should be pointed out that these are conditions representative of high frequency, high leaching fraction irrigation systems, such as drip irrigation. Under more conventional irrigation systems, the salinity and water content profiles will not probably be as uniform, and the plant response functions may be different.
View larger version (52K):
[in this window]
[in a new window]
|
Fig. 5 Mean and one standard deviation values of EC1:5 (1:5 soil:water extract) and gravimetric soil water content measured in 6 to 10 random plots of each salinity treatment (target ECiw) imposed by the drip-injection system. The percent difference between the values measured at 0–25 and 25–50 cm soil depths is also given
|
|
The temporal variability of soil salinity imposed by the DIS was analyzed by installing in each salinity treatment a four-electrode probe (FEP) to a soil depth of 30 cm. The FEPs were not calibrated before installation, since our objective was to analyze their relative changes with time and not the absolute EC values. The ECa readings of the probes were made two days per week in the period from 5 April to 25 May. Since soil texture and soil water content were quite uniform within the experimental area, these ECa readings are representative of the soil solution EC.
The temporal changes in the ECa readings within each of the nine salinity treatments are presented in Fig. 6
. After Day 32, the FEP installed in the 23 dS m-1 treatment was discarded because of its erratic readings, and the ECa of the FEP installed in the treatment of 26 dS m-1 sharply increased for no apparent reason. The ECa readings obtained with the rest of FEPs indicate that the soil salinities imposed by the DIS were relatively constant. These results are confirmed by the CV of the mean ECa readings obtained in each salinity treatment (maximum of 26.5% in Treatment 26, minimum of 14.1% in Treatment 2, and average of 18.2% for the pooled treatments). We concluded that the temporal variability of soil salinity was low to moderate and appropriate for crop tolerance evaluations in which it is desirable to obtain constant salinity values for the period studied.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 6 Four-electrode probe readings (ECa) obtained in nine salinity treatments imposed by the drip-injection system during the period 5 April to 25 May 1995
|
|
Correlations between Irrigation Water Salinity (ECiw) and Soil Salinity (ECe)
The soil salinity gradient imposed by the DIS should be consistently and significantly correlated with the irrigation water salinity gradient in order to accurately estimate ECe (saturation extract EC) from the average ECiw. Figures 7a, 7b, and 7c
present, respectively, the ECe values measured by standard methods (USSL, 1954) in each salinity treatment at 0- to 25-, 25- to 50-, and 50- to 100-cm soil depths at two sampling dates (6 April and 11 May 1995) against the corresponding ECiw (time-weighted average values measured from the start of irrigation up to the corresponding sampling date). The dilution effect of rainfall was not included in this analysis because rainfall was negligible during the irrigation season.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7 Relationships between the electrical conductivity of the soil saturation extract (ECe) and the time-weighted average ECiw (measured electrical conductivity of the irrigation water) obtained on two sampling dates (6 April and 11 May) for the (a) 0–25 cm, (b) 25–50 cm, and (c) 50–100 cm soil depths. Linear regression equation and coefficient of determination between the ECe of (d) the 0–50 cm soil depth and the measured ECiw using the pooled observations of the two sampling dates
|
|
All the regression lines depicted in these figures had correlation coefficients significant at P < 0.001, except that of the 50- to 100-cm soil depth sampled on 6 April, which was significant at P < 0.05. The regression equations obtained at the two sampling dates were different, so that the ECe of the lowest salinity treatments tended to increase with time more than the ECe of the highest salinity treatments. This is the expected response in experiments where the same amount of water is applied in all the salinity treatments, because crop evapotranspiration (and its corresponding concentration effect on soil water) decreases with increasing salinities. However, as mentioned before, and in contrast to the triple-line source system, with the DIS we could have obtained, if desired, similar leaching fractions in all the salinity treatments by adjusting the volumes of applied water relative to the corresponding crop evapotranspiration estimates.
The first 50 cm of the soil are often the most important from the point of view of the response of crops to soil salinity, especially in high-frequency irrigation systems such as the DIS. For this reason, we obtained the ECiw–ECe (0–50 cm) linear regressions for the two sampling dates. Since, as expected from the previous results, they were different, we pooled together the observations of the two sampling dates in order to obtain a single regression equation that was likely to be more representative for the irrigation season of barley (Feb.–May). Figure 7d shows that the ECiw gradient imposed by the DIS is satisfactorily reproduced in the soil, with a coefficient of determination (0.91) significant at P < 0.001 and a standard error of the ECe estimate of 1.80 dS m-1. This equation may be used to estimate the soil ECe (0–50 cm) within each irrigation sector, based on the average ECiw values measured during the growing season.
|
Conclusions
|
|---|
A new drip-injection irrigation system (DIS) was developed for the establishment of salinity gradients in field experiments. The DIS is based on the combination of two pumps in parallel: a centrifugal pump for fresh water and an injection pump for saline water. The two waters are mixed within the irrigation line to produce an irrigation water whose salinity (ECiw) depends on the discharge ratio of the pumps. We have shown that the number of emitters connected to the pumping system determines the discharge of the centrifugal pump that blends with the fixed discharge of the injection pump. Therefore, the number of emitters installed in a given irrigation sector controls the salinity of its irrigation water. The relationship between ECiw and the N was characterized by an equation
with a highly significant coefficient of determination (
0.99, P < 0.0001).
The DIS set-up requires one tank each for fresh and saline water (including an electric stirrer), various filters, the above-mentioned pumps, pipes, valves, and emitters. The system in the reported experiment was fully automated, using an irrigation computer and solenoid valves. Automation facilitated the operation of the system and gave good control of irrigation timing and application depths. Nevertheless, automation is not strictly required and should be considered as an upgrade to the DIS system. The costs of the DIS system totaled $5000 U.S. in 1998. If automation were not implemented, the cost would be on the order of $3000.
The DIS has proven to be accurate and robust in that: (i) the imposed irrigation water salinity gradient was highly consistent, with low spatial (average CV < 3.5%) and temporal (average CV < 12%) variabilities for both the ECiw and the emitter's flow rate, and with measured ECiw values similar to the target ECiw (r significant at P < 0.0001); (ii) the soil horizontal (0–50 cm depth) uniformities were acceptable for soil salinity (average CV < 23%) and high for soil water content (average CV < 8%), the 0- to 50-cm soil profiles were very uniform both for salinity (average CV = 10.4%) and water content (average CV = 6.9%), and the temporal variability of soil salinity was low to moderate (average CV = 18.2%); and (iii) the irrigation water salinity and the soil salinity were significantly correlated (P < 0.001), indicating that the ECiw gradient imposed by the DIS produced a satisfactory ECe gradient in the soil, and that the ECe of each salinity treatment may be accurately estimated from the corresponding ECiw values measured during the irrigation season.U.S. Salinity Laboratory Staff 1954
|
ACKNOWLEDGMENTS
|
|---|
We thank J. Gaudó, M. Izquierdo, T. Molina, and L. Naval for technical assistance.
Received for publication October 5, 1998.
|
REFERENCES
|
|---|
- Aragüés R., Royo A., Faci J. Evaluation of a triple line source sprinkler system for salinity crop production studies. Soil Sci. Soc. Am. J. 1992;56:377-383.[Abstract/Free Full Text]
- Blum A. Plant breeding for stress environments. Boca Raton, FL: CRC Press, 1988.
- De Malach Y., Ben-Asher J., Sagih M., Alert A. Double-emitter source (DES) for irrigation experiments in salinity and fertilization. Agron. J. 1996;88:987-990.[Abstract/Free Full Text]
- Flowers T.J., Yeo A.R. Breeding for salinity resistance in crop plants: Where next?. Aust. J. Plant Physiol. 1995;22:875-884.[ISI]
- Frenkel H., Mantell A., Vinten A., Meiri A. Double line-source sprinkler system for determining the separate and interactive effects of water and salinity on forage corn. Irrig. Sci. 1990;11:227-231.
- Magnusson D.A., Ben Asher J., De Malach Y. A continuous two-variable design using the line-source concept. Agron. J. 1988;80:132-133.
- Merriam, J.L., and J. Keller. 1978. Farm irrigation system evaluation: A guide for management. US AID Rep. Utah State University, Logan, UT.
- Ortiz R. Diseño y validación del nuevo sistema doble fuente de goteo para el establecimiento de gradientes de salinidad en el suelo. Spain: Análisis de la tolerancia a la salinidad de diez cultivares de cebada. Proyecto Final de Carrera. Universidad de Lérida, 1997.
- Pasternak D., Azoulai A., Danon A., Levi S., De Malach Y., Shalev G. Irrigation with brackish water under desert conditions. VI. Automated systems to produce a range of salt concentrations in irrigation water for experimental plots. Agric. Water Manage. 1986;12:137-147.
- Pasternak, D., and Y. De Malach. 1994. Crop irrigation with saline water. In M. Pessarakli (ed.) Handbook of plant and crop stress. Marcel Dekker, New York.
- Royo A., Faci J., Aragüés R. Utilización de la triple línea de aspersión para evaluar la respuesta de los cultivos a la salinidad. Riegos y Drenajes XXI 1987;17:48-54.
- Shannon M.C. Adaptation of plants to salinity. Adv. Agron. 1997;60:75-120.
- Shannon M.C., Qualset C.O. Benefits and limitations in breeding salt-tolerant crops. Calif. Agric. 1984;38:33-34.
- Subbarao, G.V., and C. Johansen. 1994. Strategies and scope for improving salinity tolerance in crop plants. In M. Pessarakli (ed.) Handbook of plant and crop stress. Marcel Dekker, New York.
- U.S. Salinity Laboratory Staff. Diagnosis and improvement of saline and alkali soils. USDA-USSL Agric. Handb. 60. Washington, DC: U.S. Gov. Print. Office, 1954.
TOP
ABSTRACT
INTRODUCTION
