| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Soil Science, Univ. of Tarbiat Modarres, Tehran 14155-4838, Iran
b Sub-Dep. of Water Resources, Wageningen Agricultural Univ., Nieuwe Kanaal 11, 6709 PA Wageningen, The Netherlands
* Corresponding author (Mhomaee{at}hotmail.com)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: EC, salinity value • R, reference treatment
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Soil water potentials are normally defined as potential energy divided by mass, 
(J kg-1), but potential energy divided by weight (J N-1 = L) or head equivalent, h, (h = g-1, where g = gravitational field strength) is more convenient to use (Koorevaar et al., 1983). From here on, we will use the head equivalents of matric and osmotic soil water potential, pressure head, h, and osmotic head, ho, respectively. The low osmotic head that corresponds to high negative osmotic pressure produces salinity stress for plants. The head equivalent of gravitational potential is simply the height, z, above an arbitrary reference level, z = 0.
Although root-water uptake is reduced because of low soil water pressure and osmotic heads, it is not clear how these stresses interact when they occur together, and vary with time and depth. The simple additivity of soil water pressure and osmotic heads as proposed by early investigators (Ayers et al., 1943; Wadleigh and Ayers 1945; Wadleigh et al., 1946, 1947; U.S. Salinity Laboratory Staff, 1954) is still questionable. Because soil water pressure and osmotic heads are additive in reducing the free energy of soil water, it was assumed that their effect on transpiration is also additive through reduction in the availability of water for plants.
Early studies (Ayers et al., 1943; Wadleigh and Ayers 1945; Wadleigh et al., 1946) indicated that yield reduction is a function of integrated osmotic and pressure heads. This observation led to the conclusion that the effect of excess salt on transpiration is similar to that of water stress, from which the additive concept was born. Meiri (1984) analyzed data of Meiri and Shalhevet (1973), Sepaskhah and Boersma (1979), Jensen (1982), and Parra and Romero (1980) by multilinear regression and found that the effects of osmotic and pressure heads are somewhat additive. It is important to keep in mind that in most of these investigations (polyethylene glycol) PEG rather than a real drought intrusion was used to create water stress. Shalhevet (1984)(1994) referred to the same studies and stated that the evidence suggests that the effects of salinity and water stress are identical, and that a unified function may be applied to both stress components. This would imply that h and ho are additive in their effect on transpiration, but one unit of h is not equivalent to one unit of ho (Shalhevet, 1994). Such a conclusion remains useless, however, unless (at least) an empirical proportionality coefficient can be determined. Such empirical coefficients are considered to be plant, soil, and climate specific, but have never been introduced in the literature. The main reason for this failure is that the proportionality coefficients are highly variable in time and space and are rather difficult to obtain experimentally.
Despite some similarities, there are some clear differences between plant responses to salt and water stresses. One of the more important observations is the lack of wilting under salt stress at water potentials which cause wilting under water stress. Wadleigh and Ayers (1945) and Sepaskhah and Boersma (1979) reported no wilting at low osmotic heads, whereas there was wilting at equivalent low pressure heads. These observations led to the conclusion that a decreasing pressure head is more detrimental than an equivalent decrease in osmotic head. Furthermore, there may be a difference in the nature of the solutes that contribute to osmotic adjustment. The obvious difference between salinity and water stress is in leaf turgor and the growth processes that are influenced by it. Shalhevet and Hsiao (1986) indicated that the growth rate under water stress was half as large as under salt stress in the leaf water potential range of interest. Meiri (1984) indicated that soil water pressure head had a greater influence on shoot growth and transpiration than osmotic head. However, for root growth the effect of ho was greater than that of h (Meiri, 1984). One more observation was that the permeability of roots was increased under saline conditions. Such adjustments have not been reported under water stress conditions. These arguments might be regarded as reasons for suspecting the additivity concept.
To quantify root water uptake, the microscopic and macroscopic extraction approaches are available. The most common formulation of microscopic models is based on the work of Gardner (1964). Among the microscopic extraction functions, only the model of Nimah and Hanks (1973a) deals with both h and ho. In this model, the osmotic head is simply added to the pressure head to establish the water potential gradient from the soil to the root. While Nimah and Hanks (1973b) and Childs and Hanks (1975) reported good agreement between field data for alfalfa under joint water and salinity stresses and computed water contents over the root zone and transpiration, others (Wolf, 1977; Hanks, 1984; Childs and Hanks, 1975; and Cardon and Letey, 1992) reported unsatisfactory results when the model was applied for different field conditions. Cardon and Letey (1992) showed that in its calculations of root water uptake the model is generally inconsistent with plant behavior. The insensitivity of this model is caused by the manner in which the salinity effect is incorporated in the uptake term S. The value of S is dominated by the nonlinear changes of h and K (unsaturated hydraulic conductivity) with 
(water content). In contrast, ho decreases linearly with (simple concentration-dilution). Moreover, increasing the salinity of the irrigation water while maintaining high water contents, results in relatively high K() values and plant water extraction proceeds at or near maximum levels.
The macroscopic approach deals differently with the combined stresses. The concept is generally based on the work of Feddes et al. (1978) and assumes that the extraction term under nonstress conditions is simply equal to potential transpiration over the root zone. As soon as the soil water pressure head reaches a critical value, the actual transpiration reduces linearly until the root-water uptake ceases completely (wilting point). This reduction is quantified by the so-called reduction function. Basically, the macroscopic models do not account for saline conditions. Van Genuchten (1987), Dirksen et al. (1993), Homaee (1999), and Homaee and Feddes (1999) incorporated different nonlinear osmotic head-dependent reduction functions in the Feddes et al. (1978) model. Most of these are based upon the so-called multiplicativity concept that uses the product of the separate reduction terms for soil water osmotic and pressure heads. This concept was originally proposed by van Genuchten (1987) and has been used extensively in many numerical simulation models dealing with root -water uptake. The objective of this paper is to investigate the joint influence of different levels of soil water osmotic and pressure heads on root-water uptake patterns, and to investigate which concept fits experimental data best, or what adjustments need to be made in existing concepts.
|
THEORY |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
 | [1] |
is volumetric water content (L3 L-3), t is time (T), C is differential soil water capacity (L-1) which is equal to the slope d/dh of the soil water retention curve, h is soil water pressure head (L), z is gravitational head, as well as the vertical coordinate (L) taken positive upward, K is soil hydraulic conductivity (LT-1), S is soil water extraction rate by plant roots (L3 L-3 T-1).
The macroscopic root extraction term reads:
 | [2] |
(h, ho) is the reduction function depending on soil water pressure (h) and osmotic (ho) head. The available reduction functions can be divided into two categories: additive and multiplicative. The only additive reduction function introduced by van Genuchten (1987) reads:
 | [3] |
(h) is reduced to 0.50; a1 and a2 are empirical coefficients for soil water pressure head and osmotic head, respectively; p is an empirical, presumably crop, soil, and climate-specific dimensionless parameter. The value of p was found to be 3 when the S-shaped function was applied to salinity stress data only (van Genuchten and Gupta, 1993).
The multiplicative reduction functions are limited to those proposed by van Genuchten (1987), Dirksen et al. (1993), van Dam et al. (1997), and Homaee (1999). Van Genuchten (1987) proposed:
 | [4] |
Dirksen and Augustijn (1988) and Dirksen et al. (1993) multiplied identical reduction terms for water stress and salinity stress, each with their own values for the threshold value h*, h*o, and 50% values h50 and ho50:
 | [5] |
Van Dam et al. (1997) simply multiplied the reduction functions of Feddes et al. (1978) and Maas and Hoffman (1977):
 | [6] |
For salinity and water stress, Homaee (1999) proposed separate two-threshold nonlinear functions that fit the experimental data satisfactorily. He replaced the difficult to obtain h50 and ho50 with hmax and homax, respectively, and determined the exponents without a need to have extra parameter values. Furthermore, he introduced a second threshold value from which decreasing osmotic or matric potentials do not influence the root water uptake significantly. For the combined stresses he proposed:
 | [7] |
01 and 02 are the relative transpiration at hmax and homax, respectively. The dimensionless exponents p1 and p2 can be obtained (Homaee, 1999) from:
 | [8] |
 | [9] |
Equation [7] belongs to the multiplicative category which in principal has no physical basis and cannot discriminate between its components. Different reductions because of osmotic and soil water pressure head produce the same result. For instance, the joint reduction (ho, h) due to (ho) = 0.25, and (h) = 0.50 is exactly the same as for (ho) = 0.50 and (h) = 0.25. From this point of view, multiplicativity can be regarded as an identical approach. As for the magnitude of the product, there is no evidence to support that separate reductions of 0.25 and 0.50 because of osmotic head and pressure head cause 0.875 reduction in water uptake.
In view of these shortcomings, we introduce a new combination of the reduction functions of Maas and Hoffman (1977) and Feddes et al. (1978) that differs conceptually from the additive and multiplicative approaches. Figures 1a and 1b illustrate these reduction functions, respectively. Figure 1a illustrates the generalized piecewise linear response function suggested by Maas and Hoffman (1977). According to this model, the relative transpiration Ta/Tp remains at the maximum up to a salinity threshold value (EC*) above which Ta/Tp decreases linearly with increasing salinity. The reduction function of Fig. 1b can be divided into three parts. Part I (triangle h1Ah2) represents air deficiency, Part II (rectangular h2ABh3) is the nonstress part, and Part III (triangle h3Bh4) represents water stress. The slope of Bh4 is determined by h3 and h4. The latter (wilting point) is constant for a particular plant and h3 is assumed to be only evaporative-demand dependent (Feddes et al., 1978). Since this model originally was developed for nonsaline conditions, the slope of Bh4 is valid for salinities equal to or less than the salinity threshold value EC*, as shown in Fig. 1b. On the other hand, the data base of Maas and Hoffman (1977) is collected from nonrestricted water conditions. Taking advantage of this, we apply the reduction of this model directly to the no-water stress part of Fig. 1b, as shown in Fig. 1c for alfalfa. Assuming linear reduction from B to h4, we assume further that each dS m-1 salinity beyond the threshold value shifts the wilting point 360 cm to the left. This is consistent with the observation that plants wilt at higher soil water pressure head in the presence of salinity than without salinity. The magnitude of 360 proposed here is only a preliminary guess based on the well-known empirical relation in USDA Handbook 60 to translate soil salinity to osmotic head (U.S. Salinity Staff, 1954) and will be used until further evidence provides a more precise quantity. The effect of each level of joint water and salinity stress can be obtained as illustrated in Fig. 1d.
|
 | [10] |
ho and (h4 - ho) h h3, respectively. Other general validities are the same as the original models.
This combination model is flexible in that any other (nonlinear) salinity reduction function can be used instead of the linear function of Maas and Hoffman (1977). Different salinity reduction functions should only change the height of the horizontal nonstress line segment of the Feddes et al. (1978) reduction function and not influence the slope of the line segment beyond h3 significantly. The reduction function because of salinity and water stress remains linear, and the right-hand side of Eq. [10] can be replaced with the appropriate parts of Eq. [4], [5], and [7], respectively, in the following way:
 | [11] |
 | [12] |
 | [13] |
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Soil water content was measured with time domain reflectometry (TDR) equipment. An automated TDR system with 32 probes was installed for four columns (eight position for each) in the greenhouse to obtain measurement series of the soil effective permittivity at different depths in the root zone. Every 2 h, wave forms of all probes were measured, using a Tektronix 1502B cable tester1 (Tektronix Inc., Beaverton, OR) and a multiplexer and control system developed by Heimovaara and Bouten (1990). The wave forms were stored in a personal computer. The measurement and storage of 32 wave forms took about 11 min. The stored wave forms were analyzed with procedures and programs of Heimovaara and Bouten (1990). The waveforms collected for the manual TDR measurements were also analyzed with the same program. All volumetric soil water contents were obtained based on the calibration equation of Topp et al. (1980). Soil water pressure heads h were obtained by converting to h based on the soil-water retention characteristics obtained by the Wind's evaporation method (Wind, 1966).
Soil water osmotic heads ho were obtained by converting the corresponding soil solution salinities based on the empirical relationship given in USDA Handbook 60 (U.S. Salinity Laboratory Staff, 1954). Salinity of soil solution, ECss, was measured in situ with salinity sensors and a salinity bridge (Model 5100, Soilmoisture Equipment Company, Santa Barbara, CA). All sensors were installed horizontally into the soil columns in one row at depth intervals of 5 cm in the top 30 cm and at 10-cm intervals below that. The same order of depth intervals was followed for the TDR sensors. Salinity of irrigation and drainage waters were measured by a conductivity cell (Digimeter L21; Eijkelkamp, Agrisearch Equipment, The Netherlands).
Alfalfa was grown from seed in the packed soil columns in a greenhouse under controlled environmental conditions. To fix N of air in the roots, the seeds were inoculated with Rhizobium bacteria. First the suspension of Rhizobia was mixed by Carboxyl methyl cellulose (CMC) and later four parts of seeds were mixed by one part of this mixture. Carboxyl methyl cellulose was used to fix the Rhizobial cells to the seed coat. The wet seeds were dusted with dry CaCO3 (1 g CaCO3 for 2 g of seeds). After this inoculation, alfalfa was immediately seeded at a density of four seeds per location and 20 locations per column. A week later, all locations were thinned to one plant, giving 20 plants per columns. The measurements started after healthy plants had developed. At the end of a stress period the plants were harvested, excess salinity was flushed out of all soil columns and the alfalfa was allowed to develop healthy plants under no-stress conditions, before the next stress was introduced. Depending on the climatic conditions of the greenhouse, the latter process usually took about 8 to 10 wk.
Actual transpiration (T) measurements were made by suspending the columns and weighing them with a digital balance, each column at least five times a day. The transpired water was related to the surface area of the soil columns, rather than to the plant canopy. To prevent evaporation from the soil surface, the top of each column was covered by inert granules (7-mm diam.).
The experiments were carried out in four major phases. Each experimental phase had its own specific (duplicated) reference treatment (R) without salinity or water stress. The amounts of water for the stressed treatments within each experimental phase were derived from the Reference R and all other collected data were compared with those of R. All irrigations of R were with tap water of ECiw < 0.2 dS m-1. Since no water stress was allowed for R, the irrigation intervals needed to be short; hence, the columns received irrigation water every 2 d during the entire experiment. The target leaching fraction of the salinity treatments was 0.5. Thus, a large amount of saline water in excess of potential transpiration was applied to the columns, and the same amount of excess water was given to the reference treatment. This provided a relatively similar water distribution over the root zone for all treatments. In the cases of water stress and joint salinity and water stress, no extra water was given to the reference as it was not given to the stressed treatments.
Experimental Phase I was carried out under salinity stress without water stress, consisting of five treatments. The soil was salinized by twice saturating and draining all columns, applying appropriate amounts of water and salinity. To avoid toxicity effects and precipitation-dissolution reactions of salts with the soil solid phase, salinities were created by adding equal molar (charge base) quantities of CaCl2 and NaCl to the irrigation water. The imposed salinity levels were related to the salinity threshold value of alfalfa, EC = 2 dS m-1 of saturation extract. Accordingly, the EC of the irrigation water was 1.5, 2.0, 3.0, 4.0, and 5.0 dS m-1 for the treatments S1W0, S2W0, S3W0, S4W0, and S5W0, respectively. Similar to R, these treatments were irrigated every 2 d. To minimize salt accumulation, particularly in the deeper part of the root zone, the target leaching fraction was 0.50 for all saline columns. The excess water was sucked out overnight with a vacuum pump at suctions of 60 to 80 cm. Assuming no water uptake during the dark period, all irrigation waters were applied to the columns by flood irrigation immediately before turning off the lights to allow the applied water to distribute over the root zone in the time that the plants did not transpire water. This allowed us to assume that most downward water movement occurred at night, and that during the light period relatively little downward water movement took place. To attain the target leaching fractions, the columns were saturated. Thus, it can be assumed that during the measurements, hysteresis in soil water did not occur and the main drying curve of the soil moisture retention characteristic could be used. All measurements (with few exceptions) started after switching on the lights. The light period normally was 15 h d-1 until 2100 h.
In experimental Phase II, two levels of water stress were introduced to the plants, denoted as S0W1 and S0W2. Salinity stress was not allowed, and hence the irrigation intervals were relatively long (i.e., 4 and 3 d) to let the columns dry out as much as possible. The amounts of irrigation water for these treatments were 70 and 50% of their own reference treatment, that is, W1 = 0.7R and W2 = 0.5R. Some fertilizers in solution form were added to the irrigation water to prevent any possible nutrient deficiency. Because alfalfa can fix N in its roots, no N fertilizer was added to the soil. A complete nutrient mixture (including P, K, Mg, Ca, S, Fe, Zn) was added in solution to the irrigation water. The EC of the applied water with the fertilizers was always <0.3 dS m-1, thus the salinity caused by these applications was negligible. To verify that, some salinity measurements were made after each irrigation. Until the last application, soil solution salinity was less than the lowest readable value with the salinity bridge.
In experimental Phases III and IV all possible combinations of the experimental Phases I and II with their own references were applied. Experimental Phase III with salinity stress and first level water stresses consisted of five treatments denoted as S1W1, S2W1, S3W1, S4W1, and S5W1, respectively. Since there was no leaching, after each water application the salinity in the root zone increased particularly in the upper parts. The amount of applied irrigation water for all replicates was about 0.7 of potential transpiration, calculated from the reference treatment (W1 = 0.7R). The first two irrigation intervals of 4 d were followed by 3-d intervals. The columns received the same amount of irrigation water; the only difference between the treatments was the salinity of the irrigation water which varied from 1.5 to 5 dS m -1. Since the columns did not receive enough water, no leaching occurred.
In experimental Phase IV, both salinity stress and the second level of water stress (W2 = 0.5R) were investigated, having five treatments denoted as S1W2, S2W2, S3W2, S4W2, and S5W2, respectively. The columns received the same amount of irrigation water; thus the only difference within treatments was the salinity of the applied water. At the end of each experimental growth period, plants were harvested and the wet and dry matter of each individual column determined; the latter by drying the plant for 24 h at 70°C.
To facilitate discovering any consistent relationship of the separate and combined stresses, various experimental parameters are summarized in Table 1. Similar to de Wit (1958) we assume that the relative uptake is equal to the relative transpiration:
 | [14] |
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
According to the multiplicative models, the separate reductions because of salinity and water stress can simply be multiplied. The data presented in Table 1 can potentially confirm or reject this concept. The product of Ta/Tp = of S1W0 and of S0W1 is 0.92 x 0.66 = 0.61. This product of the reduction terms because of the individual stresses S1 and W1 was smaller than the reduction term of the combined stress S1W1, for which Ta/Tp = = 0.74. Similarly, for W1 and S2, S3, S4, and S5 this product was 0.56, 0.51, 0.44, and 0.39, respectively, while the reduction term for the combined stresses S2W1, S3W1, S4W1, and S5W1 was 0.72, 0.65, 0.60, and 0.50. The same comparison yielded (0.92 x 0.50 =) 0.46, 0.42, 0.39, 0.33, and 0.30 versus 0.67, 0.64, 0.62, 0.40, and 0.26 for S1W2, S2W2, S3W2, S4W2, and S5W2, respectively. The multiplicative model predicted larger reductions and thus underestimated the actual transpiration, except for S5W2. This means that the presented experimental data did not confirm the multiplicative concept. This conclusion was drawn irrespective of any mathematical functions. This will be evaluated in the next subsection.
Application of Reduction Functions to Mean Soil Solution Osmotic and Pressure Heads
Figure 2
presents the relationships between the experimental relative transpiration Ta/Tp and the mean1/2h1/2 for different mean ECss over the root zone. Either linear or nonlinear fitting can be applied for the lowest salinity level, ECss equals 2 to 3 dS m-1, but the relations became almost linear as the mean ECss increased. The maximum relative transpiration decreased from 1.00 for mean ECss = 2 to 3 dS m-1 to 0.40 for mean ECss equals 9 to 11 dS m -1. The mean1/2h1/2 at which the plants' biological activities became minimal (wilting) changed from 8000 cm for mean ECss = 2 to 3 dS m-1 to about 2000 cm for mean ECss = 9 to 11 dS m-1.
|
|
|
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Received for publication July 30, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSNOTESRESULTSSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. M. Dudley, A. Ben-Gal, and U. Shani Influence of Plant, Soil, and Water on the Leaching Fraction Vadose Zone J., April 14, 2008; 7(2): 420 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Fujimaki, Y. Ando, Y. Cui, and M. Inoue Parameter Estimation of a Root Water Uptake Model under Salinity Stress Vadose Zone J., January 23, 2008; 7(1): 31 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. d. J. van Lier, K. Metselaar, and J. C. van Dam Root Water Extraction and Limiting Soil Hydraulic Conditions Estimated by Numerical Simulation Vadose Zone J., November 20, 2006; 5(4): 1264 - 1277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Shani, A. Ben-Gal, and L. M. Dudley Environmental Implications of Adopting a Dominant Factor Approach to Salinity Management J. Environ. Qual., August 9, 2005; 34(5): 1455 - 1460. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. T. Corey and S. D. Logsdon Limitations of the Chemical Potential Soil Sci. Soc. Am. J., June 2, 2005; 69(4): 976 - 982. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Journal of Environmental Quality |
||||
