HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (14) Citing Articles via Google Scholar Google Scholar Articles by Das, B. S. Articles by Inskeep, W. P. Search for Related Content PubMed Articles by Das, B. S. Articles by Inskeep, W. P. GeoRef GeoRef Citation Agricola Articles by Das, B. S. Articles by Inskeep, W. P.

DIVISION S-1-SOIL PHYSICS

Nitrate Concentrations in the Root Zone Estimated Using Time Domain Reflectometry

^a Land Resources and Environmental Sciences Dep., Montana State Univ., P.O. Box 173120, Bozeman, MT 59717-3120 USA

jwraith{at}montana.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
Improved ability to monitor soil water and ionic chemical distributions in field soils will contribute to better land management opportunities. We evaluated the potential to use time domain reflectometry (TDR) to simultaneously estimate volumetric soil water content (), soil solution electrical conductivity (_w), and soil nitrate concentrations in an irrigated peppermint (Mintha piperita L.) field using simple models and calibration methods. Two models and three calibration approaches for estimating _w from TDR were compared to _w obtained from soil cores and soil solution samplers at three depths, and over a wide range in applied KNO₃. The applied N rates were well within those commonly used in peppermint production. Soil nitrate concentrations were estimated based on TDR _w predictions and a simple linear calibration of _w vs. nitrate concentration and then compared to independent soil core measurements. Estimates of _w and nitrate concentrations using TDR exhibited similar pattern, magnitude, and variance to those based on direct soil measurements. The Mualem and Friedman (1991) model (MF), which is based on soil water retention measurements, provided _w and nitrate estimates in better agreement with soil cores than did the Rhoades et al. (1976) model. Our results suggest that simple physical–conceptual models combined with laboratory or field parameter estimation methods may be suitable for automated real-time field monitoring.

Abbreviations: DOY, day of year • MF, Mualem and Friedman (1991) model • RRP, Rhoades, Raats, and Prather (1976) model • SEM, standard error of mean • TDR, time domain reflectometry

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
KNOWLEDGE OF SOIL WATER AND SOLUBLE CHEMICAL DISTRIBUTIONS influences management strategies in agriculture, environmental impact assessment, and other land resources applications. Measuring or monitoring unsaturated field soils generally requires large sample number and remains time and labor intensive. Improved measurement techniques that are capable of monitoring field soil water contents and ionic solute distributions in the same sampling volume and in real time will increase opportunities for effective land management and improve our understanding of flow and transport processes in field environments.

Conventional methods to sample solute distributions in space and time include soil coring, vacuum extraction of soil solution, capillary wick samplers, and techniques based on measurement of soil electrical conductivity (e.g., four-electrode sensors; TDR). While considered to be a potentially accurate means to determine solute concentrations in soils, the destructive and labor intensive nature of soil coring severely limits its practical application. A number of potential disadvantages to vacuum extraction methodologies have been identified, including an unknown volume of soil sampled, difficulty or inability in acquiring samples as soils desaturate, the potential for non-representative sampling of pore sizes, and a potential lack of repeatability (van der Ploeg and Beese, 1977; Lord and Shepherd, 1993). Capillary wick samplers have demonstrated merit in a research context, but the destructive and labor-intensive nature of emplacement limits their widespread adoption.

During the last two decades, TDR has become an established and reliable means to determine volumetric soil water content (Topp et al., 1980) and bulk soil electrical conductivity (Dalton et al., 1984). During the past several years, TDR has been used to monitor solute mass in soil columns (Vanclooster et al., 1993; Ward et al., 1994; Vanclooster et al., 1995; Vogeler et al., 1997) and under steady water flow conditions in the field (Kachanoski et al., 1992; Noborio et al., 1994); to estimate solute transport parameters (Vanclooster et al., 1993; Risler et al., 1996) and to characterize anion exchange in laboratory leaching experiments (Vogeler et al., 1996). Recent work under controlled conditions demonstrates the potential of TDR to monitor the movement of ionic tracers through soils. In particular, Risler et al. (1996), Persson (1997), and Nissen et al. (1998) conducted transient unsaturated flow experiments in order to characterize ionic solute transport in soil columns. These studies, which made use of the ability to simultaneously measure volumetric water content and bulk electrical conductivity (_a) in the same sampling volume, in combination with the relative ease with which sensors may be concurrently measured at multiple soil locations, suggest that TDR might be used to monitor field-scale ionic solute dynamics. Although initial acquisition costs are high, Ward et al. (1994) pointed out that TDR may be less expensive than conventional techniques, considering the total cost of sampling and analyses.

Soil- and solute-specific calibrations are needed to estimate solute concentrations based on measured _a. Two main calibration approaches have been investigated with the use of laboratory experimental data (Mallants et al., 1994). In one approach, an empirical linear relationship is assumed between solute concentration and measured _a (e.g., Kachanoski et al., 1992; Mallants et al., 1996). Persson (1997) observed that this method requires elaborate calibration experiments and is not well suited for transient field conditions, as the coefficients of the linear equation depend on and must be determined for varying water contents. A second approach is based on the empirical Archie's Law, wherein _a is related to _w and tortuosity of the electrical flow path. Two steps are involved in this process. A soil-specific calibration is obtained to estimate _w from _a, following which a solute-specific calibration is obtained to estimate solute concentration from estimated _w. The latter procedures are relatively simple and can be easily evaluated both in the field and laboratory. Although several laboratory calibration experiments have shown the utility of this approach (Noborio et al., 1994; Heimovaara et al., 1995; Vanclooster et al., 1994; Risler et al., 1996; Persson, 1997; Ferré et al., 1998), it has not been evaluated under natural field conditions.

The objective of this study was to test the feasibility of employing TDR to simultaneously estimate , _w, and nitrate concentrations in a production agricultural field using simple models and calibration methods. We considered only relatively simple approaches because we deem more rigorous models to be less feasible or appropriate for some large scale and/or management applications. We compared two models and three approaches to the calibration of model parameters for their agreement with _w that was obtained from soil cores and soil solution samplers for a wide range in applied KNO₃. We also estimated soil nitrate concentrations based on TDR _w predictions and compared these with independent soil core measurements.

	Materials and methods

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
Theoretical Background and Calibration Models
Electrical conductivity of bulk soil is an integrated measure of contributions from ions on solid surfaces (surface conductance, _s), _w, and tortuosity of electrical flow paths. A number of linear and curvilinear relationships between _w and _a have been proposed in geophysics and soils applications. Each of these models is a variant of classical Archie's Law, , and can be described by a general equation:

(1)

where m is an empirical constant, is effective saturation, and X, F, and Y are functions of pore geometry, surface conductance, and cation-exchange capacity (de Lima and Sharma, 1990). A simple form of Eq. [1] is the two-conductors model (Rhoades et al., 1976), designated as RRP (following authorship of Rhoades, Raats, and Prather):

(2)

where liquid and solid phases in the bulk soil are treated as two macroscopic and parallel conductors, and T is interpreted as a soil-specific transmission coefficient to account for changes in tortuosity of the electrical current flow path with changes in soil wetness. The transmission coefficient is conventionally characterized as a linear function of ; i.e., , with a and b constants for a given soil (Rhoades et al., 1976). This simple model is highly attractive for many practical applications and is especially amenable to use with TDR, where both and _a may be measured in the same soil volume. Commonly applied linear empirical calibrations (e.g., Kachanoski et al., 1992; Mallants et al., 1996) represent particular cases of this model. For example, combining Eq. [2] with a linear relationship between solute concentration and _w (Bohn et al., 1982) and assuming constant and T yields a direct linear relationship between solute concentration and _a.

Equation [2] fails to describe the curvilinear behavior of _a - _w at low salinity levels (below _w 1 dS m^-1). For low salinity levels, Shainberg et al. (1980), Nadler and Frenkel (1980), and Rhoades et al. (1989) proposed variants of Eq. [2]. While Nadler and Frenkel (1980) proposed that the surface conductance is a function of _w, Shainberg et al. (1980) and Rhoades et al. (1989) incorporated a third conductor that consisted of solid and immobile water in series configuration. The resulting models are inherently more complex and some model parameters are difficult to measure or estimate. Hence, we will not consider these models in our present study.

An alternative to models that are based on Eq. [1] is the application of capillary bundle theory, which is based on statistical pore size distribution, along with the assumption that the flow lines for water and electrical current are similar under a given hydraulic gradient (Mualem and Friedman, 1991). In this approach, the surface conductance of the soil solids is assumed to be negligible. Similarity assumption on hydraulic and electrical flow paths suggests that the ratio between the hydraulic conductivity in tortuous flow paths and in straight capillaries can be used as the geometry factor:

(3)

where _r is residual soil water content and the geometry factor F_g is related to the soil water retention function h():

(4)

In Eq. [4] h is matric head, is effective saturation, _s is saturated soil water content, ß is a parameter related to the combined effects of the pore connectivity and the tortuosity factors in the bundle of capillaries model, and x is a dummy variable in the integral. Using the van Genuchten (1980) soil water retention function in Eq. [3], Heimovaara et al. (1995) obtained

(5)

where and . Based on examination of 45 different soils, Mualem (1976) proposed that a value of 0.5 can be used for ß, while Heimovaara et al. (1995) optimized ß for each of their soil conditions.

Estimating Ionic Concentrations from _w
The relationship of _w to specific ionic constituents has been formulated (Bohn et al., 1982) for pure solutions as

(6)

where k is the cell constant that accounts for electrode geometry, is the molar limiting ion conductivity, M is the molar concentration, v is the ion charge, and i denotes ion species. Equation [6] suggests a linear relationship between electrical conductivity and molar concentrations of ions in pure solutions. Marion and Babcock (1976) showed a linear relationship between _w and solute concentrations in the soil solution. More recently, Heimovaara et al. (1995) showed that _w was linearly related to soluble Al and H ion concentrations in soil solution. Furthermore, empirical linear calibration has been widely employed to relate _a to the solute concentration (Kachanoski et al., 1992; Vanclooster et al., 1993; Ward et al., 1994; Mallants et al., 1994). These findings support the potential to estimate ionic solute concentrations through _w estimates using simple models in combination with TDR. In this study, we evaluate the potential to estimate nitrate concentrations from _w measured or estimated using soil cores and TDR under unsaturated field conditions.

Field Experiment and Sampling Protocols
A field experiment was conducted during 15 April to 6 August 1996 at the Northwest Agricultural Research Center near Creston, Montana. Soils at this site were until recently classified as Flathead fine sandy loams (coarse-loamy, mixed Pachic Udic Haploborolls). The soil physical and chemical properties listed in Table 1 document little textural difference in the upper 0.90 m. Table 1 shows that the shallower depths (0.15 and 0.45 m) are relatively high in organic matter content, soluble cations, and cation-exchange capacity compared with deeper depths (0.9 m). Saturated hydraulic conductivity ranged from 0.43 m d^-1 for surface soil to 0.65 m d^-1 for subsurface soil, as determined with tension infiltrometers. The field area was under peppermint cultivation. Peppermint is a high value specialty crop and typically receives high levels of N fertilization and irrigation.

View larger version (17K): [in this window] [in a new window]   Fig. 1 Schematic of field experiment layout including approximate locations of time domain reflectometry (TDR) arrays (X), solution samplers (), and soil cores () in the four plots having differential KNO3 application. Soil cores were collected from five randomly selected locations in each plot at each sampling date

(7)

where Z₀ is probe impedance (), L is probe length (m), Z_L is resistive impedance load across the buried TDR probe (), and f_T is a correction factor (U.S. Salinity Laboratory Staff, 1954);

(8)

for measured soil temperature (°C) adjacent to TDR probes. The resistive impedance load across the TDR probe was calculated by subtracting the combined series impedance (Z_cable) of the multiplexer, connectors, and connecting cables from the total impedance (Z_total) measured with the cable tester (Heimovaara et al., 1995). We estimated Z_cable using standard KCl solutions and multiple combinations of connecting cable and hardware in the laboratory, following a procedure similar to that of Heimovaara et al. (1995). Readers are referred to a recent paper by Reece (1998) for an alternative approach to estimate Z_cable.

Small porous ceramic soil solution samplers were buried at depths of 0.15, 0.45, and 0.9 m at one location within each plot (Fig. 1). We attempted to withdraw 10 mL of soil solution at 1- or 2-wk intervals during the season. Soil core samples (2.5-cm diam.) were collected from depth increments of 0.10 to 0.20, 0.35 to 0.55, and 0.80 to 1.00 m in five randomly selected locations within each plot on the same time intervals as for soil solution samples. Gravimetric water content was measured in subsamples drawn from each core. To measure _w and nitrate concentrations in the soil cores, 40 g of soil was mixed with distilled and deionized water to obtain 1:1 (kg kg^-1) saturation pastes, which were then shaken for 45 min and centrifuged at 10000 RPM for 25 min. Solution electrical conductivity was measured in the filtered extracts using a conductivity meter (Accumet model 50, Denver Instrument, Arvada, CO). All samples were maintained at 25°C in a circulating water bath prior to measurement. Soil solution electrical conductivities of field core samples were estimated by multiplying the dilution factor with electrical conductivity of extracted solution, , where _x and _f are water contents of the soil extract and the field soils at sample wetness, respectively (Nadler, 1997). A nitrate autoanalyzer (AAII, Alpkem Corp., Clackamas, OR) was used to measure nitrate concentrations in the extract solutions. Measured nitrate concentrations in the extracts were then multiplied by the dilution factor to obtain nitrate concentrations at sample wetness.

Sprinkler irrigation was applied approximately weekly during the latter half of the growing season. Each irrigation followed collection of soil cores and soil solutions. The entire field area was fertilized with 18.5 kg N ha^-1 and 38.1 kg P ha^-1 as 11-52-0 (N–P₂O₅–K₂O equivalent) plus 112 kg K ha^-1 as 0-0-60 the previous October, and with 119 kg N ha^-1 plus 27 kg S ha^-1 as a mixture of 34-0-0 (NH₄NO₃) with 21-0-0-24 (NH₄SO₄) on day of year (DOY) 99, about 6 d before installation of field sensors. We hand-applied KNO₃ salt at 300, 150, 75, and 0 kg N ha^-1 to Plots I, II, III, and IV, respectively on DOY 170, followed by sprinkler irrigation. These fertilizer treatments were well within the ranges used in peppermint production, as multiple applications of N fertilizer are typically applied during the growing season.

Laboratory Calibration Procedures
Soil water retention [(h)] for surface and subsurface soils was measured in the range 0 to -80 kPa using a pressure plate apparatus (Klute, 1986). Parameters in van Genuchten's (1980) retention model were then fitted to measured data to obtain , m, n, p, and q in Eq. [4] by two-step optimization of the retention data. In the first step, van Genuchten's (1980) retention function was used to optimize and n with the constraint . In the second step, the same retention function and measured data were used to optimize and p with the constraint . Note that our approach is slightly different from that of Heimovaara et al. (1995) in that we optimized separately in both the steps, while Heimovaara et al. (1995) fixed during the second optimization.

In order to estimate the parameters required for the RRP model (Eq. [2]), a laboratory calibration experiment was conducted using repacked soil columns under unsaturated, transient water flow, following the method outlined by Risler et al. (1996). Briefly, surface and subsurface soils collected from our experimental plots were packed in 0.15-m-diam. by 0.15-m-long PVC columns to uniform field bulk density. A single horizontal TDR probe was inserted near the bottom of each column. The soil columns were preconditioned by flushing with dilute KCl solution until constant effluent _w was attained. A quasi-sinusoidal application rate of KCl solution was obtained by controlling a power relay to a syringe pump with a datalogger. This consistent but time-variable application rate established unsaturated and transient wetness conditions within the columns. After constant _w was obtained in the column effluent, soil and _a were continuously monitored over several wetting and drying cycles. Equation [2] was fitted to measured and _a to estimate RRP parameters a, b, and _s using nonlinear least-squares optimization (Wraith and Or, 1998).

A dilution series procedure was conducted to estimate field nitrate concentrations based on measured _w. Replicate field soil samples were oven dried, brought to varying soil:water ratios (0.2–2.0 kg kg^-1), then shaken for 45 min. The samples were centrifuged and filtered as described above, then analyzed for nitrate concentration and _w. Linear regression was used to relate nitrate concentration and _w.

Field Calibration Procedures
Soils in the field are often structured and more naturally heterogeneous than uniformly repacked soil columns, and the transient behavior in the field may be more pronounced than observed during systematic drying and wetting of soil columns. The issue of variable transmission coefficients a and b for repacked vs. intact field soils represents a potentially important consideration for application of laboratory results to the field situation. Consequently, in situ calibration of RRP model parameters may be preferred to obtain reliable _w estimates for field soils. A laboratory calibration approach similar to that of Risler et al. (1996) may be prohibitive to field application because large volumes of input solution may be required and equilibration time to input _w may be long, except near the soil surface. Hence, we examined a simple in situ mass balance approach to estimate RRP model parameters. This approach is similar to some previously reported TDR calibration methods (Kachanoski et al., 1992; Mallants et al., 1994; Vogeler et al., 1996; Persson, 1997), except that we parameterized a physical–conceptual, rather than empirical, model.

Assuming conservation of mass, the total mass density of field solute breakthrough curves may be equated to the total mass of KNO₃ salt applied. Field-measured _a and for vertical TDR probes were converted to _w using the RRP model, then transformed into KNO₃ concentrations using the linear calibration relationship obtained from field soil samples. The total mass recovery (m_r) was calculated from TDR-measured KNO₃ concentrations (C_i) using the relationship

(9)

where A is the area of the plot, L is soil depth (0.29 m in our case), V_w is the total volume of water applied to the plot during the calibration period, and n is the total number of observations. We calculated V_w by water balance using measured , depth of the vertical TDR probes, and A. The m_r was then equated with the applied solute mass in an iterative optimization procedure (Wraith and Or, 1998) to obtain in situ RRP parameters. Because of the relatively homogeneous nature of the field site, and to minimize the number of fitted coefficients, we held _s constant at 0.015 dS m^-1 and optimized RRP parameters a and b. This also maintained consistency with the other RRP calibration approaches, in which _s was maintained at 0.015 dS m^-1 for the 0.15- and 0.45-m depths. Model parameters were obtained for each vertical TDR probe within Plots I, II, and III. A mass balance procedure within Plot IV was not possible because no KNO₃ was added. Because _a for all vertical probes at the end of the experiment were similar to or lower (at similar ) than those measured before KNO₃ application, we assumed complete salt removal from the measurement volume of the vertical TDR probes.

	Results and discussion

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
Laboratory Model Calibrations
We conducted several cycles of sequential soil column drying and wetting for surface and subsurface soils, and for two different input _w levels. For each input _w, a different set of RRP model parameters was obtained (Table 2) . To test the model sensitivity to these parameters, we calculated _w for all field plots using measured and _a, and for each set of a, b, and _s. In agreement with earlier findings (Risler et al., 1996), we observed only very minor differences between predicted _w using parameters obtained from variable input concentrations. For 96% of all field data, the absolute difference between _w predicted using calibrated surface or subsoil model coefficients was less than 0.1 dS m^-1. Assuming the RRP model holds for the relationship between _a, _w, and , then an effective single-concentration calibration should be adequate as long as the soil wetness () varies sufficiently during the calibration process (and preferably to a similar degree to conditions under which field monitoring will be conducted) and assuming the surface conductance _s is not concentration-dependent (as is assumed in model development). Because substantial time may be required to equilibrate soils to an input solution _w, we recommend that a background input solution near the ambient _w of the soil solution be used for this calibration approach. In our subsequent analyses we therefore used parameter sets obtained from 0.75 dS m^-1 for surface soil and 0.5 dS m^-1 for subsurface soil.

View this table: [in this window] [in a new window]   Table 2 Best fit parameters for Rhoades et al. (1976)(RRP) and Mualem and Friedman (1991)(MF) models for soil solution electrical conductivity in Flathead soil. RRP parameters were obtained by sequential wetting and drying at known constant w (Risler et al., 1996). MF parameters were obtained by fitting van Genuchten's (1980) equation to measured soil water retention data

Field Water Regime
Gravimetric water contents of soil core samples were converted to volumetric water contents using soil bulk density measured in the field. Reasonable agreement was obtained between measured with TDR and soil coring for all plots (Plots II and IV shown in Fig. 2) . Substantial variation in measured with location within plots was apparent for both TDR and soil cores (e.g., Fig. 2). The mean water contents across plots were similar at a given depth, except that Plot IV was slightly wetter than Plots I, II, and III (data not shown). Figure 2 also illustrates the fluctuating at the 0.15-m depth that results from periodic rainfall and irrigation, with highly transient water regime in the surface soil compared to that at lower depths. A specific advantage of automated TDR over soil coring is its ability to capture this detailed temporal pattern.

View larger version (34K): [in this window] [in a new window]   Fig. 2 Volumetric soil water contents () measured using replicate soil cores (symbols) and TDR (lines), at 0.15- and 0.9-m depths in Plots II and IV. Abrupt increases in measured by TDR at 0.15-m depth are in response to rainfall and irrigation

TDR Estimates of Field _w Using Laboratory Calibrations

View larger version (33K): [in this window] [in a new window]   Fig. 3 Mean ± standard error of mean (SEM) soil solution electrical conductivities (dS m-1) measured using replicate soil cores , horizontal or vertical TDR probes , and solution samplers at 0.15-m depth in Plots II and IV. Horizontal probes were at 0.15-m depth, while vertical probes spanned 0–0.29 m depth

View larger version (38K): [in this window] [in a new window]   Fig. 4 Mean ± standard error of mean (SEM) soil solution electrical conductivities (dS m-1) measured using replicate soil cores , solution samplers , and TDR probes at 0.45- and 0.9-m depths in Plots I and III

Field Calibration of RRP Model
Estimates of _w predicted using the KNO₃ mass balance calibration (Table 3) were in reasonable agreement with independently measured _w of the field soil cores (Fig. 5) . Our assumption of closure of the solute breakthrough curve in the soil depth spanned by the vertical probes (based on similar _a at similar before and after) was supported by predicted _w at the beginnings and ends of breakthrough curves that have equivalent magnitudes (Fig. 5). Although we did not conduct the in situ mass balance for 0.45- or 0.9-m depths, we applied the RRP model parameters obtained from 0 to 0.29 m to and _a measured at 0.9 m. Estimated _w at 0.9-m depth based on mass balance for individual probes (0–0.29 m depth) were intermediate between the MF and RRP (laboratory wetting–drying calibration) estimates (Fig. 4), in reasonable agreement with solution samplers, and generally somewhat lower than values from soil cores. We also attempted to estimate RRP model parameters for individual field plots and for the entire study area, but the resulting mass balance fits were poor (r² < 0.5). This indicated that a single set of RRP model parameters did not adequately predict measured _w for the entire study area.

View this table: [in this window] [in a new window]   Table 3 Rhoades et al. (1976)(RRP) model parameters estimated using in situ mass balance approach equating applied KNO3 and estimated KNO3 based on field and a measurements from vertical TDR probes. Model parameter s was held constant at 0.015 dS m-1 in the optimization procedure

View larger version (32K): [in this window] [in a new window]   Fig. 5 Mean ± standard error of mean (SEM) soil solution electrical conductivities (dS m-1) measured using replicate soil cores (circles, ), horizontal TDR probes (lines, ), and solution samplers (squares, ) at 0.15-m depth in Plots I, II, and III. Time domain reflectometry (TDR) estimates are based on in situ mass balance RRP model optimizations vs. applied KNO3 (Eq. [9]) using vertical surface probes, and applied to horizontal probes at 0.15 m depth

Monitoring Field Nitrate Concentrations
Linear relationships between measured _w and nitrate concentrations were observed for each depth and for pooled soil core data (Fig. 6) . We also superimposed the relationship obtained from the dilution series using surface soil samples. The coefficients derived from the dilution series experiment were very similar to those observed with pooled soil core data. This indicates that simple dilution series experiments can be performed to calibrate estimates of nitrate concentrations from measured or estimated _w. For comparison purposes, we used the coefficients from the depth-specific soil core _w and nitrate concentrations to estimate nitrate concentrations from field (TDR) measured and _a. Some of these results are summarized in Fig. 7 , in which we apply the MF model for 0.15-m depth in Plots I–IV. We found similar agreement for estimated (TDR) and independently measured (cores) nitrate concentrations as we did for _w, at 0.15-, 0.45-, and 0.9-m depths. Because of differences in specific ionic conductance among species that may undergo exchange in the soil solution, the contributions of nitrate to total soil solution ionic concentration (Eq. [6]) will vary with time and space in field soils. Hence, any calibrated relationship between _w and nitrate concentration may provide only an approximate estimate. For some field conditions, e.g., near the bottom of plant root zones, changes in _w will mainly result from changes in highly soluble species, such as nitrate. Monitoring and _w just below the root zone may alert producers or land managers to potential nitrate leaching.

View larger version (24K): [in this window] [in a new window]   Fig. 6 Linear calibration results for soil nitrate concentrations (ppm) vs. soil solution electrical conductivities (dS m-1) in soils collected at the Creston field site. Relationships were determined for cores from each measurement depth, for all depths pooled, and using a dilution series experiment with surface soil. Numbered calibration equations correspond to parenthetical numbers in the legend

View larger version (30K): [in this window] [in a new window]   Fig. 7 Mean ± standard error of mean (SEM) soil nitrate concentrations (ppm) measured using replicate soil cores and estimated using horizontal TDR probes at 0.15 m depth in all field plots. Amount of N applied as KNO3 is indicated for each plot

Some sensitivity analyses of MF and RRP models to calibrated parameters have also been reported previously (Vanclooster et al., 1994; Risler et al., 1996). In addition to confirming the conclusions of these authors, we found that the MF model provided highly variable estimates when we used different soil water retention models that are based on the same measured retention data. Mualem and Friedman (1991) incorporated the Brooks and Corey (1964) retention model into their original formulation. For our MF model application to the field data set we used the van Genuchten (1980) retention model, slightly modified from the approach reported by Heimovaara et al. (1995). However, we also explored use of alternative retention models including those of Tani (1982; cited in Kosugi, 1994) and Kosugi (1994). When applied to our field –_a data, each of these resulted in widely varying estimates of _w. This and other aspects of the MF modeling approach will be more fully explored in a future paper.

Plot-averaged concentration changes during the season were smaller than the perceived uncertainty (SEM) of measurements for both TDR and soil cores in some cases. Yet, variation in the TDR estimates reflect differences among probe locations rather than measurement scatter for individual probes (Fig. 2). Hence, as in traditional soil core sampling, measuring at a number of locations may be required to achieve more precise mean values. For many applications, however, specific knowledge of spatial differences will be more important than a representative field mean. TDR may be well suited to this sort of measurement, as long as probe locations may be located in reasonable proximity to the cable tester. Many managed fields encompass large areas, and the cable tester may need to be moved from place to place to record periodic measurements, as opposed to the continuous monitoring scheme used in our study.

Because of the highly destructive nature of repeated core sampling, our experimental design did not provide direct comparisons of TDR and soil cores at the same measurement locations. Although vacuum solution samplers were installed 1 m from some TDR probes, lack of sufficient solution samples prevented meaningful comparisons of measured _w with those estimated using TDR. The field-averaged TDR estimates appear very reasonable given the similar temporal behavior and ranges in _w measured using soil cores and solution samplers (Fig. 3 and 4). The Creston field location exhibits little apparent soil morphological definition; hence, we attribute some of the measured spatial variation to nonuniform fertilizer salt application, and to the patchy nature of peppermint plant density with implication to water and nutrient uptake.

Time domain reflectometry in combination with physical–conceptual models can provide highly detailed information concerning the temporal dynamics of , _w, and NO₃ at multiple soil locations. While additional work is needed to evaluate and ensure accuracy of _w and NO₃ estimates under transient conditions, these time series will be very informative in their precision even without (necessarily) high accuracy in some cases. Finally, we emphasize that accurate measurements are critical to accurate _w predictions that use either the RRP (Eq. [2]) or MF (Eq. [3]) model. Any errors in measured will propagate through to calculated _w and NO₃ concentration.

	Conclusions

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
This field and laboratory study demonstrated that TDR may be used in conjunction with simple models and calibration methods to estimate ionic solute concentrations in agricultural fields. Combining real-time _w estimates with existing tools, such as climatically driven soil–plant simulation models, may enhance intensive management of agricultural production systems to achieve greater resource efficiency. Our study was not designed to critically evaluate site-specific calibration parameters for salinity models. However, results for the RRP model suggest that the efficacy of potential TDR calibration approaches may depend on the degree of field scale heterogeneity present. Additional work has evaluated calibration methods and models (e.g., Heimovaara et al., 1995; Risler et al., 1996; Mallants et al., 1996; Persson, 1997; Vogeler et al., 1997), and the results, though mixed, appear promising for application in soils monitoring and management.

Predicted _w and nitrate concentrations using TDR exhibited similar magnitude, pattern, and variation as measured using replicate soil cores, over a range in applied KNO₃. Highly detailed time series were provided by TDR in comparison with soil cores. Nevertheless, future experiments are needed to verify agreement between TDR and soil core measurements at the same field locations for specific soils and calibration approaches. This may be a considerable undertaking, because buried TDR probe locations are sacrificed upon removal of surrounding soil. Application of TDR to field-scale management may require lower absolute accuracy of _w estimates than for finer scale resolution, but field sampling schemes will need to consider the distributions of and _w in space and time. Our results suggest that rather simple physical–conceptual models, combined with laboratory or field calibration methods, may be suitable for some field scale monitoring enterprises. Characterization of spatial structure of relevant field soil properties should contribute to optimized sampling locations, but this may not be a reasonable expectation for many applications.

	ACKNOWLEDGMENTS

 
This work was supported by Montana Agricultural Experiment Station Projects 103302 and 104327, the MT Fertilizer Tax Committee, and NRI Competitive Grants Program/USDA (Agreement 95-37102-2175). We thank Dr. Hesham Gaber and Chris Wright for their assistance during the field trial, and Leon Welty for providing project support at the NWARC field location.

	NOTES

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 
¹ Mention of company names or commercial products is for the convenience of the reader and does not imply endorsement.

Received for publication July 17, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION Materials and methods NOTES Results and discussion Conclusions REFERENCES

 

• Bohn H.L., Ben-Asher J., Tabbara H.S., Marwan M. Theories and tests of electrical conductivity in soils. Soil Sci. Soc. Am. J. 1982;46:1143-1146.[Abstract/Free Full Text]
• Brooks, R.H., and A.T. Corey. 1964. Hydraulic properties of porous media. Hydrol. Pap. 3. Civil Engin. Dep., Colo. State Univ., Fort Collins.
• Dalton F.N., Herkelrath W.N., Rawlins D.S., Rhoades J.D. Time-domain reflectometry: Simultaneous measurement of soil water content and electrical conductivity with a single probe. Science 1984;224:989-990.[Abstract/Free Full Text]
• de Lima O.A.L., Sharma M.M. A grain conductivity approach to shaly sandstones. Geophysics 1990;55:1347-1356.
• Ferré P.A., Redman J.D., Rudolph D.L., Kachanoski R.G. The dependence of the electrical conductivity measured by time domain reflectometry on the water content of a sand. Water Resour. Res. 1998;34(5):1207-1213.
• Giese K., Tiemann R. Determination of the complex permittivity from thin-sample time domain reflectometry: Improved analysis of the step response waveform. Adv. Mol. Relax. Processes. 1975;7:45-59.
• Heimovaara T.J., Focke A.G., Bouten W., Verstraten J.M. Assessing temporal variations in soil water composition with time domain reflectometry. Soil Sci. Soc. Am. J. 1995;59:689-698.[Abstract/Free Full Text]
• Kachanoski G., Pringle E., Ward A. Field measurement of solute travel times using time domain reflectometry. Soil Sci. Soc. Am. J. 1992;56:47-52.[Abstract/Free Full Text]
• Klute, A. 1986. Water retention: laboratory methods. p. 635–662. In: A. Klute (ed.), Methods of soil analysis: Part 1. Physical and mineralogical methods. 2nd ed. ASA, Madison, WI.
• Kosugi K. Three-parameter lognormal distribution model for soil water retention. Water Resour. Res. 1994;30:891-901.
• Ledieu J., de Ridder P., de Clerck P., Dautrebande S. A method of measuring soil moisture by time-domain reflectometry. J. Hydrol. 1986;88:319-328.
• Lord E.I., Shepherd M.A. Developments in the use of porous ceramic cups for measuring nitrate leaching. J. Soil Sci. 1993;44:435-449.
• Mallants D., Vanclooster M., Meddahi M., Feyen J. Estimating solute transport in undisturbed soil columns using time-domain reflectometry. J. Contam. Hydrol. 1994;17:91-109.
• Mallants D., Vanclooster M., Toride N., Vanderborght J., van Genuchten M.Th., Feyen J. Comparison of three methods to calibrate TDR for monitoring solute movement in undisturbed soil. Soil Sci. Soc. Am. J. 1996;60:747-754.[Abstract/Free Full Text]
• Marion G.M., Babcock K.L. Predicting specific conductance and salt concentration in dilute aqueous solutions. Soil Sci. 1976;122:181-187.
• Miller E.E., Miller R.D. Physical theory for capillary flow phenomena. J. Appl. Phys. 1956;27:324-332.[Web of Science]
• Mualem Y. A new model for predicting the hydraulic conductivity of unsaturated porous media. Water Resour. Res. 1976;12(3):513-522.
• Mualem Y., Friedman S.P. Theoretical predictions of electrical conductivity in saturated and unsaturated soil. Water Resour. Res. 1991;27:2771-2777.
• Nadler A. Discrepancies between soil solute concentration estimates obtained by TDR and aqueous extracts. Aust. J. Soil Sci. 1997;35:527-537.
• Nadler A., Frenkel H. Determination of soil solution electrical conductivity from bulk soil electrical conductivity measurements by the four-electrode method. Soil Sci. Soc. Am. J. 1980;44:1216-1221.[Abstract/Free Full Text]
• Nissen H.H., Moldrup P., Henriksen K. Time domain reflectometry measurements of nitrate transport in manure-amended soil. Soil Sci. Soc. Am. J. 1998;62:99-109.[Abstract/Free Full Text]
• Noborio K., McInnes K.J., Heilman J.L. Field measurements of soil electrical conductivity and water content by time-domain reflectometry. Comput. Electron. Agric. 1994;11:131-142.
• Persson M. Soil solution electrical conductivity measurements under transient conditions using time domain reflectometry. Soil Sci. Soc. Am. J. 1997;61:997-1003.[Abstract/Free Full Text]
• Reece C.F. Simple method for determining cable length resistance in time domain reflectometry systems. Soil Sci. Soc. Am. J. 1998;62:314-317.[Abstract/Free Full Text]
• Rhoades J.D., Manteghi N.A., Shouse P.J., Alves W.J. Soil electrical conductivity and soil salinity: New formulations and calibrations. Soil Sci. Soc. Am. J. 1989;53:433-439.[Abstract/Free Full Text]
• Rhoades J.D., Raats P.A.C., Prather R.J. Effects of liquid-phase electrical conductivity, water content, and surface conductivity on bulk soil electrical conductivity. Soil Sci. Soc. Am. J. 1976;40:651-655.[Abstract/Free Full Text]
• Risler P.D., Wraith J.M., Gaber H.M. Solute transport under transient flow conditions estimated using time domain reflectometry. Soil Sci. Soc. Am. J. 1996;60:1297-1305.[Abstract/Free Full Text]
• Shainberg I., Rhoades J.D., Prather R.J. Effect of exchangeable sodium percentage, cation-exchange capacity, and soil solution concentration on soil electrical conductivity. Soil Sci. Soc. Am. J. 1980;44:469-473.[Abstract/Free Full Text]
• Topp G.C., Davis J.L., Annan A.P. Electromagnetic determination of soil water content: measurements in coaxial transmission lines. Water Resour. Res. 1980;16:574-582.
• U.S. Salinity Laboratory Staff. Diagnosis and improvement of saline and alkali soils. Washington, DC: USDA Agric. Handb. 60. U.S. Gov. Print. Office, 1954.
• Vanclooster, M., C. Gonzales, J. Vanderborght, D. Mallants, and J. Diels. 1994. An indirect calibration procedure for using TDR in solute transport studies. p. 215–226. In Proc. Symposium on Time Domain Reflectometry in Environmental, Infrastructure, and Mining Applications, Evanston, IL, Sept. 7–9, 1994. Pub. SP19-94, US Dep. Interior, Bureau of Mines, Washington, DC.
• Vanclooster M., Mallants D., Diels J., Feyen J. Determining local scale solute transport parameters using time domain reflectometry (TDR). J. Hydrol. 1993;148:93-107.
• Vanclooster M., Mallants D., Vanderborght J., Diels J., van Orshoven J., Feyen J. Monitoring solute transport in a multilayered sandy lysimeter using time domain reflectometry. Soil Sci. Soc. Am. J. 1995;59:337-344.[Abstract/Free Full Text]
• van der Ploeg R.R., Beese F. Model calculations for the extraction of soil water by ceramic cups and plates. Soil Sci. Soc. Am. J. 1977;41:466-470.[Abstract/Free Full Text]
• van Genuchten M.Th. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 1980;44:892-899.[Abstract/Free Full Text]
• Vogeler I., Clothier B.E., Green S.R., Scotter D.R., Tillman R.W. Characterizing water and solute movement by time domain reflectometry and disk permeametry. Soil Sci. Soc. Am. J. 1996;60:5-12.[Abstract/Free Full Text]
• Vogeler I., Clothier B.E., Green S.R. TDR estimation of the resident concentration of electrolyte in the soil solution. Aust. J. Soil Res. 1997;35:515-526.
• Ward A.L., Kachanoski R.G., Elrick D.E. Laboratory measurements of solute transport using time domain reflectometry. Soil Sci. Soc. Am. J. 1994;58:1031-1039.[Abstract/Free Full Text]
• Warrick A.W., Nielsen D.R. Spatial variability of soil physical properties in the field. In: Hillel D.I., ed. Applications of soil physics. New York: Academic Press, 1980:319-344.
• Wraith J.M., Or D. Nonlinear parameter estimation using spreadsheet software. J. Nat. Resour. Life Sci. Educ. 1998;27:13-19.

This article has been cited by other articles:

				D. L. Corwin and S. M. Lesch Application of Soil Electrical Conductivity to Precision Agriculture: Theory, Principles, and Guidelines Agron. J., May 1, 2003; 95(3): 455 - 471. [Abstract] [Full Text] [PDF]

				D. S. Long, J. M. Wraith, and G. Kegel A heavy-duty time domain reflectometry soil moisture probe for use in intensive field sampling Soil Sci. Soc. Am. J., March 1, 2002; 66(2): 396 - 401. [Abstract] [Full Text] [PDF]

				J. M. Wraith and D. Or Soil Water Characteristic Determination from Concurrent Water Content Measurements in Reference Porous Media Soil Sci. Soc. Am. J., November 1, 2001; 65(6): 1659 - 1666. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (14) Citing Articles via Google Scholar Google Scholar Articles by Das, B. S. Articles by Inskeep, W. P. Search for Related Content PubMed Articles by Das, B. S. Articles by Inskeep, W. P. GeoRef GeoRef Citation Agricola Articles by Das, B. S. Articles by Inskeep, W. P.