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) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Gasser, M. O. Articles by Lagacé, R. Search for Related Content PubMed PubMed Citation Articles by Gasser, M. O. Articles by Lagacé, R. GeoRef GeoRef Citation Agricola Articles by Gasser, M. O. Articles by Lagacé, R. Related Collections Flow Soil Physics Vadose Zone Processes and Chemical Transport Ground Water Quality Soil Models

TECHNICAL REPORTS
Vadose Zone Processes and Chemical Transport

Solute Transport Modeling under Cultivated Sandy Soils and Transient Water Regime

Département des sols et de génie agro-alimentaire, FSAA, Université Laval, Sainte-Foy, QC, Canada G1K 7P4

* Corresponding author (mogasser{at}grr.ulaval.ca)

Received for publication August 22, 2001.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Drainable lysimeters offer the possibility to integrate heterogeneous solute leaching conditions caused by row crops and transient water regime, and to conveniently measure water and solute fluxes at the drainage outlet. To compare solute leaching behavior in and around drainable lysimeters operating under a transient water regime in potato (Solanum tuberosum L.) fields, parameters of the convective lognormal transfer (CLT) function model were fitted using bromide (Br^-) flux concentrations (C^f) measured in lysimeters and from Br^- resident concentrations (C^r) measured in adjacent soil cores. Expected mean values E_z(I) obtained from C^r and C^f CLT parameters were equivalent and well correlated (R² = 0.78). However, estimated median values µ of the CLT function were smaller when derived from C^r (1.05 to 1.28) compared with C^f (1.23 to 2.14). Most µ values were also smaller than previously reported values for a 30-cm reference depth, indicating that 50% of solute mass would leach more readily in these coarse sandy soils. Higher variance and dispersion of C^r compared with those of C^f could be related to a smaller sampling support (sample size/sampling area) in the case of C^r measured by soil coring, or to disruption of solute transport mechanisms in the repacked lysimeter. Retained Br^- in the top soil layer after 12 to 17 cm of cumulative drainage was indicated by measured C^r. Neither CLT function simulated well residual topsoil C^r values, indicating that Br^- plant cycling or preferential flow probably interfered even though tuber Br^- uptake was relatively small.

Abbreviations: CLT, convective lognormal transfer

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
SOLUTE TRANSPORT MODELING under intensively cultivated field crops is needed to evaluate the risk of contaminating ground water with nitrates, pesticides, and other crop-related inputs, under a wide variety of management and climatic conditions. Solute transport is mainly affected by water flow through the soil profile. However, water flow under cultivated field conditions is highly variable in time and space due to the presence of plants, heterogeneous soil conditions, and soil tillage. Therefore, conventional modeling of water and solute fluxes in a one-dimensional homogenous soil profile with the Richards and convection–dispersion equations may not apply.

Plants play an important role in evapotranspiration, which under row crops creates important tracer leaching patterns (Kung, 1990; Timlin et al., 1992; Powers et al., 1997). Root water uptake and lower soil water contents in the crop-row zone can induce smaller downward soil water and solute fluxes below this zone (Timlin et al., 1992). Under potato crops, ridging or hilling will shed rain water away from the plant-row zone, causing greater infiltration of water and leaching in the interrow zone (Saffigna et al., 1976). As a result, solute leaching is highly variable under row-cropped soils and modeling such a heterogeneous system with the Richards and convection–dispersion equations requires a detailed knowledge of spatial and temporal changes of soil hydraulic properties, rainfall intensities, and root growth with time (Timlin et al., 1992).

Drainable lysimeters offer the opportunity to measure mean concentrations of solutes leaching through a reference surface fitted to include row and interrow zones. While differential leaching patterns created by row and interrow zones are not monitored directly in drainable lysimeters, mean solute concentrations reflect an overall effect of crop production on solute leaching dynamics at the row scale. Lysimeters offer a direct approach to measure solute flux and discharge under various pedoclimatic and cropping conditions (Prunty and Greenland, 1997; Shipitalo and Edwards, 1993; Saffigna et al., 1977). Drainage and bromide fluxes measured in drainable lysimeters have also been used recently to validate deterministic water and solute transport models under field conditions. Water fluxes from drainable lysimeters can be relatively well simulated under transient conditions with the Richards equation and some calibration (Majdoub et al., 2000; Dust et al., 2000; Vanclooster et al., 2000). However, bromide breakthrough curves are in some cases poorly simulated with the classical convection–dispersion equation due to soil heterogeneity, preferential flow, and plant interference on bromide leaching (Dust et al., 2000; Vanclooster et al., 2000). Soil heterogeneity causing preferential flow under transient water regime may be one of the causes for the classical convection–dispersion equation (CDE) to inadequately describe the solute transport process (Costa et al., 1994).

To overcome the tremendous efforts involved in characterizing local soil hydraulic properties at the field scale, Jury (1982) proposed a transfer function model with a lognormal probability density function, which obeys a stochastic–convective assumption. Under this assumption, dispersion within stream tubes is negligible and there is no mixing between stream tubes. Thus, solute macrodispersion is primarily due to variations in convective fluxes resulting from variations in stream tube velocities. Solutes moving in accordance with this assumption would be subjected to a linearly increasing dispersion with depth below the point of solute injection and the field-scale convection–dispersion model should not apply. Ellsworth et al. (1996) reviewed works where the dispersion coefficient D has been found to vary with depth in many ways: linear increase, nonlinear increase, constant decrease, and erratic fluctuations. Their study showed a near-linear increase in dispersivity with travel distance (1.7-m solute leaching depth) and a better fitting of the CLT function compared with the convection–dispersion model. Costa et al. (1994), comparing four different models, found comparable results with the convection–dispersion model and the CLT function.

Comparing the behavior of solutes leaching through drainable lysimeters and surrounding intact soil cores, Ellsworth et al. (1996) obtained 10% larger values for CLT parameter µ derived from lysimeter flux concentrations when compared with soil core resident concentrations, implying a greater volume of net applied water needed to displace the center of mass of solutes in drainable lysimeters. Drainable lysimeters offer a direct approach to monitor solute flux concentrations and water fluxes but may not represent true field conditions.

The objectives of this study are to obtain solute transport parameters for the CLT function under transient water conditions in sandy soils and to compare the behavior of solute leaching in drainable lysimeters and surrounding soil to assess the validity of drainable lysimeters as a tool for establishing solute transport modeling under field conditions.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Site Description and Instrumentation
This study was conducted in five fields cultivated by potato farmers and located in the Portneuf region, near Québec City, Canada (latitude: between 46°45' and 46°50'; longitude: between 71°35' and 72°00'). The soils under investigation were either a Morin or a Pont-Rouge sand (Humo ferric Podzol; Humic Haplorthod) developed on medium to coarse sands of fluvio–marine or deltaic origin (Raymond et al., 1976).

Fifteen drainable lysimeters were installed in the summer of 1995, three per field, at a minimum distance of 100 m between each lysimeter to reduce correlated soil properties and represent field-scale leaching conditions. At each lysimeter location, the soil was excavated in 5-cm depth increments, the size of a rectangular reservoir (1-m² surface by 1-m maximum depth) (Fig. 1) . After inserting a reservoir made of a PVC geomembrane, the soil was replaced in its original sequence and repacked to its original height with a hand pillar. The geomembrane was cut at a 30-cm depth to allow tillage and seeding operations. A 10% slope at the bottom of the reservoir allowed water to drain in a sampling reservoir (25 L) through a 1-cm-diameter hose. The sampling reservoir was accessed through a 2-m-long, 30-cm-diameter vertical well. The 45-cm upper part of this access well was retractable to allow tillage operations.

View larger version (24K): [in this window] [in a new window]   Fig. 1. Drainable lysimeter.

Solute Application
Bromide (Br^-) pulses were applied on 24 May 1996 in Fields 4, 2, and 3, and on 29 May 1996 in Fields 1 and 5. Four hundred grams of KBr were diluted in 2.2 L of water and pulses of 10 g Br^- m^-2 were hand-sprayed uniformly on soil surfaces of 4.5 x 6 m², centered over each drainable lysimeter. Prior to its application, Br^- was not detected (<0.1 mg Br^- L^-1 water).

Bromide Sampling and Analysis
Water samples were collected from drainable lysimeters at a 1-m depth on 15 occasions during 1996, following Br^- application until 14 Nov. 1996. Bromide concentrations from 181 water samples were determined by ion chromatography (Dionex [Sunnyvale, CA] 4000i) (O'Dell et al., 1984). The water samples were eluted with a mixture of 1.8 mM Na₂CO₃ and 1.7 mM NaHCO₃ through a Dionex AS4A-SC ion exchange column and Br^- was detected by electric conductivity. Water sample Br^- concentrations were interpreted as flux concentrations (C^f) in mg Br^- L^-1 water.

Soils surrounding drainable lysimeters where Br^- had been applied (4.5 x 6 m² surface) were sampled with a hand auger (7-cm diameter) on 3 July, 10 July, 24 July, 20 Aug., 2 Oct., and 18 Nov. 1996 at five depths (0–15, 15–30, 30–60, 60–90, and 90–120 cm). In potato fields, samples were taken 15 to 30 cm from the center of the hill. In the barley field, samples were taken randomly. Composite samples were made of six subsamples in the case of samples taken at the 0- to 15- and 15- to 30-cm depths and two subsamples for the 30- to 120-cm depths. Bromide resident concentrations (C^r) in soil samples were determined by ion chromatography as reported earlier (Dionex 4000i) after water extraction (O'Dell et al., 1984). Fifteen grams of moist soil (<4 mm) were mixed with 30 mL of distilled water (1:2 solution) and agitated during an hour, prior to filtration on #410 VWR (West Chester, PA) paper filters. Bromide C^r (mg Br^- L^-1 soil) values were calculated with measured soil gravimetric water content, gravel (>4 mm) content, and bulk density for each layer and lysimeter location.

Bromide concentrations in tubers were determined by harvesting tubers on 1-m² surfaces occupied by drainable lysimeters 125 d following planting in Field 1, 123 d in Field 2, 113 d in Field 3, and 125 d in Field 4. Fresh tuber mass was determined and subsamples of tubers were weighted, dried at 65°C, and weighted again to determine dry matter content. Samples of tuber dry matter were milled and sieved at 0.5 mm to determine Br^- concentrations with a phenol red indicator and colorimetry (Jones, 1993). One gram of sample was mixed with 30 mL of distilled water during 24 h. Water extract was filtered on Whatman (Maidstone, UK) #1 paper filter and filtered a second time on a Millipore (Bedford, MA) filter (45 µm) following addition of activated charcoal to remove color. In a 50-mL Erlenmeyer flask, 10 mL of extract were mixed with 0.5 mL of buffering solution (0.5 M NaC₂H₃O₂, 1.0 M NaNO₃, 0.5 M CH₃COOH, 9 mM NH₄OH), 0.5 mL chloramine T solution (3.55 mM), and 0.5 mL phenol red indicator (0.56 mM). After exactly 20 min, 0.13 mL of 2 M Na₂S₂O₃ was added and mixed. Color absorbance was read at 590 nm after 15 min and Br^- concentrations were determined with a standard curve. Tuber Br^- uptake was estimated with Br^- concentrations and tuber dry mass harvested on lysimeters.

Climatic Data
Daily precipitations were manually recorded with a funnel and cylinder at less than 1 km from Fields 1, 3, 4, and 5. An automatic weather station located less than 1 km from Field 2 and less than 21 km from the other fields recorded on an hourly and daily basis precipitation, air temperature, soil temperature at a 20-cm depth, relative humidity, and wind speed.

Convective Lognormal Transfer Function Modeling
To simplify the solute transport modeling, we assumed that during a long time period, transient water flow could be represented by a steady state flow approximation (Wierenga, 1977). This seemed reasonable due to relatively small changes in soil water content following Br^- application (Fig. 2) and to the absence of swelling clays, which can affect pore size distribution and water flux when water content varies (Dyson and White, 1986). It was also hypothesized that solute transport in these nonstructured coarse sandy soils should be less affected by soil water status than in well-structured soils (Jury and Roth, 1990).

View larger version (16K): [in this window] [in a new window]   Fig. 2. Mean and standard deviation of soil volumetric water content (0- to 90-cm soil depth) at 15 lysimeter locations during 1996.

A lognormal distribution can be used to estimate mean and variance of solute distribution related to cumulative drainage, which assumes that the solute travel times through successive layers are correlated (Jury and Roth, 1990). Following a narrow pulse ( function) input of solute, parameters of the lognormal distribution can be estimated by relating flux concentrations (C^f) measured in drainable lysimeters to cumulative drainage at depth z with the following mathematical form (Jury and Sposito, 1985):

[1]

where C^f is solute mass per volume of water flux, I is cumulaVar(ln I), and L is length.

Following the same solute pulse, resident concentrations (C^r) measured in soil cores may be related to cumulative drainage at depth z with the following mathematical form (Butters and Jury, 1989):

[2]

where C^r is solute mass per volume of soil. To estimate parameters µ, , and M_o, least squares optimization was conducted with the NLIN procedure in the SAS statistical package (SAS Institute, 1987).

Expected mean value of I, variance of I, and dispersion of f_z(I) are estimated as (Jury et al., 1991):

[3]

[4]

[5]

Average solute particle velocity is expressed in terms of the first-order time moment of f^f(z,I) (Eq. [1]) (Vanderborght et al., 1996):

[6]

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Flux Concentrations in Drainable Lysimeters
Bromide flux concentrations in 14 drainable lysimeters were monitored during 180 d, following Br^- application at the end of May until the end of November 1996 when soil-freezing conditions occurred (Fig. 3) . Lysimeter 3 on Field 3 was punctured and did not produce any drainage; therefore, no observations were recorded. Total masses of Br^- recovered in the other lysimeters were calculated with Br^- concentrations and drained water volumes. Considering 10 g Br^- m^-2 applied at the soil surface, percentage of Br^- recovered in drainage water ranged from 43 to 72% (Table 2). Bromide recovery in drainage water was on average higher in Fields 4 and 5, compared with the three other fields. Soil residual Br^- at the end of the monitoring period was apparently low, ranging from 2.7 to 7%. Bromide exported with tubers represented 1.4 to 3.4% of Br^- applied to the surface.

View larger version (29K): [in this window] [in a new window]   Fig. 3. Normalized bromide flux concentrations (Cf) measured at the bottom of drainable lysimeters (z = 100 cm) and convective lognormal transfer (CLT) functions fitted with Cf.

Considering tuber Br^- concentrations and Br^- recovered in leachates, the potato crop in Field 4 and barley crop in Field 5 seem to have less influenced Br^- leaching. The late planting of cultivar Kennebec in Field 4 (21 May) and its late emergence (24 June) could have contributed to a smaller interference of plants on Br^- transport in soil, which apparently gave higher recovery of Br^- in drainage water. Barley sown on a late date (24 May) in Field 5 and poor growing conditions due to soil acidity could also explain a better recovery of Br^- in drainage water.

Given the fact that plant interference appeared small, parameters of the CLT function were estimated to compare estimates with previously published values obtained under bare soils. Parameter optimization was conducted with Br^- flux concentrations (C^f) and drainage water measured at a 1-m depth in all three lysimeters located in each field (Table 3 and Fig. 3). Bromide masses applied (M_o) were also estimated with least squares fit of CLT model. Coefficients of determination (R²) for nonlinear models ranged from 0.72 to 0.88 and confidence in parameter estimates was greater under Fields 4 and 5. Bromide mass recovery [M_o/(10 g Br^- m^-2)] ranged from 62 to 74%, and was similar to results obtained under transient flow in soil columns by Meyer-Windel et al. (1999), which ranged from 47 to 72%.

These results indicate that under transient conditions in our drainable lysimeters, solute leaching seems to have operated faster than in most previously reported studies, even though plants might have interfered with Br^- breakthrough curves. Transient water conditions can alter Br^- breakthrough curves as shown in work by Meyer-Windel et al. (1999). In 15-cm soil columns of sandy soil surface horizons, Br^- concentrations appeared sooner under transient water conditions than under a permanent water regime considering a cumulative drainage scale. However, contrasting water regimes did not significantly alter Br^- breakthrough curves in 29-cm soil columns made of massive, single-grain sand. These authors concluded that preferential flow in surface horizons was responsible for unequivalent solute transport features between transient and steady state water regimes.

As mentioned earlier, crop uptake, the presence of a crop canopy with hills and furrows, and soil heterogeneity created by tillage and contrasting soil layers may also interfere with the Br^- transport process (Kung, 1990; Timlin et al., 1992; Powers et al., 1997).

Bromide Resident Concentrations in Soil
Soils adjacent to lysimeters (less than 2-m radius) were sampled at five soil depths six times during the season. Measured bromide resident concentrations (C^r) were used to optimize CLT function parameters with Eq. [2]. The reference depth was set at 30 cm to compare parameter estimates (µ and ) with those obtained from flux concentrations (C^f). Applied bromide masses (M_o) were also estimated by least squares fitting. Parameter optimization was conducted with bromide C^r and mean cumulative drainage water measured in all three lysimeters located in each field (Table 4). Coefficients of determination (R²) for nonlinear models ranged from 0.69 to 0.88 and confidence in parameter estimates was again greater under Fields 4 and 5. Bromide masses recovered ranged from 58 to 101%, and were not significantly different than masses recovered with C^f (nonsignificant Student test).

Solute spreading in the soil profile can be described by the variance of I (Eq. [4]) and the dispersion of f_z(I) (Eq. [5]), both of which are summarized in Table 5, along with the expected mean value of I (Eq. [3]) obtained under a CLT assumption with flux (C^f) and resident (C^r) concentrations (Jury et al., 1991). Although expected mean values of I were comparable and closely correlated (R² = 0.78) when either derived from C^r or C^f CLT functions, variance and dispersion of f_z(I) grew significantly at 1 m when calculated with C^r. Lower means, variances and dispersions of I in Fields 4 and 5 also related to a smaller interference of plants on the Br^- transport process (Table 2). Higher variances and dispersions of the CLT function calculated with C^r can also be related to sampling methodology or soil disturbance created during lysimeter installation.

Soil disturbance created during lysimeter installation can also affect solute transport parameters. Preferential flow paths generally occur in heterogeneous materials under transient water conditions, leading to early solute breakthrough times (Meyer-Windel et al., 1999). However, as soil depth increments of 5 cm were repacked in the drainage lysimeters, soil heterogeneous conditions were homogenized and preferential flow paths may have been disrupted. Also, under transient water regimes, preferential flow paths are known to increase with lower water flow rates (Meyer-Windel et al., 1999).

Convective Lognormal Transfer Function Model Assumptions
Resident concentrations (C^r) measured at five soil depths were used to verify CLT assumptions. Parameters M_o, µ, and optimized with bromide C^r measured in all fields at each depth are reported in Table 6 along with calculated expected mean of I (Eq. [3]), variance of I (Eq. [4]), dispersion of f_z(I) (Eq. [5]), and average solute particle velocity () (cm soil cm^-1 water) (Eq. [6]). Estimates of µ and for each soil depth are in the same range as µ and values obtained for each field with C^r (Table 4); however, µ estimates are lower at soil depths > 30 cm. As a result, increases with soil depth, but should be constant under the assumption of a homogenous soil profile.

View larger version (19K): [in this window] [in a new window]   Fig. 4. Variance and dispersion of fitted convective lognormal transfer (CLT) functions increasing with soil depth.

Comparing Measured and Simulated Bromide Concentrations
Bromide resident concentrations (C^r) measured at each depth in Field 5 are illustrated in Fig. 5 and compared with CLT functions either fitted with adjacent soil C^r (Table 4) or lysimeter C^f (Table 3). The CLT function fitted with C^r measured at all depths closely match measured C^r, except during the later sampling events when measured concentrations were relatively small and represented a small weight in the least squares optimization process. Since variance and dispersion of C^r surrounding lysimeters were greater than those estimated with C^f from lysimeters, the flux CLT function predicts a rapid decline of C^r in surface horizons (between 0 and 30 cm) and a more dense migration of C^r in subsurface horizons, with a greater deviation from measured concentrations.

View larger version (33K): [in this window] [in a new window]   Fig. 5. Normalized bromide resident concentrations (Cr) measured at five depths and six dates in Field 5, and convective lognormal transfer (CLT) function either fitted with Cr (—) (Table 4) or with Cf (– –) (Table 3). Cross centers and vertical bars represent mean and standard deviation of three measured Cr, and horizontal bars represent depth of soil sampling. The CLT parameters are reported in the top left graphic. Number of days following Br- application (t) and cumulative drainage (I) are also reported in every graphic.

Plant interference on Br^- transport in these fields may also contribute to the divergence between CLT functions fitted either with C^r or C^f. Plant Br^- uptake can result in Br^- cycling in the rooting zone (0–30 cm), leading to redistribution of soil Br^- concentrations in surface horizons (Kung, 1990). Sustained Br^- concentrations in the surface horizons were also noticed in our soil C^r profiles (Fig. 5), although final Br^- tuber uptake was relatively small (Table 2).

Measured resident concentrations (C^r) at 105 cm in Field 5 are compared with CLT functions either fitted with C^r measured at a 105-cm soil depth (Table 6), with C^r measured over all depths (Table 5), or with C^f measured in lysimeters (Table 4) (Fig. 6) . As expected, CLT functions fitted with C^r better predict a fast arrival of C^r at 105 cm compared with the C^f CLT fitted function. Again, this indicates that restricted C^r movement in the top soil layers does not delay solute leaching at greater depths as expected. Therefore, plant Br^- uptake is not the only reason for discrepancies between bromide C^r and C^f. In our model, water flux is assumed to be constant over the whole soil profile; however, evapotranspiration and contrasting soil hydraulic properties between the 0- to 30- and 30- to 120-cm soil layers (Table 1) probably create contrasting water fluxes in these two layers. Plant roots are mostly restricted to the 0- to 30-cm depth and a higher evapotranspiration rate in this zone would result in a greater net applied water volume (infiltration - evapotranspiration) near the soil surface than measured drainage volume at a 1-m depth. A higher volume of water displaced in the surface horizons would lead to a faster displacement of Br^- in the surface horizons; however, sustained bromide C^r values in these horizons indicate that other mechanisms are involved. The top soil layer in these sandy soils can act as a buffer zone capable of restricting solute leaching, when Br^- cycling in the root zone and contrasting soil properties between the surface and underlying soil layers (>30 cm) create preferential flow patterns.

View larger version (21K): [in this window] [in a new window]   Fig. 6. Normalized bromide resident concentrations (Cr) measured at a 105-cm depth in Field 5 (Crz = 105 cm). Convective lognormal transfer (CLT) functions either fitted with Cr measured at 105 cm (Table 6) [— f(Cr)z = 105 cm], with Cr measured at all depths (Table 4) [- - - f(Cr) Field 5], or with Cf measured in lysimeters (Table 3) [- — - f(Cf) Field 5].

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

Disagreement between flux and resident CLT functions optimized under a transient regime and plant interference undermines the use of drainable lysimeters as a tool to estimate solute transport parameters, anywhere but within lysimeter conditions. However, the assumption of a homogeneous soil profile in terms of water flux and solute transport parameters is also questionable as in these cultivated soils, the top soil layer exhibits contrasting hydraulic properties with the underlying soil. Soil disturbances related to drainage lysimeter installation as well as differences in sampling methodology between C^r and C^f also may be responsible for differences observed between solute transport parameters.

	ACKNOWLEDGMENTS

 
Funding for this research was provided by the Natural Sciences and Engineering Research Council of Canada, the Canadian Pork Council, and the Fédération des producteurs de porcs du Québec. Gratitude is expressed to M.H. April, M. Lambert, G. Thériault, and D. N'Sengiyumva for technical assistance.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Butters, G.L., and W.A. Jury. 1989. Field scale transport of bromide in an unsaturated soil 2: Dispersion modeling. Water Resour. Res. 25:1582–1589.
• Costa, J.P., R.E. Knighton, and L. Prunty. 1994. Model comparison of unsaturated steady-state solute transport in a field plot. Soil Sci. Soc. Am. J. 58:1277–1287.[Abstract/Free Full Text]
• Dust, M., P.N. Baran, G. Errera, J.L. Hutson, C. Mouvet, H. Schafer, H. Vereecken, and A. Walker. 2000. Simulation of water and solute transport in field soils with the LEACHP model. Agric. Water Manage. 44:225–245.
• Dyson, J.S., and R.E. White. 1986. The effect of irrigation rate on solute transport in soil during steady state flow. J. Hydrol. 107: 19–29.
• Ellsworth, T.R., P.J. Shouse, T.H. Skaggs, J.A. Jobes, and J. Fargerlund. 1996. Solute transport in unsaturated soil: Experimental design, parameter estimation, and model discrimination. Soil Sci. Soc. Am. J. 60:397–407.[Abstract/Free Full Text]
• Jones, D.R. 1993. Improved spectrophotometric method for the determination of low levels of bromide. Anal. Chim. Acta 271:313–321.
• Jury, W.A. 1982. Simulation of solute transport using a transfer function model. Water Resour. Res. 18:363–368.
• Jury, W.A., W.R. Gardner, and W.H. Gardner. 1991. Soil physics. 5th ed. John Wiley & Sons, NY.
• Jury, W.A., and K. Roth. 1990. Transfer functions and solute movement through soil. Birkhaüser, Basel, Germany.
• Jury, W.A., and G. Sposito. 1985. Field calibration and validation of solute transport for the unsaturated zone. Soil Sci. Soc. Am. J. 49:1331–1341.[Abstract/Free Full Text]
• Jury, W.A., L.H. Stolzy, and P. Shouse. 1982. A field test of the transfer function model for predicting solute transport. Water Resour. Res. 18:369–375.
• Kung, K.-J.S. 1990. Influence of plant uptake on the performance of bromide tracer. Soil Sci. Soc. Am. J. 54:975–979.[Abstract/Free Full Text]
• Majdoub, R., J. Gallichand, and J. Caron. 2000. Simulation of water flow based on the numerical method of lines. Can. Agric. Eng. J. 42:65–74.
• Meyer-Windel, S., B. Lennartz, and P. Widmoser. 1999. Bromide and herbicide transport under steady-state and transient flow conditions. Eur. J. Soil Sci. 50:23–33.
• O'Dell, J.W., J.D. Pfaff, M.E. Gales, and G.D. McKee. 1984. Test method: The determination of inorganic anions in water by ion chromatography. Method 300.0. USEPA Environ. Monitoring and Support Lab., Cincinnati, OH.
• Powers, W.L., R.L. Verhoeff, A.M. Parkhurst, P.J. Shea, and W. Troyer. 1997. Spatial series analysis of horizontal cores to characterize tracer patterns in soil profiles. Soil Sci. Soc. Am. J. 61:1018–1023.[Abstract/Free Full Text]
• Prunty, L., and R. Greenland. 1997. Nitrate leaching using two potato–corn N-fertilizer plans on sandy soil. Agric. Ecosyst. Environ. 65:1– 13.
• Raymond, R., G. Laflamme, and G. Godbout. 1976. Pédologie du Comté de Portneuf. Bulletin technique no 18. Ministère de l'agriculture du Québec, Québec City, Canada.
• Saffigna, P.G., C.B. Tanner, and D.R. Keeney. 1976. Non-uniform infiltration under potato canopies caused by interception, stemflows and hilling. Agron. J. 68:337–342.[Abstract/Free Full Text]
• Saffigna, P.G., D.R. Keeney, and C.B. Tanner. 1977. Nitrogen, chloride and water balance with irrigated Russet Burbank potatoes in a sandy soil. Agron. J. 69:252–257.
• Shipitalo, M.J., and W.M. Edwards. 1993. Seasonal patterns of water and chemical movement in tilled and no-till column lysimeters. Soil Sci. Soc. Am. J. 57:218–223.[Abstract/Free Full Text]
• Shnabel, R.R., W.L. Stout, and J.A. Shaffer. 1995. Uptake of a hydrological tracer (bromide) by ryegrass from well and poorly-drained soils. J. Environ. Qual. 24:888–892.[Abstract/Free Full Text]
• SAS Institute. 1987. Statistical Analysis Systems user's guide. Statistics. Version 5. 3rd ed. SAS Inst., Cary, NC.
• Timlin, D.J., G.C. Heathman, and L.R. Ahuja. 1992. Solute leaching in crop row vs. interrow zone. Soil Sci. Soc. Am. J. 56:384–392.[Abstract/Free Full Text]
• Van de Pol, R.M., P.J. Wierenga, and D.R. Nielsen. 1977. Solute movement in a field soil. Soil Sci. Soc. Am. J. 41:10–13.[Abstract/Free Full Text]
• Vanclooster, M., S. Ducheyne, M. Dust, and H. Vereecken. 2000. Evaluation of pesticide dynamics of the WAVE-model. Agric. Water Manage. 44:371–388.
• Vanderborght, J., M. Vanclooster, D. Mallants, J. Diels, and J. Feyen. 1996. Determining convective lognormal solute transport from resident concentration data. Soil Sci. Soc. Am. J. 60:1306–1317.[Abstract/Free Full Text]
• Wierenga, P.J. 1977. Solute distribution profiles with steady-state and transient water movement models. Soil Sci. Soc. Am. J. 41:1050– 1055.[Abstract/Free Full Text]

This article has been cited by other articles:

				A. Gaur, R. Horton, D. B. Jaynes, and J. L. Baker Measured and Predicted Solute Transport in a Tile Drained Field Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 872 - 881. [Abstract] [Full Text] [PDF]

				M. O. Gasser, J. Caron, R. Lagace, and M. R. Laverdiere Predicting Nitrate Leaching under Potato Crops Using Transfer Functions J. Environ. Qual., July 1, 2003; 32(4): 1464 - 1473. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Gasser, M. O. Articles by Lagacé, R. Search for Related Content PubMed PubMed Citation Articles by Gasser, M. O. Articles by Lagacé, R. GeoRef GeoRef Citation Agricola Articles by Gasser, M. O. Articles by Lagacé, R. Related Collections Flow Soil Physics Vadose Zone Processes and Chemical Transport Ground Water Quality Soil Models