JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Agricola
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Related Collections
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Nutrients
Right arrow Other Models
Right arrow Nitrogen
Journal of Environmental Quality 32:1455-1463 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Vadose Zone Processes and Chemical Transport

Estimating Nitrate Leaching with a Transfer Function Model Incorporating Net Mineralization and Uptake of Nitrogen

Li Rena, Junhua Maa and Renduo Zhang*,b

a Department of Soil and Water Sciences, China Agricultural University, and Key Laboratory of Plant–Soil Interactions, MOE, Beijing, 100094, China
b Department of Renewable Resources, University of Wyoming, Laramie, WY 82071-3354 and Department of Water Resources, Wuhan University, Wuhan 430072, China

* Corresponding author (renduo{at}uwyo.edu)

Received for publication April 7, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Because of the complex interaction of chemical and biological processes of nitrogen (N) in soils, it is difficult to estimate the leaching of nitrate with various N transformations in porous media. In this study, a transfer function model was developed to simulate the outflow concentration of nitrate in soils during the growth of winter wheat (Triticum aestivum L.), taking into account the main N transformations using source and sink terms. The source and sink terms were treated as inputs to the solute transport volume and incorporated into the transfer function model to characterize their effects on nitrate concentration in the outflow. A field experiment was conducted in three nonweighing lysimeters for 181 d. Nitrate concentrations were measured along the 2-m soil profile of each lysimeter at different times. Comparison between the experimental data and simulated results with the transfer function showed that the model provided reasonable prediction of the nitrate leaching process as well as the total amount leached. Results also indicated that considering the N transformations in the transfer function significantly increased the estimation accuracy. The relative errors of total amount leached were <7% with the N transformations included, but up to 17% without including the transformation processes.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
WITH THE USE OF N FERTILIZERS increasing steadily for agricultural productivity, nitrate is being transported to surface and ground water systems. Recent public awareness has lead to the perception that NO3–N contamination of surface and ground water resources is closely related to the large increase in N fertilizer use over the past five decades (Keeney and DeLuca, 1993). Nitrate is the most ubiquitous chemical contaminant in the world's aquifers and the levels of contamination continue to increase (Spalding and Exner, 1993). The relationships between N fertilizer application rates, crop yields, and the amount of NO3–N leaching need to be quantified for the development of soil and crop management practices that are economically and environmentally sustainable (Jaynes et al., 2001). However, nitrate transport in soils is a complicated process, which is affected by physical, chemical, and biological properties of the porous media. Therefore, it is very challenging to estimate the leaching of nitrate and various associated N transformation processes in soils.

Simulation models provide an effective means to study the complex and interactive processes of N in soils (Johnsson et al., 1987; Godwin and Jones, 1991; Quemada and Cabrera, 1995; Pang and Letey, 1998). Johnsson et al. (1987) investigated mineral N dynamics and N losses in a layered agricultural soil, including the major processes that determined the inputs, transformations, and outputs of N in the soil. Lafolie (1991) simulated the N cycle in soils with a simple model, in which mineralization of organic matter, nitrification of ammonium, and denitrification were characterized as first-order processes. Besides the N cycle within the soil, N uptake by plants is also an important process of N transformation. The Michaelis–Menten formula is a common method to simulate N uptake by plants (Cabon et al., 1991; Lafolie, 1991) and needs parameters, such as the maximum uptake rate of N, the equilibrium constant for the competitive ion interaction, and others. The effects of N on crop growth were also considered in many crop growth models (Groot and de Willigen, 1991; Kersebaum and Richter, 1991).

Stochastic models offer an alternative approach to simulate N transport processes when random variables are considered or where input parameters are difficult to obtain. To represent the process of chemical transport in soils, transfer function models (TFMs) are used based on a probability density function of the travel time of chemical molecules (White et al., 1986; Jury et al., 1990; Jury and Scotter, 1994). Transfer functions are applied to model complex systems in a simple way by characterizing the output flux as a function of the input flux. White (1987) applied the TFM for prediction of nitrate leaching under field conditions. However, the study was only applied in shallow well-drained fields. In addition, the studied time period was relatively short and nitrate was treated as a nonreactive solute.

Heng et al. (1994) simulated sulfate sulfur, considering the effect of transformations of the reactive solute as source–sink terms in a transfer function approach. Heng and White (1996) applied a simple analytical form of the TFM to model sulfate sulfur leaching in a drained pasture soil. The combined effect of the sources and sinks was treated as an input to the initial resident concentration in the transport volume. White et al. (1998a) summarized the previous work and indicated that the combined effect of sources and sinks could be calculated effectively. White et al. (1998b) adopted the TFM to simulate nitrate leaching from a drained sheep-grazed pasture and the results supported the approach of Heng and White (1996).

The main objective of this study was to develop a transfer function model to characterize nitrate transport in deep soils with the main N transformations, such as immobilization, mineralization, and plant uptake. Subsequently, the TFM was used to estimate nitrate leaching in a field with a wheat crop. The simulated results based on the model were compared with experimental data and with simulated results using a transfer function without considering the N transformations (White, 1987).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Theoretical Background
The transfer function can be written in the form of (Jury et al., 1986):

[1]
where Qex(t) is the loss rate of chemical mass (g d-1) through the exit surface at t (d), Qin(t') is proportional to the input rate of chemical mass (g d-1) at the soil surface and at an earlier time t' (d), and g(t - t') is the probability density function (pdf) of chemical travel time (d-1). The pdf can be estimated from the effluent volume and concentration at each time step as follows:

[2]
where C(ti) is the chemical concentration (g m-3) measured at time ti and {Delta}(Vex)i is the volumetric water flux (m3) during the time interval {Delta}ti = ti - ti-1. Field and lab experiments (Jury et al., 1986; White et al., 1984, 1986; Dyson and White, 1987) show that the travel time pdf can be characterized by a lognormal pdf:

[3]
where µ and {sigma}2 are the mean and variance of ln t, respectively.

Given that nitrate resident in a defined soil volume is leached in response to a pulse input of water, the quantity of NO3–N leached can be calculated from a particular sample realization of the stochastic leaching process. Assuming that during the observed time interval, NO3–N concentrations in the outflow vary slowly, we can express the leaching mass of NO3–N at the outlet as follows:

[4]
where D is the drainage rate (m3 d-1) at time t (d), C is the flux-averaged concentration (g m-3) in the drainage at time t, and {Delta}t is the time interval measured (d). The leaching mass can also be written in the form of (White, 1987):

[5]
in which {Delta}C is the change in nitrate concentration within the transport volume during {Delta}t and Vst is the transport volume (m3). Integrating Eq. [5] gives the leaching concentration (g m-3):

[6]
where C0 is the initial concentration of nitrate in the transport volume (g m-3) and D0 is the average drainage rate (m3 d-1) during the time interval {Delta}t. Then, the leaching mass can be estimated by:

[7]

The derivation of Eq. [6] and [7] is based on the assumption that nitrate is a nonreactive solute, that is, without considering N transformations in soils. Generally speaking, transformations of N in soils include adsorption, mineralization, nitrification, denitrification, volatilization, and plant uptake. In this paper we only consider the main N transformations (immobilization, mineralization, and plant uptake) using the following source–sink terms:

[8]
where Cm is the net mineralization (g m-3) (i.e., the combination of immobilization and mineralization) and Cu represents plant uptake of N (g m-3). Since nitrification is a much faster process than mineralization, it was assumed that NO3–N was the direct product of mineralization and nitrification was ignored in the calculation of source–sink terms under the condition of our experiments. To estimate Cu, it is assumed that the rate of plant uptake of N is proportional to the evapotranspiration rate of the wheat. The N uptake at any time per unit of evapotranspiration is approximated by:

[9]
where Ut is the total amount of plant uptake of N (g m-2), ETc is the cumulative evapotranspiration (m), ET0 is the cumulative evapotranspiration (m) during the period in which there is no plant uptake of N, and ET(t) is the cumulative evapotranspiration (m) from the beginning to time t.

White (1987) treated N as a resident solute once fertilizer N entered the soil transport volume. The leaching of the resident solute from the transport volume is dependent on hydrodynamic dispersion, convection, and source–sink terms. The sum of the probability of solute resident leaching from an exit surface and the probability of solute disappearance attributable to the source–sink terms should be equal to 100% (White, 1989). By treating the source–sink terms as an input to the initial concentration in the transport volume, assuming soil water flow is approximately steady state, the disappearing concentration of solute attributable to source–sink terms (g m-3) is related by:

[10]
where P(t) is the cumulative probability function (cpf) of solute leaching from exit surface and:

[11]

Here the cpf is used to describe the travel time of a conservative solute (White, 1987). If the drainage depth is shallow and rainfall intensity is large, such as the conditions described by White (1987), nitrate in the effluent may be treated as a conservative solute. However, the condition in our study was much different, where the drain was deep (2 m) and the study period was long (181 d). In such conditions, nitrate cannot be treated as a conservative solute. We consider the transformation processes of nitrate using the source–sink terms according to the linear superposition of transfer functions (White et al., 1998a). Combining Eq. [6] (considering the initial concentration) and Eq. [10] (considering the transformation processes), we have:

[12]

The issue of how to choose the value of C0 in Eq. [12] is a critical one that was discussed in the latter section. Using Eq. [12], we calculated the leaching concentration of N at any time. With the drainage volume {Delta}Vi in the time interval {Delta}ti = ti - ti-1, we estimated the nitrate amount leached by:

[13]

Field Experiment
To study N transport in soils, a field experiment was conducted at the experiment station of the China Agricultural University, Beijing. The experiment was performed in three nonweighing lysimeters, called SE2, SE3, and SE4 in the following analyses. The volume of each lysimeter was 2 x 2 x 2 m3. The texture of the soil was sandy clay loam (Aquic Cambisol) from a depth of 0 to 60 cm and sandy loam (Aquic Cambisol) below 60 cm.

During the whole growing period of winter wheat, 240, 120, and 0 kg N ha-1 of urea was applied to SE2, SE3, and SE4, respectively. The amount of 240 kg ha-1 of fertilizer N is the conventional application of N fertilizer for winter wheat in the region. Half of the amount was applied as basal fertilizer (uniformly mixed into 0 to 10 cm of the topsoil) at the beginning of the experiment (8 Oct. 1998), and the other half was applied as topdressing on 11 Apr. 1999. After fertilization, wheat seeds were sown and 75 mm of water was applied on the same day. During the growing season, two further irrigations (75 mm water each time) were applied on 14 Apr. 1999 and 7 May 1999. At depths of 0.2, 0.4, 0.6, 0.8, 1.0, 1.4, and 1.8 m, suction cups were installed to sample the soil solution. Soil solution was also collected from the bottom of the lysimeters. At the beginning, the solution samples were collected every 3 or 4 d. During the winter (from December 1998 to February 1999), the drainage decreased gradually; therefore, the solution samples were taken every 8 to 10 d. In the spring, we only took solution samples on 10 Mar. 1999 and 8 Apr. 1999 because of little outflow. Concentration of NO3–N was measured using the samples of soil solution with a continuous-flow analytical system (TRAACS 2000; Bran+Luebbe, Norderstedt, Germany). After 11 Apr. 1999, there was no drainage, thus we selected the study period from 8 Oct. 1998 to 8 Apr. 1999, totaling 181 d. Therefore, only the basal fertilizer was applied in the lysimeters during the study period and the three different N treatments were referred to as N120, N60, and N0, respectively, for SE2, SE3, and SE4 in the following analyses.

For each lysimeter, tensiometers were installed at locations with 20-cm intervals in the 2-m soil profile. During the experiment, soil water matric potential values were recorded with the tensiometers and used to estimate soil water contents based on soil water retention curves. The soil water retention curves were measured using soil samples with the pressure membrane method (Klute, 1986). Soil properties of SE2, such as saturated soil water content, field capacity, soil water content at wilting point, bulk density, organic matter content, and total porosity, are listed in Table 1 . The soil properties for SE3 and SE4 were similar to those in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Soil properties of the SE2 lysimeter.

 
Weather data were also collected. Potential evapotranspiration was estimated from the crop coefficient and reference crop evapotranspiration using the Penman–Monteith equation. Actual evapotranspiration was evaluated by the product of the potential evapotranspiration and water stress coefficient. The average annual rainfall in the region is about 640 mm (mainly in summer). However, the study year was much drier than regular years. The rainfall was 374 mm from 1 Oct. 1998 to 30 Sept. 1999, and 51 mm during the study period (Fig. 1) .



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1. Rainfall distribution during the study period.

 

   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Figure 2 shows the cumulative water input (irrigation and rainfall), and cumulative actual and potential evapotranspiration vs. time, indicating that the amount of evapotranspiration was larger than the water input in most of the study period. The vertical water content distribution in lysimeter SE2 is shown for different times in Fig. 3 . Obviously, the change in water content over time was small. The maximum variation of total water storage in SE2 was <6%. Similar results were observed in SE3 and SE4. Therefore, in the following analyses, we assumed a steady state flow in the lysimeters.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Cumulative water input (irrigation and rainfall) and cumulative actual and potential evapotranspiration vs. time.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Vertical water content distribution in lysimeter SE2 for different times.

 
To account for the N transformation processes, we needed to estimate the source–sink terms in Eq. [8]. As mentioned above, in this study we only consider the main N transformations: immobilization, mineralization, and plant uptake. The consideration was justified as follows. As shown in Table 1, with the increasing of depth, the organic matter content decreased rapidly. Because the processes of mineralization–immobilization mediated by microbes take place mainly in organic-matter-rich soil layers, it was reasonable to assume that mineralization–immobilization was active only within the top-1-m depth of soil and the mineralization rate was not constant, but a function of time during the study period. The term Cm (i.e., net mineralization) in Eq. [8] could be a sink or source, depending on whether immobilization or mineralization was dominant. The amount of net mineralization was estimated by an N-balance method using field experimental data. Since it was difficult to measure some terms in the balance equation in the field, the calculation of net mineralization amount was simplified. First, the net leaching amount of NO3–N at the top-1-m depth was ignored because of the frequently upward and downward transport of NO3–N within the depth with the recharge and discharge of soil water. Second, denitrification and volatilization were neglected during the calculation as well. Fenn and Miyamoto (1981) reported that the potential for ammonia losses from urea decreased rapidly with increasing quantities of water applied and ammonia losses from urea were minimal if injected into a 2.5-cm soil depth or deeper. In this experiment, after urea was mixed into the top 10 cm of soil, 75 mm of water was applied immediately; therefore, ammonia volatilization should be minimal. As for denitrification, as shown in Fig. 4 , the air-filled porosity within the top 1 m of soil is much larger than the critical value for denitrification, 0.10 m3 m-3. Thus, denitrification was not considered in this study. The net mineralization amount at a time interval between t1 and t2 was calculated as follows:

[14]
where Np is the plant uptake of N (kg ha-1), N1 is the mineral N (kg ha-1) at time t1, N2 is the mineral N (kg ha-1) at time t2, and Nf is the fertilizer N (kg ha-1). If the calculated result was positive, net mineralization occurred in the soil; if the result was negative, then net immobilization occurred. Because of the deep drainage (2 m) and the small drainage volume, to avoid overestimating Cm, net mineralization was estimated using the soil water storage from the soil surface to a 1-m soil depth rather than the drainage. In other words, Cm was equal to the net mineralization amount (Eq. [14]) divided by the water storage in the top-100-cm soil profile.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Vertical distribution of air-filled porosity in lysimeter SE2 for different depths.

 
During the study period, a field trial with 15N microplots was also conducted in the adjacent field to study the plant uptake of N. The results showed that the total amounts of plant uptake of N were 86.7, 71.6, and 63.1 kg ha-1, respectively, for the treatments of N120, N60, and N0. Among the total amounts of plant uptake, N uptake from the fertilizer was 21.9 and 30.8 kg ha-1 for the treatments of N60 and N120, respectively. During the early period (from seeding to the first-leaf appearance), when the plant depended on nutrition in the seed, there was little plant uptake of N from the soil. The term of Cu (i.e., plant uptake of N) in Eq. [8] was estimated using Eq. [9] with an early plant period about 14 d. In Eq. [9] the cumulative evapotranspiration was calculated from the cumulative potential evapotranspiration, multiplied with a coefficient in the form of (Jensen et al., 1971):

[15]
in which Av = (Wt - Wr)/(Wf - Wr) x 100 (%), Wt is the actual water storage in the root zone at time t (d), Wf is the water storage at the field capacity (the soil water content at -0.03 MPa) (m3 m-3), and Wr is the water storage at the wilting point (the soil water content at -1.5 MPa) (m3 m-3). The terms Wt, Wf, and Wr were determined from the water content distribution across the 2-m soil profile in the lysimeter, from which we had Wf = 0.51 m and Wr = 0.41 m.

To apply the transfer function model, we needed to estimate the C0 value in Eq. [12]. White (1987) defined C0 as NO3–N concentration in the transport volume at t = 0. Jury et al. (1990) suggested that it was appropriate to take the uniform resident concentration in the soil solution at t = 0 as C0. For studies in regions with shallow drainage systems (Heng et al., 1994; Heng and White, 1996; White et al., 1998b) or with uniform initial concentration distributions or both, using the concentration of the first outflow sample as C0 should give sufficiently accurate simulation results (White, 1987, 1989). However, in our study, the initial concentration distributions were not uniform within the deep soil profiles (Table 2) . The higher NO3–N concentration of the first outflow samples in the three lysimeters may be caused by bypass flow in the lysimeters. Different from White (1987) and White et al. (1998b), in which rainfall was frequent and the drainage system was shallow (0.45 m), the depth of profiles in this experiment was much deeper (2 m), and water movement was much slower because of the small water input. Therefore, in our case the C0 value suggested by White (1987)(1989) may characterize poorly the effect of the initial concentration on the outflow concentration dynamics. Considering the contribution to NO3–N leaching from bypass flow and the effect of initial concentration distributions (sampled by suction cups) along the profiles, we proposed the following equation to estimate C0:

[16]
where Ci is the initial nitrate concentration (g m-3) at depth zi (m), n is the number of initial concentration measurements down the soil profile, and Cf is the nitrate concentration of the first outflow sample (g m-3). The values of Ci and Cf for the different treatments are listed in Table 2. According to Eq. [16], we obtained the C0 values as 59.53, 29.55, and 15.95 mg L-1 for the treatments of N120, N60, and N0, respectively.


View this table:
[in this window]
[in a new window]
  		Table 2. Initial nitrate concentrations within the lysimeters (Ci) and the nitrate concentration of the first outflow sample (Cf) with the 120, 60, and 0 kg N ha-1 treatments.

 
Using the outflow flux data and Eq. [2], we calculated probability density function (pdf) values of NO3–N at the exit surface (z = 2 m). Based on the experimental concentration data and the moment method (Jury and Sposito, 1985), we estimated the parameters of Eq. [3], that is, the mean (µ) and variance ({sigma}2) of ln t. As shown in Fig. 5 , the lognormal pdf function (Eq. [3]) fitted the pdf data reasonably well. The shapes of the pdfs of the three lysimeters are very similar, indicating that the pdfs are dependent on the transport characteristics of the soils, but not on the amount of applied fertilizer.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 5. Measured probability density function (pdf) data of NO3–N and fitted lognormal pdf functions for (A) 120, (B) 60, and (C) 0 kg N ha-1 treatments.

 
The dynamics of net mineralization vs. time are shown in Fig. 6 . At the beginning, net mineralization was promoted by tillage. After about 10 d of the fertilizer application, the soil N presented strong immobilization. In the N0 treatment, without fertilizer application, the net mineralization was larger than that of N120 and N60 treatments, to provide more mineral N for plant growth. In the early spring, the higher amount of net mineralization may result from stimulating activities in the wet–dry and freeze–thaw cycles. Using Eq. [9], we calculated plant uptake of N as a linear function of the cumulative evapotranspiration. The combined effects (Cb) of the N transformations, or the source–sink terms, vs. time are presented in Fig. 7 . At the beginning, Cb was positive, attributable to small plant uptake. Then, Cb became negative because of strong immobilization. After that, although the net mineralization was positive, Cb was still negative attributable to much more plant uptake of N. The total plant uptake of N, total net mineralization amount during the study period, and grain yield at harvest are summarized in Table 3 . The results showed that the total plant uptake of N increased with the increasing fertilizer N rate; the total net mineralization amount was positive and the plant absorbed a considerable amount of soil mineral N during the growing period; and the higher fertilizer N rate did not result in a higher production, thus the local agricultural practice in the region may need to be improved.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 6. Dynamics of net mineralization vs. time in lysimeters for (A) 120, (B) 60, and (C) 0 kg N ha-1 treatments.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 7. Combined effects (Cb in Eq. [8]) of the main N transformations (immobilization, mineralization, and plant uptake) vs. time for the 120, 60, and 0 kg N ha-1 treatments.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Total plant uptake of N, total net mineralization amount during the study period, and grain yield at harvest for the 120, 60, and 0 kg N ha-1 treatments.

 
Using Eq. [12], we estimated the outflow concentration of NO3–N. We also used the method of White (1987) (Eq. [6]), which treated nitrate as a conservative solute, to estimate the outflow concentration of NO3–N. The comparison of the estimated results with the measured data showed that Eq. [12] provided better results than Eq. [6].

After using Eq. [12] and Eq. [6] to estimate the outflow concentration of NO3–N, we also calculated the total amount of NO3–N leached in kg ha-1. Table 4 shows that with increasing fertilizer N rate, the amount of nitrate leached also increased. As shown in Fig. 8 , the estimated results of Eq. [12] matched the measured data much better than those by Eq. [6]. Compared with the measured total leaching amount, the relative estimated errors of Eq. [12] range between 1 and 7%, whereas the relative estimated errors of Eq. [6] range between 8 and 17% (Table 4). The root mean square errors also indicated that incorporating the main N transformations in the model improved the estimation accuracy significantly. The advantages of the transfer function model (Eq. [12]) were further demonstrated by the goodness of fits between the simulated results using the model and the measured leaching amount of NO3–N (Fig. 9) . The coefficients of determination (r2) between the simulated results and the measured data were greater than 0.99 for the different N treatments.


View this table:
[in this window]
[in a new window]
  		Table 4. Comparison of the estimated results using Eq. [6] and Eq. [12] for total leaching amount of NO3–N with the 120, 60, and 0 kg N ha-1 treatments.

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 8. Measured leaching amount of NO3–N and estimated results using the White (1987) method (Eq. [6]) and Eq. [12] for (A) 120, (B) 60, and (C) 0 kg N ha-1 treatments.

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 9. Goodness of fits between simulated (Eq. [12]) and measured leaching amount of NO3–N for (A) 120, (B) 60, and (C) 0 kg N ha-1 treatments.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
To study nitrate leaching in deep unsaturated soils, a field experiment was conducted in three soil lysimeters with plant growth at the experiment station of the China Agricultural University, Beijing. A transfer function model was developed to simulate nitrate transport with N transformations in soils. With reasonable assumptions, N transformations in soil and plant uptake were incorporated in the model through source–sink terms, which were treated as inputs to the transport volume. Different from previous studies, the source–sink terms are more easily estimated using the relationship between N uptake and evapotranspiration and the relationship between net mineralization amount and soil water storage. To account for nonuniform distributions of initial nitrate concentration in deep drained soils, an approach was proposed to select an average value of nitrate concentration in the transport volume, in which potential bypass flow and the characteristic of deep drainage were incorporated. Therefore, combined with the transfer function model, the average value was used to characterize the effect of initial concentration on the nitrate outflow variation. Good agreement was obtained between the experimental data and the estimated results using the transfer function model. Statistical analyses showed that the transfer function model has significantly improved the estimation accuracy, compared with the model without considering the N transformation processes. Therefore, the transfer function model provides a useful tool to study nitrate transport with transformation processes and plant uptake in deep soils.


   ACKNOWLEDGMENTS
 
This research was partly supported by the National Key Basic Research Special Funds (NKBRSF; no. G1999011803), China. The data of plant uptake and net mineralization of N were provided by the Department of Plant Nutrition, China Agricultural University.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 1167-1172. [Full Text]  



This article has been cited by other articles:


Home page
 Soil Sci.Home page
L. Ren, J. Ma, and R. Zhang
Estimating the Influence of Nitrogen Transformations on Nitrate Leaching in Soils
Soil Sci. Soc. Am. J., August 9, 2007; 71(5): 1460 - 1468.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Agricola
Right arrow Articles by Ren, L.
Right arrow Articles by Zhang, R.
Related Collections
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Nutrients
Right arrow Other Models
Right arrow Nitrogen

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		 Soil Science Society of America Journal