VZJ sign up for etocs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 13 May 2005
Published in Vadose Zone J 4:329-336 (2005)
DOI: 10.2136/vzj2004.0102
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
Related Collections
Right arrow Batch Studies
Right arrow Kinetics
Right arrow Nutrient Management

SPECIAL SECTION: ZNS'03 VADOSE ZONE RESEARCH

Characterization of Nitrogen Transformations, Sorption and Volatilization Processes In Urea Fertilized Soils

S. Bolado Rodrígueza,*, A. Alonso-Gaiteb and J. Álvarez-Benedíb

a Dep. de Ingeniería Química, Univ. de Valladolid, Spain
b Inst. Tecnológico Agrario de Castilla y León, Valladolid, Spain
* Corresponding author (silvia{at}iq.uva.es)

Received 23 June 2004.



   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Nitrogen Transformation Model
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The fate and transport of nitrogen compounds in urea-fertilized soils has historically been a topic of great interest. At present, several models exists for simulating the various transformations involved. A major constraint in the application of these models is the assignment of numerical values to the coefficients that quantify these transformations. This study attempts to characterize the processes linked to the hydrolysis of urea, adsorption and volatilization of ammonia, and nitrification in agricultural soils. We followed the temporal sequence of volatilization of ammonia in two different soils fertilized with urea under controlled conditions in the laboratory. The amounts of nitrate and ammonia in soil were also measured at the end of each trial. Several modeling alternatives were analyzed in order to accurately reproduce the results. Among these alternatives, the most successful approaches included the kinetics of urea hydrolysis and ammonia adsorption and desorption processes. Urea hydrolysis required a parameter accounting for an activation time, which was found to be greater at relatively high urea concentrations and low experimental temperatures. The adopted methodology in this study can provide rate data for N transformation and coupled sorption processes critically needed in N fate and transport studies. Values were estimated for kinetic coefficients which may be useful in general-purpose models. Rate data presented in this work quantify the effects of temperature, soil moisture, and initial concentration of applied urea on the dynamics of urea hydrolysis, nitrification, and ammonia sorption and volatilization.

Abbreviations: LS, loamy sand • SCL, sandy clay loam


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Nitrogen Transformation Model
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE DYNAMICS OF NITROGEN in soil–water–air systems is of considerable interest in many environmental and agricultural applications, including the development of cost-effective and environmentally sound nutrient management practices. Historically, the main problem with the use of urea as a nitrogen source has been the control of ammonia losses by preventing urea hydrolysis since these two processes are strongly coupled in soils (e.g., Sturnpe et al., 1984; Al-Kanani et al., 1991).

An extensive list of models exists for studying the use of nitrogen in agriculture and minimizing N losses to the environment. These models differ from each other in terms of their representation of the processes involved, the invoked numerical algorithms, and applicable working scales. Examples given by Shaffer (1995) include EPIC (Williams et al., 1984), GLEAMS (Knisel, 1993), NLEAP (Shaffer et al., 1991), NTRM (Shaffer and Larson, 1987), and LEACHM-N (Wagenet and Hutson, 1989). While most of these models were initially developed for use at the point and field scales, several, such as NLEAP and LEACHM-N, have also been applied at the farm and regional scales through the use of GIS and related techniques (Wylie et al., 1994; Bleecker et al., 1990). Another group of models was developed primarily to estimate transport of nitrogen and other chemicals in surface runoff. These include general models such as CREAMS (Knisel, 1980) and more detailed two-dimensional models such as AGNPS (Young et al., 1989). Initially, most of the models had a fixed objective, and focused on specific processes such as volatilization of ammonia (e.g., Parton et al., 1981), or lixiviation of nitrates (e.g., Addiscott, 1981). The focus of these various studies gradually broadened to include more complex simulation algorithms, such as coupling the nitrogen dynamics with energy and water balances as in the case of CERES-N (Godwin and Jones, 1991), SWATNIT (Vereecken et al., 1990, 1991), WAVE (Vanclooster et al., 1996), LEACHN within the LEACHM model (Wagenet and Hutson, 1989; Hutson and Wagenet, 1992), and CropSyst (Campbell and Stockle, 1995).

With greater sophistication of a model, the more likely that the model will be able to accurately represent the actual nitrogen transformation process within a given application. Unfortunately, the number of required parameters also increases with increased complexity of the model. Parameter uncertainty can also lead to inaccurate applications. Thus, the main limitations in the use of N fate and transport models are inadequate representation of the coupled processes involved (Diekkrüger et al., 1995) and difficulty of obtaining input parameters for the models (De Willigen, 1991; Schmied et al., 2000). The parameters may be obtained independently from correlations or laboratory assays, or by means of field experiments and parameter optimization using inverse simulation techniques.

When the objective of a study involves the application of a predictive model at the field scale, the use of inverse simulation techniques in conjunction with field data is probably the more efficient alternative (Ritter, 2002). With such a method, data from a field experiment are typically combined with a forward model and an appropriate optimization algorithm to estimate the model parameters. There are, however, few good published data, and the processes may vary considerably from one scenario to another. Also, experimental work at the field scale is costly in terms of time and money. Laboratory studies can provide an alternative for analyzing the processes of interest under controlled conditions close to those at the field scale. Through controlled experiments, estimates of important kinetic parameters of the model could be achieved. Also, experimentation under controlled conditions allows one to establish the dependence of nitrogen dynamics on variables which are difficult to control at the field scale such as moisture content, temperature, and wind velocity. Studies of the effects of these variables also allow for the establishment of ranges of applicability for the model parameters, and their ability to represent processes under field conditions.

The ultimate similarity among the parameters obtained under laboratory conditions and those in the field will depend upon the experimental procedures used to duplicate natural conditions. Still, when experimental field data are not available, prior estimation of the main model parameters may be more desirable than using correlations. Where possible, the results could always be refined later using inverse methods.

The objective of the present study was to generate and analyze rate data for N transformations in soils fertilized with urea. This required the characterization of the main nitrogen transformation processes (i.e., urea hydrolysis, ammonia adsorption, volatilization, and nitrification), and analyzing the effects of temperature, moisture, and initial concentrations of urea on the transformations. Our experimental study focuses on the relative importance of each transformation in a system in which several coupled processes operate simultaneously. A conceptual model was developed for analyzing the experimental results. The model is based on descriptions used in several general-purpose models. The model was subsequently used to obtain values for kinetic parameters characterizing the main nitrogen transformation processes at the field scale.


   Nitrogen Transformation Model
TOP
 ABSTRACT
 INTRODUCTION
 Nitrogen Transformation Model
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The model used in this study is a modification of the model used by Álvarez-Benedí et al. (1999) for simulating the coupled processes of adsorption and volatilization. The approach was previously applied successfully to the volatilization of organic compounds such as pesticides (Tabernero et al., 2000). As part of this study, we incorporated urea hydrolysis (generation of ammonia) and nitrification (transformation of ammonia to nitrate) into the model. The model considers an isotropic soil, while concentration gradients (and therefore transport) were not considered, consistent with the experimental system employed here. The nitrogen transformations were simplified based on Fig. 1 , so as to consider first-order kinetics for hydrolysis, nitrification, and volatilization processes (Wagenet et al., 1977). Mineralization (release of mineral-N by decaying organic matter) was considered negligible compared with urea hydrolysis because of the short duration of the experiments, the low organic matter content, the absence of fresh organic matter, and the relatively high concentration of applied urea. Denitrification (reduction of nitrate to gaseous products) was also considered negligible under the imposed aerobic conditions. The resulting model hence provides the simplest possible situation from a conceptual point of view.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 1. Schematic of the two models used: (a) considers linear local equilibrium and reversible NH4+ sorption, (b) considers NH4+ sorption and desorption as independent kinetics processes. Both models consider first-order kinetics for urea hydrolysis and pseudo-first order kinetics process for ammonia volatilization. KD is a distribution coefficient and kV, kN, kHU, kads, kdes are rate coefficients (volatilization rate, nitrification rate, urea hydrolysis rate, adsorption rate, and desorption rate, respectively).

 
Hydrolysis of urea in the model was assumed to follow the reaction: CO2 + 3H2O -> 2NH4+ + 2OH– + CO2. Hydrolysis has traditionally been described using first-order kinetics (e.g., Godwin and Jones, 1991; Vanclooster et al., 1996):

[1]
where U is the concentration of N-urea expressed as mg N/cm3 solution, {theta} is the volumetric moisture content (cm3/cm3), t is time, and kHU (per hour) is the kinetic rate constant for hydrolysis. Factors affecting the hydrolysis constant (such as type of soil, soil moisture, and temperature) were maintained constant during each experiment in order to avoid dependence on other experimental variables in Eq. [1]. Also, if the soil water content is constant, {theta} can be removed from both sides of Eq. [1]. Since urea hydrolysis implies the net consumption of H+, an increase in soil pH is expected during relatively high rates of degradation. However, this effect is generally smoothed by the buffering capacity of the soil (Ferguson et al., 1984).

Volatilization has been described as a pseudo first-order kinetic process by various authors. For example, Hengnirun et al. (1999) proposed and tested a model for simulating ammonia in soil (VOLAT) with losses due to volatilization governed by first-order kinetics. The WAVE model (Vanclooster et al., 1996) also represents volatilization using first-order kinetics. This same approach has been used also for volatilization of pesticides assuming experimental conditions similar to those presently employed (Álvarez-Benedí et al., 1999; Tabernero et al., 2000). In our experiments, urea hydrolysis could generate small variations in pH, which could have some influence on the equilibrium between NH3 and NH4+ (and hence on the amount of N volatilization). This could be taken into account in a refined model that also simulates soil pH and then uses a pH-dependent volatilization coefficient kV. Nevertheless, this would imply the simulation of an additional soil variable (pH) and at least one additional model parameter. This refinement was not justified because of the limited pH variations related to urea hydrolysis in our experiments. Thus, the possible influence of pH variations due to urea hydrolysis was not explicitly modeled here, but indirectly lumped in the macroscopic parameter kV.

Nitrification may be important in this study because of the availability of ammonia generated by hydrolysis of urea in our samples, and because of the aerated conditions employed. Factors influencing this transformation, such as the type of soil, moisture content, aeration, or temperature, were carefully controlled in our experimental system. As in the case of hydrolysis or volatilization, nitrification can be described using constant kinetics for each experiment. Although nitrification-ammonification processes have been described using Michaelis Menten kinetics (Paulson and Kurtz, 1970), an apparent first-order kinetic process can also be used (Starr, 1983), especially when relatively high levels of ammonium are present.

Denitrification (conversion of nitrates to NO, N2O, and N2) is an important process under anaerobic conditions, and hence potentially is a major part of the nitrogen cycle in soils below certain depths. However, its relevance declines in more aerated surface horizons, especially in comparison with volatilization of ammonium produced by urea hydrolysis. Because of this, the contribution of denitrification was neglected in our study, thus further minimizing the number of possible coupled processes. This reasoning was also used to omit N mineralization and immobilization from our model.

Following the flow diagram of Fig. 1, the dynamics of N-NH4+ in our soil samples was represented by

[2]
where C is the concentration of N-NH4+ in the soil solution (mg N/cm3 solution), S is the concentration of N-NH4+ adsorbed to the solid phase (mg N/g soil), {rho} is the soil bulk density (g/cm3), kV is the volatilization rate constant [includes the NH4+ {leftrightarrow} NH3 and Henry transformation NH3(aq) -> NH3(g) constants], and kN is the nitrification rate constant. The terms in Eq. [2] represent, in this order, variations in N-NH4+ in the aqueous phase, variations in NH4+ adsorbed to the soil, losses through volatilization, losses through nitrification, and generation of NH3 from urea hydrolysis. Equation [2] contains three dependent variables (C, S, and U), which suggests that two additional equations must be formulated. Description of the sorption process and urea hydrolysis will complete the required set of equations.

Ammonia is subject to cationic exchange processes that reflect the potentially strong retention of this ion by the solid phase. This exchange appears in Fig. 1a as reversible adsorption equilibrium represented by a linear isotherm:

[3]
where KD is the distribution coefficient. In addition to the equilibrium assumption, we also tested the possibility of using first-order kinetic rate constants (kads and kdes) for the adsorption and desorption process, respectively (Fig. 1b). This variation allows us to assume the existence of nonequilibrium sorption, as well as consider situations involving irreversible adsorption of the ammonium ion. Mathematically, this nonequilibrium process is given by

[4]
The total amount of nitrogen in the system per unit mass of soil is given by the expression

[5]
where NNO3– is the concentration of N-NO3– expressed as mg N/cm3 solution. The required third equation describing urea hydrolysis is discussed below (see Eq. [6–8]).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 Nitrogen Transformation Model
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experiments
Samples studied included two different agricultural soils from the upper 15 cm of the Ap horizon in experimental plots of the Instituto Tecnológico Agrario de Castilla y León at Valladolid (Spain). The first of these was a sandy clay loam (SCL) soil, and the second a loamy sand (LS). Soil samples were air-dried and sieved to 2 mm. Table 1 summarizes the main physical and chemical characteristics of the soils.


View this table:
[in this window]
[in a new window]
  		Table 1. Selected physical and chemical properties of the soils used in this study.

 
Experiments were carried out in 500-mL flasks containing 400 g of dry soil treated with commercial urea (98.5%) at concentrations of 0.2 and 2 g urea/kg soil (corresponding to normal and high field application rates, respectively). The SCL experiments at 0.2 g urea/kg did not provide quantifiable amounts of volatilized ammonium using our experimental procedures, and therefore all SCL tests in this study were performed using soil samples treated with 2 g urea/kg soil. In order to study the effect of soil moisture on ammonia volatilization, treated soil samples were homogenized with known volumes of distilled water. Once prepared, the containers were sealed using septa with input-output gas connections and incubated in a temperature-controlled water bath for 12 d. Experiments carried out at the lower temperature were allowed to continue for three additional days. Each series of experiments was carried out for three different moisture levels (see Table 2). The effect of temperature on ammonia volatilization was studied for the SCL soil using two sets of experiments at 18° and 28°C. These values represented extremes of the temperature in agricultural soils during the summer in Central Spain. All LS trials were run at 28°C. Table 2 summarizes the various experimental conditions, including observed values for the moisture content and the urea concentration in the soil.


View this table:
[in this window]
[in a new window]
  		Table 2. Experimental conditions employed.{dagger}

 
The experimental system was designed to obtain a constant air flow rate of {approx}5 L/min over the urea-fertilized soils (Fig. 2) . Air was introduced to the flasks using a diaphragm pump which forced air to first pass through a safety trap to prevent any reflux of liquid in case of systems failure. The safety trap was followed by a system for humidification of the air, calibrated to maintain a stable level of humidity in the samples during the experiment, and a trap containing boric acid in order to prevent entry of any ammonium. All air exiting the soil-containing test flasks was passed through 200 mL of an aqueous 20 g/L boric acid solution in order to collect the volatilized ammonium. In order to increase the amount of volatilized ammonium available for detection, a group of flasks was arranged in parallel under identical working conditions. Periodic tests showed no air losses from the containers and connections in the experimental array. Emission of volatilized ammonia was determined by periodic titration of the boric acid solution at its original pH, using 0.1 M HCl (O'Halloran, 1993). Preliminary tests with ammonium sulfate showed that 100% of the volatilized ammonia was collected in this solution. At the end of the experiments, final concentrations of nitrate and ammonia were determined in the soils. For this we used high resolution liquid ionic chromatography with ultraviolet detection and a colorimetric method for nitrate and ammonia, respectively (Sparks et al., 1996).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 2. Schematic of the experimental system.

 
Analysis of Experimental Results
The simulation model was coded in VBasic and coupled to a least-squares, nonlinear Levenberg-Marquardt algorithm (Press et al., 1992). The quantity of ammonium generated by the hydrolysis of urea was first calculated for different times. Using these values, Eq. [2] and [3] (in the case of local equilibrium), and Eq. [2] and [4] (when considering kinetic sorption) were used to estimate the quantities of ammonia adsorbed, volatilized, and transformed into nitrate for each time period. A numerical discretization scheme based on finite differences was used to evaluate the material balance (Eq. [5]) for each time interval (Álvarez-Benedí et al., 1999).

Some of the observed volatilization curves were found to exhibit a time lag, which could not be reproduced with the proposed model. Two alternative modeling assumptions were used for this situation: (i) occurrence of irreversible urea adsorption, and (ii) adaptation of the microbial population for degradation of the urea. In the first case, irreversible adsorption of urea would decrease its availability for degradation, which occurred in the dissolved phase. Results obtained with this option did not lead to lag in the degradation curve, but rather only to a lower effective degradation rate due to the smaller amount of urea available. To enable the second alternative, an effective activation coefficient {alpha} was included in the first-order hydrolysis kinetics to account for the activation time:

[6]
where

[7]
In which tact is an effective activation/adaptation time for the microbial population. Integrating Eq. [6] and using Eq. [7] gives the following equation:

[8]
where U0 is the initial concentration of urea. Ammonium volatilization curves obtained with this model did show the experimentally observed time lags.

Once the volumetric moisture content ({theta}) and initial concentration of urea (U0) were established, the value of the nitrification constant, kN, was in all cases independently estimated from the measured final nitrate concentrations. Initial values for the hydrolysis constant of urea were taken from the literature (Godwin and Jones, 1991; Ma et al., 1999). Values for kV and KD or the kV, kads, and kdes parameters were subsequently fitted to the data. The activation time was estimated from an analysis of the shape of the measured sigmoidal volatilization curves.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 Nitrogen Transformation Model
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The equilibrium sorption model, schematically presented in Fig. 1a, was not able to reproduce the experimental results because of its inability to account for irreversible sorption (curve for adsorbed ammonia, Fig. 3) . Adsorption-desorption kinetics (Fig. 1b) better represented the experimental results since this process maintained solid-phase ammonium concentrations, even at very low solution concentrations as shown by the curves of dissolved and adsorbed-N in Fig. 4 . Table 3 shows values of the optimized parameters, and their corresponding 95% confidence intervals. Results for some of the SCL experiments at 18°C and 28°C and for moisture contents of 0.33 and 0.20 cm3/cm3 are shown in Fig. 5 . Both temperature and soil moisture content had important effects on the total amount of volatilized ammonium during the trials. Figure 6 shows four experiments carried out with the LS soil. Notice a similar effect of the moisture content as in Fig. 5. Differences between the two sets of curves (Figs. 5 and 6) are substantial since they compare experiments carried out with initial urea concentrations that differed by an order of magnitude.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. Evolution of the different nitrogen forms described with the model assuming local equilibrium adsorption and the presence of an activation (lag) time for urea hydrolysis. Parameters used for this example are for the loamy sand (LS) soil at 28°C and {theta} = 0.15 cm3/cm3: U0 = 9.30 mg N/cm3, kV = 0.0095/h, KD = 0.025, tact = 200 h, kHU (hydrolysis rate constant) = 0.03/h. The value of kN (nitrification rate constant) was set to 0.001/h in order to reproduce the nitrification curve in the graph.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Evolution of the different nitrogen forms described with the model assuming ammonium adsorption-desorption kinetics and the presence of an activation (lag) time for urea hydrolysis. Parameters used for this example are for the loamy sand (LS) soil at 28°C and volumetric moisture content ({theta}) = 0.15 cm3/cm3: initial concentration of urea (U0) = 9.30 mg N/cm3, volatilization coefficient (kV) = 0.0095/h, adsorption rate coefficient (kads) = 0.055/h, desorption rate coefficient (kdes) = 10–5/h, effective activation/adaptation time for the microbial population (tact) = 200 h, and hydrolysis rate constant (kHU) = 0.03/h. The value of kN (nitrification rate constant) was set to 0.001/h in order to reproduce the nitrification curve in the graph.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Values of the optimized model parameters and corresponding 95% confidence intervals.{dagger}

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Experimental results for volatilized ammonia in the Sandy Clay Loam (SCL) soil at 28°C (circles) and 18°C (squares). Open symbols represent experiments carried out at volumetric moisture content ({theta}) = 0.20 cm3/cm3; solid symbols represent experiments at {theta} = 0.33 cm3/cm3. Lines show model response with parameters presented in Table 3.

 


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 6. Experimental results for the loamy sand soil at 28°C using two different initial levels of urea. Squares represent volatilized amounts of ammonia for experiments using an initial urea concentration of 2 g/kg, and circles represent volatilized amounts of ammonia for experiments using an initial urea concentration of 0.2 g/kg. Open symbols show experiments carried out at volumetric moisture content ({theta}) = 0.15 cm3/cm3, and solid symbols represent experiments at {theta} = 0.22 cm3/cm3. Lines show model response using the parameters presented in Table 3.

 
The estimated volatilization coefficients were found to increase with decreasing soil moisture contents, as shown by the numerical values in Table 3. This dependence of volatilization on soil moisture was found to be statistically significant for the LS soil (experiments at {theta} = 0.15 cm3/cm3 and {theta} = 0.22 cm3/cm3). Temperature increased volatilization, especially at the higher soil moisture contents, although the differences between the volatilization constant (kV) at the 95% confidence level intervals were significant only for the experiments carried out at {theta} = 0.33 cm3/cm3 (28 and 18°C). Also, the volatilization coefficients of the LS soil were higher, although not sufficiently so as to justify the observed differences in volatilization between the two soils. An apparent dependence of the volatilization coefficient kV on the available ammonium concentration is noticeable from the data in Table 3. This table shows that the values kV are systematically lower for the trials with lower initial concentrations of urea in the LS soil, for similar temperature and soil moisture conditions. This dependence may complicate the representation of the processes at the field scale. The results suggest that a model based on first-order volatilization kinetics may not be realistic for practical situations. Also, since the volatilization process is modeled by means of a lumped parameter, kV will depend on the prevailing soil properties. Lumped coefficients of this type hence should be viewed as being site-specific parameters.

Regarding ammonia sorption, a large number of studies (e.g., Wagenet et al., 1977) exist which assume reversible, linear equilibrium sorption with distribution coefficients (KD) ranging between 1 and 10 (mg N/g soil)/(mg N/cm3 solution). However, the equilibrium model between the dissolved phase and the adsorbed phase (Fig. 3) did not accurately represent our data. Representing adsorption-desorption as a kinetic process gave a better description of our experimental results. The values in Table 3 show that almost no desorption occurred during the experiments (coefficients approaching zero). Exceptions were the LS experiments with the lower initial urea concentrations. For the same initial concentrations of urea and the same temperature (28°C), ammonium adsorption on the SCL soil was higher than adsorption on the LS soil. Differences were statistically significant in all cases for the adsorption rate coefficient kads. The adsorption coefficients showed no clear dependency on soil moisture content (only the SCL experiments carried out at the lower temperature showed less adsorption with soil moisture). To evaluate net adsorption, it is necessary to compare the ratio of the kinetic adsorption and desorption coefficients. As expected, the greatest net adsorption in all cases was observed for the clay soil because of its greater specific surface area. Al-Kanani et al. (1991) obtained first-order coefficients for volatilization of 0.013/h, which was of the same order of magnitude as in our study (Table 3). Small differences may be due to the better control of evaporation of water in our experimental system. Al-Kanani et al. (1991) also showed a decrease in volatilization upon increasing the proportion of clay in their soils, which they attributed to adsorption of ammonia. Our experimental results, further supported by the proposed conceptual model, confirmed their findings. Also, the role of nearly irreversible adsorption of ammonium in our soils was identified.

The activation or lag time for urea hydrolysis (Eq. [8]) constituted a critical parameter for accurate representation of the experimental results. We note that the employed model includes first-order kinetics as a specific case when tact = 0. This limiting case did not, however, provide satisfactory results for our data. The tact values in Table 3 indicate that lower temperatures require longer times for microbial adaptation. Also, the lag time was found to strongly depend upon the initial concentration of urea applied to the soil.

Hongprayoon et al. (1991), among others, previously showed that adsorption of urea was minimal (a distribution coefficient KD of 0.03 for a medium texture soil), which is in agreement with the limited improvement we obtained when urea sorption was included in our conceptual model. Hongprayoon et al. (1991) observed that the urea hydrolysis constants increased with incubation time from 0.0036 to 0.288/h for flooded soil columns, and from 0.00072 to 0.0144/h in the supernatant solution. It is likely that a first-order kinetic description of their data would also have revealed the need for an activation or lag time.

Our values for the first-order nitrification rate coefficient for the SCL soil were very close to those presented by Wagenet et al. (1977). No decrease in the nitrification rate coefficients with an increase in soil moisture (anaerobic conditions) was evident. A possible explanation for this is that the continuous aeration of our systems and the relatively low volume of soil in each flask provided sufficient aeration in all of our trials. We note, however, that nitrification decreased at the lower temperature. Finally, no nitrate was detected in the LS soil (kN = 0) at the end of the experimental test periods.

We emphasize that extrapolation of the results obtained in this study to different spatial and temporal scales (e.g., the field scale) may require several other considerations. Among these, the mineralization of N is perhaps the most important. This process is often associated with the decomposition of organic matter, which presumably is a function of the C/N ratio. According to Johnsson et al. (1987), who compared several N transformation models, the decomposition rates of different fractions of organic matter (typically three) should be considered. Schmied et al. (2000) obtained mineralization constants in the order of 10–5/d, three to four orders of magnitude lower than their nitrification rate coefficients. Also, denitrification may become much more important for deeper soils and anaerobic conditions, as compared with the experiments discussed in this paper.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 Nitrogen Transformation Model
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The present study attempted to describe the main processes which occur upon the application of urea, including sorption and volatilization of ammonia. We experimentally determined the release of volatilized ammonia, and the initial and final contents of ammonium and nitrate in two soils, for different temperatures, moisture contents, and initial concentrations of urea. The results were analyzed in terms of a simplified conceptual model accounting for the coupled processes of urea degradation, ammonium adsorption, nitrification-ammonification reactions, and volatilization of ammonia. We tested several kinetic equations for the different processes, also in attempts to obtain coefficients for future use in more complex general-purpose models.

The hydrolysis of urea was represented using first-order kinetics with an activation or lag time to account for microbial dynamics. The activation time was found to increase at relatively high urea concentrations and low temperatures. The kinetic volatilization rate coefficient increased with temperature and decreased with soil moisture, being higher for the coarse-textured soil. A dependence of this coefficient on initial urea concentration was also observed, thus casting doubt on the appropriateness of first-order kinetics in representing larger-scale volatilization. Concerning ammonium adsorption, poor results were obtained by applying an equilibrium distribution coefficient. The process of adsorption was found to be essentially irreversible. The incubation experiments did not show any evidence of nitrification (no NO3– production) for the LS soil. On the other hand, appreciable rates for this process were observed for the SCL soil. First-order kinetic rate coefficients for nitrification were found to be of the same order of magnitude as those reported in the literature, with slightly higher values at higher temperatures.

The methodology adopted in this study should provide a convenient way of estimating rate data for N transformation and coupled sorption processes sorely needed in N fate and transport studies. Rate data presented in this work quantify the effects of temperature, soil moisture, and initial concentration of applied urea on the dynamics of urea hydrolysis, nitrification, and ammonia sorption and volatilization.


   ACKNOWLEDGMENTS
 
This study was financed by the Instituto Nacional de Investigación Agraria y Tecnología de los Alimentos (INIA), project CAL-01-029 and by the Junta de Castilla y León, project JCYL 023/03. The authors acknowledge Dr. Louis DiSalvo for his advice with the English language. Special thanks are due also to the editorial staff of Vadose Zone Journal for their careful editing of this paper.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Nitrogen Transformation Model
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





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 Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Rodríguez, S. B.
Right arrow Articles by Álvarez-Benedí, J.
Related Collections
Right arrow Batch Studies
Right arrow Kinetics
Right arrow Nutrient Management

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome