Agronomy Journal 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 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 HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Agricola
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Related Collections
Right arrow Wheat
Right arrow Crop Models
Right arrow Soil Fertility and Productivity
Published in Agron. J. 96:1050-1057 (2004).
© American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA

FERTILIZER MANAGEMENT

The Balance Sheet Method as a Conceptual Framework for Nitrogen Fertilization of Wheat in a Pampean Agroecosystem

R. Alvareza,*, H. S. Steinbacha, S. M. Grigeraa, E. Cartiera, G. Obregona, S. Torria and R. Garcíab

a Facultad de Agronomía, Univ. de Buenos Aires, Av. San Martín 4453 (1417), Buenos Aires, Argentina
b EEA Pergamino INTA, CC 31 (2700), Pergamino, Argentina

* Corresponding author (ralvarez{at}agro.uba.ar).

Received for publication March 5, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Yield response curves to fertilizer are needed to determine N rates for wheat (Triticum aestivum L.) in the Argentine Rolling Pampa. Relationships between yield and available N have not been developed for this region. The balance sheet method can be use to define N fertilization strategies when yield data are not available. The shortcoming of this method is that profitability cannot be assessed. This study was conducted to: (i) determine N requirements for wheat in the Rolling Pampa, (ii) develop algorithms for predicting crop available N, and (iii) develop a model for making N fertilizer recommendations. Fifty-eight field plots were selected with contrasting soil and management conditions. Soil mineral N content and N in residues of preceding crop were determined at initial and harvest stages of wheat from 1997 to 1999. Plant N accumulation was measured, and N mineralized from organic soil pools was calculated using a mass balance approach. Nitrogen available in the mineral and organic soil pools limited N uptake and yield. As available N increased, crop use efficiency and agronomic efficiency decreased. The model prediction agreed well with measured plant N uptake (R2 = 0.82) and explained 51% of yield variability. The model estimated the impact of increasing fertilizer N on yield in soils differing in N mineralization capacity. The economic optimum N requirement in the Rolling Pampa is 140 kg ha–1 for soil having an average capability to supply N from organic pools and at a mean fertilizer price/grain price ratio. This optimum increased or decreased 30 kg ha–1 across the range of N mineralization rates observed. Despite the empirical nature of the algorithms developed, the concepts may be used in other regions where response functions to N are not available.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
SOUND N FERTILIZATION strategies have agronomic, economic, and environmental implications. Farmers desire high yields and optimal economic returns. Efficient use of N fertilizers that maintain low residual soil nitrate levels reduce the potential for nitrate leaching to groundwater. Yield response functions to applied N have been used extensively for determining optimal fertilizer rates (Black, 1993, p. 392–404; Colwell, 1994). Response models allow quantification of yield increments as a function of increased fertilizer N and determination of N rate for maximum profit (Schlegel et al., 1996). Since yield functions are site specific, soil characteristics such as residual nitrate (Vanotti and Bundy, 1994a), potential N mineralization (Johnson, 1991), or N credits for previous crops (Bullock and Bullock, 1994) are included to improve fertilizer recommendations. Developing a yield curve requires multiple years of agronomic and soils data for each site. This demand restricts the use of the yield curve approach where information is not available.

The balance sheet method is an alternative for defining N fertilization strategies under scenarios where yield data are scarce (Vanotti and Bundy, 1994b). The balance sheet method integrates crop production factors and soil N availability factors into an overall N evaluation (Meisinger, 1984). Yield goal and crop N requirement are estimates of N demand while N release during residue decomposition and soil organic matter mineralization are combined with measured soil mineral N to estimate N supply. A mass balance is developed to provide the crop N requirement for an expected yield using soil N, N mineralized during the growing season, and N fertilizer applied (Neeteson, 1995). Uncertainties in this approach include N losses from the soil–plant system due to leaching, denitrification, or volatilization; N immobilized in microbial biomass; and the amount of unutilized N remaining in the soil at harvest. Estimates of these uncertainties can be made but their experimental determination is extremely difficult (Meisinger, 1984; Neeteson, 1995). The balance sheet equation for N in the crop at harvest is (see Table 1 for definition of variables)

[1]


View this table:
[in this window]
[in a new window]
  		Table 1. Abbreviations, definitions, and units of the variables included in the model.

 
Crop N demand for an attainable yield goal is supported by changes in the magnitude of soil N sources minus losses. Since differences in soil humified organic N between sowing and harvest change little and are difficult to determine, and N from straw decomposition is only partly available to the following crop (Parton et al., 1993), the balance sheet model can be rearranged to (see Table 1 for definition of variables)

[2]

The term (NHI – NHH – NL) represents the difference between net soil N mineralized and N loss. It may be calculated by difference if N crop, NMI, NMH, NRI, and NRH are measured. This term is similar to apparent N mineralization (Engels and Kuhlmann, 1993), differing only because N release from decomposing crop residue is not included.

A major shortcoming to the mass balance approach is that it does not include a yield response or a product quality curve, and profit cannot be assessed using fertilizer/product prices ratios (Makowski et al., 1999). Additionally, N rates developed using the balance methods are usually overestimates as farmers tend to overestimate expected yield (Black, 1993, p. 392–404). This can lead to higher-than-optimal N rates resulting in negative economical and environmental consequences (Vanotti and Bundy, 1994a, 1994b).

The Argentine Humid Pampa region is one of the world's best regions for grain crop production (Satorre and Slafer, 1999). Climate is humid temperate with Mollisols as the predominant soil (Hall et al., 1992). The Rolling Pampa in the northeast portion covers about 10 Mha, and wheat is grown on 1.5 Mha annually. This area is the main agricultural region of the country, and adequate N inputs are essential for obtaining high wheat yields. In this region, site-specific response functions to N fertilization are unavailable, and farmers fertilize crops without the benefit of soil testing or profitability estimates. The balance sheet method has been used to estimate fertilization requirements by some agronomists, but the agreement between model predictions and observed yields was usually low (R2 = 0.2) (Alvarez et al., 2000). Our aims were to: (i) determine wheat N requirements in relation to attainable yields in the Rolling Pampa, (ii) develop algorithms for predicting crop available soil N, and (iii) integrate this information in a model for fertilizer recommendation.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
The Rolling Pampa (32 to 35°S and 58 to 61°W) is an extensive plain with slightly rolling relief and long slopes of 0.5 to 1%. Mean annual rainfall is 940 mm (1900–1990 period), with 35% received during the wheat growing cycle of June to November. Mean annual temperature is 17°C and ranges from an average monthly high of 24°C in January to an average monthly low of 10°C in July. Predominant soils are Typic Argiudolls developed over aeolian sediments, with illite as the most common clay mineral (Hall et al., 1992). Soil texture is usually silty clay loam with {approx}250 g kg–1 clay. Spring wheat is sowed in winter (June–July) and harvested at the beginning of summer (December).

During the growing seasons of 1997, 1998, and 1999, 58 field plots were selected with contrasting soil and management conditions. Twenty-six sites received some degree of tillage (moldboard plow, chisel plow, or harrow disk), and 32 were no-tilled. The previous crops were soybean [Glycine max (L.) Merr.] at 26 sites and corn (Zea mays L.) at 32 sites. All fields received P fertilization at sowing (60–200 kg diammonium phosphate ha–1), and many of them received N fertilization usually at sowing or immediately after (55–210 kg urea ha–1). At a wheat growth stage of two expanded leaves (beginning of August), experimental plots of about 400 m2 were delimited in the fields. Soil samples (0- to 30-cm depth) were collected by composing 6 to 12 cores sampled on the row and at different distances from the row. Surface residue was collected from an area of 625 cm2 in 6 to 12 sites too. Ammonium and nitrate were determined on fresh soil samples by steam distillation (Mulvaney, 1996). Incorporated crop residue was isolated by wet sieving on a 500-µm mesh sieve. Organic C and N were determined by wet digestion (Amato, 1983; Bremner, 1996), extractable P by the Bray method (Kuo, 1996), and pH in a soil/water ratio 1:2.5. Total N content of residues was analyzed by the Kjeldahl digestion (Jackson, 1960, p. 183–204). Soil bulk density was determined gravimetrically using 4-cm-diam. cores taken from 0- to 30-cm depth. Mineral N in the 0- to 60-cm depth was estimated using the concentration in the upper 30 cm (Alvarez et al., 2001). Soil N mineralization capacity was determined using dry soil samples passed through a 500-µm mesh sieve. Samples of 100 g of dry soil were incubated in 400-mL flasks at 30°C and 50% soil water-holding capacity for 17 d. Ammonium and nitrate were determined at the end of incubations and results expressed on a per-hectare basis. Carbon in the soil light fraction was determined by centrifuging (1000 x g) 5 g of soil in 30 mL of a bromoform-ethanol mixture having a density 2 g mL–1 (Alvarez and Alvarez, 2000). At crop physiological maturity (December), soil and residue sampling was repeated as describe above. Aboveground wheat biomass was determined using 1 m of row at 6 to 12 locations within each plot. Wheat roots and buried decomposing residue were separated by hand during wet sieving using a 500-µm mesh sieve. Grain and straw were separated, weighed, and total N in plant materials determined by the Kjeldahl method (Jackson, 1960, p. 183–204). The N factor (kg of N required per Mg grain) was calculated as the ratio between N in plant biomass (aboveground organs + roots + rhizodeposition) and yield (Makowski et al., 1999). All determinations were performed in duplicate and results expressed on an oven-dry basis. Rainfall was recorded in each plot, and air temperature was recorded within 50 km of the plots.

Nitrogen released during residue decomposition and available for crop utilization was estimated as the difference between N in buried and surface residue at the two-leaf stage and maturity, assuming that 30% of the N not found in surface residue and 70% of the N in buried residue at maturity became microbial biomass (Parton et al., 1993). Total N uptake was the sum of N in aboveground biomass, roots, and rhizodeposition. We assumed this pool was 6% of aboveground + root N for wheat at harvest (Merbach et al., 1999). The term (NHI – NHH – NL) was calculated by difference for each field. Data were analyzed by two-sample t test, linear regression, and correlation methods. Multiple linear regressions were performed stepwise using R2 as the decision criterion for selection of the best functions. Inclusion–exclusion of variables was accepted at P = 0.05, tested by the F test. Different transformations of variables were performed, and some empirical nonlinear models were also fitted (Colwell, 1994). Categorical variables like tillage system and previous crop were tested in the models as dummy variables. Rainfall and temperature for different periods during wheat growing season were also tested as independent variables. Combination of regression equations was employed for estimating the impact of soil and environmental factors on wheat yield. Regressions were tested by the F test and accepted at P = 0.01. Observed vs. predicted values were compared by linear regression. Intercepts and slope values were tested comparing with 0 and 1, respectively, by the t test.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Soil properties, rainfall, and yield exhibited high variability across sites (Table 2). When compared with the 1900–1990 average, rainfall during the growing season was greater in 1997 (+58%), similar in 1998 (+3%), and lower in 1999 (–20%). Tillage system and previous crop had significant effects (P = 0.01) on residue dry matter (DM) and N concentration at the beginning of the wheat growing season. Field plots managed under tillage had lower residue levels than no-till plots (7.0 vs. 12.6 Mg DM ha–1), but N concentration in residue was similar (1.01 vs. 1.09%). Residue DM was lower when soybean was the previous crop compared with corn (7.5 vs. 11.2 Mg DM ha–1). Soybean residue had greater N concentration than corn residue (1.22 vs. 0.90%). Management did not affect other site properties such as organic C and N, light fraction C, or N mineralized in vitro.


View this table:
[in this window]
[in a new window]
  		Table 2. Soil, climate, and yield variability in wheat crops (n = 58).

 
The relationship between N crop and aboveground DM production was curvilinear (Fig. 1A) . The relationship between grain yield and crop N was also curvilinear, with maximum yield above 300 kg N crop ha–1 (Fig. 1B). The Mitscherlich equation provided a better fit (R2 = 0.38) than the linear-plateau (R2 = 0.30) and quadratic functions (R2 = 0.33). As crop N increased, more N was allocated to vegetative parts and less to grain, changing the wheat straw/grain ratio. As rainfall increased, the straw/grain ratio also increased. Crop N and rainfall, combined in a multiple-regression equation, explained the major portion of the wheat straw N/grain N ratio variability (R2 = 0.79). The consequence of these changes was that under the most fertile and wet scenarios, wheat was less efficient in producing grain and the N factor increased (Fig. 2A ; Eq. [3]). On a DM basis, the N factor ranged from 28 to 82 kg N Mg–1 grain, with a mean value of 45 kg N Mg–1 grain. Variability of the N factor was not related to yield, nor to other soil or environmental characteristics. Grain N concentration was positively correlated with crop N (R2 = 0.32) but not with yield. When the N factor was regressed against grain N, a good fit was obtained using the linear-plateau model (Fig. 2B).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1. (A) Aboveground biomass and (B) yield of wheat as a function of N in the crop.

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2. Relationships (A) between observed and estimated values for N factor and (B) between N concentration in grain and N factor. For definition of variables, see Table 1.

 

[3]

As the plant accumulated more N by unit of grain produced, N concentration in grain increased up to 44 kg N kg–1 DM grain ratio. Above this ratio, no change was found.

Pool size changes of the balance sheet (Eq. [2]) components could be predicted well using regression. The term (NMI – NMH) was very closely related to NMI (Fig. 3A ; Eq. [4]). A saturation plateau was not detected up to 250 kg NMI ha–1.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 3. (A) Changes in soil mineral N content during the growing season as a function of initial mineral N level and (B) relationship between N crop and N crop + NMH. For definition of variables, see Table 1.

 

[4]

At harvest, NMH averaged 37 kg N ha–1, with a range of 19 to 76 kg N ha–1 and was not correlated with NMI. Increases in residual nitrate were not detected in these experiments when NMI was high. At harvest, crop N represented 70 to 90% of the total N available for crop uptake (N crop + NMH) (Fig. 3B).

Mineralization–immobilization of N in decomposing plant residues was governed by mass and composition of residues (Fig. 4A ; Eq. [5]). As mass or N content increases, liberation of N also increased.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 4. (A) Observed vs. predicted N liberated from decomposing plant residues and (B) observed vs. predicted N apparently produced by mineralization from the soil organic matter. For definition of variables, see Table 1.

 

[5]

In most cases, residues provided net mineralization of N with no period of immobilization. In three cases (5%), immobilization was of agronomic significance (>10 kg N ha–1). Net mineralization averaged 16 kg N ha–1 regardless of whether corn or soybean was the previous crop. Corn residues had lower N concentration but double the mass of soybean residues. Residues liberated 4 kg N ha–1 under tilled and 32 kg N ha–1 under no-tillage systems (P = 0.01). This was a consequence of the greater levels of residues remaining in the conservation tillage parcels. Consequently, the f factor (Table 1) of Eq. [2] ranged between 0.30 and 0.64, depending on the fraction of total residue incorporated at each site. The term (NHI – NHH – NL) could not be predicted by a simple equation, and many variables must be combined to attain a good fit (Fig. 4B; Eq. [6]).


[6]

The mean for this N pool was 34 kg N ha–1 and ranged from –80 to +220 kg N ha–1. This component may be a source or a sink for N. When NMI, clay + silt content, pH, or residue mass increased, the (NHI – NHH – NL) value decreased. Conversely, in vitro N mineralization capacity of the soil and rainfall increases were associated with a greater contribution of this pool to crop N. At sites where soybean was the previous crop, (NHI – NHH – NL) doubled when compared with sites where corn was the previous crop (47 vs. 23 kg N ha–1), but the effect was not significant. Under tillage, (NHI – NHH – NL) was greater (P = 0.05) than under no-tillage (48 vs. 17 kg N ha–1). This seemed to be an indirect effect as plant residues and NMI were greater under no-tillage and these two factors had a strong influence on this component. In multiple-regression equations, inclusion of the dummy variable tillage did not improve the correlation coefficient. Differences in NMI between tillage systems (180 vs. 130 kg N ha–1, P = 0.01) resulted from contrasting fertilization strategies as farmers usually applied more N to no-till fields. As immobilization of previous crop residues was uncommon, negative values of the (NHI – NHH – NL) component may be associated with immobilization in soil microbial biomass and greater N losses from the soil.

Results showed that crop N requirement for grain production is regulated by environmental and soil conditions and that N availability from soil organic matter is also affected by these factors. Consequently, the use of fixed values for the N factor or the soil mineralization capacity would produce errors in the estimation of N rates for wheat under Pampean conditions using the balance sheet method. Combining Balance Sheet Eq. [2] and Regression Eq. [3], [4], [5], and [6] resulted in an empirical model for predicting wheat yield under contrasting scenarios of soil fertility, management, and rainfall (Fig. 5) . Using experimentally determined soil and environmental variables, the terms (NMI – NMH), [(NRI – NRH)f], and (NHI – NHH – NL) can be estimated and used to simulate crop N, the N factor, and yield under contrasting scenarios. The model predicted crop N well (Fig. 6A) and explained 51% of yield variability (Fig. 6B). The regression between observed and predicted yields agreed well with the 1:1 line. Yield estimations of the model were within 20% of observed yields in 85% of the cases (RMSE = 0.64 Mg DM grain ha–1). The model can be used for predicting wheat yield response to changing soil mineral N contents under contrasting fertility and rainfall scenarios and for calculating variation in the agronomic efficiency as mineral N increases (Fig. 7) . Agronomic efficiency decreased at higher mineral N levels and was lower when soil N mineralization capacity increased. Plant residue DM or N concentration did not affect agronomic efficiency. Rainfall, texture, pH, and N mineralization rate in incubation tests affected agronomic efficiency though their effects on the soil mineralization capacity. Rainfall averaged 330 mm from 1900 to 1990 for the June-to-November period and had a normal distribution. Fertilizer recommendations may be made assuming average rainfall conditions. Soil texture usually does not change, and soil pH changes slowly. These properties may be treated as constant for many years. Using these assumptions, the initial mineral N content of the soil, and N mineralization capacity for a specific soil, fertilizer recommendations may be developed. An economic optimum targeted N supply can be determined using the fertilizer price/grain price ratio. The mean price ratio for wheat has been 7 to 8 $ kg–1 N/$ kg–1 grain in the Pampas. This means that the average mineral N level of the soil must not exceed a targeted N supply where agronomic efficiency became lower than 7 to 8 kg grain kg–1 mineral N. Under medium fertility and rainfall scenarios, this N requirement is 140 kg N ha–1, and it increases or decreases about 30 kg N ha–1 as soil mineralization capacity declines or increases.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Diagram of the model developed to predict wheat yield in the Pampean agroecosystem. For definition of variables, see Table 1.

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 6. Relationship between observed vs. predicted (A) N crop and (B) yield. For definition of variables, see Table 1.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 7. Agronomic efficiency as function of initial mineral N content of the soil for a typical rainfall scenario and three soil mineralization capacity levels. For definition of variables, see Table 1.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The Mitscherlich model described the relationship between biomass production or yield and N uptake for Pampean conditions. As N absorbed by wheat increased, the DM straw/DM grain or N straw/N grain ratios also increased. As rainfall increases, DM straw/DM grain and straw N/grain N ratios also increase. Wheat N uptake was linearly related to N rate at low values and followed a hyperbolic relationship at high rates (Makowski et al., 1999) while yield was a curvilinear function of N uptake (van Keulen and Stol, 1991). These relationships show that the crop is less efficient in producing grain at high crop N levels and use of a fixed value for the N factor in the balance equation will introduce bias in the determination of the crop N requirement. The mean value for the N factor of 45 kg N Mg–1 DM grain obtained in our work was somewhat greater than previous reports, partially because previous studies neglected rhizodeposition. Reported average values for the N factor ranged from 35 to 42 kg N Mg–1 DM grain (González Montaner et al., 1997; Makowski et al., 1999). These estimates were developed assuming that N in roots accounts for 20% of that accumulated in aboveground biomass (González Montaner et al., 1997). In our experiments, N in washed roots averaged 22% of total N in aboveground biomass. Addition of N via rhizodeposition increased belowground N allocation to 29%. The linear-plateau function described the relationship between grain N concentration and the N factor, in the range of 30 to 80 kg N crop kg–1 DM grain. This function indicates that at high crop N or low yields, N in grain changes little with the N factor. Above a value of 44 kg N Mg–1 DM grain, N concentration in grain did not change as the N factor increased. Consequently, N concentration in grain was not susceptible to yield dilution effects when the crop utilized greater than 44 kg N Mg–1 DM grain. Site x year interaction effects had little influence on this relationship. A linear relationship had been described previously between the N factor and grain N concentration in the range 20 to 60 kg N crop kg–1 DM grain (Makowski et al., 1999).

Equation [4] estimates residual N under the contrasting scenarios of this study. The model predicted a slow rise from 30 to 40 kg NMH ha–1 when NMI increased from 50 to 250 kg N ha–1. This suggests that wheat crops were N limited and nearly all mineral N was utilized under these conditions. When utilization of soil N is high, there is little potential for negative environmental consequences. Fertilizer N recovery usually decreases when N rates increase due to crop limitations for uptake and this results in higher residual nitrate levels under high fertilization scenarios (Schlegel et al., 1996). In other agroecosystems, the plateau-linear model has been successfully used to describe residual nitrate, with reported values for the break point of the model ranging from 60 kg N ha–1 (Johnson and Raun, 1995) to over 200 kg N ha–1 for wheat crops (Makowski et al., 1999). Nitrogen uptake by wheat is about 85% complete at flowering stage (Fisher, 1993). Thereafter, uptake decreases as a consequence of declining root activity and exhausted available mineral N. Wheat flowering usually occurs during October in Pampean agroecosystems. Mineralization of N from organic matter would accumulate in soil as uptake by the crop decreased. Integration of field measurements of organic matter mineralization indicated that 0.6% of N in the A horizon is mineralized from late November to December (Alvarez, 1999). This represented an average of 40 kg N ha–1 for the soils in our study and agrees well with observed values of NMH. Consequently, NMH values reported in this study appear to be the result of postflowering soil N mineralization.

In our study, we observed a net release of N during residue decomposition even with crop residue N concentrations lower than 20 g kg–1. Dynamics of the N immobilization–mineralization process are mainly regulated by the crop residue N concentration, with a net release of N during the initial stages of decomposition if the concentration is above 20 g N kg–1 and immobilization below this concentration (Seneviratne, 2000; Trinsoutrot et al., 2000). Mineralization results in an increase in soil solution N whereas immobilization results in a decline in soil mineral N available to the crop due to microbial biomass demands for limited available N (Nicolardot et al., 2001). As a consequence, when a leguminous crop precedes a nonleguminous crop, a N credit is often realized because of the greater N content (N'Dayegamiye and Tran, 2001). The warm and humid environment of the Rolling Pampa ecosystem leads to very rapid decomposition of crop residues (Alvarez et al., 1991). If an initial phase of immobilization occurred, it was not detected, and no differences among previous crops were observed as sources of N for wheat.

Many variables must be included to explain variation in the term (NHI – NHH – NL) of the balance equation. The effects of some variables on apparent mineralization are well understood while others require further elucidation. The protective effect of fine particles on soil organic matter is well known (Parton et al., 1993), and lower in situ N mineralization rates in fine-textured soils has been observed (González Montaner et al., 1997). Also in vitro tests for mineralization have been used successfully in the past as predictors of the N-supplying capacity of the soil (Rohde, 1996). In Pampean soils, a negative relationship had been observed between in situ N mineralization and initial soil mineral N content (González Montaner et al., 1997). Interactions between inorganic soil N and organic matter mineralization had been included in models for predicting crop response to N and explain inhibition of net mineralization as mineral N rises (Johnson, 1991). The inhibition effect on net mineralization may result from increased immobilization of N in microbial biomass or organic pools or to greater N losses from the soil–plant system, which can result in low or negative values for net N mineralization when N rates are very high (Engels and Kuhlmann, 1993; Blankenau et al., 2000). In fields where more straw was present at the beginning of the wheat growing season, (NHI – NHH – NL) was lower. Immobilization in crop residue was rare, and an indirect effect of residue is more probable. At greater straw levels, more N was released from residues, and this could have an inhibitory effect on net mineralization. It is also possible that immobilization occurred in the soil light fraction or in the <500-µm size particulate soil organic matter. These fractions are sinks for N under conditions that favor immobilization (Recous et al., 1999; Whalen et al., 2000). In our plots, the soil light fraction was weakly correlated with crop residue DM content (R2 = 0.14). Rainfall enhanced (NHI – NHH – NL) possibly because at greater rainfall level, more aboveground biomass increased crop N absorption (R2 = 0.32) and decreased soil mineral N content with a subsequent increase in mineralization. The effect of pH on (NHI – NHH – NL) is unclear.

Application of N fertilizers can increase N losses from the soil–plant system as fertilizer N is sometimes more susceptible to volatilization and leaching than indigenous N (Meisinger, 1984). To meet the crop N requirement, it may be necessary to apply more fertilizer N than the difference between the N requirement predicted by our model using the fertilizer/grain prices ratio and initial soil mineral N levels. Nitrogen rates can be adjusted by an efficiency factor (E) that accounts for losses of fertilizer N after application at the appropriate fertilizer/grain prices ratio [e.g., N rate = (140 – NMI)/E]. Quantifying gaseous losses and leaching of N in wheat crops in the Pampas has shown that losses from these processes for fertilizer N varied between 0 and 15% depending on the type of fertilizer, incorporation technique, and rate (Alvarez, 1999). These results lead to E factors varying from 1 to 0.85. Nitrogen rates can be calculated using a combination of the model based on the balance sheet method presented here and the efficiency factor.


   ACKNOWLEDGMENTS
 
We acknowledge Dr. B. Wienhold for his help in the preparation of the manuscript. This work was granted by the projects UBACYT AG01 and TG01 from the University of Buenos Aires.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
L. M. Arregui and M. Quemada
Strategies to Improve Nitrogen Use Efficiency in Winter Cereal Crops under Rainfed Conditions
Agron. J., February 26, 2008; 100(2): 277 - 284.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
S. Takahashi, M. R. Anwar, and S. G. de Vera
Effects of Compost and Nitrogen Fertilizer on Wheat Nitrogen Use in Japanese Soils
Agron. J., June 26, 2007; 99(4): 1151 - 1157.
[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 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 HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Agricola
Right arrow Articles by Alvarez, R.
Right arrow Articles by García, R.
Related Collections
Right arrow Wheat
Right arrow Crop Models
Right arrow Soil Fertility and Productivity

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science Vadose Zone Journal
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