Agronomy Journal 92:303-308 (2000)
© 2000 American Society of Agronomy
SOIL MANAGEMENT
Wheat Nitrogen Use Efficiency in a Bed Planting System in Northwest Mexico
Agustin Limon-Ortegaa,
Kenneth D. Sayrea and
Charles A. Francisb
a CIMMYT, A.P. 6-641, Mexico D.F. 06600, Mexico
b Univ. of Nebraska–Lincoln, 225 Keim Hall, Lincoln, NE 68583-0949 USA
k.sayre{at}cgiar.org
 |
ABSTRACT
|
|---|
Wheat (Triticum aestivum L.) in the Yaqui Valley, northwest Mexico, is planted as a winter crop using a raised-bed, furrow-irrigated system and high fertilizer N rates. Wheat residues are usually burned before planting maize (Zea mays L.) as a summer crop. The N use of wheat planted following conventional tillage using a raised-bed system (CTB) incorporating both wheat and maize residues was compared with wheat planted using permanent raised beds (PB) under four residue management treatments: all straw (wheat and maize) left as stubble, straw partly removed (maize residues removed; wheat residues retained), all straw removed, and all straw burned. Each wheat plot was split into seven N fertilizer (Nf) treatments: five applied at planting (0, 75, 150, 225, and 300 kg ha-1) and two at the 1st node stage (150 and 300 kg ha-1). Maize received a uniform Nf application of 150 kg ha-1. The N use efficiency of wheat with 150 kg Nf ha-1 at the 1st node stage was superior to basal applications at the same rate. Permanent bed–all straw left as stubble and PB–all straw burned had the highest average wheat grain yields (5.57 and 5.52 Mg ha-1, respectively), N use efficiency (28.2 and 29.1 kg grain kg-1 of N supply, respectively), and total N uptake (133 and 137 kg ha-1, respectively). Total N uptake for 150 and 300 kg Nf ha-1 at the 1st node stage was 14 and 8% greater, respectively than at planting. In most tillage–straw treatments, 21% of the difference in wheat grain yields was due to the N supply component at low N rates; at high N rates, 97% was due to N use efficiency.
Abbreviations: CTB, conventional tillage, raised-bed system • Gw, wheat grain yield • Nf, nitrogen fertilizer • Ns, nitrogen supply • Nt, total nitrogen uptake • PB, permanent raised beds
|
INTRODUCTION
|
|---|
THE YAQUI VALLEY is located in the state of Sonora in northwest Mexico. Wheat production systems in the Yaqui Valley are characterized by burning crop residues after harvesting and heavy N fertilizer applications at or prior to planting (Meisner et al., 1992). According to Sowers et al. (1994), the application of high N rates may result in poor N uptake and low N use efficiency due to excessive N losses. Inefficient N use contributes to greater use of energy resources, increased production costs, and possible pollution of water by nitrates (Sharpe et al., 1988). The use of a permanent raised-bed system is an alternative practice in wheat that allows for the timing of N fertilizer application to increase N use efficiency and to lower production costs. A permanent bed system allows the bed to be reused for each succeeding crop. A superficial cultivation to reshape the sides of the beds is necessary before planting the next crop following harvest of the previous crop. The term bed is used instead of ridge, to differentiate this system from ridge tillage as practiced in the U.S. Corn Belt. Bed planting systems using conventional tillage and irrigating furrows have already been adopted by the majority of growers in the Yaqui Valley (Meisner et al., 1992).
Developing cropping systems that use N efficiently is important for reducing costs of N fertilizer inputs and for minimizing nitrate contamination (Huggins and Pan, 1993). Proper timing of N application and adequate N rates are critical in meeting plant needs and in improving N use efficiency. Nitrogen use efficiency can be defined as grain yield per unit N fertilizer applied (Gw/Nf) (Sowers et al., 1994). However, this definition is useful only within a specific crop system. If the evaluation of N use efficiency is to be expanded to include comparisons between cropping systems, tillage regimes, and other practices that affect N supply, this definition is no longer adequate (Huggins and Pan, 1993). Alternatively, Moll et al. (1982) presented a more comprehensive definition of N use efficiency that overcomes such weaknesses: grain yield per unit N supply (Gw/Ns), where both variables are expressed in the same units. This definition provides a basis for evaluating soil and plant physiological processes, including productivity with respect to N use, and can be used to compare management practices that are supposed to differentially affect the amount of residual or mineralized N (Huggins and Pan, 1993). Nitrogen supply, on the other hand, has been defined in different ways. For example, Sowers et al. (1994) estimated N supply as the sum of (i) preplant inorganic soil N, (ii) N fertilizer, and (iii) mineralized N. Huggins and Pan (1993) and Lory et al. (1995) estimated N supply as the sum of (i) post-harvest soil nitrate in control plots (0 N applied), (ii) N fertilizer, and (iii) aboveground plant N uptake in control plots (0 N applied). Campbell et al. (1993) estimated N supply as the sum of (i) soil nitrate measured in fall, and (ii) N fertilizer. In any case, N supply estimates do not consider possible effects of N fertilizer on mineralization or immobilization (Sowers et al., 1994).
Our objectives were (i) to determine the effects of N fertilization and residue management practices on grain yield and N use efficiency of wheat grown on permanent beds and (ii) to compare a permanent bed-planting system with the conventional-tilled bed system now in use.
|
Materials and methods
|
|---|
The experiment was initiated in 1992 at CIANO (Centro de Investigaciones Agrícolas del Noroeste) research station, near Ciudad Obregón in Sonora, Mexico (27.33° N, 109.09° W; 38 m above sea level). The soil type was a coarse sandy clay (mixed, montmorillonitic Typic Calciorthid), low in organic matter and slightly alkaline (pH 7.7). This report presents results obtained in four wheat seasons between 1993 and 1997.
The experiment comprised two crops in rotation: spring wheat as a winter crop, planted in the optimum period of late November to early December and harvested in late April to early May, and maize as a summer crop, planted in June and harvested in October. The experiment was designed as a randomized complete block with three replications and a split-plot treatment arrangement. Main plots consisted of five tillage–straw treatments:
- CTB–all straw incorporated: conventional-tillage beds with both wheat and maize residues incorporated by disking and subsoiling.
- PB–all straw left as stubble: permanent raised beds with no tillage operation, except a cultivation between each crop to reshape the beds, with both wheat and maize residues left as stubble.
- PB–straw partly removed: tillage as in (2) with maize residues removed and wheat residues retained.
- PB–all straw removed: tillage as in (2) with both wheat and maize residues removed.
- PB–all straw burned: tillage as in (2) with both wheat and maize residues burned.
Subplot treatments for wheat consisted of seven N fertilizer applications of urea. The first five N fertilizer (Nf) treatments were a single basal application at planting from 0 to 300 kg Nf ha-1 (0, 75, 150, 225, and 300 kg Nf ha-1). The sixth and seventh N treatments were 150 and 300 kg Nf ha-1, respectively, both topdressed at Zadoks stage 31 (Zadoks et al., 1974), 1st node stage. Before sowing of each crop, all plots received an application of 46 kg ha-1 P2O5 banded in furrow and incorporated through cultivation in the furrows during reshaping of the beds. Maize in all plots received a uniform application of 150 kg Nf ha-1, irrespective of the previous N treatment to wheat. In both crops, urea fertilizer was banded in the furrow and incorporated through irrigation.
Plots in the winter season were planted to the bread wheat cultivar Rayon 89 at a seeding rate of 100 kg ha-1, resulting in 56 to 113 plants m-2 over seasons. Plots measured 78 m2 and consisted of eight raised beds, each 75 cm wide (center furrow to center furrow) and 13 m long. Beds were planted with an Aitchison-SeedMatic 2112C drill, with planter units modified for planting wheat on top of a bed with two seed rows seeded 20 cm apart, and maize in one row. Irrigations were applied in furrows between beds when 50% of available water determined gravimetrically was depleted in the top 60 cm of soil.
Wheat was combine-harvested from the four center beds (3 by 11 m; 33 m2) in each plot when the crop reached about 160 to 180 g kg-1 seed moisture content. A random sample of 100 entire tillers cut at soil level was taken after physiological maturity. This sample was oven-dried for 48 h at 65°C and threshed to determine harvest index and to collect a subsample of grain and straw to estimate N content by Kjeldahl procedure.
We defined N use efficiency as wheat grain yield (Gw) per unit of N supply (Ns), (Gw/Ns). The latter term was defined as the sum of (i) N applied as fertilizer, and (ii) total N uptake (Nt) in control (0 N applied) plots. Total N uptake was the sum of straw N plus grain N. Differences in grain yield between tillage–straw treatments were partitioned into N use efficiency, N supply, and N utilization efficiency (Gw/Nt) components, as outlined by Huggins and Pan (1993). The relationships between grain yield and N supply, grain yield and total N uptake (needed for grain yield partitioning), and total N uptake and N fertilizer were fitted to the Mitscherlich function (Overman et al., 1994) given by
where Y is the independent variable (grain yield, Mg ha-1, or total N uptake, kg ha-1); A is maximum grain yield, Mg ha-1, or maximum total N uptake, kg ha-1; b is the intercept parameter; c is the response coefficient to N supply, total N uptake, or N fertilizer, all in ha kg-1; and N is N supply, total N uptake, or N fertilizer, all in kg ha-1. Since there is not an exact statistical test for nonlinear parameters, the differences among tillage–straw treatments were tested with an approximate F-test from the residual sum of squares, as shown by Mead et al. (1993), using SAS software with the PROC NLIN option (SAS Inst., 1989).
Nitrogen use efficiency and total N uptake data were subjected to analysis of variance assuming year effects were random and tillage–straw and fertilizer N treatments were fixed (Carmer et al., 1989). The treatment x year interaction for N use efficiency was tested using stability analysis by means of regression of the N use efficiency of each N fertilizer treatment on year mean N use efficiency (Raun et al., 1993).
|
Results and discussion
|
|---|
On average, PB–all straw burned and PB–all straw left as stubble treatments had greater wheat grain yield (5.49 and 5.48 Mg ha-1, respectively), N use efficiency, (29 and 28 kg grain kg-1 Ns, respectively), and total N uptake (137 and 133 kg ha-1, respectively) than the other tillage–straw treatments. Analysis of variance for N use efficiency and total N uptake indicated that the N x year and N x tillage–straw interactions were significant (Table 1)
. The average N use efficiency and total N uptake for all tillage–straw and N treatments was 28 kg grain kg-1 Ns and 131 kg ha-1, respectively.
Nitrogen Use Efficiency
Stability analysis for N x year interaction applied to each N fertilizer treatment regressed on the year mean N use efficiency showed that N use efficiency in all years decreased as N rate increased. Regression coefficients (Table 2)
from stability analysis showed that the response of N use efficiency to N fertilizer treatments, as a function of the year mean N use efficiency, differed between timing N fertilizer applications (P = 0.004). The N fertilizer application at the 1st node stage had greater regression coefficients (>1) compared with basal N applications (<1). The size of the regression coefficient for the application of 150 kg Nf ha-1 at the 1st node stage suggests a distinct advantage in years of high N use efficiency over all N treatments. However, in years when N use efficiency was low, the N use efficiency was similar for both timing applications of 150 kg Nf ha-1. The changes in N use efficiency along year means for basal application rates from 150 to 300 kg Nf ha-1 were similar, as shown by regression coefficients (P = 0.96; Table 2). Nitrogen use efficiency for the application of 300 kg Nf ha-1 at either the 1st node stage or basal at planting was also similar (P = 0.51; Table 2), suggesting that this fertilizer N rate was beyond the response curve.
View this table:
[in this window]
[in a new window]
|
Table 2 Linear regression equations for the N use efficiency of each N fertilizer (Nf) treatment regressed on the year mean N use efficiency combined over tillage–straw treatments
|
|
Wheat Grain Yield vs. N Supply
Grain yield as a function of N supply for each tillage–straw treatment increased from an average of 3.05 to 6.25 Mg ha-1 as N supply increased from 61 to 361 kg ha-1 (Fig. 1a)
. This indicates that N use efficiency decreased as N supply increased. The test of the Mitscherlich model applied to the relationship between grain yield and N supply showed that model parameters A, b, and c (Table 3)
varied among tillage–straw treatments (Fig. 1a), but F-tests did not indicate which of these was statistically different. However, the PB–straw partly removed treatment had the lowest maximum grain yield parameter (A). The differential response to N supply by tillage–straw treatments (the c parameter in Table 3) is an indication that removing crop residues by burning, as opposed to baling, may have altered the pattern of N mineralization (Rasmussen and Rohde, 1988). Permanent bed–all straw as stubble consistently had the largest N supply, followed by PB–straw partly removed. In contrast, CTB–all straw incorporated and PB–all straw burned had the lowest N supply. The differential amounts of N supply for tillage–straw treatments with crop residues suggests that residue management is more important than residue quantity. At low N supply, grain yield was greater for tillage–straw treatments with crop residues as stubble (Fig. 1a). Nitrogen use efficiency was greater for the PB–all straw left as stubble and PB–all straw burned treatments than for the other tillage–straw treatments (Fig. 1a). Compared with basal N fertilizer application, 150 kg Nf ha-1 applied at the 1st node stage increased the N use efficiency from 26 to 27.5 kg grain kg-1 Ns. Application of 300 kg Nf ha-1 at the 1st node stage and at planting resulted in N use efficiency of 17.3 kg grain kg-1 Ns.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 1 Wheat grain yield response to (a) N supply and (b) total N uptake on basal N fertilizer rates fitted to the model Y = A/[1 + exp(b - cN)], with N application at the 1st node stage for each tillage–straw treatment
|
|
View this table:
[in this window]
[in a new window]
|
Table 3 Parameter estimates from fitting the model {Y = A/[1 + exp(b - c x N]} to grain yield vs. N supply, grain yield vs. total N uptake, and total N uptake vs. N fertilizer
|
|
Changing tillage–straw from CTB to PB tended to increase grain yield at both low and high N supply (0 and 300 kg Nf ha-1, respectively). An exception was PB–straw partly removed, where grain yield at high Ns was lower than for CTB–all straw incorporated (Table 4
; Fig. 1a). Partitioning differences in wheat grain yield between CTB and PB into N use efficiency and N supply components, as outlined by Huggins and Pan (1993), showed that an average of 79% of the differences in grain yield at low N supply can be attributed to the N use efficiency component. This average does not include the PB–straw partly removed treatment, where 65% of the increase in grain yield was due to the N supply component (Table 4). Thus, in a wheat–maize cropping system planted on beds with low N supply, removing only the maize residues results in a greater amount of residual or mineralized N from wheat residues than occurs with leaving or removing both wheat and maize residues. When considering the shift from CTB to PB at high N supply, more than 95% of the increase in grain yield is attributed to N use efficiency component (except for the PB–straw partly removed, where grain yield decreased).
View this table:
[in this window]
[in a new window]
|
Table 4 Wheat grain yield (Gw) differences at low and high N supply (Ns) between tillage–straw treatments (CTB subtracted from PB treatments) attributed to N use efficiency (Gw/Ns) and N supply (Ns) components
|
|
Wheat Grain Yield vs. Total N Uptake
The relationship between grain yield and total N uptake was similar for all tillage–straw treatments (Fig. 1b). This suggests that the differences in grain yield found when shifting from CTB to PB, at either low or high total N uptake (measured at 0 and 300 kg Nf ha-1, respectively), cannot be attributed to differences in the N utilization efficiency component (Huggins and Pan, 1993). However, grain yields at high total N uptake (Fig. 1b) for PB–all straw left as stubble and PB–all straw burned seem to result in a wheat crop with more efficient physiological N use. Conversely, PB–straw partly removed at low total N uptake was less physiologically efficient in using N. On average, basal fertilizer applications at planting of 150 kg Nf ha-1 made the wheat crop more physiologically efficient in N use (44 kg grain kg-1 Nt), compared with N fertilizer applications at the 1st node stage (40 kg grain kg-1 Nt). Treatments consisting of crop residues left as stubble showed a tendency to take up more N at 0 N rates (Fig. 1b).
Total N Uptake vs. N Fertilizer
Total N uptake as a function of N fertilizer increased as N rate increased. Tests of the Mitscherlich model fitted to this relationship indicated differences among the tillage–straw treatments (Fig. 2)
, with the intercept parameter (b) being significantly different. The PB–straw partly removed treatment had the lowest intercept parameter, followed by PB–all straw left as stubble (Table 3). The resulting total N uptake due to partial removal of crop residues (only maize residues removed) denotes that, at low N fertilizer rates, the amount of N mineralized from wheat residues is greater than when both wheat and maize residues are left as stubble. The opposite was apparently true for PB–all straw burned, where the intercept parameter (b) denotes a low total N uptake, probably because of the partial absence of N mineralization sources (only roots). On average, both PB–all straw left as stubble and PB–all straw burned treatments resulted in higher total N uptake, which is of environmental interest because N losses into the environment are smaller. This benefit can even be enhanced for all tillage–straw treatments through the application of N fertilizer at the 1st node stage. For example, the application of 150 and 300 kg Nf ha-1 at the 1st node stage recovered 14 and 8% more N than when N fertilizer was basal-applied at planting.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2 Response of total N uptake to (a) basal N fertilizer fitted to the model Y = A/[1 + exp(b - cN)]; and to (b) N application at the 1st node stage for each tillage–straw treatment
|
|
The maximum total N uptake parameter (A) of the Mitscherlich model for tillage–straw treatments consisting of straw residue retention (incorporated, all left as stubble, and partly removed) was lower than with treatments in which straw was removed by either burning or baling (Table 3). The differential response of tillage–straw treatments at high N fertilizations indicates that, when residues in the wheat–maize cropping system are retained, a larger portion of N fertilizer is immobilized, thus lowering the amount of N losses to the environment.
|
Conclusions
|
|---|
On average, N fertilizer banded at the 1st node stage increased the N use efficiency by 3% and total N uptake by 10%, compared with basal application at planting. Permanent bed–all straw burned and PB–all straw left as stubble had the greatest average N use efficiency and total N uptake.
Our results agree with those of Rasmussen and Rohde (1988), who found that burning crop residues caused no short-term wheat grain yield reduction. Wheat grain yields in all tillage–straw treatments increased 5% with the application of 150 kg Nf ha-1 at the 1st node stage compared with N fertilizer application at planting. The increase in grain yield found when shifting from a conventional-till bed to a permanent-bed planting system was due to a higher N use efficiency. However, within the range of 56 to 186 kg ha-1 Nt, the grain yield increase cannot be attributed to differences in N utilization efficiency. The use of model parameters to explain the relationship of total N uptake vs. N fertilizer indicated that the permanent bed planting system, with full retention of wheat and maize residues as stubble, and the application of N fertilizer at the 1st node stage, can reduce N losses to the environment.SAS Institute 1989
|
ACKNOWLEDGMENTS
|
|---|
The authors gratefully acknowledge the expert assistance of Jaime Cruz and Saul Sánchez.
|
NOTES
|
|---|
The senior author was a graduate student at the Univ. of Nebraska-Lincoln sponsored by the Mexican government through CONACYT (Consejo Nacional de Ciencia y Tecnología).
Received for publication March 13, 1999.
|
REFERENCES
|
|---|
- Campbell C.A., Zentner R.P., Selles F., McConkey B.G., Dyck F.B. Nitrogen management for spring wheat grown annually on zero-tillage: Yields and nitrogen use efficiency. Agron. J. 1993;85:107-114.[Abstract/Free Full Text]
- Carmer S.G., Nyquist W.E., Walker W.M. Least significant differences for combined analyses of experiments with two- or three-factor treatment designs. Agron. J. 1989;81:665-672.[Abstract/Free Full Text]
- Huggins D.R., Pan W.L. Nitrogen use efficiency component analysis: An evaluation of cropping system differences in productivity. Agron. J. 1993;85:898-905.[Abstract/Free Full Text]
- Lory J.A., Russelle M.P., Randall G.W. A classification system for factors affecting crop response to nitrogen fertilization. Agron. J. 1995;87:869-876.[Abstract/Free Full Text]
- Mead R., Curnow R.N., Hasted A.M. Statistical methods in agriculture and experimental biology, 2nd ed London: Chapman & Hall, 1993.
- Meisner C.A., Acevedo E., Flores D., Sayre K., Ortiz-Monasterio I., Byerlee D., Limon A. Wheat production and grower practices in the Yaqui Valley, Sonora, Mexico. Mexico City: Wheat Spec. Rep. 6. CIMMYT, 1992.
- Moll R.H., Kamprath E.J., Jackson W.A. Analysis and interpretation of factors which contribute to efficiency of nitrogen utilization. Agron. J. 1982;74:562-564.[Abstract/Free Full Text]
- Overman A.R., Wilkinson S.R., Wilson D.M. An extended model of forage grass response to applied nitrogen. Agron. J. 1994;86:617-620.[Abstract/Free Full Text]
- Raun W.R., Barreto H.J., Westerman R.L. Use of stability analysis of long-term soil fertility experiments. Agron. J. 1993;85:159-167.[Abstract/Free Full Text]
- Rasmussen P.E., Rohde C.R. Stubble burning effects on winter wheat yield and nitrogen utilization under semiarid conditions. Agron. J. 1988;80:940-942.[Abstract/Free Full Text]
- SAS Institute. SAS/STAT: User's guide. Version 6, 4th ed Cary, NC: SAS Inst, 1989.
- Sharpe R.R., Harper L.A., Giddens J.E., Langdale G.W. Nitrogen use efficiency and nitrogen budget for conservation tilled wheat. Soil. Sci. Soc. Am. J. 1988;52:1394-1398.[Abstract/Free Full Text]
- Sowers K.E., Pan W.L., Miller B.C., Smith J.L. Nitrogen use efficiency of split nitrogen applications in soft white winter wheat. Agron. J. 1994;86:942-948.[Abstract/Free Full Text]
- Zadoks J.C., Chang T.T., Konzak C.F. A decimal code for the growth stages of cereals. Weed Res. 1974;14:415-421.
This article has been cited by other articles:
|
|
|
|
 
Y. Gan, S. S. Malhi, S. Brandt, F. Katepa-Mupondwa, and C. Stevenson
Nitrogen Use Efficiency and Nitrogen Uptake of juncea Canola under Diverse Environments
Agron. J.,
February 26, 2008;
100(2):
285 - 295.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
