| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Soil Science, Univ. of Manitoba, 362 Ellis Bldg, Winnipeg, MB, Canada, R3T 2N2
b Agriculture and Agri-Food Canada, Brandon, MB, Canada R7A 5Y3
c Western Co-operative Fertilizers Limited, Calgary, AB, Canada
* Corresponding author (Don_Flaten{at}umanitoba.ca)
Received for publication November 8, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: DCD, dicyandiamide • NBPT, N-(n-butyl) thiophosphoric triamide • NHU, nitrification heat units • RFN, recovery of fertilizer nitrogen • SHU, soil heat units
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Numerous reviews and studies have reported that the recovery of fall-applied N is increased by banding or nesting, compared with broadcast and incorporation, especially in soil zones where moisture supply is relatively high (i.e., Malhi and Nyborg, 1985; Malhi et al., 2001; Yadvinder-Singh et al., 1994). Banding or nesting of fall-applied N slows the nitrification rate of fertilizer N by reducing the exposure of the fertilizer to the soil. In addition, the high pH, NH3 concentration, and osmotic pressures found within the fertilizer band create a toxic environment for soil microorganisms (Harapiak et al., 1993). In fact, Malhi and Nyborg (1990) questioned whether fall-banded N fertilizers require delaying of application date to improve their efficiencies, as recommended for broadcast and incorporated N fertilizers. For example, Malhi et al. (1989) and Nyborg et al. (1990) reported no significant differences in the recovery of 15N in the crop or soil when the application of banded urea was delayed from mid to late October in Alberta. Gomes and Loynachan (1984) quantified the effect of time and temperature on nitrification of anhydrous NH3 applied at three different times in the fall (9 October, 27 October, and 14 November) in central Iowa, and observed a highly significant linear relationship (R2 = 0.84) between the recovery of NH4+ in the band row and total accumulation of soil heat units during the fall and early spring. To minimize the nitrification of ammoniacal fertilizers, current guidelines in Manitoba (Manitoba Agriculture and Food, 2001) recommend that fall N fertilizer applications are banded and delayed until soil temperatures have cooled to 5°C, when microbial activity is slow. While there is a large quantity of literature describing the effect of application date on broadcast and incorporated N fertilizers, however, there is a relatively little information about the effect of delayed application as a tool for reducing potential N losses from banded N. In fact, this was the first study to measure the efficiency and potential losses of early fall-banded N fertilizers in Canada. In addition, no previous studies have provided detailed information about the rate of transformation of fall-banded N as influenced by early application dates in western Canada.
In a previous study, Tiessen et al. (2005) reported that landscape position played a major role in the efficiency of fall-banded urea for spring wheat production in the Red River Valley of Manitoba. On these near-level fields, yields from early fall-applied N were approximately 30% greater in the microhigh than the microlow landscape positions. Therefore, landscape position is another factor that could affect the rate of nitrification and the risk of loss of fall-banded N fertilizers. After heavy rains in the fall, water often collects in the lower, convergent areas of the field (Hanna et al., 1982), potentially increasing nitrification rates because nitrifier activity is generally highest at soil water contents near field capacity (Havlin et al., 1999). Malhi and McGill (1982) reported that nitrification rates increase with increasing soil moisture potential from –1500 to –33 kPa, with appreciable nitrification at –1500 kPa and no nitrification at 0 kPa due to the shortage of O2 in the soil caused by excess water. The current study is the first to investigate the interactive effects of application date and landscape position on the transformation of fall-banded N in western Canada.
Inhibitors of ammonification and nitrification may be useful in reducing fertilizer losses associated with fall applications. A promising urease inhibitor is NBPT, which reduces urease activity and slows the hydrolysis of urea to NH4+ by competing with urea for active sites on the urease enzyme complex (Rawluk et al., 2001). Numerous nitrification inhibitors have been used in research trials to improve the efficiency of fall-applied N (e.g., Malhi and Nyborg, 1983a, 1988; Malhi et al., 1992b), with increased crop yields reported in situations when spring N losses were high. Recently, there has been renewed interest in the use of the nitrification inhibitor DCD (or Didin), because it is less volatile than other nitrification inhibitors (e.g., nitrapyrin) and can be easily blended with solid N fertilizers (Guiraud and Marol, 1992). Under field conditions, the effectiveness of DCD in retarding the nitrification of urea was reported to be highest on coarse-textured soils, in poorly drained soils where conditions are conducive to NO3– losses, and when N rates were not in excess of crop requirements (Malzer et al., 1989; Yeomans, 1991). In England, DCD was as effective as nitrapyrin in slowing the overwinter nitrification of injected aqueous urea, applied late in the fall when soil temperatures were <5°C (Ashworth and Rodgers, 1981). In Ontario, fall-applied large urea granules containing low rates of DCD slowed nitrification (Yadvinder-Singh and Beauchamp, 1987, 1988a) and reduced overwinter losses in winter wheat (Yadvinder-Singh and Beauchamp, 1988b); however, the effectiveness of a double inhibited formulation of urea containing NBPT and DCD in slowing the transformation of urea fertilizer has not been investigated in fall banding trials in western Canada.
The objective of this study was to document the transformation of banded urea fertilizer during the fall and generate information on the effect of application date, landscape position, soil temperature, and fertilizer additives on the rate of ammoniacal N transformation into NO3– via the nitrification process.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Experimental Design and Treatments
A detailed description of the experimental design and treatments used in this project was reported in Tiessen et al. (2005). Three of the four sites were located in the near-level lacustrine landscape of the Red River Valley of southern Manitoba (Kane in 2000–2001 and 2001–2002, and Rosser in 2001–2002) with shallow depressions (microlows) and typical elevation differences of <1 m km–1 within each site. The topography at the Brandon site was slightly more undulating, and representative of glacial till landscapes in the Black soil zone of southwestern Manitoba. All sites are classified as having a continental, subhumid, midlatitude climate with cold winters (Raddatz, 1997).
A nested split-plot design was used at all sites, with landscape position main plots and fertilization treatment subplots. At Kane in 2000–2001 and 2001–2002 and Rosser in 2001–2002, topographical maps were created to choose individual high and low landscape positions. It was not necessary to create a topographical map at Brandon in 2001–2002, because the slightly more undulating topography made it easy to choose individual high and low landscape positions. At each site, eight main plots, consisting of four plots in high areas and four plots in low areas, were selected throughout the field. The landscape positions studied in this experiment were defined as high and low based on their relative elevations to one another within the field. Separate main plots were located in an individual high or low landscape position (four of each) throughout the field. Each main plot contained five, 2 by 10 m fall fertilization treatment subplots, with all fertilization treatments assigned randomly to the subplots within the main plot. Fall fertilization treatments consisted of urea fertilizer (46–0–0), banded at a rate of 80 kg N ha–1, on 40-cm spacing, and at a depth of 7.5 cm, applied at three dates in the fall between mid-September and mid-October (i.e., early fall, midfall, and late fall). In addition, urea formulated with a urease and nitrification inhibitor (IMC-Agrico Super Urea containing NBPT and DCD) was applied once in the early fall and there was a control where no N fertilizer was applied.
The three target dates for fall application of the urea fertilizer were 15 September (early fall), 30 September (midfall), and 15 October (late fall) of each year. In Year 1, however, excess moisture in the fall caused a delay in application dates to 29 September and 12 and 26 October at Kane. In Year 2, treatments were applied at Kane on 26 September and 9 and 19 October, at Rosser on 19 September and 1 and 19 October, and at Brandon on 15 September and 1 and 15 October. As a result, all fertilizer applications occurring between 15 and 29 September were classified as early fall, applications occurring between 30 September and 14 October as midfall, and any applications that occurred after 15 October as late fall. The corresponding soil temperatures at 7.5-cm depth for the three fall application dates at each site were 11.3, 8.8, and 7.8°C at Kane in 2000–2001; 10.4, 7.8, and 6.6°C at Kane in 2001–2002; 13.9, 12.5, and 5.6°C at Rosser in 2001–2002; and 12.2, 11.6, and 5.7°C at Brandon in 2001–2002. Each fertilizer band row was marked with small wooden stakes and pin flags to allow precise sampling of the banded areas.
Soil Sampling and Analyses
To characterize the soil properties for each main plot, the soil was sampled to 120 cm in mid-September before fertilization (Tiessen et al., 2005). In addition, soil samples of 0 to 15 and 15 to 30 cm were taken three times in the fall from the band zone and between band zones, at approximately 2-wk intervals, to monitor the transformation of banded fertilizer. The band zone was subsampled 20 times within each subplot, with five cores taken at each of four different band locations, using a JMC Backsaver probe (Clements Assoc., Newton, IA) with a 2-cm-diameter coring tube. The sampling pattern for the band zone was in a W shape, with one subsample taken from the center of the band and the other subsamples taken at 2 and 4 cm on both sides of the band. Zones between bands were also sampled at four locations within each subplot. The between-band samples were also taken as 2-cm-diameter cores, but at distances of 10, 15, 20, 25, and 30 cm from a fertilizer band (i.e., the subsamples that are 25 and 30 cm from one band were 15 and 10 cm away, respectively, from the adjacent band). The control treatment was sampled with 20 random cores throughout each subplot. Subsamples were mixed into one composite for each combination of zone, sample depth, and treatment in the field.
Weather and soil conditions dictated when the fall soil samples were collected at the individual sites. In Year 1 at Kane, soil samples were collected on 12 and 26 October, but the third fall sampling period was missed because of snow and frozen soil conditions. In Year 2, fertilized subplots were sampled three times at Kane (9 and 23 October, and 1 November) and Brandon (1 and 15 October, and 1 November), and twice at Rosser (2 and 30 October). Wet soil conditions forced the cancellation of the mid-October sampling period at Rosser in 2001–2002.
A detailed description of the procedures used to analyze the soil samples is reported in Tiessen et al. (2005). Moist soil samples were refrigerated and stored at 2°C until air dried, at 
30°C for 2 to 3 d, and ground to pass a 2-mm sieve. Ground soil samples were analyzed for water-soluble NO3– and NO2–, exchangeable NH4+ and urea-N using 2 M KCl and phenyl mercuric acetate (Douglas and Bremner, 1970). Nitrate was determined by subtracting the NO2– from the NO3– + NO2– value.
Environmental Measurements
Mean monthly air temperature and total monthly precipitation for the Red River Valley and Brandon regions are reported in Tables 1 and 2, respectively. Rainfall data were also collected at each site located in the Red River Valley using a tipping bucket rain gauge and at Brandon in 2001–2002 using an on-site weather station. Gravimetric soil moisture contents (percentage by weight) at increments of 0 to 7.5, 7.5 to 15, and 15 to 30 cm were measured on a weekly basis during the fall and early spring at Kane in 2000–2001 and 2001–2002 and at Rosser in 2001–2002. Gravimetric moisture contents at Brandon in 2001–2002 were not measured due to resource limitations. The rainfall data and gravimetric soil moisture contents for the respective field sites are reported in Tiessen et al. (2005).
|
|
Time Temperature Equations
Gomes and Loynachan (1984) used average weekly soil temperatures to accumulate heat units to predict the rate of nitrification of anhydrous NH3, in the presence and absence of a nitrification inhibitor. A similar method was employed by Malhi and Nyborg (1990), who used soil degree days (an accumulation of the average daily soil temperature from date of application to the first day that the soil reached 0°C) to predict the relative efficiency of fall broadcast and incorporated urea. Using the concepts of Gomes and Loynachan (1984) and Malhi and Nyborg (1990), in conjunction with a model developed by Sands et al. (1979) for determining the physiological age of potato (Solanum tuberosum L.), cumulative SHU were estimated in the current project as an accumulation of average daily temperature (monitored on an hourly basis, Ti) at 7.5-cm soil depth from the date of application to the date of soil sampling:
 | [1] |
In addition, to account for the exponential changes in nitrification rates at various soil temperatures, we applied Malhi and McGill's (1982) equation for a Black Chernozemic soil [y = 0.059exp(0.21x), R2 = 0.97 at P 
0.001) to the cumulative SHU equation and defined this new parameter as cumulative NHU (A = 0.059, B = 0.21, Ti = hourly soil temperature at 7.5 cm):
 | [2] |
Data Analyses
Two weeks after application of each fertilizer treatment, the majority of fertilizer N was found within a zone 5 cm to each side and 7.5 cm above and below the fertilizer band (Tiessen, 2003). By the final sampling period in the fall, however, some of the fertilizer N applied in September and early October had already moved downward and laterally from the band zone. Therefore, to account for all banded fertilizer N in the soil, statistical analyses were performed using the bulked band zone and between-band zone soil samples to a depth of 30 cm.
A detailed description of the statistical analyses used in this experiment can be found in Tiessen et al. (2005). Statistical analyses were conducted using the general linear model procedure of the SAS package (SAS Institute, 1999). Descriptive statistics were used to test the data for normality and skewness using the Proc Univariate function of SAS. Fisher's (protected) LSD test was used to compare the subplot (fertilization) and main plot (landscape position) treatment means (Steel et al., 1987). The LSMEANS test was used to compare the fertilization treatment effects within each landscape position. For the fertilization treatment means and the fertilization treatment means within each landscape position, a probability level (
) of 0.05 was used as the significance threshold for soil and plant variables. Due to the high variability inherent in field-based landscape experiments, however, a higher probability level of 0.10 was used for landscape position effects and for landscape position x fertilization treatment interactions. This higher probability level is within the typical range of probability values (P 0.10–0.20) used in many previous landscape studies (e.g., Corre et al., 1996; Jowkin and Schoenau, 1998; Pennock et al., 1992).
In addition, simple linear regression analysis (r2) was used to test the relationship of the %RFN as NH4+ as a function of the date of fertilizer N application, the average soil temperature at 7.5 cm on the day of application, cumulative SHU, and cumulative NHU. The %RFN as NH4+ was estimated as:
 | [3] |
Linear regression analysis was also used to test the relationships of %RFN as NH4+ in the absence and presence of NBPT and DCD inhibitors, as a function of cumulative SHU and cumulative NHU. The Proc GLM procedure was used to compare the slopes of the linear relationships between early fall-banded urea with and without inhibitors.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
14 d at each site. The most obvious and consistent effects on transformation, however, in terms of landscape position, application date, and fertilizer additives, were found at the end of the fall in both years of the study. Therefore, the early and midfall sampling periods will not be discussed in detail, but can be found in Tiessen (2003). In Year 1, there was a significant landscape position main effect at Kane, as the low landscape positions had significantly more NH4+ than the high landscape positions (Table 3), suggesting that the rate of nitrification of fertilizer N may have been slower in the low landscape positions than in the high landscape positions at this site. High soil moisture contents in the low landscape positions at Kane (2000–2001) during the fall (Tiessen et al., 2005) may have reduced the availability of O2 to nitrifying microorganisms. This could also be interpreted as a faster rate of ammonification, since there were no significant differences between high and low landscape positions in terms of NO3– and Ni at this site.
|
The recovery of NH4+ was also strongly influenced by N application date and the presence of NBPT and DCD (Table 3). In Year 1, midfall-banded N had significantly higher concentrations of NH4+ and Ni than the early-fall-banded treatment without inhibitors at Kane. The addition of NBPT and DCD also resulted in significantly greater concentrations of NH4+ at Kane in 2000–2001, compared with early-fall-banded urea without inhibitors. There were no significant differences in NO3–-N or Ni between the two early fall-banded treatments, however. Of note, at Kane in 2000–2001 there was a 62% increase in Ni in the control plots from mid- to late October, suggesting that substantial mineralization of soil N occurred during the fall period at this site (Tiessen, 2003).
Fertilizer treatment effects in Year 2 were similar to those from Year 1, as the transformation of banded urea fertilizer into NO3– was significantly reduced by delaying fertilization at Kane, Rosser, and Brandon (Table 3). At each site in Year 2, late fall-banded N consistently had the highest concentrations of recovered NH4+ and the lowest concentrations of recovered NO3– at the final sampling period in late October to early November. At Rosser and Brandon in 2001–2002, there was also a significant landscape position x treatment interaction for NH4+; however, at both sites, late-fall-banded N had significantly more NH4+ than early- and midfall-banded N in either landscape position.
At Kane in 2001–2002, there were no significant differences in Ni among application dates, suggesting that there were few gains from mineralized soil N or losses of fertilizer N during the fall at this site. At Brandon in 2001–2002, late-fall-banded N significantly increased Ni compared with early-fall- and midfall-banded N in the high landscape positions but not the low landscape positions. There is no simple explanation why the apparent recoveries of fertilizer N were so large in the high landscape positions at this site, especially for the late fall application.
At Rosser in 2001–2002, late-fall-banded N significantly increased Ni compared with early-fall- and midfall-banded N in the low landscape positions; in the high landscape positions, there were no differences in Ni. This suggests that some N losses from the early- and midfall-banded N may have occurred in the low landscape positions during the fall at Rosser in 2001–2002. During the week following application of the midfall treatment on 19 October, this site received 35 mm of rain (Tiessen et al., 2005), creating saturated soil conditions in the low landscape positions and potentially denitrifying some of the nitrified fertilizer N; however, the apparent fertilizer N recovered from the late fall treatment was greater than the amount of fertilizer N applied, so sampling error could also have affected the results.
Overall, the NBPT and DCD inhibitors appeared to be less effective in the second year of the study than in the first, but the trend still indicates that the inhibitors slowed urea hydrolysis and nitrification (Table 3). At Rosser and Brandon in 2001–2002, the NBPT and DCD treatment had significantly lower concentrations of NO3––N in the soil at the final fall sampling period than the early-fall-banded N without inhibitors. In addition, at Rosser in 2001–2002, early-fall-banded N with inhibitors had significantly higher concentrations of NH4+ than early-fall-banded N without inhibitors in the low landscape positions. There were no other significant differences, however, between the two early-fall-banded treatments, at any of the three sites in either landscape position.
Predicting Recoveries of Ammonium: Time of Application Effect
Predicting NH4+ recoveries as a function of application date, soil temperature on the date of application, and time x temperature interactions, with and without the addition of fertilizer additives, would aid producers in determining when they should apply N fertilizer and what proportion of the applied fertilizer is at risk to loss in the spring. Manitoba Agriculture and Food currently recommends that producers who wish to apply ammoniacal N fertilizer in the fall delay application as late as possible until the soil temperature has declined to 5°C or less, to minimize the conversion of fertilizer N to NO3– before winter. The regression analyses generated from our data support this recommendation. There is a strong linear relationship between the %RFN as NH4+ at the final fall sampling period before freeze and the date of application in the fall (adjusted r2 = 0.88 at P 0.001; Fig. 1a
). The %RFN as NH4+ increased from a minimum of 6% when urea fertilizer was banded on 17 September to a maximum of 96% when banded on 19 October. This suggests that, in a typical fall in southern Manitoba, producers who band ammoniacal fertilizer N in mid- to late September can expect that the majority of the fertilizer will have converted to NO3– before freeze and therefore be vulnerable to leaching and denitrification losses before crop establishment.
|
0.001; Fig. 1b). Soil temperature on the date of application produced the weakest coefficient of determination of all the environmental measures that were tested, but this measure was still highly significant. Delaying application in the fall until soil temperatures had reached 5 to 7°C resulted in 87 to 96% of the fertilizer N remaining as NH4+ at the beginning of November, as opposed to 6 to 32% when fertilizer N was applied when soil temperatures were 12 to 14°C.
From a practical point of view, the date of fall N application or the soil temperature on the date of fall N application adequately describes the transformation of fertilizer N into NO3– and is easy for producers to use when planning a fall fertilization program. Due to variations in soil temperature from 1 yr to another, however, the date of application may not adequately reflect current environmental conditions for nitrification. Furthermore, day-to-day variability in soil temperature may result in soil temperatures that are much different after fertilizer application than before. Malhi and Nyborg (1990) reported that soil temperature on the day of fall fertilizer application had a low correlation with increases in grain yield (r = –0.55) and attributed the low correlation to day-to-day variability in soil temperatures instead of a steady decline during the fall to the day the soil froze. To account for this year-to-year and day-to-day variability in weather, two time–temperature interactions were examined: cumulative SHU and cumulative NHU. Regression analysis indicated that there is a very strong negative linear relationship between the %RFN as NH4+ and both cumulative SHU (adjusted r2 = 0.90 at P 0.001; Fig. 1c) and cumulative NHU (adjusted r2 = 0.92 at P 0.001; Fig. 1d). Therefore, the use of either of these time–temperature parameters to account for nitrification rates was superior to using the date of application and soil temperature on the date of application.
Overall, the %RFN as NH4+ at the end of the fall increased from 22 to 87% when application was delayed from mid-September to mid-October and only 50 SHU accumulated as opposed to 300 SHU (Fig. 1c). This also suggests that producers who band fertilizer N in the fall, even after soil temperatures have declined to a given level, must consider the date of application and the overall length of time that the fertilizer will be exposed to soil processes before the soil freezes.
The highest coefficient of determination was generated using NHU, with the %RFN as NH4+ decreasing substantially as NHU increased (Fig. 1d). The regression equation predicted that 88% of the fertilizer N had converted to NO3– after 18 NHU had accumulated in the soil before the soil froze. If only 2 NHU had accumulated in the soil, however, merely 14% of the fertilizer N had converted to NO3–. We suspect that NHU best described the %RFN as NH4+ because nitrification rates slow linearly with decreasing temperatures to 4°C (Chandra, 1962), at which point rates begin to slow exponentially (Malhi and McGill, 1982). We also suspect that if more sites or site years had been included in this experiment, the marginal improvement in accuracy by using SHU or NHU instead of the date of application or the soil temperature on the date of application in the fall would have increased due to increased variability in weather conditions after application. For producers to use SHU or NHU to predict the proportion of fall-banded fertilizer N remaining as NH4+ at freeze, however, they would require access to accurate forecasts of soil temperature. Therefore, SHU and NHU are best suited for historical monitoring, while the date of application and the soil temperature at the date of application are more practical tools for predicting the best time to apply N fertilizer in the fall.
The improved regression coefficient of determination using cumulative SHU and NHU, compared with the date of fall application and the soil temperature on the date of fall application, are similar to the results of Malhi and Nyborg (1990). Using broadcast and incorporated urea fertilizer, they reported that grain yield increases were correlated best with soil-degree-day values (r = 0.78), compared with the date of application (r = 0.54) and the soil temperature on the day of N application (r = 0.55). Our findings also agree with those of Gomes and Loynachan (1984) in central Iowa, who report that recoverable NH4+ from anhydrous NH3 application was highly correlated with the accumulation of SHU. Gomes and Loynachan (1984) reported, however, that 1000 soil heat units calculated on an average weekly basis were necessary before 80% of the fertilizer N had been nitrified. Conversely, our correlation analysis shows that, under Manitoba conditions, only 300 soil heat units were necessary before 80% of the banded urea-N was converted to NO3–. This difference in the amount of soil heat units supports the claim by Malhi and McGill (1982) that the soil's microbial population is able to adapt to local climates, and suggests that the transformation of banded ammoniacal fertilizers is more rapid in western Canada than the rate that would be predicted on the basis of research in warmer climates.
Similar relationships between the %RFN as NH4+ and both cumulative SHU and cumulative NHU were observed when linear regression analyses were performed using the data collected from all the fall sampling periods at each site (two or three sampling dates per site depending on weather conditions; Tiessen, 2003). For both SHU and NHU, there was a strong negative linear relationship describing the %RFN as NH4+ (adjusted r2 = 0.85 and 0.83, respectively, at P 0.001). Although these coefficients of determination were slightly lower than those generated when only the last fall sampling period was used, this strong relationship suggests that the time–temperature relationships with nitrification were generally stable during the fall, regardless of when measurements were taken. The slightly lower coefficient of determination generated using all sampling periods may be attributed to the initial hydrolysis of the urea fertilizer and short-term increase in NH4+ during the first week after the urea fertilizer was applied to the soil. This was not a factor when only the final sampling period at the end of the fall was used.
Predicting Recoveries of Ammonium: Fertilizer Additive Effects
To be agriculturally useful, additives must maintain inhibitory action for periods ranging from several weeks to months after fertilizer application (Yeomans, 1991). The long-term benefits from the inhibitor must also be sufficient to justify the added costs to the producer (Grant and Bailey, 1999). Using either SHU or NHU as a measure of thermal time, the rate of conversion of NH4+–N in the urea banded with NBPT and DCD was approximately half of that for the conventional urea (Fig. 2
). The two slopes were significantly different for cumulative NHU (P = 0.02), but not for cumulative SHU (P = 0.069).
|
0.001 with NBPT and DCD and 0.80 at P 0.001 without NBPT and DCD for SHU; adjusted r2 = 0.45 at P 0.001 with NBPT and DCD and 0.78 at P 0.001 without NBPT and DCD for NHU; Tiessen, 2003). Similar to the regression equations developed for the late fall sampling period only, when all the fall sampling dates were used, the rate of conversion of NH4+–N in the urea banded with NBPT and DCD was approximately half of that for the conventional urea. In addition, increasing the sample size to include all fall sampling periods reduced the overall variability within the data set and resulted in significant differences between the slopes of the two early-fall-banded treatments, with and without inhibitors, for both cumulative SHU and NHU (P = 0.002 for both; Tiessen, 2003). Nonetheless, the coefficients of determination were slightly lower than those for the final sampling period only, presumably due to the short-term rise in NH4+ following the initial hydrolysis of the urea. The initial increase in NH4+ during the first 2 wk after application was most obvious in the inhibitor treatment; no appreciable quantities of urea were detected in soil from any treatments beyond 2 wk after application (Tiessen, 2003).
Overall, the addition of NBPT and DCD to the urea extended the persistence of NH4+ by 200 SHU or 7 NHU compared with conventional urea (Fig. 2). With more frequent soil sampling, the degree to which this persistence was extended by the NBPT or DCD may have been determined by the extent to which urease inhibition delayed the conversion to NH4+ (a lateral shift for peak NH4+ concentration) or nitrification inhibition delayed the conversion to NO3– (a shallower slope for NH4+ disappearance); however, our 2-wk sampling interval did not provide sufficient temporal detail to separate those processes.
The effectiveness of NBPT and DCD in retarding the conversion of urea fertilizer to NO3– during the fall has been shown conclusively in this study. Therefore, in an environment where the risk of denitrification and leaching losses of NO3– in the spring is high, using NBPT and DCD would increase the flexibility of a fall fertilization program and enable producers to apply fertilizer N earlier in the fall; however, slowing the conversion of fertilizer N in the fall with fertilizer additives does not necessarily reduce overall N losses in the spring or translate into increased grain yields and N uptake by the crop. Also, since urea hydrolysis was inhibited from the NBPT, a late fall application with this additive followed by a fast freeze could enhance N losses by leaching (since urea is highly mobile) in certain soils. In addition, our data suggests that delaying application of fall-banded N until mid-October is at least as effective in slowing nitrification as using the NBPT and DCD inhibitors. In the end, it is probably easier for most producers in Manitoba to simply delay application date in the fall to improve efficiencies of fall-banded N than to incur the added expense of using fertilizer additives.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Delaying application of banded urea fertilizer into the late fall delayed conversion of urea to NO3– and increased the %RFN as NH4+ in the soil before freeze. The date of application had a positive linear relationship with the %RFN as NH4+ at the last fall sampling period (adjusted r2 = 0.88 at P 0.001). This relationship suggested that the majority of urea fertilizer, if banded in the early fall without inhibitors, will convert to NO3– before the soil freezes and is therefore susceptible to loss in the spring. Soil temperature at application showed a negative linear relationship with the %RFN as NH4+ at the last fall sampling period (adjusted r2 = 0.69 at P 0.001), but had the lowest coefficient of determination of the four approaches used. The lower coefficient was probably due to day-to-day variations in soil temperature during the remainder of the fall after application instead of a smooth decline toward 0°C. The proportion of the %RFN as NH4+ accounted for by regression analysis were highest when cumulative SHU (adjusted r2 = 0.90 at P 0.001) and cumulative NHU (adjusted r2 = 0.92 at P 0.001) were used. Nonetheless, the regression equations generated for each of the four parameters are similar, and each was highly significant.
The addition of a double inhibitor also slowed the conversion of urea to NO3– and increased the %RFN as NH4+ in the soil before freeze. Regression analysis comparing early-fall-banded urea with early-fall-banded urea that included NBPT and DCD showed that the inhibitors slowed nitrification by 50% and increased retention of fertilizer N as NH4+ in the fall. Therefore, this product might allow a producer increased flexibility in a fall fertilization program, and may translate into reduced losses of N in the spring.
The interactions between time, temperature, and nitrification demonstrate that producers in the Northern Great Plains region must consider the overall length of time that fall-applied N fertilizer will be exposed to the soil before the soil freezes, even after the soil temperature has reached a given level and even if the fertilizer is banded.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. H. D. Tiessen, D. N. Flaten, P. R. Bullock, C. A. Grant, R. E. Karamanos, D. L. Burton, and M. H. Entz Interactive Effects of Landscape Position and Time of Application on the Response of Spring Wheat to Fall-Banded Urea Agron. J., May 7, 2008; 100(3): 557 - 563. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
