Published in Agron J 91:969-974 (1999)
© 1999 American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
Agronomy Journal 91:969-974 (1999)
© 1999 American Society of Agronomy
WHEAT
Use of Anhydrous Ammonia in Single-Pass Seeding Operations of Spring Wheat at Varied Landscape Positions
Alejandro Matusa,
Fran Walleya,
G. Hnatowichb,
Chris van Kesselc and
J.D. Knighta
a Dep. of Soil Science, Univ. of Saskatchewan, Saskatoon, SK, S7N 5A8, Canada
b Saskatchewan Wheat Pool, Saskatoon, SK, S7N 3R2, Canada
c Univ. of California, Davis, CA 95616 USA
knight{at}sask.usask.ca
 |
ABSTRACT
|
|---|
Adoption of no-till seeding practices by many farmers has increased interest in using anhydrous ammonia (AA) in single-pass seeding operations. It is expected that crop response to different fertilizer N sources and fertilizer placement will vary at the landscape level, because of inherent differences in soils related to topography. Two openers (side-band and sweep wing tip) were evaluated for use in a single-pass seeding operation of spring wheat (Triticum aestivum L.). The openers differed in fertilizer placement and soil disturbance. The effect of AA on grain yield and protein content as influenced by application rate and opener was investigated at varied landscape positions. The experiment was conducted at six sites in two years using a randomized complete block design. Averaged across year and location, wheat grain yield was higher on footslopes than on shoulders; however, the effect of landscape position was not consistent at all locations. Landscape position did not affect protein content. Application of AA with either opener resulted in grain yields and protein contents comparable to granular urea and ammonium nitrate (AN) fertilizers. Even at the highest AA application rate tested (105 kg N ha-1), no crop damage was expressed in final yield.
Abbreviations: AA, anhydrous ammonia • AN, ammonium nitrate • SB, side-band • SW, sweep wing tip
|
INTRODUCTION
|
|---|
IN MOST SEMIARID REGIONS
characterized by undulating or hummocky terrain, soil properties vary across landscape position (Pennock et al., 1994). The redistribution of water toward convergent areas can exert an important landscape-scale control on crop productivity (Pennock et al., 1987). Lower landscape positions (e.g., footslopes) usually have the greatest soil water and nutrient contents, and frequently produce the highest grain yields (Stevenson and Van Kessel, 1996). However, landscape position is not always a predictable indicator of crop yield and has been reported to either influence (Jowkin, 1997; Stevenson and Van Kessel, 1996) or not influence (Solohub, 1997) spring wheat grain yield.
Direct seeding or one-pass placement of seed and fertilizer into untilled land is increasing in popularity (Hnatowich, 1995). Granular urea is a popular N fertilizer used in direct-seeding operations. Although anhydrous ammonia (AA) generally is considered the most economical N fertilizer form available, it has not been considered compatible with one-pass applications, because of concerns about ammonia damage to germinating seedlings. In recent years, however, a number of trials conducted throughout western Canada indicate that placement of AA at seeding may be possible (Hnatowich, 1995; Johnston et al., 1997), as long as adequate separation between the seed and fertilizer band is achieved. Varvel (1982) observed that AA can be safely side-banded with seeds of spring wheat and barley (Hordeum vulgare L.) if a separation distance of 5 to 11 cm is used.
Recently, several new openers and packing systems have been developed for direct seeding. These openers need to be evaluated for their suitability in AA placement and their subsequent influence on grain yield and grain protein content of spring wheat. Application of N fertilizer treatments across shoulder and footslope positions allowed us to evaluate the magnitude of yield and protein responses to N fertilizer sources, rates, and types of openers across landscapes.
Improvements to agronomic practices such as the selection of fertilizer source, application rate, and type of opener requires knowledge of how the specific agronomic practice performs in different environments (e.g., landscape position, location, and year). A significant treatment (agronomic practice) x environment interaction may be either (i) a noncrossover interaction, in which case the ranking of treatments remains constant across the different environments and the interaction is significant because of changes in the magnitude of the response, or (ii) a crossover interaction, in which case the ranking of the treatment changes from one environment to another. With the crossover interaction, different treatments must be chosen for each environment (Baker, 1988).
The objectives of this study were to evaluate the effect of AA on spring wheat grain yield and protein content as influenced by application rate and opener (side-band vs. sweep wing tip) and to evaluate crop yield and grain protein response to N sources on landscapes varying in topography. It was of particular interest to determine if high rates of AA reduced grain yields and/or affected grain protein contents. An additional objective was to determine the type and magnitude of interactions influencing grain yield and protein content among the N treatments and landscape positions.
|
Materials and methods
|
|---|
Field trials were conducted on Chernozemic (Typic Boroll) soils at three sites in Saskatchewan, Canada, during the 1996 growing season and at three sites during the 1997 growing season. All sites were characterized by a hummocky surface in a complex assemblage of knolls and depressions, with slopes ranging from 4 to 6%. The experimental area at each site was surveyed and the topographical information used to develop a digital elevation model of the surface. Salinity measurements were made with a field salinity monitor at the time of surveying. None of the sites selected had any saline areas identified. The study sites were classified into two landscape element complexes according to the model (Pennock et al., 1994). Landscape element complexes with convex profile curvatures and limited catchment areas were classified as shoulders; areas with concave profile curvatures and catchment areas less than 91 m2 were classified as footslopes.
Prior to seeding, soil was sampled to a depth of 15 cm at each sampling point for determination of pH, organic matter, inorganic N, moisture content, cation exchange capacity, particle size, bulk density, and Ap horizon thickness (Table 1)
. A portion of each air-dried soil sample was screened using a 2-mm sieve and pH was measured using a soil-to-solution ratio of 1:2. Organic C was determined on finely ground soil samples using a carbon analyzer. Soil samples were extracted with 2 M KCl, and NH+4 and NO-3 were measured using an Auto Analyzer system. Because the amounts of NH+4 measured were very small relative to the NO-3 (and within the experimental error of the apparatus), the two values were combined and reported as available inorganic N. Moisture content was determined gravimetrically. Cation exchange capacity was determined by the exchangeable hydrogen and exchange capacity using BaCl2–triethanolamine methodology (McKeague, 1978). A modified pipette procedure was used to determine particle size (Indorante et al., 1990) and bulk density was assessed with the core method (Blake and Hartge, 1986). Growing season precipitation, sowing date, sowing depth, and maturity date are reported in Table 2
.
View this table:
[in this window]
[in a new window]
|
Table 2 Precipitation, sowing and harvest dates, and sowing depth for spring wheat grown at six locations in Saskatchewan
|
|
Two openers, contrasting in the amount they disturb the soil, were evaluated. The side-banding bolt-on opener is a minimal soil disturbance dual-delivery system. Fertilizer is placed below and to the side of the seed row, with a minimum separation distance of 2.5 cm in both directions. The sweep wing tip system uses a high-disturbance 35 cm shovel with AA tubes split and extended to the shovel tip. Every second shank is equipped with a wing tip system, thus simulating a midrow fertilizer application. The AA is applied at the same depth as the seed. Shovels are equipped with Froc air seeder boots (Froc Industries Ltd., Melfort, SK) that scatter the seed, creating a seed row 12.5 to 20 cm wide. The Froc boot minimizes seed scatter and damage from striking bolts and splitters common in many other air seeder boots, resulting in precise placement of the seed and fertilizer. In both seeding operations, seed was sown at a depth of 3.0 cm in 1996 and 3.5 cm in 1997 (Table 2).
Spring wheat (cv. Pasqua) was seeded at a rate of 80 kg ha-1. The experiment consisted of nine fertilizer and opener treatments (Table 3)
, replicated five times, using a randomized complete block design. Phosphorus at 8.7 kg P ha-1 (0–45–0 N–P–K) was seed-placed in all plots. Treatment areas measured 2.5 by 15 m and were oriented to extend across both landscape positions. Grain yield and protein content were measured for both positions.
View this table:
[in this window]
[in a new window]
|
Table 3 Analysis of variance and contrast analysis for grain yield and grain protein of spring wheat grown at six locations in Saskatchewan
|
|
In 1996, wheat was seeded on canola (Brassica napus L.) stubble at Wakaw, on pea (Pisum sativum L.) stubble at St. Benedict, and on flax (Linum usitatissimum L.) stubble at St. Louis. In 1997, wheat was seeded on flax stubble at St. Louis and Watrous and on wheat stubble at Hepburn. Glyphosate [N-(phosphonomethyl)glycine] was applied at the recommended rate prior to seeding. At maturity, 2- by 5-m areas from within each shoulder and footslope, within each treatment plot were harvested using a small plot combine. Grain samples were air dried under forced air at 40°C to 14% moisture and ground using a cyclone mill (0.4-mm screens). Grain N content was measured with a continuous-flow isotope ratio mass spectrometer (Walley et al., 1996) and multiplied by 5.7 to obtain crude protein content (Williams et al., 1998).
The statistical model includes sources of variation due to landscape position, N treatment, year, location within year, replication, and their interactions. Year, location, and replication were regarded as random effects, whereas N treatment and landscape position were fixed effects. Least significant difference (protected LSD) and linear contrast analysis were used to compare means. The Azallini–Cox test was used to determine crossover interactions (Baker, 1988). This test for crossover interactions is conservative (Cornelius et al., 1993), and the actual number of crossover interactions may be higher than reported here.
|
Results and discussion
|
|---|
Grain Yield
Location, landscape position, and N fertilizer treatment all had a significant effect on wheat grain yield (Table 3). Better growing conditions (Tables 1 and 2) accounted for the higher grain yields at St. Louis in 1996 as compared to 1997 (Table 4) . Averaged across year, location, and landscape position, the range of grain yields was 0.74 t ha-1 (Table 5)
. The linear response of grain yield to increasing AA rates observed with both openers, along with the absence of an AA application rate x opener interaction, indicates that increasing the AA application rate was safe, to the extent of not reducing final yields, in both one-pass direct-seeding systems (Table 3). Had either AA application been toxic to seedlings to such a degree as to interfere with final yields, a yield reduction at high AA rates would be expected.
View this table:
[in this window]
[in a new window]
|
Table 5 Mean spring wheat grain yield and grain protein content, averaged across locations, years and landscape position
|
|
Grain yield of wheat fertilized with 70 kg N ha-1 applied as AA, regardless of the opener, was similar to the yield of wheat fertilized with the same rate of urea and AN when yields were averaged across landscape position and location (Table 5). Johnston et al. (1997) also reported that wheat yields were comparable with different forms of N fertilizer applied.
Averaged across year and location, spring wheat grain yield was higher on the footslopes than on the shoulders (Table 6)
. However, the significant location x landscape position interaction suggests that landscape position is not a reliable predictor of grain yield. The interaction was a noncrossover type, in that while grain yields were consistently higher on the footslopes compared to shoulders, the differences were not always significant. Drier growing conditions in 1997 than 1996 (Table 2) probably accounted for the more extreme differences in grain yields observed between the landscape positions in 1997. Solohub (1997) reported no differences in the grain yield of spring wheat grown on landscape positions with soil properties similar to those at St. Louis. In contrast, Stevenson and Van Kessel (1996) observed higher wheat grain yields on the shoulders compared to footslope positions where low harvest indices were observed.
View this table:
[in this window]
[in a new window]
|
Table 6 Mean grain yield and grain protein of spring wheat grown on the footslopes and shoulders at six locations in Saskatchewan
|
|
The location x N treatment interaction for grain yield (Table 3) was a crossover interaction primarily caused by the side-band 70 kg N ha-1 treatments. For example, at Hepburn the 70 kg N ha-1 side-band application of AA resulted in lower grain yield than the same rate of AA applied with the sweep wing tip opener. However, at Watrous and the 1997 St. Louis site, the reverse occurred (Table 5). Four individual crossover interactions were observed in the location x N treatment interaction, indicating that the ranking of N treatments changed across location. Because locations were considered a random effect in this study, any number of factors (including disease, soil and air temperatures, soil physical and chemical factors, and precipitation) could account for these crossover interactions (Brandle and McVetty, 1988). None of the soil factors measured (Table 2) explained the interactions.
Grain Protein
Location and fertilizer treatment affected grain protein content, whereas landscape position and year had no effect (Table 3). Grain from the 1997 St. Louis site had the highest grain protein content and grain from the Wakaw site had the lowest (a range of 64 g kg-1) (Table 7)
. The linear grain protein response to N fertilization (Table 3) provides further evidence that AA application was not detrimental to the final crop in these direct-seeding operations. Furthermore, as was observed for grain yield, AA was equivalent to the granular fertilizer forms (urea and AN) in terms of grain protein content (Table 3).
The significant location x landscape position interaction for grain protein content (Table 3) was caused by changes in the rank of the response to landscape position at the different locations (i.e., crossover interaction) (Table 6). At the Wakaw, 1997 St. Louis and Watrous sites, protein contents were higher on the footslopes than on the shoulders (Table 6). At the Hepburn site the reverse occurred, and at the 1996 St. Louis and St. Benedict sites protein content was not different for the two landscape positions. As was observed with grain yield, any number of random factors associated with the different locations could be responsible for these interactions. None of the soil physical or chemical characteristics measured in this study (Table 2) explained this interaction. Fiez et al. (1994) reported a significant location x landscape position noncrossover interaction for the protein content of soft spring wheat grain. In addition to the response to N fertilization, they observed that grain protein was influenced by preplant deep residual soil N. Late-season N uptake may increase grain protein significantly under dryland conditions.
Although no quantitative measurements were made, it was observed that the amount of residue cover on the fields affected the placement of the AA and varied from site to site. Deep residue cover caused the AA to be placed more shallowly in the soil than was intended and probably resulted in higher losses compared to sites with little or no residue accumulation. Heavy residue accumulation was a problem only at the Hepburn site and may have accounted for the slightly lower grain protein contents (and hence, lower N recovery) of spring wheat crops receiving 70 kg N ha-1 as side-banded AA compared to the other N sources (Table 7). In addition to the Hepburn site, losses of AA were observed at the time of application at the Watrous site in the footslope positions. It appears that the relatively high clay contents combined with the higher water accumulation in the footslopes compared to shoulders (Table 1) at this site interfered with sealing of the fertilizer band. Based on grain protein contents, this appears to have been more of a problem with the sweep wing tip applicator than the side-banding applicator for AA (Table 8)
. Although it was not observed directly in the field, high clay contents combined with water accumulation in the footslopes may also have interfered with fertilizer band sealing in the sweep wing tip application at the St. Benedict site (Table 8). These observations further serve to illustrate our reasoning in attributing site as a random effect in the statistical analysis.
View this table:
[in this window]
[in a new window]
|
Table 8 Mean grain protein of spring wheat grown on the footslopes and shoulders at six locations in Saskatchewan
|
|
Unlike the case with grain yield, an interaction among location, landscape, and treatment within year occurred for grain protein (Table 3), indicating that the ranking of grain protein caused by the fertilizer and opener treatments changed with landscape position at different locations. For example, on the footslopes at Hepburn, 105 kg N ha-1 AA applied with the sweep wing tip opener resulted in lower grain protein contents than the comparable 70 kg N ha-1 application. At both landscape positions at Wakaw, and the shoulder position at Hepburn, virtually no difference in protein content was observed, whereas at both landscape positions at all of the remaining sites the higher application rate (105 kg N ha-1 AA with the sweep wing tip applicator) resulted in higher protein contents. The absence of a landscape position x fertilizer treatment interaction indicates that the ranking of the treatments remained constant across landscape position within a location; the three-way interaction indicates that the ranking of treatments changed with landscape position at different locations.
Soil temperature, degree of soil erosion, quantity of snow present, and available soil moisture are influenced by landscape position (Pennock et al., 1994). Differences in environmental conditions such as these can cause differences in spring wheat growth among slope positions. Although our study shows grain yield to vary with landscape position, landscape position alone is not a consistent indicator of grain yield (Jowkin, 1997; Solohub, 1997). Furthermore, the relative productivity of a landscape position can vary from location to location (Ciha, 1984; Fiez et al., 1994) and year to year (Jaynes and Colvin, 1997). The spatial variability of grain yield is controlled by soil properties, salinity, microclimatic conditions, weeds, diseases, and the unpredictability of precipitation among locations and years.
Anhydrous ammonia was an acceptable alternative to either granular urea or AN as a N fertilizer source. Typically, spring wheat grain yields and protein contents were comparable irrespective of the N fertilizer source. Even at the highest AA application rate tested (105 kg N ha-1), no crop damage was expressed in final yield, regardless of the opener used. The influence of landscape position on grain yields was site-specific but showed a distinct tendency for yields to be higher on footslopes than shoulders. Grain protein content was not affected by landscape position. However, even though excessive N losses from AA applications were not evident in grain yields, nor in grain protein contents, field observations indicated that site-specific factors such as heavy crop residues, wet soils, and high clay contents may have interfered with fertilizer band sealing.
|
ACKNOWLEDGMENTS
|
|---|
The authors acknowledge the support of the Saskatchewan Agriculture Development Fund, the Saskatchewan Wheat Pool, and the Saskatchewan Conservation Learning Centre. The technical assistance of Mr. K. Vanthuyne and Mr. G.R. Parry are appreciated. SCSR Contribution No. R852.
Received for publication July 29, 1998.
|
REFERENCES
|
|---|
- Baker R.J. Test for crossover genotype–environmental interaction. Can. J. Plant Sci. 1988;68:405-410.
- Blake G.R., Hartge K.H. Bulk density. In: Klute A., ed. Methods of soil analysis. Madison, WI: Agron. Monogr. 9. ASA, 1986:363-375 Part 1. 2nd ed..
- Brandle J.E., McVetty P.B.E. Genotype x environment interaction and stability analysis of seed yield of oilseed rape grown in Manitoba. Can. J. Plant Sci. 1988;68:381-388.
- Ciha A.J. Slope position and grain yield of soft white winter wheat. Agron. J. 1984;76:193-196.[Abstract/Free Full Text]
- Cornelius P.L., Van Sanford D.A., Seyedsadr M.S. Clustering cultivars into groups without rank–change interaction. Crop Sci. 1993;33:1193-1200.[Abstract/Free Full Text]
- Fiez T.E., Miller B.C., Pan W.L. Winter wheat yield and grain protein across varied landscape positions. Agron. J. 1994;86:1026-1032.[Abstract/Free Full Text]
- Hnatowich, G.L. 1995. Anhydrous ammonia application and seeding in a single pass system. p. 104–113. In Proc. Western Canada Agronomy Workshop, Red Deer, AB. 5–7 July 1995. Potash and Phosphate Inst. of Canada, Saskatoon.
- Indorante S.J., Follmer L.R., Hammer R.D., Koenig P.G. Particle-size analysis by a modified pipette procedure. Soil Sci. Soc. Am. J. 1990;54:560-563.[Abstract/Free Full Text]
- Jaynes D.B., Colvin T.S. Spatiotemporal variability of corn and soybean yield. Agron. J. 1997;89:30-37.[Abstract/Free Full Text]
- Johnston A., LaFond G., Harapiak J., Head K. No-till spring wheat and canola response to side banded anhydrous ammonia at seeding. J. Prod. Agric. 1997;10:452-458.
- Jowkin, V. 1997. Impact of tillage management and landscape on nitrogen availability in cereal fallow cropping systems. Ph.D. diss. Univ. of Saskatchewan, Saskatoon (Diss. Abstr. NQ23943).
- McKeague, J.A. (ed.) 1978. Manual on sampling and methods of analysis. 2nd ed. Can. Soc. Soil Sci., Ottawa, ON.
- Pennock D.J., Anderson D.W., de Jong E. Landscape-scale changes in indicators of soil quality due to cultivation in Saskatchewan, Canada. Geoderma 1994;64:1-19.
- Pennock D.J., Zebarth B.J., de Jong E. Landform classification and soil distribution in hummocky terrain, Saskatchewan, Canada. Geoderma 1987;40:296-315.
- Solohub, M.P. 1997. Variable rate nitrogen fertilization in the Black soil zone of Saskatchewan. M.Sc. thesis. Univ. of Saskatchewan, Saskatoon.
- Stevenson F.C., Van Kessel C. A landscape-scale assessment of the nitrogen and non-nitrogen rotation benefits of pea. Soil Sci. Soc. Am. J. 1996;60:1797-1805.[Abstract/Free Full Text]
- Varvel G.E. The effects of anhydrous ammonia at planting on spring wheat and barley. Agron. J. 1982;74:1081-1083.[Abstract/Free Full Text]
- Walley F.L., Tomm G.O., Matus A., Slinkard A.E., van Kessel C. Allocation and cycling of nitrogen in an alfalfa–bromegrass sward. Agron. J. 1996;88:834-843.[Abstract/Free Full Text]
- Williams, P., D. Sobering, and J. Antoniszyn. 1998. Protein testing methods. p. 37–47. In D.B. Fowler et al. (ed.) Wheat production and marketing. Proc. Wheat Protein Symp., Saskatoon, SK. 9–10 Mar. 1998. Univ. Extension Press, Univ. of Saskatchewan, Saskatoon.
This article has been cited by other articles:
|
|
|
|
 
F. Walley, T. Yates, J.-W. van Groenigen, and C. van Kessel
Relationships Between Soil Nitrogen Availability Indices, Yield, and Nitrogen Accumulation of Wheat
Soil Sci. Soc. Am. J.,
September 1, 2002;
66(5):
1549 - 1561.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
