| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Plant Sciences, North Dakota State Univ., Fargo, ND, 58105
* Corresponding author (Mohamed.Mergoum{at}ndsu.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AACC, American Association of Cereal Chemists • GPC, grain protein content • GVW, grain volume weight • HRSW, hard red spring wheat • NDSU, North Dakota State University • NUE, nitrogen use efficiency • RCBD, randomized complete block design • TKW, thousand kernel weight • UAN, urea ammonium nitrate
Department of Plant Sciences, North Dakota State Univ., Fargo, ND, 58105
* Corresponding author (Mohamed.Mergoum{at}ndsu.edu)
Received for publication January 2, 2007.
Seeding rate, N level, and N application timing are key management factors for spring wheat (Triticum aestivum L.) production in North Dakota. Experiments were conducted under dryland (Casselton, ND) and irrigated (Carrington, ND) conditions in 2003 to 2005 to determine the optimum combination of seeding rate and N management to maximize yield of hard red spring wheat (HRSW). Treatments consisted of a factorial combination of HRSW genotypes (‘Briggs’, ‘Alsen’, ‘Granite’, and ND 740), seeding rates (2.9 and 4.2 million seeds ha–1), N levels (140 and 224 kg ha–1 for the dryland site; 168 and 280 kg ha–1 for the irrigated site), and N application timing (preplant, two-split, and three-split). Nitrogen level, N timing, and seeding rate showed no significant effect on grain yield across environments. However, genotype significantly influenced grain yield. Increasing seeding rate failed to increase grain yield of the three cultivars, but ND 740 was most productive at the lowest seeding rate. Increasing the level of N significantly increased grain protein content (GPC) over all environments. Grain volume weight (GVW) and thousand kernel weight (TKW) decreased with increasing N level and were influenced by genotype. Increased seeding rate significantly increased plant stand and tiller number while increasing N levels significantly increased head count. These data indicate that although genotype was the primary factor in determining grain yield, GPC, and agronomic traits, individual genotypes responded differently to varying seeding rates and N management practices.
Abbreviations: AACC, American Association of Cereal Chemists • GPC, grain protein content • GVW, grain volume weight • HRSW, hard red spring wheat • NDSU, North Dakota State University • NUE, nitrogen use efficiency • RCBD, randomized complete block design • TKW, thousand kernel weight • UAN, urea ammonium nitrate
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cultivar selection plays a very important role in determining grain yield and quality. Genotypes vary widely in yield potential, GPC, and agronomic traits. Previous studies (Terman, 1979; Fowler et al., 1990; Coventry et al., 1993) have emphasized the importance of genotype in determining grain yield and protein content.
Depending on the environment and genotype, seeding rate can impact wheat tillering and grain yield. Higher seeding rates compensate for reduced tiller development and promote more main stem spikes (Coventry et al., 1993; Staggenborg et al., 2003), which can be favorable, especially to cultivars that tend to produce fewer tillers. Wiersma (2002) however, emphasized that cultivars differ genetically for yield components at a given yield level and that individual cultivars should be tested at different sowing rates to determine their optimum seeding rates. Optimum seeding rates for wheat to achieve maximum grain yield vary across growing regions of North America. Staggenborg et al. (2003) reported inconsistent winter wheat grain yield responses to seeding rate, which was highly dependent on environmental conditions.
Nitrogen represents one of the most expensive inputs; yet, it is one of the most critical inputs affecting yield and quality. The response of spring wheat to N fertilizer can vary with rate and timing of application relative to plant development (Mossedaq and Smith, 1994). Nitrogen demand increases sharply just before onset of stem elongation, the most rapid phase of crop growth (Darwinkel, 1983). Other studies have found the grain yield of spring sown cereals is not as responsive to N application timing, primarily due to rapid growth and development resulting in high N demand during early growth stages (Easson, 1984). Westfall et al. (1996) found that N rate was more important in determining grain yield under dryland conditions than N placement, whereas N placement is more important when moisture was not a limiting factor.
Foliar application of N has often been used as a way to increase nitrogen use efficiency (NUE) and GPC in wheat. Bly and Woodard (2003) found post-pollination foliar N applications increased grain protein, especially when the planned yield goal was exceeded. Schatz et al. (1991) showed that postanthesis application of N increased grain protein by 0.95% points at 33.6 kg ha–1 of urea ammonium nitrate (UAN). Similarly, Staggenborg et al. (2003) found in 3 yr of research that grain protein increased as applied N rates increased. They also reported that response to N depended on previous crop. Different seeding and N rates are required to optimize wheat yields when planted after grain sorghum [Sorghum bicolor (L.) Moench] and soybean [Glycine max (L.) Merr.]. Current NUE is estimated to be only 33% in wheat (Mullen et al., 2003), indicating that much of the N is susceptible to loss. Applying N when a crop response to N is expected can help improve NUE. In recent years intensive wheat management, including N management and seeding rates, have been promoted. The use of split N applications to improve NUE and higher seeding rates to promote more main stem spikes in an effort to increase grain yields are the main areas of intensive wheat managements. However, limited information regarding the effects of intensive wheat management on spring wheat yield and quality has been published. Therefore, this study was conducted to determine the effects of seeding rate, N rate, and timing of N application on major agronomic and quality traits of modern HRSW genotypes under North Dakota conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experiments were set up as a randomized complete block design (RCBD) with split-split plot arrangements where N level was the main plot, N application timing was the subplot, and genotype by seeding rate was the sub-subplot treatments. All experiments were replicated four times. Each treatment consisted of a 7-row plot, 1.22 m wide, and 2.44 m long. Plots of equal size were planted on the edges of blocks to minimize border effects.
Four genotypes were used in this study: Briggs (PI 632970), Alsen (Frohberg et al., 2006), Granite, and ND 740. Briggs, released by South Dakota State University in 2002, has high grain yield and protein, high tillering potential, medium early maturity, and medium straw strength. Alsen, released by North Dakota State University (NDSU) in 2000, has average grain yield, high grain protein, medium tillering potential, medium early maturity, and strong straw. Alsen is the most popular cultivar in North Dakota because of its relative resistance to Fusarium head blight (FHB) {caused by Fusarium graminearum Schwabe [telomorph Gibberella zeae (Schwein.) Petch]}. Granite, released by WestBred, LLC (Bozeman, MT) in 2002, exhibits high grain yield, high grain protein, very strong straw, medium late maturity, and low tillering potential. ND 740, an experimental line developed by NDSU HRSW breeding program, has shown high grain yield potential, medium straw strength, average grain protein, and high tillering potential. Except Alsen, the other genotypes are susceptible to FHB.
Two seeding rates, 2.9 million (low seeding rate) and 4.2 million (high seeding rate) seeds ha–1 were used in this experiment. The low seeding rate is based on the current NDSU recommended seeding rate for spring wheat while the high seeding rate is recommended by intensive wheat management groups to promote less tillering and more main stem spikes to obtain maximum grain yield.
Two N levels included in the study were determined based on the N requirement of an average and highest likely attainable yield for each location. At Casselton, N rates were 140 and 224 kg ha–1, targeting yields of 3362 and 5380 kg ha–1, respectively. At the irrigated site in Carrington, 168 and 280 kg ha–1 N were applied targeting yields of 4034 and 6725 kg ha–1, respectively. Based on available soil N, adjustments in N applied were made as necessary to each N-level treatment. The actual amount of N applied to each treatment of each experiment was adjusted for N (nitrate N) levels found in the top 610 mm of soil.
Three N timings were used in the study: preplant, two-way split, and three-way split. For the preplant N treatment, all N was applied as dry urea (46–0–0) and incorporated into the soil before planting. In the two-way split, N was applied at preplant and at the five leaf stage (Zadoks 15 stage) (Zadoks et al., 1974), at a 1:1 ratio. In the three-way split, N was applied at preplant, the five-leaf stage (Zadoks 15 stage), and postanthesis (Zadoks 69 stage) at a 1:1:1 ratio. Post-planting N applications were made with UAN solution (280 g N kg–1). Nitrogen streamer bars were used to minimize leaf burn for the Zadoks 15 application. A flat-fan nozzle was used for the Zadoks 69 stage application to ensure adequate spike and flag leaf coverage for maximum absorption.
Weeds were initially controlled with an application of bromoxynil octanoate ester (3, 5-dibromo-4-hydrozybenzonitrile) and MCPA isooctyl ester [(4-chloro-2-methylphenoxy) at 0.28 kg a.i. ha–1, and fenoxaprop-p plus mefenpyr (ethyl 2-[4-[(6-chloro-2-benzoxazoly) oxy] phenoxy] propanoate at 0.067 kg a.i. ha–1 at Zadoks 15 growth stage. Additional hand weeding at later stages was also used as needed. Fungicide to control leaf spotting diseases was applied at Zadoks 39 stage using 0.126 kg a.i. ha–1 propiconazole (1-[[2-(2,4-dichlorophenyl)-4-propyl-1,3-dioxolan-2-yl] methyl]-1H-1,2,4-triazole). A split application (0.063 kg a.i. ha–1) of tebuconazole (a-(2-(4-chlorophenyl)ethyl)-a-(1-1-dimethylethyl)-1H-1,2,4-triazole-1-1-ethanol) to control FHB was made at the beginning of flowering (Zadoks 61 stage) and 7 d later.
Data were collected on plant stand, tiller number, spike number, heading date, plant height, and lodging. Plant stand was measured at Zadoks growth stage 12 by averaging the total number of plants in two random 61-cm-long row sections. The number of total stems at Zadoks growth stage 37 and spike number were taken on the same row section used for plant stand. Tiller number was determined by subtracting the stand count from total stem number. Spike number was determined by averaging the total number of fully developed spikes. Heading date was determined by counting the number of days from planting to 50% of spikes fully emerged from the boot taken at Zadoks growth stage 55. Lodging scores were rated on a scale of 0 to 9 with 0 = erect and 9 = flat. Plant height was determined at Zadoks growth stage 91 by measuring the distance from soil surface to the tip of spikes, excluding awns.
Grain samples were cleaned on a Clipper grain cleaner following harvest. Grain samples for each plot were weighed to determine grain yield (kg ha–1). The GVW was determined according to AACC standard method 55-10 (AACC, 2000). The TKW (g) was determined based on weight of 250 seeds counted on an electric seed counter (Seedburo Equipment Co., Chicago, IL). Grain protein content (g kg–1) was determined for each plot sample on a whole grain basis using Tecator Infratec 1226 Grain Analyzer (Foss, Eden Prairie, MN).
Data analysis was performed using SAS (SAS Institute, 2004). In the combined analysis, environment was considered a random effect, while genotype, seeding rate, N level, and N timing were considered fixed effects. F-tests were conducted using the appropriate denominator for the error term. Mean separation tests were conducted using an F-protected LSD (P = 0.05) as described by Steel and Torrie (1997).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
The significant SR x G interaction can be attributed to ND 740 producing significantly greater yield at the lower seeding rate (Table 5 ) while the yield of the other genotypes was not significantly affected by seeding rate. The grain yield of ND 740 increased by 124 kg ha–1 at the lower seeding rate compared with the higher seeding rate. ND 740 along with Briggs has been found to produce many tillers (Table 6 ). Briggs, Alsen, and Granite were nonresponsive to seeding rate changes.
|
|
The effect of the T x E interaction on GPC means is shown in Table 3. In 2003, the preplant N treatment produced the greatest GPC over both split N applications at Carrington, but was equal to the three-split application and 2 g kg–1 higher than the two-split application at Casselton (2003). In 2004 and 2005, however, the three-split N application produced the greatest GPC among treatments.
The G x E and SR x E interactions were significant for GPC (Table 2). Nevertheless, the ranking of the genotypes for GPC was identical in all environments. Similarly, the higher seeding rate produced a higher GPC across all environments, though to a greater extent in some environments than others (data not shown).
Genotypes differed significantly for GPC with Granite producing 17 g kg–1 more GPC than the experimental genotype ND 740, while Briggs and Alsen were nearly equal in GPC (Table 7 ). As expected, increasing the N level significantly increased GPC across all environments with a 7 g kg–1 increase in GPC at the highest N level.
|
The G x E interaction and environment, genotype, and N level main effects were significant for GVW (Table 2). The G x E interaction was significant due to Granite producing a 19 and 22 kg m–3 GVW increase over the next closest genotype at Carrington 2004 and Casselton 2005, respectively (data not shown).
Genotype significantly affected GVW (Table 2). Among genotypes, Granite had the greatest GVW across five environments at 819 kg m–3, followed by ND 740, Alsen, and Briggs (Table 7). Nitrogen level also significantly influenced GVW with the lowest N level producing a 4 kg m–3 increase in GVW compared with the higher N level (Table 7).
The N timing x N level (T x N) interaction was significant for TKW (Table 2). At the lower N level, TKWs were similar among N timing treatments. However, at the higher N level, the three-split and preplant N timing treatments produced 0.8 and 0.7 g greater TKW compared with the two-split N application (data not shown). The G x E interaction for TKW was significant though the ranking of genotypes was similar in all environments (data not shown). Genotype and N level main effects also influenced TKW. Among genotypes, ND 740 had the greatest TKW followed by Briggs, Alsen, and Granite (Table 7). At the lowest N level, a significant increase of 0.9 g in TKW occurred.
Plant height was influenced by a G x E interaction (Table 2). However, there was no change in ranking between genotypes for plant height in all environments. Seeding rate, N level, and N timing main effects had no effect on plant height. However, genotypes differed significantly for plant height (Table 2) with the cultivars Briggs and Alsen being the tallest (83 cm), followed by Granite and the genotype ND 740 (80 cm) (Table 6).
Heading date was influenced by the interactions of T x E, G x E, and SR x E (Table 2). These interactions were related to differences in magnitude; the ranking of the various treatments of interest did not differ between environments (data not shown). Heading date was also influenced by genotype and N level main effects (Table 2). Among genotypes, Briggs had the earliest heading date, followed by ND 740, Alsen, and Granite (Table 6). Increasing the N level slightly delayed heading date across environments (Table 6).
No lodging occurred at Casselton (dryland) in 2003, 2004, or 2005 regardless of genotype, seeding rate, N level, or N timing. However, significant lodging occurred at Carrington (irrigated). Lodging was primarily influenced by genotype, with Granite being the most resistant cultivar to lodging, followed by Alsen, Briggs, and ND 740 (Table 6). Significant N x E, and T x E interactions were due to differences in magnitude; treatment rankings did not change between environments (data not shown).
Stand count was influenced by genotype and seeding rate (Table 2). Among genotypes, Alsen produced the highest average stand count of 400 plants m–2, followed by Briggs, ND 740, and Granite with 388, 381, and 362 plants m–2, respectively (Table 6). Stand count increased by 91 plants m–2 from the lowest to highest seeding rate. The T x E, G x E, and SR x E interactions were significant, but due to differences in magnitude; rankings did not change between environments (data not shown).
Tiller number was influenced by N x E, T x E, T x N, G x E, SR x E, and SR x G interactions and environment, N timing, genotype, and seeding rate main effects (Table 2). The T x N interaction was significant for tiller number due to a 9 and 14% increase in tiller number with the two-split and preplant N timing treatment, respectively, when compared with the three-split treatment at the lower N level. Tiller number was equal for the preplant and two-split N timing treatments at the higher N level, and increased by 14% compared with the three-split N timing treatment (data not shown). Genotype interacted with seeding rate for tiller number with varying results. Briggs showed no difference between seeding rates while Alsen, ND 740, and Granite produced 15, 14, and 13% more tillers at the lowest seeding rate, respectively. Also, N level interacted significantly with N timing to affect tiller number. At the lowest N level, each successive split application reduced tiller number by approximately 8%. However, at the highest N level, tiller numbers remained unchanged between the preplant and two-split applications, whereas a 12% reduction in tillers occurred with the three-split application. The N x E, T x E, G x E, and SR x E interactions were all significant but were due to differences in magnitude; rankings did not change between environments. The cultivar Briggs produced the greatest tiller number averaging 485 tillers m–2 across all environments. ND 740, Alsen, and Granite produced an average of 423, 371, and 311 tillers m–2, respectively (Table 6).
Spike counts were taken before harvest. A significant SR x G interaction occurred (Table 2) due to a 12 and 5% increase in Briggs and Alsen spike counts, respectively, when seeding rate was increased whereas no increase occurred in other genotypes. The number of spikes m–2 was also influenced mainly by genotype and N level main effects. Among genotypes, Briggs produced the greatest spike count, followed by ND 740, Alsen, and Granite (Table 6). Increasing the N level increased spike counts by 7% across all environments. Although N timing did not affect total spike count (Table 7), tiller number was reduced with the three-split N application suggesting survival of spikes may be greater with three-split N applications.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Grain yield was not improved by increasing the seeding rate. Three of the four genotypes had no response to increasing seeding rate, whereas the yield of the experimental genotype ND 740 actually declined at the higher seeding rate. ND 740 has been found to produce many tillers. Thus, this characteristic may allow it to produce greater yields at the lower seeding rate. Briggs, Alsen, and Granite were nonresponsive to seeding rate changes for grain yield. Granite, a cultivar known to produce very few tillers, did not respond significantly to increase seeding rate for grain yield. These results agree with those of Wiersma (2002), suggesting that cultivars are genetically different for yield components and may require different seeding rates to optimize grain yield.
Nitrogen timing and environment interacted to produce mixed results on grain yield. The three-way split yielded more than the preplant N timing under irrigated conditions at Carrington 2003, while at the same location in 2004, the three-split treatment yielded less than the preplant and two-split N treatments. These results are somewhat inconsistent with Mossedaq and Smith (1994), who reported the greatest response when N was applied just before stem elongation and the lowest response when N was applied at anthesis. Under dryland conditions, the preplant N treatment produced a greater grain yield compared with the three-split N treatment in two of three environments. These results agree with Easson's findings (Easson, 1984) that spring-seeded cereals may not respond to N fertilization timing due to the short growing season where N demand is relatively high soon after emergence. Therefore, economic factors associated with each N-timing application needs to be considered in determining which N timing is used. Additional precautions also need to be considered for the environmental conditions necessary for post-applied N with streamer bars since adequate rainfall is needed to move the N into the soil profile to be taken up by the plant roots.
Grain protein content increased as N levels increased and was consistent with previous findings (Terman, 1979; Fowler, 2003; Staggenborg et al., 2003). Across environments, an increase in GPC was achieved as N level was increased from the lowest to highest level. These results reinforce the importance of N management in improving spring wheat quality. Maintaining adequate N levels is important since many of the newer high-yielding cultivars commonly have lower GPC, requiring careful N management to meet grain quality standards. Post-pollination (three-split) N treatment gave the highest grain protein concentration in three out of five environments and was in agreement with previous studies (Bly and Woodard, 2003; Woolfolk et al., 2002).
Genotype consistently influenced yield components in this study. Granite produced the greatest GVW across environments whereas ND 740 produced the greatest TKW. Stand count, tiller number, and spike count were all significantly influenced by genotype and seeding rate. Alsen and the highest seeding rate significantly increased stand count, while Briggs and the lowest seeding rate produced the greatest tiller number. Briggs produced the greatest spike count among genotypes. Other studies (Coventry et al., 1993; Hucl and Baker, 1989) have reported similar differences among genotypes for yield components. Plant height was influenced by genotype with Briggs and Alsen the tallest, whereas Granite and ND 740 were the shortest. Lodging was also influenced by genotype, with Granite being the most resistant, and ND 740 the most susceptible to lodging.
Overall, all factors included in this study showed an important role in any intensive wheat management program. Previous studies (Coventry et al., 1993) showed that cultivar selection is an important factor for maximizing wheat grain yield. Most new HRSW cultivars offer the latest and best in disease protection, high yield potential, and acceptable quality traits. Cultivar selection should not be based on just yield potential alone, but also quality attributes such as GPC, straw strength, disease resistance, GVW, and certainly milling and baking traits. Previous findings (Fowler et al., 1990; Terman 1979) have emphasized the important role of genotype and environment in determining the level of GPC in wheat.
Increasing the seeding rate did not produce a higher grain yield with any of the genotypes used in this study. Therefore, there appears to be no advantage to increasing seeding rates from the current recommended rate for HRSW. Nitrogen levels in this study were based on average and highest attainable yield goals. Increasing the N level did not increase grain yield but did increase GPC, a very important criteria for quality, and therefore, wheat grain price. Nitrogen timing also did not influence grain yield but produced mixed results on GPC. Therefore, growers need to consider N cost, time constraints, environmental conditions, and equipment needs when deciding on N fertilizer timing.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, ND) from 2003, 2004, and 2005.