|
|
||||||||
Crop Dev. Cent., Univ. of Saskatchewan, Saskatoon, SK, S7N 5A8, Canada
* Corresponding author (Brian.Fowler{at}usask.ca)
Received for publication April 15, 2002.
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: CPSR, Canadian prairie spring red CPSW, Canadian prairie spring white GPC, grain protein concentration HRS, hard red spring HRW, hard red winter ESS, extra-strong spring SWS, soft white spring SWW, soft white winter
| INTRODUCTION |
|---|
|
|
|---|
Several studies have suggested that the close relationship between GPC and the amount of available soil N may allow GPC to be used as a postharvest indicator of the adequacy of N management (Pierre et al., 1977; Goos et al., 1982). The critical GPC for N sufficiency has been reported to be 8.8% for Stephen's soft white winter (SWW) wheat grown in Oregon (Glenn et al., 1985) and between 11.1 and 12.0% for dryland winter wheat produced on summer fallow in eastern Colorado (Goos et al., 1982). Selles and Zentner (2001) reported that a grain protein concentration of 12.8% was a reliable indicator of N sufficiency in hard red spring (HRS) wheat grown in southwestern Saskatchewan. In contrast, Fowler and Brydon (1989) reported that the N requirements for maximum grain yield are normally met when the GPCN response curve for Norstar winter wheat reaches approximately 13% under average to good weather conditions in Saskatchewan. Similarly, the critical GPC for spring wheat has been reported as 13.5% for both the eastern prairies (Flaten and Racz, 1997) and Montana (Long and Engel, 1998). This wide range of critical GPC values supports the conclusion drawn by Fowler et al. (1990) that there are important differences in GPCgrain yield relationships that depend on production area (environment) and cereal species and genotypes within species.
The objectives of this study were to quantify the important genotypic and environmental responses that determine GPC and yield potential in common wheat and to determine if GPC can be used as a practical indicator of crop N deficiencies grown under the variable environmental conditions of western Canada.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Trials that included spring wheat were grown under partial irrigation at Saskatoon in 1992, 1993, 1994, 1995, 1996, 1997, and 1998 and on dryland at Saskatoon in 1996 and 1997 and Clair in 1996, 1997, and 1998. Trials that included winter wheat were grown under partial irrigation at Saskatoon in 1993, 1995, 1997, and 1998 and on dryland at Saskatoon in 1997, Yorkton in 1997 and 1998, and at two locations at Clair in 1997. The cultivars AC Reed, Katepwa, BW90, Roblin, and AC Taber were included in the spring wheat trials starting in 1992. Glenlea replaced BW90 in 1995, AC Barrie was substituted for Roblin in 1997, and AC Vista replaced AC Reed in 1998. CDC Ptarmigan, CDC Kestrel, S86-101, Norstar, and Winalta were included in all winter wheat trials up to 1996 and in Saskatoon and Clair dryland trials in 1997. The winter wheat cultivars CDC Kestrel, CDC Clair, and CDC Osprey were grown in trials under partial irrigation at Saskatoon and dryland at Clair and Yorkton in 1997. The 1998 winter wheat cultivars were Norstar, Winalta, CDC Harrier, CDC Osprey, and CDC Clair.
All trials were direct-seeded into standing stubble from a previous crop (no-till) with a small-plot hoe-press drill. Experimental design was a four-replicate split plot, with N fertilizer rates as the main plots and cultivars as the subplots. Each plot was 5.5 m long and 1.2 m wide. Optimum seeding dates were achieved in all trials, and PO4 fertilizer was applied with the seed at recommended rates. Nitrogen fertilizer was added as early springbroadcast ammonium nitrate (3400) at 0, 40, 80, 120, 160, and 240 kg N ha-1. Other soil nutrients were not considered limiting. After removing approximately 30 cm from each end at maturity, the plots were direct-cut with a self-propelled small-plot combine. The outside two rows of each plot were not harvested. Exact plot lengths were recorded before harvest. The GPC was determined from Leco N x 5.7 (Leco Corp., St. Joseph, MI) (Am. Assoc. of Cereal Chem. Method 46-30) for each plot in each trial. Wheat grain yield and GPC have been reported at 135 g kg-1 moisture.
Analyses of variance were conducted to determine the level of significance of differences due to N levels and cultivars in each trial. Regression analyses of treatment means were used to plot curves that best described the response of grain yield and GPC to N fertilizer application. The peak four-parameter Weibull equation was employed to describe the grain yield response:
![]() | [1] |
The sigmoidal four-parameter Gompertz equation was used to describe the GPC response to N fertilizer applications:
![]() | [2] |
The peak three-parameter log normal equation was used to describe the relationship between grain yield and GPC using the individual plot data for each cultivar at each location:
![]() | [3] |
Nonlinear regression procedures outlined by SigmaPlot (SPSS, Chicago, IL) were used to provide least-squares estimates of the regression coefficients in these equations.
Two methods were employed to identify critical grain yieldGPC relationships: (i) Maximum grain yield and the N rates required to achieve maximum grain yield and 90 and 80% of maximum grain yield were estimated using the peak four-parameter Weibull equation. These N rates were then used to estimate the GPC at maximum grain yield and 90 and 80% of maximum grain yield using the sigmoidal four-parameter Gompertz equation (Fig. 1) . (ii) Grain protein concentration at maximum grain yield and 90 an 80% of maximum grain yield was also estimated using the peak three-parameter log normal equation (Fig. 2) . Because this study was only concerned with GPC at grain yields that were 80% or more of the maximum, initial decreases in GPCN responses at low levels of applied N were disregarded to increase the accuracy and simplify curve fitting (Fig. 1 and 2). Estimates of maximum and 90 and 80% of maximum grain yield and GPC at maximum grain yield and 90 and 80% of maximum grain yield for each cultivar in each trial were then subjected to analysis of variance using the General Linear Model procedure of Minitab 13 (Minitab, State College, PA). Adjusted means for these variables are reported.
|
|
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
The typical nonlinear GPCN response pattern (Fowler, 1998a), which includes the three phases designated as zones of minimum percentage, poverty adjustment, and luxury consumption by Macy (1936), were observed in these trials (Fig. 1). Low GPC was associated with low levels of residual soil available N and favorable growing conditions. In these instances, N fertilization stimulated large increases in grain yield that produced a lag phase (zone of minimum percentage) in the GPCN response curve. The lag phase was longest when cultivars with high grain yield potentials were grown under low levels of available soil N. Under these conditions, the correction of severe N stress by the addition of fertilizer N often produced an initial decrease in the GPCN response curve (Partridge and Shaykewich, 1972; Terman, 1979; Goos et al., 1982; Bole and Dubetz, 1986; Fowler and Brydon, 1989) that extended beyond the 40 kg ha-1 N level. The most extreme example of an initial decrease in GPC associated with the lag phase in this study occurred in the 1993 partial irrigation trial at Saskatoon. Very low residual N levels in the soil and favorable growing conditions resulted in a decrease in GPC from 119 to 101 g kg-1 with the first 40 kg ha-1 fertilizer N for both Winalta winter wheat and AC Taber spring wheat. Nitrogen fertilizer applications of >120 kg ha-1 were required before the GPC of these two cultivars exceeded the level found at the 0 kg ha-1 N rate in this trial (data not shown). In contrast, the lag phase of the GPC response curve became shorter as environmental limitations increased or cultivar grain yield potential decreased, and it often disappeared entirely in trials grown in fields with high levels of residual soil available N and/or under moderate or high drought stress.
Once cultivar yield potential or environmental factors other than available N became limiting to plant growth, excess N was utilized mainly for grain protein production, and the GPCN response curve entered an increase phase (zone of poverty adjustment). During this phase, GPC increased rapidly, even under favorable growing conditions. However, the response curve turned up at lower N levels and tailed off at higher GPC under poor compared with good growing conditions (data not shown).
The GPC response to increased N quickly diminished to near zero when cultivar yield potential or environmental factors, such as moisture, limited grain yield. The end of the increase phase and the start of the maximum phase (zone of luxury consumption) of the GPCN response curve usually occurred at approximately the same N rate as maximum grain yield was achieved (Fowler et al., 1990). A detrimental effect (Goos et al., 1982; Fowler et al., 1989) that resulted in yield depression was observed at high N levels.
Different GPCgrain yield relationships were associated with each of the three phases of the GPCN response curve. The lag phase of the response curve often had a negative slope. From the end of the lag phase to the point of maximum grain yield there was a positive correlation between grain yield and GPC that was due to increased N availability. Beyond the point of maximum grain yield, the correlation between GPC and grain yield was once again either nonsignificant or negative. Consequently, the relationship between GPC and grain yield depended on the region of the response curve sampled.
The peak four-parameter Weibull equation was employed to describe the grain yield response, and the sigmoidal four-parameter Gompertz equation was used to describe the GPC response to N fertilizer applications for each of the 99 genotypetrial comparisons made in this study (examples given in Fig. 1). Average reductions in sums of squares due to model were 97.2 and 98.9%, respectively, indicating that these equations provided an excellent fit to the observed data. Maximum grain yield (Table 1) and the N rates required to achieve maximum grain yield and 90 and 80% of maximum grain yield were estimated for each cultivar in each trial using the peak four-parameter Weibull equation. These N rates were then used to estimate the GPC at maximum grain yield and 90 and 80% of maximum grain yield using the sigmoidal four-parameter Gompertz equation. The GPC at maximum grain yield and 90 an 80% of maximum grain yield was also estimated by fitting the grain yield and GPC data for each of the 99 genotypetrial comparisons to the peak three-parameter log normal equation (examples given in Fig. 2). In this instance, average reduction in sums of squares due to model was 66.8%. The two approaches used the same database and arrived at similar estimates of maximum grain yield and GPC at maximum grain yield and 90 and 80% of maximum grain yield (Tables 1 and 2).
|
|
There were large differences in GPC and maximum grain yield (Tables 1 and 2) among the genotypes considered in this study. For example, average GPC at maximum grain yield ranged from 107 g kg-1 for CDC Ptarmigan to 159 g kg-1 for AC Barrie while average maximum grain yield ranged from 3906 kg ha-1 for Katepwa to 5844 kg ha-1 for CDC Ptarmigan (Table 1). The often reported strong negative correlation between cultivar GPC and grain yield (Terman et al., 1969; Terman, 1979) was especially evident in these comparisons, translating into more than a one-third tonne-per-hectare reduction in maximum potential grain yield for every 10 g kg-1 increase in GPC [Ymax (kg ha-1) = 9593 - 35 x Pmax (g kg-1); r2 = 0.76]. Among western Canadian wheat market classes, the GPC at maximum grain yield was lower for the high-yielding SWW wheat cultivar CDC Ptarmigan than for the cultivars that represented the SWS, HRW, CPSR, and CPSW classes. In turn, the GPC at maximum grain yield was lower for cultivars that represented the HRW, CPSR, and CPSW classes than for the HRS cultivars, which had a very low maximum grain yield potential.
The observations made in this study help us to understand the limitations in our ability to select for both grain yield and GPC in wheat breeding programs. Wheat quality standards in western Canada have been maintained by a restrictive cultivar registration system that only allows the commercial release of lines with quality characteristics that are equal to or better than designated check cultivars for each quality class. The HRS class has been the mainstay of the western Canadian wheat industry, and there has been nearly 100 yr of intensive breeding effort concentrated on improving cultivars in this class (DePauw et al., 1995). The breeding efforts on the rest of the quality classes have been much less intensive, and the intermediate-GPC classesESS, CPSR, and CPSWhave only been seriously pursued in the last 15 to 20 yr (DePauw, 1995). Interestingly, in spite of all the attention in the last century, modern cultivars of the HRS wheat class have the lowest maximum grain yield potential (Tables 1 and 2). In contrast, the SWW wheat cultivar CDC Ptarmigan, which originated from a group of three lines selected from a single hard wheat x soft wheat cross in a HRW breeding program, had the highest maximum grain yield potential of all the cultivars considered in this study. These observations indicate that, while improvements have been made within classes, even small increases in grain yield are extremely difficult to achieve when breeding programs are also expected to maintain or improve GPC. Certainly, the production of high-yielding cultivars with high GPC is more complicated than simply stacking yield genes in a high-GPC genetic background or vice versa.
Part of the explanation for the negative relationship between cultivar grain yield potential and GPC lies in the fact that the amount of N available to a plant for protein production depends on a substrate-inducible, relatively unstable enzyme, nitrate reductase, which is regulated by the level of available soil N (Eilrich and Hageman, 1973). Cultivars growing side by side have access to similar amounts of available soil N, and GPC is determined by the ratio of grain protein yield to total grain yield. As a result, higher-yielding cultivars will have lower GPC than lower-yielding cultivars unless they have an increased ability to extract N from the soil, translocate it to the grain, or use it in protein synthesis. However, these observations do not explain the strong negative correlation between cultivar GPC and maximum potential grain yield observed in the present study where plant available N was not limiting. The higher energy requirements for protein compared with carbohydrate synthesis (Penning de Vries et al., 1974) and/or critical genetic adjustments that limit grain yield potential may provide possible explanations for this negative relationship. Whatever the cause, the general rule in effective breeding programs appears to be that the higher a cultivar's relative GPC is, the lower its maximum grain yield potential.
Grain Protein Concentration as a Postharvest Indicator of Crop Nitrogen Deficiencies
This study focused on important genotypic and environmental interactions that determine the GPC and yield potential in common wheat. From a practical standpoint, the results indicate that GPC may be a useful postharvest indicator of N deficiencies for crops grown under high N stress, but caution must be used when the goal is to optimize N management systems. The following limitations should be kept in mind:
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. M. Carr, Glenn. B. Martin, and R. D. Horsley Wheat Grain Quality Response to Tillage and Rotation with Field Pea Agron. J., October 21, 2008; 100(6): 1594 - 1599. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. N. Otteson, M. Mergoum, and J. K. Ransom Seeding Rate and Nitrogen Management on Milling and Baking Quality of Hard Red Spring Wheat Genotypes Crop Sci., March 19, 2008; 48(2): 749 - 755. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. N. Otteson, M. Mergoum, and J. K. Ransom Seeding Rate and Nitrogen Management Effects on Spring Wheat Yield and Yield Components Agron. J., November 6, 2007; 99(6): 1615 - 1621. [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 | |||