Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published in Agron J 100:681-689 (2008)
DOI: 10.2134/agronj2007.0179
© 2008 American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Agricola
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Related Collections
Right arrow Soybean
Right arrow Crop Growth and Development
Right arrow Crop Physiology & Metabolism
Right arrow Seed Quality
Right arrow Plant and Environment Interactions

SOYBEAN

Canopy Nitrogen Reserves: Impact on Soybean Yield and Seed Quality Traits in Northern Latitudes

Seth L. Naeve*, Tracy A. O'Neill and Jill E. Miller-Garvin

Dep. of Agronomy and Plant Genetics, Univ. of Minnesota, 411 Borlaug Hall, 1991 Upper Buford Cir., St. Paul, MN 55108. Research supported in part by the Minnesota Agricultural Experiment Station

* Corresponding author (naeve002{at}umn.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soybean [Glycine max (L.) Merr.] seed produced in the upper Midwest of the United States is lower in protein concentration than soybean produced throughout much of the Corn Belt, the southern United States, and Brazil. As protein comprises on average more than 400 g kg–1 of the soybean seed, yields in northern latitudes may be limited by seed protein accrual. Our objective was to examine the role of canopy N reserves at maximum vegetative growth (R5.5) on seed yield, and seed protein, oil, and sulfur concentrations in the upper Midwest. Six commercial cultivars with similar yield potentials and maturities were grown in 10 central Minnesota environments. Mean cultivar seed yields ranged from 3.1 to 3.2 Mg ha–1, while mean environment yields ranged from 1.7 to 5.0 Mg ha–1. High-yielding environments produced soybean plants with greater R5.5 canopy N reserves and total dry matter (DM) and seed with low protein and S concentration. Mean R5.5 canopy N quantities ranged from 117 to 309 kg ha–1. These quantities were sufficient to satisfy seed N requirements which ranged from 43 to 91 kg ha–1. While R5.5 canopy N and DM may have a positive influence on seed yields in these upper Midwestern environments, seed protein concentration effects appear to be mediated through effects on seed yield. A positive correlation between yield and seed N to S ratio noted here indicates that sulfur may be more limited than protein at high yield levels. Oil concentration was not affected by yield.

Abbreviations: DM, dry matter


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Received for publication May 29, 2007.
   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SOYBEAN YIELDS in northern United States (North Dakota, South Dakota, and Minnesota) averaged 24% lower from 2002 to 2006 than yields from states near the center of the Corn Belt (Iowa, Indiana, Illinois, and Ohio) (USDA-National Agricultural Statistics Service, 2006). Protein concentrations in soybean crops from these northern states were ~1% lower as well (Hurburgh et al., 1990; Brumm and Hurburgh, 2006), although U.S. soybean crops overall are deficient in protein when compared with those from Brazil (Mounts et al., 1990; Hurburgh et al., 1990). Soybean seed yield can be thought of in terms of yield components such as seeds per node, nodes per plant, and plants per hectare, or in terms of accumulation of individual seed constituents, such as protein, oil, and carbohydrates. As protein comprises more than 400 g kg–1 of soybean seed dry weight (Wilcox and Guodong, 1997), it is likely that protein deposition in developing soybean seed is a primary limiter to higher seed yields in the North.

Although soybean is capable of fixing large quantities of atmospheric N (Harper, 1987; Imsande, 1989), the developing seed's demand for N is so great that atmospheric and soil N sources are not large enough to fulfill this requirement (Sinclair and de Wit, 1976; Imsande, 1989). Leaf senescence, which includes a nearly complete cellular disassembly and the subsequent mobilization of released materials, is controlled largely by the nuclear genome (Thomas and Stoddart, 1980). Zeiher et al. (1982) and Egli et al. (1983) found from one-third to the entire seed's N comes from mobilized N. Others (Hanway and Weber, 1971; Loberg et al., 1984; Imsande and Edwards, 1988) found that soybean seed N is composed of approximately half mobilized N under normal field conditions. Egli et al. (1983) concluded that the proportion of redistributed N that is found in the seed at harvest is more dependent on the amount of N stored in the vegetative tissues than it is on the amount of N either fixed or obtained from the soil during seed filling. Sinclair (1998) states that increased vegetative N storage is essential for increased harvest index and ultimately yield in high N seed crops like soybean. Loberg et al. (1984) and Vasilas et al. (1995) found soybean to mobilize 61 to 82% and 66 to 79%, respectively, of its vegetative N to seed. Loberg et al. (1984) concluded that at these levels it is unlikely that N mobilization efficiency could be further optimized. Egli and Bruening (2007) examined seed growth rates for both DM and N and found that N accrual in the seed does not limit total DM accumulation. Since final seed N concentration did not affect the rate or duration of seed DM accrual, total plant N supply to the seed may be the primary limiter to seed protein and yield accrual (Egli and Bruening, 2007).

Several groups have examined cultivar effects related to mobilized N and seed yield, each across only two site years. Zeiher et al. (1982), Loberg et al. (1984), Vasilas et al. (1995), and Kumudini et al. (2002) examined 8, 15, 10, and 4 genotypes, respectively. Results from Zeiher et al. (1982), Vasilas et al. (1995), and Kumudini et al. (2002) indicated that mobilized N was weakly correlated with seed yield. Loberg et al. (1984) found a strong correlation between mobilized leaf N and yield in 1 yr (r = 0.73), but a weaker one the following year (r = 0.43).

Shibles and Sundberg (1998) examined a set of 63 cultivars, ancestral lines, and plant introductions in 2 yr near Ames, IA. Total leaf N at R5 (Fehr and Caviness, 1977) was correlated with seed yield each year (r = 0.37 and 0.31). While these genotypes had wide-ranging yields (2.4–4.4 Mg ha–1 and 2.2–3.8 Mg ha–1 in each of the 2 yr), there were significant differences in lodging and maturities. Adjusting for these two factors, leaf N at R5 accounted for 40 and 34% of the genotypic variation among lines for yield in the 2 yr of this study.

As with N, Naeve and Shibles (2005) found S mobilization to seed dependent on the quantity of S stored in leaf tissue. The concentration of the S-containing amino acids cysteine and methionine in soybean protein is considered limiting to growth of humans and monogastric livestock (Krishnan, 2005). Perhaps, additional vegetative protein storage may promote mobilization of both N and S to the seed.

Soybean seed oil concentration may be affected by genotype, and oil and protein are negatively correlated (Wilcox and Guodong, 1997; Helms and Orf, 1998). Various environmental factors appear to affect oil concentration directly; however, ambient temperature during seed filling may be the best characterized factor. Piper and Boote (1999) found soybean seed oil concentration to increase with increasing temperatures at a rate of about 10 g kg–1 °C–1 at 15°C to about 2.1 g kg–1 °C–1 at 25°C.

Due to the abbreviated seed-filling period of northern grown soybean crops, it is possible that whole-plant vegetative N storage may be more important for N mobilization and yield formation than previously described. All of the studies mentioned above were conducted south of 42° north latitude. Soybean canopies of northern soybean tend to be shorter with fewer nodes than those grown in the central Corn Belt and in southwestern United States. We, therefore, hypothesized that canopy development and N storage is crucial for protein deposition and yield formation in northern latitudes. Our study was intended to examine the contribution of mobilized N to seed yield and quality traits in six commercial soybean varieties with similar maturities and yield potentials across a broad range of Minnesota environments.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Ten field locations across four counties (Dakota, Ramsey, Sherburne, and Stevens) in central Minnesota were selected for studies in 2003, 2004, and 2005 (Table 1 ). Individual locations were selected to provide a wide range of soil types. While locations were not entirely unique to commercial soybean production in Minnesota, as a whole, these locations should not be considered as representative of commercial soybean production in the state. All field locations were considered as unique environments as none were located on identical field sites across years. Soils differed considerably across environments. Six commercial soybean cultivars were chosen for similarity of maturity date, yield, and variation in protein and oil concentrations, based on University of Minnesota Soybean Variety Testing Program in 2001 (Orf et al., 2002). The six cultivars were Mallard 1511 (genotype #1), Asgrow AG1401 (#2), Pioneer 91B03 (#3), KCS Challenger K121 (#4), Mustang 153 (#5), and High Cycle 2131 (#6).


View this table:
[in this window]
[in a new window]

  		Table 1. Longitude, latitude, soil type, and daily minimum, maximum, and average air temperatures during the seed-filling period (15 August –14 September) for the 10 locations used in the 2003–2005 studies in Minnesota.

 
Experimental design at each location was a randomized complete block with four replications. Soybean cultivars were planted in eight-row plots with 76-cm row spacing using a small-plot, cone-type research drill. Plots measured 7.6 m in length. Plots were seeded at 395,000 seeds ha–1. All locations were planted between 16 and 28 May each year. Weeds were controlled by hand weeding and two applications of glyphosate [N(phosphonomethyl)glycine] at 0.84 kg a.e. ha–1 as needed.

The Becker site is located on a coarse loam soil requiring irrigation. Irrigation was provided twice weekly throughout the season via a riser and sprinkler system. Irrigation was delivered by a checkbook ET system developed for this site (Wright and Bergsrud, 1991).

At all locations 1.52 m2 areas were harvested for total DM and total N stored at R5.5 (approximately 7 d after R5.0) (Fehr and Caviness, 1977). This stage falls midway between R5 and R6 and is identified in the field by the development of terminal leaflets that appear as several small leaves with very short internodes at the apex of the plant. It is our experience that vegetative growth continues past R5, thus R5.5 captures the maximum vegetative growth of the soybean plant. Maximum plant height, node number, and leaf area are thought to occur midway between R5 and R6 (Pedersen, 2004). While some early seed tissue is present at R5.5, we believe that the seed development at that time is in ‘phase I’ (Adams and Rinne, 1980), a phase in which all of the pod and seed structures are formed, but rapid accumulation of storage reserves has not yet begun (Egli, 1998). Further, seed growth before R6 constitutes <20% of total harvested dry matter (data not shown). R5.5 canopy N harvests were taken from bordered sections of rows two and three of each plot. One meter lengths of the two rows were harvested by hand at ground level with pruning shears. Wet weights were recorded and 10-plant subsamples were weighed and returned to the lab for analysis. Whole plants were dried at <60°C, weighed, and ground with a Wiley mill (Thomas Scientific, Swedesboro, NJ) to 6 mm. Subsamples were subsequently ground to 1 mm in a Foss/Tecator Cyclotec mill (model 1093, Foss North America Inc., Eden Prairie, MN) and analyzed for crude protein by Near Infrared Spectroscopy (NIR) with a Foss full scanning 6500 monochrometer (Foss North America, Eden Prairie, MN) fitted with calibration equations this project developed using a total of 791 samples from 2003, 2004, and 2005 (after Halgerson et al., 2004). Crude protein concentration was measured by combustion analysis (LECO method, Leco Corp., St. Joseph, MI) at the Experiment Station Chemical Laboratories, University of Missouri, using AOAC Official Method 990.03 (AOAC, 2000). A validation survey using 25 independent samples found a standard error of cross-validation (SECV) of 0.74 g kg–1 and a 1-VR (1 minus the ratio of unexplained to explained variance) of 0.996.

Rows six and seven of each plot were harvested at physiological maturity with a small-plot combine. Seed mass and moisture concentration were recorded in-field, and a 1-kg subsample was retained for further analysis. Protein and oil analyses were conducted using NIR with the same instrument described above but fitted with equations developed for whole soybean seed.

Sulfur analysis was completed by the Research Analytical Lab at the University of Minnesota. For every sample, 0.100 to 0.150 g was weighed into a ceramic boat and covered with about 0.5 g of tungsten oxide (ComCat accelerator, Leco Corp., St. Joseph, MI). Samples were combusted in a furnace at 2500°F (1371°C) in an oxygen rich atmosphere; this produced SO2 which was carried by the oxygen stream to an infrared detector where it was quantified (model LECO S-144DR, Leco Corp., St. Joseph, MI). Compositional data are presented on a dry matter basis throughout.

Analyses of variance using Type II sums of squares and regression analyses were conducted using PROC MIXED and PROC REG, respectively, in SAS v. 9.1 (SAS Institute, Cary, NC). Homogeneity of variance was tested in the Analyst Function of SAS using a one-way ANOVA Levene's Test. Replications were considered as random effects. If treatment effects were significant at the 0.05 probability level, LSD0.05 values were calculated for comparison of means. Correlations were computed using PROC CORR in SAS. Because of the strong positive correlation between R5.5 canopy N and yield, R5.5 canopy N was employed as a covariate to examine its role in the G x E for yield and seed quality parameters. The GGEbiplot software (Yan, 2001) was used to generate a trait by environment plot as described in Yan and Tinker (2005).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Temperatures during the growing season differed slightly among environments within years, with the most northerly location (Morris) tending to be the coolest, and the most urban (St. Paul) tending to be the warmest. Year effects, however, were greater and more complex. For clarity, Fig. 1 shows only mean temperatures of years across environments. Rainfall differed across environments within years and across years, so rainfall data from environments are shown in Fig. 1, with the exception of BECK03 and BECK04, which were supplied with irrigation water throughout the season. Although not readily discernable from Fig. 1, unfavorably-timed seasonal rainfall distribution likely resulted in significantly depressed yields at the RSMT04 environment. This site is predisposed to drought stress in years with irregular rainfall patterns as the topsoil is shallow and is underlain by coarse gravel.


Figure 1
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 1. Rainfall and average ambient temperature profiles for 10 environments in Minnesota. Solid or broken-line curves represent mean temperature profiles for environments within each year [2003 (STP03, BECK03, and MORR03); 2004 (RSMT04, STP04, BECK04, MORR04); 2005 (RSMT05, STP05, and MORR05)]. Bars represent rainfall distribution shown as totals for early-, mid-, and late-season precipitation by environment. These seasonal totals represent 15 May through 30 June (Days 135–181), 1 July through 14 Aug (Days 182–226), and 15 August through 30 September (Days 227–273). Two environments, BECK03 and BECK04, are not shown due to bi-weekly irrigation from 1 June through 15 September.

 
Among the six cultivars, cultivar had relatively little effect on yield or seed quality traits when compared with environmental effects (Table 2 ). Mean location yields ranged from 1.65 to 4.98 Mg ha–1, while mean cultivar yields had a range of only 3.11 to 3.23 Mg ha–1. Protein ranges were 396 to 442 g kg–1 across locations and 399 to 420 g kg–1 across cultivars, and similarly, oil ranges were 187 to 223 g kg–1 and 205 to 217 g kg–1. Likewise, all other traits measured differed greatly across environments. Most notable are the sum of protein and oil concentration (range of 608–637 g kg–1) and the quantity of N found in R5.5 canopy tissue (range of 117–309 kg ha–1).


View this table:
[in this window]
[in a new window]

  		Table 2. Mean yield, canopy N at R5.5, and seed quality trait data for six soybean cultivars grown at 10 locations in Minnesota.

 
When all cultivars and environments were combined for correlative analysis using individual plot values, yield and quality relationships tended to align with conventional wisdom (Wilcox and Guodong, 1997; Table 3 ). Yield was negatively correlated with protein but positively correlated with oil. Protein and oil were strongly negatively correlated with each other (r = –0.64, P < 0.001). The sum of protein and oil (P+O) was negatively correlated with yield, but appeared to be driven entirely by its numerically greater constituent, protein, as oil was not correlated with P+O. Seed sulfur concentration was positively correlated with protein and P+O (data not shown), and negatively correlated with yield. Seed N to S ratio is a measure of legume protein quality (Radford et al., 1977). Here the N to S ratio was not correlated with seed protein; however, it was highly negatively correlated with S (r = –0.90, data not shown). While seed protein and S were both negatively correlated with yield, the N to S ratio was positively correlated with yield. Protein production on an area basis (protein yield) was driven exclusively by yield, and protein concentration did not contribute positively to protein yield. In fact, protein yield was negatively correlated with protein concentration. Yield was positively correlated with the quantity of N found in aboveground plant biomass at full vegetative growth (R5.5 canopy N). Despite the positive R5.5 canopy N and yield correlation, R5.5 canopy N did not correlate with any quality traits measured.


View this table:
[in this window]
[in a new window]

  		Table 3. Correlation matrix for seed yield, canopy N at R5.5, and seed quality trait data across six soybean cultivars and 10 environments and within environments and cultivars.

 
In addition to the overall correlations among yield and quality traits across cultivars and environments that include any cultivar by environment interactions (G x E), we have included the number of environments and cultivars exhibiting a significant correlation among traits when the traits are examined by environment and cultivar, respectively (Table 3). This arrangement facilitates the visualization of the influence of E and G individually on the relationships among the traits. For example, when analyzed across cultivars within environments, quality parameters did not correlate well with yield; however, when analyzed across environments, data from five of the six cultivars showed a negative correlation between protein and yield and S and yield. Further, data from all six cultivars revealed significant positive correlations between yield and R5.5 canopy N, yet within only two environments were yield and R5.5 canopy N positively correlated.

Seed S concentration differed greatly across environments (Table 2) but it is unlikely that S availability limited yields (Hitsuda et al., 2004). Seed S concentration was positively correlated with protein at BECK03, BECK04, MORR05, and RSMT04. These environments represented both high- and low-yielding sites as well as environments with both high and low protein, S, and R5.5 canopy N values.

Analysis of variance using a model that included only main effects identified significant G x E for all traits except S concentration (Table 4 ). Table 4 shows the ANOVA results for the fixed effects, environment and cultivar, the covariate, R5.5 canopy N, and their interactive effects on seed yield and quality traits. Inclusion of the covariate and its interactive effects changed the G x E greatly. The G x E, now adjusted by R5.5 canopy N, R5.5 canopy N x G, and R5.5 canopy N x E interactions, was no longer significant for any traits examined. The R5.5 canopy N appeared to have an effect on protein, protein yield, S, and the P to O ratio without interactions.


View this table:
[in this window]
[in a new window]

  		Table 4. F test results from the mixed model analysis of variance for yield and seed quality traits in 10 environments.

 
Environment main effects remained significant for yield, protein yield, S, and N to S, but an E x R5.5 canopy N interaction was present for yield. This interaction would appear to render the main effect of E meaningless; however, Fig. 2 illustrates the basis of the interaction between R5.5 canopy N and E on yield. Within the environments MORR04 and RSMT04, R5.5 canopy N was positively correlated with yield, with r2 = 0.56 and 0.71, respectively. It is noteworthy that these were the two lowest-yielding locations in this study. Data from these two environments were separated from the other eight environments and both sets of environments were then subjected to the same analysis as described earlier. Removing MORR04 and RSMT04 eliminated the R5.5 canopy N x environment interaction for yield for the larger set of environments, but retained the main effect of R5.5 canopy N (P = 0.059) (Table 5 ). The R5.5 canopy N by location interaction for yield remained for the two segregated environments, MORR04 and RSMT04. The slopes of the linear regression lines between R5.5 canopy N and yield are significantly different (P < 0.05) in these two environments.


Figure 2
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 2. Relationship between R5.5 canopy N and seed yield in 10 environments in Minnesota. Linear regressions are provided for visualizing separation of individual environments; dotted lines do not imply relationships between canopy N and seed yield. For RSMT04 and MORR04, solid regression lines indicate a significant correlation between traits (P < 0.05). Regression equations and coefficients of determination are noted.

 

View this table:
[in this window]
[in a new window]

  		Table 5. F test results from the mixed model analysis of variance for yield and seed quality traits in eight environments (STP03, BECK03, MORR03, STP04, BECK04, RSMT05, STP05, and MORR05) and in two environments (RSMT04 and MORR04).

 
Since G contributes relatively little to the G x E, we have included a GGE biplot (Fig. 3 ) for environments and measured traits, or a trait x environment plot as in Yan and Tinker (2005). The plot is drawn with vectors to individual traits to indicate their relative ability to discriminate between environments (vector length), as well as their interrelationships (vector angles) and relationships with environments. The cosine of angles between two traits approximates their correlation such that acute angles (between 0 and 90 degrees) represent positive correlations, obtuse angles ( >90 degrees, <180 degrees) represent negative correlations, and right angles (90 degrees) represent independence. Note that the GGE biplot-derived correlations do approximate those shown in Table 3, with a positive yield/R5.5 canopy N correlation and a negative yield/protein correlation. Oil concentration appeared to be relatively unrelated to yield and canopy N. The P+O and S appeared to be negatively correlated with yield and R5.5 canopy N, while the N to S ratio was positively related to yield and R 5.5 canopy N, negatively related to S, and unrelated to protein. As vector length indicates, yield, protein yield, protein, oil, and P to O are equally strong at differentiating environments. The environments were not as efficiently separated on the basis of P+O, S, or R5.5 canopy N. It is likely that these traits do not assist in environmental differentiation (e.g., to form mega environments) because they interact with one another in manners too complex to capture on a two dimensional plot. Integrating all traits into a single plot allowed a separation of environments by year that was not achieved by analyzing GGE plots for individual traits (data not shown). The 2003 environments tended to produce higher oil concentrations than the other years. Likewise, 2005 environments tended to pool around high yield, R5.5 canopy N, and N to S. The environments BECK04, STP04, and RSMT04 resided at the center point of the GGE biplot, indicating that these environments produced approximately average values for most traits. Meanwhile, MORR04 was near P to O and opposite oil.


Figure 3
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 3. Trait x environment biplot describing the G + G x E (genotype + genotype x environment interaction) of 10 environments and nine traits for six cultivars. This plot is based on nontransformed, scaled (by dividing values by the standard deviation of their corresponding environment), environment-centered, and environment-focused single value partitioned data. The biplot explained 75.3% of the total G + G x E.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Studies examining G and E impacts on soybean quality parameters have tended to be highly confounded by variation in yield potentials of the genotypes examined (Primomo et al., 2002; Rao et al., 2002). Soybean seed quality is greatly affected by realized yields; therefore, with this study, we strove to minimize confounding effects of yield. Commercial soybean cultivars with a range of quality factors but similarity across other notable agronomic traits (yield, maturity, and lodging) were used to examine the effects of R5.5 canopy N on seed yield and quality traits, and to examine relationships between quality characteristics. Selection of cultivars with similar yield potentials likely led to the inadvertent selection of cultivars with similar vegetative growth patterns and canopy N storage potentials. Lack of variation for R5.5 canopy N among cultivars allowed the examination of environmental effects on yield and seed quality factors but it hindered the examination of within-location variation due to cultivar. Similarity between correlations across all the data and across only environments, together with a lack of correlation within environments, supports the notion that environmental variation drove the yield/quality relationships in this study. This is likely due to the narrow range in yields among cultivars and the broad yield range across environments. There was, however, sufficient variation for quality traits among cultivars, such that correlations among those traits within environments were similar to the overall (across cultivars and environments) correlations.

Soybean R5.5 canopy N played a role in determining seed yield across a wide range of environments, as other research has indicated (Shibles and Sundberg, 1998; Loberg et al., 1984). However, contrary to our initial hypothesis that R5.5 canopy N would directly affect seed protein concentration in northern latitudes, its effects on seed quality traits appear to be secondary at best. There appears to be sufficient N stored in the canopies of the soybean crops examined in this study to fully supply the seed with N for seed protein accrual. In fact, canopies contained two to four times the N as was present in the mature seed. Assuming a 70% mobilization efficiency [a conservative estimate from Loberg et al. (1984) and Vasilas et al. (1995)], all environments demonstrated the capacity to fulfill their seed N requirement entirely through mobilization of vegetative N.

While it is possible that R5.5 canopy N affects seed protein concentrations indirectly through its positive effect on yield, it is likely that canopy N and yield are simply coincident. The effect of canopy N on S may be through the same indirect mechanism; however, a clear understanding of this interaction was not determined. Contrary to Paek et al. (1997), our results support those of Wilcox and Shibles (2001), who found no correlation between protein and N to S. A negative correlation between S and yield found here indicates that soybean produced in high-yielding environments may have difficulty acquiring seed S in proportion to overall seed DM accrual. The positive relationship between yield and N to S indicates that S concentrations fall faster with increased yields than do protein concentrations. Therefore, S may be more limiting than reduced N on a land-area basis at high yield levels.

Although protein and oil were negatively correlated, the wide range in P+O noted across environments indicates that protein and oil accrual may occur somewhat independently. For example, at MORR03 oil concentrations were high, at MORR04 protein concentrations were high, and at MORR05 both oil and protein were low. In support of others (Piper and Boote, 1999; Yaklich and Vinyard, 2004), our data indicated that temperature during seed filling played a large role in determining seed oil concentration. In 2003, when temperatures during the seed-filling period were above average (Table 1), oil concentrations were higher than for 2004 and 2005 (Table 2 and Fig. 3), years when temperatures during the seed-filling period were lower than average and average, respectively. As location yields were not well correlated with temperature during seed filling, it is possible that protein is determined to a greater extent by yield, while oil is driven by temperature. This would lead to the wide range in P+O values noted here. Also, high-yielding environments tended to produce seed with a larger nonprotein, nonoil fraction (higher fiber) than low-yielding ones.

It appears that R5.5 canopy N affects yield potentials of environments and likely plays a large role in affecting rank order of cultivars across environments (G x E). However, the canopy N values may have been driven primarily by total canopy DM at R5.5 as tissue N concentration differences tend to be very small, as noted by others (Loberg et al., 1984; Shibles and Sundberg, 1998). When the data were reanalyzed using R5.5 canopy DM in the place of R5.5 canopy N, nearly identical results were obtained (data not shown). This indicates that much of the canopy N effect noted here may simply be related to overall canopy growth, leaf area development, and DM accrual. Our results support those of Zeiher et al. (1982) who found that mobilized N was primarily dependent on vegetative DM at R5, as well as those of Kumudini et al. (2001) who found DM accumulation throughout seed filling to be the most important factor measured that was involved with increased yield potentials of new vs. old varieties. It appears that canopy N (or DM) accrual impacts seed yield, but has little direct impact on seed quality traits.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Increased canopy N does not appear to promote increased seed protein concentration in northern-grown soybean, nor does it promote increased yields through increased seed protein accrual. Although canopy N does play a role in affecting yield, it is likely that this effect is a result of increased canopy DM rather than as a direct result of increased N reserves. The positive N to S ratio/yield relationship found in this study indicates that the accrual of S amino acids may be more limiting on a land-area basis than protein as a whole.


   ACKNOWLEDGMENTS
 
This project was made possible only though the financial generosity of the Minnesota Soybean Research & Promotion Council. We thank Arthur Killam for his expert technical assistance, and Sheri Huerd for her critical review.

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Agricola
Right arrow Articles by Naeve, S. L.
Right arrow Articles by Miller-Garvin, J. E.
Related Collections
Right arrow Soybean
Right arrow Crop Growth and Development
Right arrow Crop Physiology & Metabolism
Right arrow Seed Quality
Right arrow Plant and Environment Interactions

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