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


     

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 Similar articles in ISI Web of Science
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 HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Agricola
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Related Collections
Right arrow Soybean
Right arrow Crop Genetics
Agronomy Journal 92:780-784 (2000)
© 2000 American Society of Agronomy

SOYBEAN

Agronomic Changes from 58 Years of Genetic Improvement of Short-Season Soybean Cultivars in Canada

Malcolm J. Morrison, Harvey D. Voldeng and Elroy R. Cober

Agric. and Agri-Food Canada, Eastern Cereal and Oilseed Res. Cent., Central Exp. Farm, K.W. Neatby Bldg., Ottawa, ON, Canada, K1A 0C6

morrisonmj{at}em.agr.ca


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results
 Discussion
 Summary
 REFERENCES
 
Yield progress of short-season soybean [Glycine max (L.) Merr.] cultivars in Canada has been approximately 0.5% per year since the early 1930s. Our objective was to identify changes in agronomic traits associated with yield increase within a selection of historical cultivars. Where applicable, we measured phenotypic stability of these traits. At Ottawa, ON, we grew 14 cultivars, representing seven decades of breeding and selection (1934–1992), in a randomized complete block design with four replicates, across 6 yr. Data were collected on seed yield, seed weight, plant height, plant population, lodging susceptibility, and foliar disease symptoms. Seed number per plant was calculated from yield, seed weight, and plant population. Seed protein and oil concentration were measured. The increase in seed yield with year of release was associated with a significant increase in the number of seeds produced per plant. There was no relationship between seed yield and seed weight. A significant decrease in seed protein concentration with year of release was offset by a significant increase in seed oil concentration. Newer cultivars were more phenotypically stable for plant height than older cultivars. Modern cultivars were more efficient at establishing, supporting, and filling seeds on a per-plant basis than older cultivars.

Abbreviations: CV, coefficient of variability • MG, maturity group


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results
 Discussion
 Summary
 REFERENCES
 
EXPERIMENTS USING historical soybean cultivars grown in the same environment, offer plant researchers the opportunity to measure yield progress and determine its cause. In the USA, Luedders (1977) used 21 cultivars spanning over 50 yr of breeding and selection, and found a 1.0% per year yield improvement, attributable in part to increased lodging resistance. Wilcox et al. (1979) examined the yield of five cultivars in maturity group (MG) II and five in MG III ranging in release date from 1921 to 1974. They found a 0.5% yield improvement without an appreciable change in phenotypic stability or lodging resistance, but with an associated decrease in protein concentration with year of release. Also in MG II and MG III, Boyer et al. (1980) used old and new cultivars to show that breeding resulted in a yield improvement of 0.6% per year, with the newer cultivars suffering less from water stress then older ones. In MG VI, VII, and VIII cultivars, Boerma (1979) found a 0.7% increase in yield per year for cultivars released from 1947 to 1970 without a consistent relationship with the agronomic traits studied. In Canada, we examined the yield response of 41 cultivars from MG 0, 00, and 000, representing seven decades of breeding and selection (Voldeng et al., 1997). Plant breeding in short-season regions resulted in a yield improvement of 0.5% per year with an associated decrease in protein concentration and some improvement in lodging tolerance. In a related experiment, we grew 14 cultivars selected from the original 41 to examine physiological changes from 58 yr of genetic improvement (Morrison et al., 1999). We found that increased seed yield was significantly correlated with decreased leaf area and increased photosynthetic and stomatal conductance rates per unit leaf area.

Our objective in the current study was to examine the changes in agronomic traits associated with 58 yr of soybean breeding in short-season cultivars. Specifically, we examined plant height, lodging score, foliar damage score, plant stand, seed weight, seeds per plant, seed yield, and protein and oil concentration. We determined the phenotypic stability of these agronomic traits.


   Materials and methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results
 Discussion
 Summary
 REFERENCES
 
Cultivar selection, soil type, and cultural practices are the same as those described in Morrison et al. (1999). Briefly, 14 cultivars, representing seven decades of cultivar development (1934–1992), were grown in a randomized complete block design with four replicates from 1993 to 1998 at the Central Experimental Farm, Ottawa, ON, Canada (45°23' N lat). Individual plots were 4.8 m wide and 5 m long, comprised of twelve 40-cm-wide rows. Seed was inoculated with Bradyrhizobium japonicum and mechanically sown to a depth of 2 cm at a rate of 50 seeds m-2. The experiment was seeded on 18 May 1993, 24 May 1994, 19 May 1995, 23 May 1996, 23 May 1997, and 22 May 1998. Phenological measurements were made using the Fehr and Caviness (1977) growth stage key.

Foliar disease ratings were done on plants at the R5 stage of development in 1993 and 1994. There was no distinction made between disease types. Ratings (minimal [1] to severe [5]) were based on the number of leaves affected per plant and the degree of leaf damage. Lodging was rated on a 1 to 5 scale (erect [1] to prostrate [5]) at the R5 to R6 growth stage.

At maturity (R8), plant height was measured and the number of plants in a 2-m length of row were counted in each plot. Plant height was not measured in 1997 and plant stand was not counted in 1997 and 1998. After trimming 0.5 m from both ends, four rows were combine-harvested, and the seed was cleaned, weighed, and adjusted to 70 g kg-1 moisture content and the 1000-seed weight was determined (an average of two 500-seed samples multiplied by 2). In all years except 1997, the protein and oil concentrations were measured using a 300-g subsample per plot of whole seed using near infrared transmittance spectroscopy. The number of seeds per plant was calculated using seed yield, seed weight, and plant stand {[seed yield (g m-2)/seed weight (g seed-1)]/plants m-2}.

For each year, analysis of variance was done for each trait. Error mean squares from each of the years were tested for homogeneity of variance to ensure the combined analyses across years were appropriate. Combined analyses were found to be appropriate. Years, cultivars, and replicates within years were considered to be random effects. To observe cultivar changes across time, the cultivar means of the agronomic traits were plotted against the year of cultivar release. A straight line was fitted through the points using simple linear regression. The degree of association between the year of cultivar release and each trait was examined by calculating the simple linear correlation coefficient (r) with n - 2 degrees of freedom. Cultivar means across years were used to calculate correlations among traits. Correlations were done among traits measured in the same years. For example, seed protein, measured in five of the six years, was correlated with yield from the same 5-yr period. Phenotypic stability of agronomic traits were determined by calculating the coefficient of variability (CV) of cultivar means from 1993 to 1996 (Francis and Kannenberg, 1978). The smaller the CV, the more stable the cultivar during that time period. To separate cultivars, an unprotected LSD was calculated from the standard deviation of the CV across cultivars multiplied by t(n-2) degrees of freedom. To observe changes in cultivar phenotypic stability across time, the cultivar CVs of each agronomic trait were plotted against the year of cultivar release. A straight line was fitted through the points using simple linear regression. The degree of association between the year of cultivar release and each trait was examined by calculating the simple linear correlation coefficient (r) with n - 2 degrees of freedom.


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
Discussion
 Summary
 REFERENCES
 
The cultivar and year x cultivar interactions were significant (P <= 0.05) for all of the traits examined (Table 1) . While significant, the magnitude of the mean square of the interaction between year and cultivar was 1.5 to 21 times smaller, depending upon the trait, than the mean square for cultivar. In multiyear tests, when the interaction mean square is considerably smaller than the cultivar mean square, the cultivar rankings are expected to be relatively stable, eliminating the need to investigate the interaction term (Gomez and Gomez, 1984, p. 332).


View this table:
[in this window]
[in a new window]
  		Table 1 Degrees of freedom (df) and mean squares for cultivar and the interaction of year x cultivar from the combined analysis of variance for agronomic traits. All mean squares are significant at the 0.05 probability level

 
When data were combined across 6 yr, there were significant yield differences among cultivars (Fig. 1a) . There was a 26.14% yield improvement across 58 yr of plant breeding and selection, or 0.45% per year as calculated from the linear regression equation. Although there were significant differences in seed weights among cultivars, there was no consistent relationship between year of cultivar release and seed weight (Fig. 1b). There were significant differences in seeds per plant among cultivars, and a positive correlation with year of cultivar release (Fig. 1c). Seeds per plant increased by 0.49% per year across the 58-yr period.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1 Relationship between year of cultivar release and (a) seed yield, (b) weight of 1000 seed, and (c) number of seeds per plant. The error bar represents differences between cultivars (LSD 0.05). *,** Linear correlation coefficient significant at the 0.05 and 0.01 probability levels, respectively

 
Seed protein concentration decreased significantly by 7.5% over the 58-yr period or by 0.13% per year (Fig. 2a) . During the same time, seed oil concentration increased significantly with year of release by 0.23% per year (Fig. 2b). When protein and oil were summed and analyzed, there were significant differences among cultivars, but no relationship with year of release (Fig. 2c). While there were significant differences among cultivars for seed quality, the association with year of release was not significant (Fig. 2d).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2 Relationship between year of cultivar release and (a) protein concentration; (b) oil concentration; (c) protein + oil concentration; and (d) seed quality. The error bar represents differences between cultivars (LSD 0.05). ** Linear correlation coefficient significant at the 0.01 probability level

 
There were significant plant height differences among cultivars, but no association with year of release (Fig. 3a) . While cultivars differed for lodging susceptibility, the association with year of release was not significant (Fig. 3b). There were significant differences among cultivars for foliar leaf damage symptoms (Fig. 3c). Foliar disease ratings (1–5) decreased with year of cultivar release.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3 Relationship between year of cultivar release and (a) plant height, (b) lodging rating, and (c) foliar disease rating. The error bar represents differences between cultivars (LSD 0.05). ** Linear correlation coefficient significant at the 0.01 probability level

 
Simple correlation of agronomic traits showed that as seed yield increased, the number of seeds per plant, oil concentration, and disease tolerance increased, while the protein concentration decreased (Table 2) . As seed weight increased, the number of seeds per plant decreased. Lodging increased as plant height increased. Protein concentration was negatively correlated with oil concentration. Seed quality decreased as foliar disease increased.


View this table:
[in this window]
[in a new window]
  		Table 2 Correlation coefficients among agronomic traits of 14 short-season soybean cultivars representing 58 yr of genetic improvement

 
When yield and the agronomic traits were examined for phenotypic stability across environments, we found a significant association with year of cultivar release for plant height (Table 3) . The CV for plant height in newer cultivars was larger than it was for older cultivars. The agronomic trait with the largest CV was lodging, followed by yield, seed quality, and plant height. While seeds per plant had a lower CV than seed weight, the range in stability among cultivars was greater (from 3.2 to 26.0 for seeds per plant and 12.8 to 23.2 for seed weight), but no relationship existed with year of release for either trait. Protein and oil concentration had a low CV for all cultivars, indicating that seed composition changes were not influenced by environmental differences as much as other traits.


View this table:
[in this window]
[in a new window]
  		Table 3 Phenotypic stability of agronomic traits for cultivars as measured by the coefficient of variability (CV) of the means across four station-years (1993–1996)

 

   Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
Summary
 REFERENCES
 
Our data indicated that seed yield gain from 58 yr of plant breeding and selection in short-season regions was 0.45% per year. This was slightly lower then the yield improvement per year we found using a smaller data set (1993–1996) of the same cultivars grown at the same location (Morrison et al., 1999). By including data from 1997 and 1998, yield gain per year was reduced. In an earlier experiment using 41 short-season cultivars, we reported a yield gain of 0.5% per year (Voldeng et al., 1997). Studies conducted in the USA on several different maturity groups have found from 0.5 to 1.0% yield improvement per year (Luedders, 1977; Wilcox et al., 1979; Boerma, 1979; and Boyer et al., 1980). The small variation in yield improvement per year among these studies was likely the result of differences in the period in soybean breeding history that the studies encompassed, the time span between the oldest and most recently released cultivar in each study, and the geographic region where the study was done.

While there were differences in phenotypic stability among cultivars for yield, there was no significant association with year of release. Yield in new cultivars appears to be as influenced by variations in the environment as in the older landrace-selected cultivars such as Mandarin and Pagoda. Similar results were reported by Wilcox et al. (1979) and Voldeng et al. (1997). Breeding methods have selected lines primarily on yield performance, not yield stability.

Our data indicated that plant breeders in short-season regions have increased yield by increasing the number of seeds per plant, not by increasing seed size. While there was considerable variation in seed size among cultivars, it did not appear that the increase in seeds per plant occurred at the expense of seed size. Seed size was more phenotypically unstable than seed number per plant. Egli et al. (1978) also found that yield was correlated with seed number per plant and not seed size. They stated that seed size was more likely to fluctuate with changes in environment than seed number per plant. Boerma (1979) reported no significant correlation of yield improvement across time with either yield component. Gay et al. (1980), using only two cultivar comparisons, reported that in MG III the new cultivar had higher yield then the old cultivar, primarily due to an increase in seed size, while in MG IV the reverse was true. From our data, we concluded that selection for higher yield in short-season regions has favored cultivars that produced more seeds per plant. This may have been the result of selecting lines with more flowers per plant, or lines with a greater capacity to support the existing number of seeds. There is also the likelihood that breeding and selection has resulted in lines with greater stress (including disease) tolerance that are more successful at producing and filling seeds than older cultivars. Foliar disease symptoms decreased with year of release. Improved leaf health may have been one of the reasons that we previously found that leaf photosynthetic rate increased with year of release (Morrison et al., 1999).

The three highest-yielding new cultivars in our study, Maple Glen (1987), AC Bravor (1990), and AC Harmony (1992), had three different strategies for obtaining high yield. AC Harmony had the smallest seed size but the greatest number of seeds per plant; Maple Glen had a large seed size but fewer seeds per plant; AC Bravor was between the two in both yield components. Three different strategies for obtaining high yield indicates that there is considerable exploitable genetic variability in short-season germplasm.

By selecting for higher-yielding cultivars only, plant breeders in short-season regions have reduced the protein and increased the oil concentration of the seed. Across time, seed yield has increased at the expense of protein concentration. Similar results were obtained by Wilcox et al. (1979) and Voldeng et al. (1997). Burton (1987) reported that yield and protein concentration were negatively correlated, as were protein and oil concentration. Seed protein and oil concentration can be reduced to the amount of glucose required for their production (Bhatai and Rabson, 1976). One gram of glucose produces 0.83 g of carbohydrate, 0.40 g of protein, or 0.33 g of oil. From our data, 58 yr of soybean breeding decreased protein from 1045.8 to 968.0 g kg-1 glucose (a difference of 77.8 g kg-1), while oil concentration increased from 594.0 to 672.8 g kg-1 glucose (an increase of 78.8 g kg-1). On an energy basis, the decrease in protein was equivalent to the increase in oil concentration, proving that yield gains across time were not the result of a diversion of photoassimilate from protein to additional seed production. The bioenergetic balance in protein and oil seen in this experiment supports our previous conclusion that new cultivars have increased photosynthetic efficiency (Morrison et al., 1999). Future research may determine if lower seed protein concentration and higher yield resulted from the diversion of N from seed protein per se to the production of additional seeds or improved seed support and filling mechanisms.

Plant breeders in short-season regions have released cultivars with reduced plant height and improved lodging tolerance. Data from our experiment showed that plant height and lodging scores were positively correlated. If AC Bravor (1990), which had a relatively high lodging score, was removed from the data set, the association between year of cultivar release and lodging score would have been significant. Wilcox et al. (1979) and Voldeng et al. (1997) also found a trend toward decreased lodging with year of cultivar release.


   Summary
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Summary
REFERENCES
 
In the short-season region, soybean yield gain over time was associated with increased number of seeds per plant. The decrease in seed protein concentration with year of release was accompanied by an increase in oil concentration. Two years of data showed that foliar disease tolerance has improved over time. Further research is required to determine which specific diseases are most important in new cultivars. In order to determine if seed protein has decreased across time because of a diversion of N to new seeds, future research should explore seed N dynamics and the relationship between protein concentration, seed number, and yield. Although extremely labor-intensive, an examination of flower numbers and their sequential development may reveal information on photoassimilate supply differences between old and new cultivars.


   ACKNOWLEDGMENTS
 
The authors thank Brian Couture, Pat Bonnilla, Ron Guillemette, and Andrew Bird for their technical support.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Summary
 REFERENCES
 
ECORC contribution no. 001482.

Received for publication October 6, 1999.
   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 Summary
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
D. B. Egli
Comparison of Corn and Soybean Yields in the United States: Historical Trends and Future Prospects
Agron. J., May 7, 2008; 100(Supplement_3): S-79 - S-88.
[Abstract] [Full Text] [PDF]

Home page
 Plant Physiol.Home page
E. A. Ainsworth, A. Rogers, and A. D.B. Leakey
Targets for Crop Biotechnology in a Future High-CO2 and High-O3 World
Plant Physiology, May 1, 2008; 147(1): 13 - 19.
[Full Text] [PDF]

Home page
 Crop Sci.Home page
Y.-B. Fu, G. W. Peterson, and M. J. Morrison
Genetic Diversity of Canadian Soybean Cultivars and Exotic Germplasm Revealed by Simple Sequence Repeat Markers
Crop Sci., September 1, 2007; 47(5): 1947 - 1954.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
M. U. Haq and A. P. Mallarino
Response of Soybean Grain Oil and Protein Concentrations to Foliar and Soil Fertilization
Agron. J., May 13, 2005; 97(3): 910 - 918.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
E. R. Cober, M. J. Morrison, B. Ma, and G. Butler
Genetic Improvement Rates of Short-Season Soybean Increase with Plant Population
Crop Sci., May 6, 2005; 45(3): 1029 - 1034.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
S. Seiter, C. E. Altemose, and M. H. Davis
Forage Soybean Yield and Quality Responses to Plant Density and Row Distance
Agron. J., July 1, 2004; 96(4): 966 - 970.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
B. L. Ma, L. M. Dwyer, C. Costa, E. R. Cober, and M. J. Morrison
Early Prediction of Soybean Yield from Canopy Reflectance Measurements
Agron. J., November 1, 2001; 93(6): 1227 - 1234.
[Abstract] [Full Text] [PDF]

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 Similar articles in ISI Web of Science
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 HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Agricola
Right arrow Articles by Morrison, M. J.
Right arrow Articles by Cober, E. R.
Related Collections
Right arrow Soybean
Right arrow Crop Genetics

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