| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Agric. Res. Stn., Fort Valley State Univ., Fort Valley, GA 31030
b Coastal Plain Exp. Stn., Univ. of Georgia, Tifton, GA 31793
c Agric. Res. Stn., Virginia State Univ., Petersburg, VA 23806
d Dep. of Plant and Soil Sci., Alabama A&M Univ., Normal, AL 35762
e Dep. of Agric., Univ. of Maryland Eastern Shore, Princess Anne, MD 21853
* Corresponding author (raom{at}mail.fvsu.edu)
Received for publication October 13, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
10% and up to 35% by ‘Hutcheson’. The genotype effects were significantly larger than the location x year effects for plant height, seed weight, and HI, but not for biomass or seed yield. Biomass and HI were important determinants of seed yield. S90-1056, V81-1603, V71-370, ‘Enrei’, ‘Nakasennari’, ‘Ware’, and ‘York’ were stable for seed weight across years. Hutcheson, S90-1056, York, MD86-5788, Nakasennari, and BARC-8 showed yield stability across environments and years. S90-1056, York, and Nakasennari were stable for both seed weight and seed yield; therefore, they could be used for commercial production in the USA or for breeding soybean cultivars suitable for tofu preparation.
Abbreviations: AAMU, Alabama Agricultural and Mechanical University • FVSU, Fort Valley State University • GE, genotype x environment • HI, harvest index • SI, stability index • UMES, University of Maryland Eastern Shore • VSU, Virginia State University
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
75.0 million metric tons of soybean, accounted for >50% of the world soybean production in 1998 to 1999 (Golbitz, 2000). Soybean meal is mainly used as a source of protein in livestock and poultry feed, whereas soybean oil is used for human consumption and industrial purposes. Prices for conventionally grown soybean in both domestic and international markets have been declining worldwide because of greater competition from South American countries. Therefore, the identification of soybean for specialty uses and diversification of utilization may be pivotal to the future of the soybean industry in the USA. Specifically, selection of soybean genotypes for soyfoods offers potential for expanding the market. In many Asian countries, soybean is an integral part of the diet. The demand for food-grade soybean is growing both in the national and international markets (Carter and Wilson, 1998; Boerma and Mian, 1998). Soyfoods have been reported to provide protection against heart disease, cancer, menopausal symptoms, and other diseases (Carter and Wilson, 1998; Messina and Messina, 1991). Because of the nutraceutical value of soybean, there is a growing demand for soyfoods, such as soymilk, several types of tofu, soybean sprouts, and soynuts. Soyfood sales in the USA have been growing steadily at a rate of 10 to 25% yr-1 (Kuhn, 1996) and exceeded $600 million in 1998. About 9 to 10 million t of soybean seed are imported into Asia for manufacturing soyfoods and for oil extraction. Tofu, a cottage cheese-like soybean curd, has high nutritional value and is rich in proteins, vitamins, and minerals, particularly Ca. Tofu consumption is growing at an annual rate of 20% in the USA and among health-conscious people around the world (Carter and Wilson, 1998). Thus, the increasing market for soyfoods and health benefits associated with them indicate the economic potential and emphasize the need for the identification and development of high-yielding U.S. soybean cultivars suitable for food processing and human consumption.
The soyfood industry utilizes soybean cultivars based on seed physical traits, chemical constituents, and processing quality (Brar and Carter, 1993). Therefore, a multistate, multidiscipline regional soybean research project, entitled Improvement of Soybean for Food Uses, was initiated in 1994 under the aegis of the Association of Research Directors of 1890 Institutions with funding from USDA/CSREES. The objectives of the project were to evaluate selected soybean genotypes for their agronomic performance at different locations and subsequently test their suitability for soyfoods, such as tofu and roasted green beans. In this study, twelve soybean genotypes selected on the basis of their attributes such as seed color and protein content which are considered suitable for tofu preparation were evaluated.
Gene expression is subject to modification by the environment; therefore, genotypic expression of the phenotype is environmentally dependent (Kang, 1998). The development of new cultivars involves breeding of cultivars with desired characteristics such as high economic yield, tolerance or resistance to biotic and abiotic stresses, traits that add value to the product, and the stability of these traits in target environments. Inconsistent genotypic responses to environmental factors such as temperature, soil moisture, soil type, or fertility level from location to location and year to year are a function of genotype x environment (GE) interactions. Genotype x environment interactions have been defined as the failure of genotypes to achieve the same relative performance in different environments (Baker, 1988). Identification of yield-contributing traits, and a knowledge of GE interactions and yield stability are important for breeding new cultivars with improved adaptation to the environmental constraints prevailing in the target environments. Currently, there is a need for increasing soybean genetic diversity in the USA (Carter, 1987; Boerma and Mian, 1998) so that new cultivars suitable for manufacturing soyfoods can be developed (Carter, 1987). To avoid genetic vulnerability associated with the narrowing of the genetic base of any crop, the GE interactions of the germplasm are important (Kang, 1998).
Because of the narrow genetic base available for developing food-grade soybean cultivars in the USA (Boerma and Mian, 1998; Wagoner et al., 1998), there is a need for introduction and evaluation of exotic food-grade soybean germplasm from Asian countries, where soybean cultivars with traits desirable for soyfood production have been developed. This is likely to benefit American soybean breeding programs in generating new genetic combinations that will facilitate breeding high-yielding food-grade soybean cultivars for sustainable production in target environments in the USA (Kang, 1998). With the availability of improved statistical tools to analyze and understand GE interactions, it is now possible to develop improved cultivars for target environments by exploiting GE interactions and marker-based selection integrated with traditional plant breeding (Kang, 1998; Boerma and Mian, 1998).
Twelve genotypes used in this study included three cultivars of Japanese origin, six advanced breeding lines, and three U.S. cultivars. Cultivars Hutcheson, Ware, and York from the USA, and Nakasennari, Enrei, and Suzuyataka from Japan are currently being used by tofu manufacturers in the respective countries. The seed of the Japanese cultivars and the U.S. breeding lines, V71-370, V81-1603, and MD86-5788, have a cream color and a clear hilum, which are preferred by tofu manufacturers (Brar and Carter, 1993). At present, V71-370 has been contracted for export to Japan for tofu production (M. Rangappa, personal communication, 1999). The seed of lines BARC-8, BARC-9, and S90-1056 have relatively high protein content (Rao et al., 1998) and may be suitable for preparing high protein tofu (Bhardwaj et al., 1999; Wang et al., 1983) with superior textural qualities (Evans et al., 1997). The data on seed oil and protein contents, fatty acid profiles, tofu yield, and tofu quality of these 12 genotypes have been published elsewhere (Rao et al., 1998; Bhardwaj et al., 1999).
The objectives of the study were to (i) identify food-grade soybean genotypes with superior agronomic traits important for adaptation to the southeastern USA; (ii) assess the presence and nature of the GE interactions for plant height, biomass, seed yield, harvest index (HI), and seed weight; and (iii) determine yield stability of the 12 genotypes across four locations in the USA.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
20 plants m-1 row-1. Fertilizer (on the basis of recommendations for soybean) and herbicide trifluralin [2,6-dinitro-N,N-dipropyl-4-(trifluoromethyl) benzenamine] at a rate of 1.8 L ha-1 were incorporated into the soil prior to planting. During the crop growing season, Permethrin [(3-phenoxyphenyl) methyl(±)-cis,trans-3-(2,2-dichloroethenyl)-2,2-dimethylcyclopropanecarboxylate] at 585 mL ha-1, and Carbaryl (1-naphthyl N-methylcarbamate) and Acephate (O,S-Dimethyl acetylphosphoroamidothioate) at 1.12 kg ha-1 were applied as needed to control insect pests. At FVSU, irrigation was applied as needed by an overhead sprinkler system.
Weather Data
Data on daily air temperatures and rainfall were recorded from weather stations at the experimental sites in AAMU, FVSU, and UMES, and from the local Meteorological Station in VSU (Fig. 1) . The AAMU, UMES, and VSU sites had fairly good rainfall and conducive temperatures throughout the growing season. At FVSU, although temperatures were conducive for crop growth, rainfall was generally insufficient and necessitated application of irrigation.
|
Harvest Index
Apparent HI was calculated as the ratio of seed yield to biomass (excluding roots and leaf debris).
Statistical Analyses
The data were subjected to statistical analyses using linear model procedures of SAS software, Proc GLM (SAS Institute, 2000) and Proc MIXED (Littell et al., 1996). For each location, the model included genotypes as fixed effects (least square means), and year, rep (year), and year x genotype as random effects (variance component effects). Test for genotype differences used year x genotype as the error term. In the model that included locations, genotypes were regarded as a fixed effect, whereas all other sources of variation were regarded as random effects. Test for genotype differences used location x genotype, year x genotype, or location x year x genotype as the error term when any interaction was significant and had the largest random effect.
Because genotype variances were heterogeneous and square root transformation did not correct the problem, the weighted analysis using the formula Yw = Y x Gi (Y = measured data, Yw = weighted data, Gi = inverse of genotype variance) as per Timm and Mieczkowski (1997)(p. 142) was carried out. The weighted data were then analyzed, so that variances among genotypes could be made homogeneous. The resulting mean separation of the genotype means was an indication of the genotype's actual performance without the extraneous variance. Mean separation technique as described in Steel and Torrie (1960)(p. 481) was by Kramer-Tukey adjustment to the Fisher's LSD (multiple t-test in Proc Mixed). Coefficients of correlation were computed using Proc CORR (SAS Institute, 2000).
Multiple (Stepwise) Regression
Multiple regression procedures [stepwise, Proc STEPWISE (SAS Institute, 2000)] were used to determine the relative importance of yield components in determining seed yield. Three models were used: (i) raw data; (ii) adjusted for genotypes (effect being discussed); and (iii) adjusted for location, year, repetition, and genotype effects (model used).
Stability Analysis
The stability analysis computer program developed in BASICA by Kang (1993) was converted to run in the Data Step of the SAS Institute (2000). This was necessary because the version of STABLE could not be run in the BASIC version available. The data furnished with the BASICA program were used to test the accuracy of the converted SAS computer program by comparing the results with those given by Kang (1993).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
|
Plant Height
The actual genotypic mean plant height was 56 cm and ranged from 44.7 cm for Suzuyataka to 88.4 cm for BARC-9 (Table 2). BARC-9 had significantly taller plants than the rest of the genotypes. The plant height of S90-1056 was significantly higher than all genotypes except Hutcheson, York, and Nakasennari. The plant height was greater than the average for Hutcheson, S90-1056, BARC-9,York, and Nakasennari, but it was lower for all other genotypes.
|
Biomass
The genotypic differences for actual biomass were not significant (Table 2). The genotypes produced similar amounts of biomass which ranged from 5.3 to 7.0 Mg ha-1. The genotypes produced an average of 6.2 Mg ha-1 of biomass across locations and years. A similar level of biomass was reported for MG VII and VIII soybean genotypes in Georgia (Rao and Bhagsari, 1998).
When subjected to weighted analysis, the genotype effects were significant. Means from weighted analysis of plant biomass showed that V71-370 had the greatest weighted value and improved in rank positions in weighted relative to unweighted analysis. BARC-9, V81-1603, Nakasennari, Suzuyataka, and MD86-5788 also improved in rank positions. Hutcheson, S90-1056, BARC-8, Enrei, and York lost rank positions, whereas Ware did not change rank position.
Seed Weight
The 100-seed weight, averaged across locations and years, ranged from 14.0 g for BARC-8 to 26.8 g for V81-1603 (Table 2). Breeding line V81-1603 had significantly heavier seeds than all other genotypes. Among the U.S. genotypes, breeding lines V81-1603 and V71-370 generally produced heavier seeds (>20 g 100-1) than most other genotypes except Japanese cultivar Enrei. The mean 100-seed weight (19.67 g) for Japanese cultivars was relatively higher than that (17.97 g) of the U.S. genotypes. Brar and Carter (1993) also reported that Japanese cultivars generally produced bigger seeds than many American cultivars. Generally, large seed size (>20 g 100-1 on a dry weight basis) is preferred by tofu manufacturers. Soybean genotypes with greater seed weight were associated with higher tofu yields (Bhardwaj et al., 1999). The three Japanese cultivars were bred for tofu production in Japan (Carter and Wilson, 1998).
Means from weighted analysis of 100-seed weight showed that BARC-9 had the highest value and improved in rank positions in weighted analysis compared with that in unweighted. Suzuyataka also improved in rank position. MD86-5788 had the smallest value and lost in rank position. The six genotypes, V81-1603, V71-370, Enrei, Nakasennari, BARC-8, and Ware with generally heavier seeds lost rank positions, whereas the rest of the genotypes with relatively smaller seeds gained in rank position.
Seed Yield
The mean seed yield, averaged across locations and years, ranged from 2.0 Mg ha-1 for Ware to 3.0 Mg ha-1 for Hutcheson (Table 2). The seed yields in the present study were similar to those reported earlier (Rao and Bhagsari, 1998). Hutcheson produced a greater seed yield (P < 0.05) than Enrei, BARC-9, Suzuyataka, and Ware, but was similar to the rest of the genotypes. In addition to Hutcheson and S90-1056, three other genotypes York, MD86-5788, and V71-370 produced more than 2.5 Mg ha-1 seed yields.
The mean seed yield of the Japanese cultivars was 10% lower than that of the U.S. genotypes. Hutcheson, the elite American cultivar, outyielded the Japanese cultivars by 35%. Japanese cultivars, albeit their bigger seed size, have been reported to give lower seed yields than many American cultivars (Brar and Carter, 1993).
Weighted analysis of seed yield increased the level of significance from P = 0.01 to P = 0.001. The weighted means ranked differently from those of the actual means. Means from weighted analysis of seed yield showed that York, Nakasennari, Ware, MD86-5788 and BARC-9 improved rank positions, whereas Hutcheson, S90-1056, V71-370, BARC-8, V81-1603, Suzuyataka, and Enrei lost rank positions.
Harvest Index
The mean HI was generally >0.40 for all genotypes except BARC-9 (Table 2). Harvest index ranged from 0.37 to 0.45 with a genotypic mean of 0.43. Hutcheson and Suzuyataka had a significantly higher HI than BARC-9, but were similar to the rest of the genotypes. Rao and Bhagsari (1998) reported a similar range of HI for soybean grown in Georgia.
Similar to the trends observed for other traits, weighted analysis of HI showed changes in rank positions of the genotypes. BARC-8 was the only genotype that gained significantly in value and rank position. BARC-9 did not change in rank position, whereas the rest of the genotypes lost rank positions.
The loss of rank position in weighted analysis indicates the magnitude of variance associated with unweighted value. No genotype consistently gained or lost rank positions for all five traits. York, BARC-9, and Suzuyataka gained rank positions for at least three traits. Hutcheson had the greatest loss of rank position with the only gain of position occurring with seed weight. Similarly, Ware, Enrei, and S90-1056 lost rank positions for all traits except seed yield, plant height, and seed weight, respectively.
Relationships among Seed Yield, Biomass, and Harvest Index
Seed yield and biomass were significantly correlated at the three locations (Table 3). A linear relationship between biomass and seed yield was reported for soybean in Puerto Rico (Ramirez-Oliveras et al., 1997) and Australia (Mayers et al., 1991). Board et al. (1996) reported strong positive correlations between yield and top dry matter at the R5 stage for late planted soybean. Further, branch dry matter was more closely related to seed yield than main stem dry matter. Seed yield was significantly correlated with HI at three locations, but r-values were relatively small, indicating that other factors also affected yield and yield components. Seed yield is a component of biomass and HI is the ratio of seed yield to biomass. Thus, these three components are closely related. The relationship between seed yield and 100-seed weight was not consistent, which confirms the reports of Carter and Boerma (1979), Board (1987); Board et al. (1996), and Shukla et al. (1998).
|
|
For 100-seed weight, the location x year and location x year x genotype effects were significant, indicating that seasonal conditions at each of the locations in a given year influenced seed weight. However, the genotype effects were greater than these effects. Seed development occurs as the last stage in the yield formation sequence. Thus, seed weight is determined towards the end of the season when the seeds are maturing. Seasonal effects would be smaller on seed weight than on the number of seeds which is determined early during the season. Changes in environment and other stress factors influence the number of seeds per pod or per unit area across a longer period of time than seed weight which could result in greater GE interactions (Board et al., 1997).
Location x year effects were larger than residual effects for seed yield. Genotype effects were highly significant when tested against location x genotype interaction because location x genotype effects were very small and not significant. For HI, location, location x year, and location x year x genotype effects were high but not greater than genotype effects.
Components Contributing to Seed Yield
Multiple regression (stepwise) analyses were performed on the raw data using three models (Table 5). The first model is the classic approach to stepwise regression. All four variables are independently related to yield with plant height (P < 0.05), HI (P < 0.001), and biomass (P < 0.001) having significant positive relationships with yield, whereas seed weight (P > 0.15) was positive but not significant.
|
In the third model, all the effects used in the analysis of yield were removed. All four variables are independently related to yield with HI (P < 0.001) and biomass (P < 0.001) having significant positive relationships with yield, whereas plant height (P > 0.99) and seed weight (P > 0.20) showed a positive relationship but were not significant. The t-values for HI and biomass changed from 49 and 99 to 32 and 59, respectively, indicating a decrease in importance in their relationship with seed yield. Seed weight showed an increase in slope value, but the relative importance remained the same as indicated by similar t-values. Plant height relationship changed from being significant to nearly not significant. Therefore, the effect of locations and years are even greater than those indicated by GE effects in Table 4. The analysis showed that biomass and HI are the important determinants of seed yield. Path coefficient analyses of soybean yield and yield components, involving 80 genotypes in India, showed that HI and plant dry matter were the most important yield-determining traits (Shukla et al., 1998). Board et al. (1990) also reported that branch dry matter at the R5 stage in narrow-row late-planted soybean in the southeastern USA was an important yield determinant.
The results from stepwise regression are in agreement with the significant correlations observed for seed yield with biomass and HI (Table 3). A study involving 81 soybean genotypes belonging to maturity groups III through VIII showed similar positive correlations between seed yield and biomass and HI (Rao and Bhagsari, 1998). Our previous research, involving a total of >200 soybean genotypes in four field experiments conducted during the past several years, showed significant positive correlations (r = 0.22 to 0.68) between seed yield and HI. Higher correlations ranging between 0.59 and 0.92 were observed between seed yield and biomass (Bhagsari and Rao, unpublished data, 1994). Board et al. (1997) suggested that path coefficient analysis is a more informative and useful method for determining selection criteria for soybean yield improvement than simple correlation coefficient. In this study, the relationships of seed yield to biomass and HI were highly consistent across locations and years, even though soybean is photoperiod sensitive.
Stability Analysis
Yield stability analysis was carried out on unweighted data, and the results are shown in Table 6. Stability analysis based on the criteria set forth in Kang (1993) examines the behavior of each genotype using the location x year x genotype means. The first criteria used is the distance a genotype is from the overall mean using its own variance to the LSD for all genotypes from the ANOVA at P = (not significant, 0.10, 0.05, 0.01) and assigns points (0, 1, 2, 3, + if above mean, - if below) to be added to the original ranking. The second criteria used is the relationship of each genotype's variance to the average variance (ANOVA error mean square). The further a variance is away from this average, the more negative points are assessed. Again, this is determined using P = (>0.10, > 0.05, >0.01, <0.01) from an F-test and assigning negative points (0, -2, -4, -8) that are subtracted from the adjusted ranking obtained from the first criteria. The higher the stability index (SI) value, the more stable the trait. Acceptance (receives a Yes) is determined based on where the mean SIs fall. Those above are accepted, while those below are rejected. Kang (1993) stressed that the method proposed selects those genotypes that exhibit the highest stability. S90-1056, the only genotype to be accepted for all five traits, had a net loss of 5 points. Hutcheson received four acceptances and lost 18 points, while York also received four acceptances and lost fourteen points. Suzuyataka was selected once and lost thirteen points. From the two worst performers, Ware lost 47 points but was selected once, and Enrei lost 41 points and was selected twice. If HI was not considered, since it is a linear combination of yield and biomass, York would be similar to S90-1056. Hutcheson, being small-seeded, was not selected for seed weight, but it was the highest ranked for all other traits. Hutcheson produces high yields, has a large plant size, a greater proportion of its biomass is seed yield, and it has a slightly shorter plant height than many other genotypes. However, because of smaller seed size, Hutcheson may not be suitable for tofu production. It was assessed a -5 penalty points because of its high variation in seed yield, but was assessed +2 points because of its fairly consistent (lower variation) plant height.
|
Hutcheson, S90-1056, York, MD86-5788, Nakasennari, and BARC-8 were selected for yield stability across locations and years. Biomass and HI are more consistently correlated with seed yield (Rao and Bhagsari, 1998). In this study, all genotypes selected for yield stability, except Nakasennari, were also selected for either biomass, HI, or both. The results of stability analysis were consistent with coefficients of correlation and stepwise regression, suggesting that improvement of HI and biomass could lead to an increase in soybean yields as reported for wheat (Triticum aestivum L.) (Gifford and Evans, 1981; McVetty and Evans, 1980).
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
35% lower than the elite American cultivar Hutcheson. However, the seed weight of Hutcheson is smaller than the Japanese cultivars, which renders this cultivar less desirable for tofu preparation. In this study, further analyses were made to determine GE interactions and yield stability to elicit information concerning the type of adaptation of the 12 genotypes. Most of the genotypes studied showed crossover interaction (differential response in different locations) as their ranks changed with environment, indicating specific adaptation. Hutcheson, York, S90-1056, MD86-5788, Nakasennari, and BARC-8 showed noncrossover interactions (Kang, 1998) because they were selected for yield stability across locations and years. Thus, these genotypes may have broad adaptation and make good sources for breeding tofu-type soybean cultivars. Three of these genotypes, Hutcheson, York, and MD86-5788 were generally high-yielding as well. This indicates the possibility of simultaneous selection for high yield and broad adaptation to diverse environments, a feature considered desirable for conserving germplasm resources (Evans, 1993; Kang, 1998). York showed both specific and broad adaptability, as it was stable across years within each location. The importance of GE interactions, yield stability analysis, and stepwise regression analysis in determining adaptability of genotypes to a specific location or several locations was clearly reflected in this study. S90-1056, V81-1603, V71-370, Enrei, Nakasennari, Ware, and York showed stability across years within a location for seed weight. This type of GE interaction is often preferred by farmers, as they rely on cultivars that perform consistently from year to year (Evans, 1993). This type of stability (temporal adaptation) has implications for food-grade soybean because seed size is one of the important selection criteria for the determination of its use for a particular type of soyfood. For example, large-seeded soybean are preferred for fresh vegetable purposes and for tofu preparation, and small seeded soybean genotypes are preferred for natto or soy sprouts (Carter and Wilson, 1998). Nakasennari, which was selected for seed weight and seed yield, may be a good source for combining tofu characteristics with agronomic traits. In conclusion, this study showed the presence of and the type of GE interactions among the 12 soybean genotypes and their yield components. High-yielding genotypes with broad adaptation and some genotypes with specific adaptation were identified. Further investigations on GE interactions at important crop growth stages for yield components and biochemical profiles would help develop strategies that integrate traditional plant breeding with modern molecular marker-based selection for tailoring soybean cultivars for soyfoods and target environments. Among the Japanese cultivars used in this study, Nakasennari may be adapted for cultivation in the USA and its seed exported to Japan. In the long term, Nakasennari and Enrei could be used as a source of genetic diversity in the U.S. soybean breeding programs to produce novel hybrids possessing processing qualities of the Japanese cultivars and the agronomic productivity of U.S. cultivars, as suggested by Brar and Carter (1993).
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. L. Naeve, T. A. O'Neill, and J. E. Miller-Garvin Canopy Nitrogen Reserves: Impact on Soybean Yield and Seed Quality Traits in Northern Latitudes Agron. J., May 7, 2008; 100(3): 681 - 689. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mebrahtu, T. E. Devine, P. A. Donald, and T. S. Abney Registration of 'Owens' Vegetable Soybean Journal of Plant Registrations, September 1, 2007; 1(2): 95 - 96. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mebrahtu, T.E. Devine, P. Donald, and T.S. Abney Registration of 'Randolph' Vegetable Soybean Crop Sci., October 27, 2005; 45(6): 2644 - 2645. [Full Text] [PDF]
|
|
|
|
|
|
T. Mebrahtu, T. E. Devine, P. Donald, and T.S. Abney Registration of 'Asmara' Vegetable Soybean Crop Sci., January 1, 2005; 45(1): 408 - 409. [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 | |||
