| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Soil and Crop Sciences, Colorado State Univ., Fort Collins, CO 80523-1170
b Dep. of Agronomy and Plant Breeding, Tehran Univ., Karaj, Iran
c Dep. of Statistics, Colorado State Univ., Fort Collins, CO 80523-1877
* Corresponding author (patrick.byrne{at}colostate.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: AGB, above-ground biomass • GA, gibberellic acid • PCR, polymerase chain reaction • QTL, quantitative trait locus • RIL, recombinant inbred line
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The Rht-B1b and Rht-D1b alleles, which occur at homoeologous loci on chromosomes 4B and 4D, respectively, reduce sensitivity to gibberellic acid (GA), which is necessary for stem elongation (Flintham et al., 1997). In favorable environments, the reduced demand for assimilates by a shorter stem results in improved assimilate partitioning to the developing head, leading to higher spikelet fertility and more but smaller grain per head. Semidwarf wheats have smaller leaves, but compensate with increased photosynthetic rates resulting in a biomass similar to that of tall lines (LeCain et al., 1989; Morgan et al., 1990; Flintham et al., 1997).
The relative yield advantage of dwarf and semidwarf cultivars varies with spring or winter habit, genetic background, and environmental conditions. The benefits of the dwarfing alleles are more pronounced in high yielding winter wheat environments (Flintham et al., 1997) and in high yielding spring wheat locations at latitudes less than 40° (Fischer and Quail, 1990). However, under heat and drought stress, there may be no benefit of the dwarfing alleles in spring wheat (Flintham et al., 1997; Nizam Uddin and Marshall, 1989; Richards, 1992a,b). Richards (1992a) concluded that grain yield does not depend on the presence of dwarfing genes per se, but rather on an optimum height for a given environment. In his study of rainfed Australian environments, lines with a single dwarfing gene consistently fell within the optimum height range in all genetic backgrounds. According to the crop plant ideotype concept advanced by Reynolds et al. (1994), shorter plants are better adapted to irrigated, high input environments, while taller plants are considered to have better yield stability under adverse conditions.
In the western Great Plains of North America, dwarfing alleles are commonly used to develop semidwarf spring wheat cultivars. In Saskatchewan and Montana evaluations, semidwarf lines with either Rht-B1b or Rht-D1b generally outyielded dwarf lines and tall lines (Knott, 1986; McNeal et al., 1972). An exception was in the lowest yielding location (<1352 kg ha–1) of the Montana study, where tall lines yielded more (McNeal et al., 1972). Because the Rht constitution of the semidwarf lines was not determined in either of those studies, effects of Rht-B1b and Rht-D1b could not be distinguished.
Determination of the Rht genotypes of wheat plants is nearly impossible on the basis of plant height alone as an indicator. The major genes determining plant height are normally classified into two groups depending on their reaction to gibberellic acid (GA) (Worland et al., 1998). Rht-B1b and Rht-D1b are insensitive to GA, so exogenous application of GA to plants with those alleles does not restore the tall phenotype (Rebetzke and Richards, 2000). The presence of Rht-B1b and/or Rht-D1b dwarfing alleles can be determined by lack of seedling response to GA (Gale and Gregory, 1977; Richards, 1992a); however, this method does not distinguish between the two alleles. In addition, while the GA test is relatively easy, it is time consuming and not always reliable (Ellis et al., 2002). Other experiments have depended on testcrosses (Fischer and Quail, 1990), another time- and labor-intensive method, to determine the presence of dwarfing alleles. The recent development of PCR-based markers for the Rht-B1b and Rht-D1b alleles allows for relatively easy and definitive genotyping of wheat plants (Ellis et al., 2002).
The objective of this study was to determine the effects of alleles at the Rht-B1 and Rht-D1 loci on plant height and agronomic characteristics in a spring wheat RIL population under irrigated, partially irrigated, and rainfed field conditions in the west central Great Plains. The use of a RIL population allowed for an analysis of the dwarfing genes segregating in the same genetic background, thereby reducing the confounding effects of different backgrounds. The availability of PCR-based markers permitted more accurate genotyping of RILs containing either Rht-B1b, Rht-D1b, both, or neither dwarfing allele.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Field Trials
Field trials were conducted at the Central Great Plains Research Station, USDA-Agricultural Research Service, Akron, CO, in 2001, and at ARDEC in Fort Collins in 2001 and 2002. The trials were designed as 15 x 10 
-0,1 rectangular lattices with two replicates (Patterson et al., 1978). Entries included 144 RILs, and two occurrences of each parent and a check variety, ‘Butte 86’. To control Russian wheat aphid, all seed was treated with Gaucho 480 [Gustafson, Calgary, AB; active ingredient: imidacloprid, (EZ)-1-(6-chloro-3-pyridylmethyl)-N-nitroimidazolidin-2-ylideneamine] before planting at a rate of 540 mL product per 100 kg seed.
In 2001, one trial was planted at Fort Collins (fully irrigated by furrow irrigation) and the other was planted at Akron, a rainfed site with no supplemental irrigation. Plots in Fort Collins were 3.4 m long on two raised beds 0.76 m apart, with two rows 20 cm apart planted on each bed. Plots at Akron were 3.35 m long and six rows wide, with 25-cm spacing between rows. In 2002, two trials were planted in Fort Collins in plots 3.35 m long and six rows wide, with row spacing of 23 cm. Sixty-five grams of seed per plot were planted in all trials. Two moisture levels, fully irrigated and partially irrigated, were imposed on the 2002 trials. The fully irrigated trial was watered at approximately weekly intervals throughout the growing season with a linear overhead sprinkler system. The partially irrigated trial received the same treatment until heading. Water was then withheld for 3 wk, after which time a final irrigation was given to prevent premature plant death. Weeds were controlled with post-emergence herbicides, supplemented by manual control as needed. Jointed goatgrass (Aegilops cylindrica Host) infested part of the trial in Akron. To indicate the year and the moisture stress treatment, we abbreviate the fully irrigated environments as FC01I and FC02I and the drought-stressed environments as AK01D and FC02D (Table 1).
Data were recorded for days to heading, plant height, grain yield, test weight, above-ground biomass (AGB), and kernel weight. Days to heading was the number of days from planting to the day on which 50% of the spikes were fully visible above the flag leaf collar. Plant height was recorded approximately 2 wk after heading and measured from the soil surface to the tip of the awns on five primary tillers per plot. Grain yield was determined with a small plot combine that harvested the center four rows at Akron and the entire plot in the other environments. A subsample of grain from each plot was cleaned and used to determine test weight. Above-ground biomass samples were collected immediately before harvest. A 1-m section of a single row was cut at the soil surface, placed in a large plastic bag, and weighed. The biomass samples were threshed and kernel weight was estimated by counting 200 kernels from each sample.
The stress intensity of yield trial environments was calculated in a manner similar to the drought intensity score of Fischer and Maurer (1978), i.e., intensity = 1 – (X/Xp), where X = the mean yield of the population in a given environment and Xp = yield potential of the population under nonlimiting conditions. We used the yield of the fully irrigated site FC01I as an estimate of yield potential.
Molecular Marker Analysis
Genomic DNA was extracted from leaves of two-week old greenhouse-grown F6:8 plants, according to the procedure of Ma and Sorrells (1995), with minor modifications. The Rht-B1 and Rht-D1 genotypes of the RILs were obtained by PCR on the basis of primer pairs developed by Ellis et al. (2002). The primer pair designated BF-MR1 was used for identification of Rht-B1 genotypes and primer pairs DF-MR2 and DF2-WR2 were used to identify Rht-D1 genotypes. The PCR conditions were the same as specified by Ellis et al. (2002), except that the annealing temperature for the BF-MR1 primer pair was 62.4°C. Primer sequences for the microsatellite marker gwm494 were used to amplify DNA with the protocols specified by Roder et al. (1998). Amplified DNA was loaded into a 4% (w/v) high-resolution agarose gel (Agarose SFR, Amresco Inc., Solon, OH) and electrophoresed at 85 V and 70 W for 2 h 30 min. The gels were stained with ethidium bromide for 30 min, rinsed for 10 min in deionized water, then digitally photographed under ultraviolet light with the AlphaImager imaging system (Alpha Innotech Corp., San Leandro, CA). Marker bands were evaluated with AlphaEase software (Alpha Innotech Corp., San Leandro, CA) by scoring each RIL in relation to the parental marker classes.
Data Analysis
Unless otherwise specified, all statistical analyses were conducted using SAS/STAT software, version 8.2 (SAS Institute Inc., 1999). Because of suspected spatial variability, especially in the drought-stressed trials, we conducted a series of analyses to determine the best method for adjusting trait values to account for this variability. For each trait in each environment, we ran three models in SAS PROC MIXED: (i) a model that adjusted for incomplete block effects based on the rectangular lattice design; (ii) a model that accounted for variation both in incomplete blocks (rows) and columns (i.e., the dimension perpendicular to the incomplete blocks); and (iii) an anisotropic model that adjusted trait values on the basis of spatial covariance analysis of adjacent plots (Table 41.4, SAS/STAT User's Guide, http://v8doc.sas.com/sashtml/stat/chap41/sect20.htm#mixedrepeat, accessed 27 Sept. 2004; verified 14 December 2004). In the anisotropic spatial power model, the covariance between two plot values is an exponentially decreasing function of the distance between plots, with separate parameters fit for column and row distances. For Models 2 and 3, the Satterthwaite option was used to calculate denominator degrees of freedom. Incomplete blocks and columns were considered to be random effects. For Model 1 in the two drought-stressed environments, an additional fixed variable designated "field" was also included. In the Akron trial, plant growth was noticeably reduced in the western 40% of the field because of jointed goatgrass infestation and previous cropping history that was different from the eastern 60% of the field. Therefore, the variable "field" was assigned a value of 1 for plots in the western 40% and 0 for plots in the eastern 60% of the site. Similarly, in the FC02D trial, a low-lying strip of land 3 m wide produced poor stands and late maturing plants when compared with the rest of the field. Plots in that strip were assigned a value of 1 for "field" and the remaining plots were assigned the value of 0.
For each model, we obtained least squares means and standard errors of each entry for each trait and environment. To determine the most appropriate model, we evaluated the results of the analyses based on the three models, and considered the most appropriate model to be one in which the spatial variable(s) were significant at = 0.05 and the mean standard error of the differences between entry means was smaller than for the other models.
Frequency distributions of least squares means for plant height were plotted separately for each environment with Excel 2000 (Microsoft, Redmond, WA). Normality of trait distribution was evaluated visually from these plots and by conducting the Shapiro-Wilk test of normality with the SAS UNIVARIATE procedure. Spearman rank correlation coefficients were calculated among plant height and grain yield values of the RILs in pairwise combinations of the four environments as an indication of the level of crossover genotype-by-environment interaction.
We performed a chi-square analysis to detect significant deviation from expected Mendelian segregation ratios. Expected ratios were 1:1 for segregation at the individual Rht-B1 and Rht-D1 loci, and 1:1:1:1 for the four genotypic classes defined by allelic combinations at the two loci.
To evaluate the magnitude of the effects of alleles at the Rht-B1 and Rht-D1 loci on the agronomic traits, we conducted an analysis of variance with SAS PROC GLM. The model included both loci as main effects and the interaction of the two loci to test for epistasis. The relative stability of Rht classes over environments was determined by analyzing the four classes separately in SAS PROC GLM models with environments and genotypes as main effects and genotype-by-environment as an interaction effect. The ratio of the interaction Type III sums of squares to genotype sums of squares was interpreted as an indication of relative stability.
To analyze the relationship between plant height and grain yield, we plotted the data for each RIL, with plant height on the x axis and percentage of the environment mean for yield on the y axis, with a separate graph for each environment. This analysis, which is modeled after the examples in Richards (1992a), allowed comparison of all environments on the same scale. We regressed yield on plant height for each environment, fitting linear and quadratic models in SAS PROC REG and testing for significance at = 0.05. When the quadratic parameter was not significant, we used the linear model if the linear term was significant. When the linear term was also nonsignificant, we concluded that there were no changes in the trait with increasing plant height. The graphs were drawn with SigmaPlot Version 8.0 software (Point Richmond, CA). Homogeneity of regression coefficients was determined according to Gomez and Gomez (1984).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
= 0.05 for 20 of 24 trait–environment combinations, and the mean standard error of the differences between entry means for that model was consistently smaller than for the other models. In the few cases where the mean standard error was lower for Model 1 or 2, the difference among models was very small. Based on these results, we concluded that Model 3 was the most appropriate model for obtaining spatially adjusted entry means for these trials and used it for all the analyses. The different moisture conditions in the four environments resulted in a range of environmental means for plant height and grain yield (Table 1). All trials experienced considerable heat stress during the critical pre-fertilization through early grain-filling periods; mean daily maximum temperatures were 31.9 to 32.7°C (Table 1), compared with an optimum temperature for wheat grain growth of 15°C (Paulsen, 1994). The highest mean yield (5067 kg ha–1) was obtained at FC01I, where mean plant height (95.2 cm) was also the highest of the four environments. Some lodging occurred in that environment, especially in plots with the tallest RILs, requiring manual assistance at harvest. The next highest yielding trial was FC02I, which was irrigated at regular intervals, but experienced extended periods of high temperatures from heading to maturity. The trials at AK01D and FC02D had similarly low mean yields (1452 and 1358 kg ha–1, respectively), both environments enduring limited moisture and high temperatures after heading. Mean plant height in FC02D was 16% less than in FC02I, but yields were 39% lower (Table 1). This is explained by the fact that those environments shared the same irrigation treatments until heading, by which time plant height was largely established, but they had very different moisture treatments during the grain-filling period, which is crucial for determining kernel weight. Stress intensities calculated for the three lower yield environments were as follows: AK01D, 0.71; FC02I, 0.56; and FC02D, 0.73, indicating yield reductions in the 50 to 75% range.
When plant heights of the RILs were compared in pairwise combinations of environments, rank correlation coefficients were high (r = 0.65 to 0.91, P < 0.001, n = 144). In contrast, grain yield ranks of RILs in different environments were poorly correlated (r = 0.13, P = 0.12 to r = 0.30, P < 0.001, n = 144). These results indicate that plant height expression was relatively consistent, even across a broad range of mean environment heights (Table 1), but that there was considerable crossover-type genotype-by-environment interaction for yield.
Frequency plots of the RILs for plant height revealed approximately normal distributions (Fig. 1) , although three of the four environments deviated from normality (P < 0.01) according to the Shapiro-Wilk statistic. Normality for plant height was not necessarily expected in this population because of the suspected segregation of major genes affecting the trait, rather than the segregation of many genes of small effect.
|
Marker-Trait Associations
Analysis of variance confirmed that both Rht-B1 and Rht-D1 had highly significant (P < 0.0001) effects on plant height, and that the two loci together accounted for major portions of the phenotypic variance for the trait, up to 70% in FC01I (Table 2). Of the other traits evaluated, the two loci had the most consistent and significant effects on test weight (P < 0.001 in all environments) and above-ground biomass (P < 0.001 in three of four environments). For grain yield, kernel weight, and days to heading, the results were more variable, yet still indicate the association (P < 0.05) of one or both Rht loci with these traits in at least three of the four environments. For all traits except days to heading, the a (tall) allele increased the trait value.
|
Statistical association of the Rht loci with traits other than plant height could be due either to pleiotropy or linkage. Pleiotropy is suggested by the following considerations. First, for all significant trait-locus associations, the effects of the a (tall) and b (dwarf) alleles were in a consistent direction at each locus. For example, test weight was always greater with the a allele at both the Rht-B1 and Rht-D1 loci. Second, the positive relationships between plant height and grain traits are physiologically logical and some of these relationships have been reported in other studies (e.g., Flintham et al., 1997; Nizam Uddin and Marshall, 1989; Richards, 1992a). Plants with taller stems presumably have greater reserves of assimilates available for remobilization to the developing grain under stress conditions (Blum et al., 1997a), resulting in higher test weight and kernel weight. Even under full irrigation at FC01I, there may have been sufficient heat stress during grain filling (Table 1) to reduce current photosynthesis and give an advantage to plants with greater stem reserves (Blum et al., 1994; Yang et al., 2002). An alternative explanation was provided by Richards (1992b), who attributed improved performance of tall wheats in dry environments more to water-use efficiency than to stem reserve mobilization. Third, the effect of plant height on days to heading was due solely to delayed heading in the dwarf class (Table 3). Slower growth of the expanding stem in this class resulted in delayed or partial emergence of the head above the flag leaf collar.
|
= 0.05. Compared with the tall class, the semidwarf classes showed an average height reduction of 14.8% (range of 11.4–17.8% for the four environments). The dwarf class was on average 35.7% shorter (range of 30.3–40.5%) than the tall class. Thus, although the environmental conditions and mean plant heights differed dramatically in our trial environments (Table 1), the relative effects of the two Rht loci on plant height were remarkably consistent across environments. Our results are similar to those of Flintham et al. (1997) on the basis of near-isogenic lines in mostly high yielding locations. Those authors reported that Rht-B1b or Rht-D1b alone reduced plant height by an average of 15.5%, whereas the combined dwarfing alleles caused a 42% height reduction. Richards (1992a) found greater height reductions than we did (23% for single dwarfing alleles and 47% for the combination), indicating the importance of genetic background and environment in determining the magnitude of height reduction. In contrast to plant height, relative grain yield of the Rht classes varied across environments (Table 3). The most consistent finding was that the average yield of the dwarf class (Rht-B1b + Rht-D1b) was significantly lower (P < 0.10) than the other classes in all environments. Yield of the tall class was similar to or higher than either semidwarf class in each environment, although the Rht-B1b + Rht-D1a class yielded slightly more (P = 0.06) in the highest yielding environment, FC01I. The two semidwarf classes did not differ from each other in three of the four environments, but the Rht-B1b + Rht-D1a class yielded more (P < 0.05) in FC01I. Averaged over all environments, the Rht-B1b + Rht-D1a class yielded 103% and the Rht-B1a + Rht-D1b class yielded 99.3% of the environment mean.
The consistently poor yield performance of the dwarf class is likely attributable at least in part to the later heading date of that group (Table 3). In all environments, the Rht-B1b + Rht-D1b class headed later (P < 0.05) than the other classes; the difference between the dwarf class and the average of the three other classes ranged from 3.0 d in AK01D to 4.5 d in FC01I, with an average difference of 3.5 d. Later heading in our environments would have exposed plants to greater heat and drought stress during the critical anthesis and postanthesis stages, thus resulting in yield reductions. This association of Rht mutant alleles with later heading dates was also reported by Richards (1992a). However, in our study the effect was observed only for the dwarf class, whereas Richards also reported delayed maturity for the single alleles Rht-B1b and Rht-D1b.
For test weight, the tall class outperformed the other classes (P < 0.01) in each environment, the dwarf class was inferior to all other classes (P < 0.01), and the semidwarf classes were intermediate and nearly equal to each other (no difference at = 0.05) (Table 3). The superiority of the tall class was particularly evident in drought-stressed AK01D, where tall lines averaged 103% of the environmental mean for test weight and the semidwarf classes averaged 99.1% of the mean. This result is consistent with the report by Guttieri et al. (2001), who found that test weights of two tall hard red spring cultivars were less affected by severe drought stress than were two semidwarf hard red spring cultivars. Results for kernel weight and AGB revealed similar trends, although not as clear-cut as for test weight, in which the taller classes showed higher values for the two traits.
To determine whether grain yield stability over environments differed among Rht classes, we calculated ratios of genotype x environment sums of squares to genotype sums of squares. The results for the four classes were not dramatically different; ratios were 2.21, 2.52, 2.82, and 1.98 for the tall, Rht-B1b + Rht-D1a, Rht-B1a + Rht-D1b, and dwarf classes, respectively. While the stability of the dwarf class appears somewhat better according to this measure, its low mean yield overrides the importance of a modest improvement in stability.
To estimate the reduction in performance in response to stress, the most appropriate comparison in our study is between environments FC02I and FC02D. These trials were planted on the same date at the same location, experienced the same air temperatures, and had the same mean heading date. However, they received different irrigation treatments (see Materials and Methods) and were grown on different fields approximately 500 m apart. Sensitivity to stress, calculated as reduction in plant height from FC02I to FC02D, was greatest for the tall class (20%), lowest for the dwarf class (6.4%), and intermediate for the semidwarf classes (13.1 and 14.5%). Nevertheless, plant height in the stress treatment remained greatest for the tall class (Table 3). These findings are consistent with the report of Blum et al. (1997b) that shorter wheat plants are more tolerant of growth reduction under stress, but absolute performance was better in taller plants because of their greater growth potential. For grain yield in our study, the percentage reduction between FC02I and FC02D was 41.9% for the dwarf class, 35.0% for the tall class, 35.2% for the Rht-B1b + Rht-D1a class, and 39.8% for the Rht-B1a + Rht-D1b class. Despite its larger percentage plant height reduction, the tall class experienced a smaller reduction in yield and, most importantly, had the highest yield under stress (Table 3).
Regression of Agronomic Traits on Plant Height
When grain yield was plotted against plant height and regression curves overlaid (Fig. 2)
, the varied relationships between the traits in different environments became clearer. In the high yield environment (FC01I, Fig. 2A), the shape of the curve approximates results from other studies (Richards, 1992a; Flintham et al., 1997), with optimum yield for intermediate plant heights and reduced yields for taller and shorter plants. In FC02I and FC02D (Fig. 2C, D), even though the quadratic coefficients were significant, the curves are mostly linear for the range of observations. The curves reveal a gradual yield increase with height for FC02I and a much sharper increase for FC02D. Linear regression coefficients were 1.2 and 2.6% of environment mean yield per cm height for FC02I and FC02D, respectively (significantly different at = 0.001). In the latter, severely stressed environment, taller plants had a clear advantage. Neither linear nor quadratic regression coefficients were significant (P > 0.05) for AK01D (Fig. 2B), probably because of variable field conditions that were not completely resolved by spatial data analysis at that location. As mentioned previously, the height–yield relationships portrayed in these graphs should be interpreted cautiously, keeping in mind the unequal representation of Rht classes and the fact that dwarf lines headed an average of 3.5 d later than the other classes.
|
For moisture stress conditions (the other three environments of our study), the optimum allele combination is less obvious. Both tall lines and Rht-B1b + Rht-D1a semidwarf lines were among the highest yielding lines in those environments, suggesting that either of these allelic classes may be desirable. In line with Richards' (1992a) observations, the specific allelic combination for these environments appears to be less important than an appropriate plant height. The best choice for stressed environments may be either taller lines within the Rht-B1b + Rht-D1a semidwarf class or shorter lines within the tall class. Because of the considerable variability we observed within each of these groups, selection for yield, yield stability, test weight, and other traits would be possible and necessary. This strategy will likely lead to acceptable performance under typical conditions, while providing some buffering against the effects of unusually wet or dry weather conditions.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication May 27, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in Crop Science:
This article has been cited by other articles:
|
|
 |
|
  S. R. Clayshulte, S. D. Haley, P. F. Byrne, and X. Shan Trait Associations at the Xgwm 261 and Rht-B1 Loci in Two Winter Wheat Recombinant Inbred Line Populations Crop Sci., November 7, 2007; 47(6): 2346 - 2355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Mathews, S. C. Chapman, R. Trethowan, R. P. Singh, J. Crossa, W. Pfeiffer, M. van Ginkel, and I. DeLacy Global Adaptation of Spring Bread and Durum Wheat Lines Near-Isogenic for Major Reduced Height Genes Crop Sci., February 1, 2006; 46(2): 603 - 613. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | 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 | |||
0.05. A, FC01I; B, AK01D; C, FC02I; and D, FC02D. Data points are distinguished by shape for allelic class (
, Rht-B1a + Rht-D1a; 
, Rht-B1a + Rht-D1b; 
, Rht-B1b + Rht-D1a; •, Rht-B1b + Rht-D1b).