| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Dep. of Plant Sciences, North Dakota State Univ., Fargo, ND 58105-5051
* Corresponding author (Mohamed.Mergoum{at}ndsu.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DYL, direct yield loss • FHB, Fusarium head blight • FS, field kernel shattering • GS, glume strength • IRI, induced random impact • KS, kernel shattering • rS, Spearman's rank correlation • SS, kernel shattering from spikes • VS, visual score
Dep. of Plant Sciences, North Dakota State Univ., Fargo, ND 58105-5051
* Corresponding author (Mohamed.Mergoum{at}ndsu.edu).
Kernel shattering (KS) is known to cause yield loss in wheat (Triticum aestivum L.). The introduction of Fusarium head blight (caused by Fusarium graminearum Schwabe) resistant genotypes has recently elevated KS importance. This study aimed to identify appropriate methods to measure and evaluate KS variation among spring wheat genotypes. Four field and two laboratory methods were developed and compared using 24 wheat genotypes in five environments. Large variation for KS among genotypes was observed. The field kernel shattering method and the kernel shattering from spikes method (SS), both field methods, showed large CVs and were inconsistent across environments. Similarly, the direct yield loss method and the visual score method, also field methods, had relatively low range/LSD0.05 ratio values, leading to difficulty in discriminating between genotypes. Among the field methods, SS appears to be the most suitable method due to its high range/LSD0.05 values and fewer input requirements. All field methods, however, were time and labor consuming. The laboratory methods, induced random impact and glume strength, were moderately to strongly correlated with the field methods. The laboratory methods were more consistent across environments, and required only a small sample size. Hence, they were more effective and efficient in evaluating KS and could be of interest for wheat researchers to indirectly screen for KS resistance.
Abbreviations: DYL, direct yield loss • FHB, Fusarium head blight • FS, field kernel shattering • GS, glume strength • IRI, induced random impact • KS, kernel shattering • rS, Spearman's rank correlation • SS, kernel shattering from spikes • VS, visual score
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Most research on KS in wheat was conducted during the 1930s to 1950s (Dunkle, 1934; Vogel, 1938, 1941; Chang, 1943; Harrington and Waywell, 1950; Porter, 1959). Since then, limited studies addressing this problem have been conducted. Similarly, few evaluation methods have been developed to estimate KS. In early work by Vogel (1941), wheat cultivars were classified as highly resistant, resistant, moderately susceptible, or susceptible to KS based on direct observation in the field. Porter (1959), however, determined shattering percentage by scoring spikes, which were selected randomly from the field when the susceptible genotype shattered >30%. More recent work (Clarke, 1981; Clarke and DePauw, 1983) estimated KS by collecting dislodged grain on the ground with a window screen placed between rows. Clarke and DePauw (1983) also studied temporal changes of KS and found that a single measurement after maturity when the grain was at 145 g/kg moisture could be predictive, since genotype rankings did not change significantly after this stage. Visual estimates of KS, however, are still commonly used by researchers.
Evaluation of KS is difficult, particularly under field conditions where various environmental conditions interfere. It may also be hampered by the absence of weather conditions conducive to KS (Harrington and Waywell, 1950). Therefore, controlled laboratory tests are needed that provide accurate estimates of KS. In the past, several laboratory devices were designed to measure KS in wheat. Dunkle (1934), as reviewed by Porter (1959), built an apparatus with two rotating brushes that alternately brush spikes to cause KS. Later, Porter (1959) found that Dunkle's apparatus broke a large portion of the rachis, which inspired him to build another device called "Amarillo," whose correlation with a field shattering score was r = 0.66. Earlier, Chang (1943) invented a paddle device, which hit the spikes uniformly with a rubber paddle. The results generated from this experiment showed significant cultivar difference for KS; however, there were no data showing its correlation with field shattering scores. In another study, Vogel (1941) found a negative correlation between KS and the force required to break the outer glume (GS). This general correlation was confirmed later by Harrington and Waywell (1950) and Ghassem (1970). Similar results were also obtained in barley (Hordeum vulgare L.) and rice (Oryza sativa L.) (Chapman and Hockett, 1976; Ichikawa et al., 1990). Thus, GS appears to be a good indicator for predicting KS.
The majority of methods described previously are time and labor consuming. Also, all reported devices for evaluating KS under laboratory conditions are presently unavailable for breeding programs. Therefore, alternative and new devices or methods need to be identified or invented. In addition, very limited information has been published on the efficiency and effectiveness of different evaluation methods in screening modern wheat genotypes for KS under modern crop management practices. Accordingly, the objectives of this study were to (i) develop and compare different methods for measuring KS, (ii) identify appropriate methods for measuring KS, and (iii) evaluate variation in KS among genotypes.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
-[2-(4-chlorophenyl)ethyl]--(1,1-dimethylethyl)-1H-1,2,4-triazole-1-ethanol) fungicide at 0.31 L/ha at Feekes Stage 10.5.1 to minimize the possible confounding effects of FHB on KS because FHB is well documented to result in shriveled kernels (Rudd et al., 2001), whereas plump kernels tend to increase KS (Platt and Wells, 1949; Clarke and DePauw, 1983). The fungicide was sprayed as described by Hofman et al. (2000).
Kernel Shattering Evaluation Methods
Field Evaluation Methods
The field kernel shattering (FS) method consisted of counting kernels that had fallen within an area of 30 by 30 cm2 per plot. The area was selected randomly and the shattered kernels were counted immediately after the combine harvest (3 wk after Feekes Stage 11.4). Grain yield was also determined for each plot and kernel weight was measured on 250 kernels. Grain yield loss percentage due to KS within each plot was estimated by multiplying the kernel weight by the shattered kernel number, and then converted to percentage of total yield.
The kernel shattering from spikes (SS) method consisted of counting the number of shattered kernels per spike on 20 fully developed spikes selected at random among the main spikes or primary tiller spikes. In 2003, the spikes were cut 2 wk after Feekes Stage 11.4. Based on 2003 data, however, the sampling time in 2004 and 2005 was modified to 3 wk after Feekes Stage 11.4. This delay allowed better expression of KS. The yield loss due to KS was estimated as the percentage of shattered kernels per spike.
The direct yield loss (DYL) method was tested in 2005 and aimed to directly measure yield loss due to KS by comparing the yields at two harvest dates (early vs. late). The early harvest was done at Feekes Stage 11.4, while the late harvest was done 3 wk later. The early harvest represented a normal harvest, while the late harvest represented a typical situation when harvest is delayed for various reasons, including climatic conditions such as rain. The yield difference between the two dates was used as a direct measure of yield loss due to KS. The yield loss was expressed as the percentage of the early harvested yield.
The visual score (VS) method consisted of assigning a score on a 1 to 5 scale 3 wk after Feekes Stage 11.4 (just before the late harvest) at Prosper in 2005. The score was based on visual estimation of the percentage of shattered kernels from spikes (1 = 0%, 2 = 20%, 3 = 50%, 4 = 80%, 5 > 80%). This method was of great interest, as it may represent an efficient and quick selection method that breeders might prefer.
Laboratory Evaluation Methods
Fully developed, intact spikes sampled from the field experiments were used for the laboratory methods. In 2003, spikes were sampled 2 wk after Feekes Stage 11.4, while in 2004 and 2005 spikes were sampled at Feekes Stage 11.4 to get intact spikes before KS occurred. The spikes were sampled from each plot in the field experiments, put in paper bags, and stored for 2 mo at room temperature to ensure that spikes dried to uniform moisture.
The induced random impact (IRI) method was designed to simulate the natural forces that cause kernels to shatter in the field. A barrel of 20-cm diameter was used to rotate spikes together with steel balls. Preliminary trials including four genotypes (two resistant and two susceptible, based on the field data) were performed to determine the optimal conditions (rotating time, ball size and number, and the number of spikes) for inducing KS. To clearly discriminate between the genotypes, preliminary trials showed that 100 steel balls of 1-cm diameter rotating at 33 rpm with 10 spikes for 20 s were the optimum conditions. Then the IRI method was applied to all genotypes. Kernel shattering was determined as the percentage of shattered kernels per spike.
The glume strength (GS) method consisted of measuring the strength required to break down the outer glume with an electronic force gauge (Model DFM10, AMETEK Inc., Largo, FL). The apparatus was fixed on a stand and attached with a small hook. To measure the force required to break down the glume, the hook was inserted between the glume and lemma. The spike was then pulled down evenly until the glume was loosened. The equipment records the maximum force required to loosen the glume. Preliminary trials using two resistant and two susceptible genotypes were conducted to determine what part of the spike would be representative of the whole spike for GS. Harrington and Waywell (1950) suggested that the average of the second and third spikelets from the top of the spike together with the third and fourth spikelets from the bottom of the spike was representative. Lebsock (1950) and Longwell (1952) measured five central spikelets in durum wheat. Our preliminary results (data not shown) indicated, however, that both the fourth spikelet counting downward from the top and the middle spikelet were highly correlated with the whole spike. The glume strength was tested for all genotypes using the fourth spikelet in 2003 and the middle spikelet in 2004 and 2005, since the fourth spikelet in some genotypes was found to be easily broken before measuring. Glume strength was determined as the average of 10 spikes in 2003 and of three spikes in 2004 and 2005.
Statistical Analysis
Analysis of variance was conducted for each evaluation method in each environment by using the GLM procedure of SAS (SAS Institute, 2003). Error variance homogeneity was determined by using Bartlett's 
2 test at P = 0.005. Combined ANOVA was not performed since error variance was not homogeneous for FS, SS, or IRI due to the large error variance observed in some environments. The value of the range/LSD0.05 ratio (Guner et al., 2002; Shetty et al., 2002; Gabriele and Wehner, 2005) was calculated for each method in each environment and used as an indicator of discrimination ability. The discrimination ability is proportional to the ratio of range/LSD0.05, so the greater this ratio, the better the discrimination between genotypes. Based on the individual genotype's mean, the CORR procedure of SAS (SAS Institute, 2003) was used to calculate the correlation between evaluation methods in each environment and correlation between environments for evaluation methods FS, SS, IRI, and GS, which were evaluated across five environments. Homogeneity of correlation between FS, SS, IRI, and GS from individual environments was tested and the results were pooled if homogeneous (Steele et al., 1997).
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
In general, both methods had large CVs in 2003 and 2004 (70.5–108.6%, Tables 1 and 2); however, CVs were generally lower in 2005 (40.7 and 33.6% for FS and SS, respectively). The values of range/LSD0.05 were relatively low (3.1 and 3.4 for FS and SS, respectively) at Carrington in 2003, where the lowest levels of KS occurred. Values of range/LSD0.05 were the highest at Prosper in 2005 (5.3 and 8.0 for FS and SS, respectively), where the most KS occurred. These data indicate that a favorable environment for KS allows these two field methods to better differentiate between genotypes. All Spearman's rank correlations (rS) between environments for both methods were significant (P < 0.01). For the FS method, the lowest correlation (rS = 0.54, P < 0.01) was observed between Carrington 2003 and Prosper 2005, while the highest correlation (rS = 0.83, P < 0.001) was observed between Carrington and Prosper in 2003 (Table 3 ). Similarly, for the SS method, the correlation coefficients ranged from 0.61 to 0.90, which were slightly higher than those for FS. The lowest correlation for SS occurred between Prosper 2005 and Carrington 2003 (rS = 0.61, P < 0.01) while the highest correlation (rS = 0.90, P < 0.001) was observed between Carrington 2004 and Prosper 2003.
|
Laboratory Methods
Induced Random Impact
The IRI method generally generated wider ranges than the field methods when low KS occurred in the field, including Carrington 2003 and 2004, for which ranges from 0 to 11% and from 0.2 to 10.9%, respectively, were observed (Table 1). The average KS values generated by the IRI method in 2004 and 2005 were similar (5.6–6.7%), while they were higher at Prosper 2003 (11%) and lower at Carrington 2003 (3.9%) (Tables 1 and 2). The resistant (low scores) and susceptible (high scores) genotypes were generally consistent across 2003 and 2005 trials. Amidon, Hanna, Parshall, Glenn, and Butte86 were consistently among the lowest scores, while 03NZ4153, Granite, Granger, ND734, and ND4197 were among the highest scores. Additionally, most of these resistant or susceptible genotypes showed shattering reactions similar to those in the field in 2005. This may indicate the effectiveness of the IRI method in screening genotypes for KS resistance.
The CVs of the IRI method in 2003 and 2005 were low (30.5–43.7%), while they were higher (59.5 and 93.4%) in 2004. Similarly, the values of range/LSD0.05 varied from 4.2 to 5.8 in 2003 and 2005, while low values (2.3 and 2.9) were associated with the 2004 environments. All Spearman's rank correlation coefficients across five environments were highly significant (P < 0.001, Table 4 ). This suggests that the IRI method could be a reliable method for evaluating KS. The highest correlation (rS = 0.93) was recorded between 2003 and 2005 at Prosper, while the lowest (rS = 0.71) was recorded between the two sites in 2004. Additionally, all correlation coefficients for the IRI method between environments in 2003 and 2005 were >0.90.
|
The GS method had relatively small CVs, ranging from 17.0 to 23.0% (Tables 1 and 2). The values of range/LSD0.05 for GS were generally larger than those reported for other methods, ranging from 6.5 to 9.6. This indicates that the GS method tended to have strong discriminatory ability. All Spearman's rank correlation coefficients between environments were highly significant (P < 0.001, Table 4). This indicates that the GS method appears to be less dependent on environmental conditions than the field methods. The highest correlation (rS = 0.91) was recorded between the Prosper and Carrington sites in 2003, while the lowest (rS = 0.69) was between 2003 and 2004 at Carrington.
Correlations among Different Evaluation Methods
The correlation coefficients among evaluation methods FS, SS, IRI, and GS were homogeneous across all five environments. The pooled correlation coefficients were all highly significant (P < 0.001; Table 5
). The correlation between field methods FS and SS was the strongest (r = 0.90 and rS = 0.84) compared with the other methods. Laboratory methods IRI and GS were moderately to strongly correlated with the field methods. The Pearson correlation coefficients between IRI and the field methods (FS and SS) were slightly lower than the Spearman's rank correlation coefficients. The GS method was negatively correlated with FS and SS, with Pearson correlation coefficients of –0.41 and –0.45, which were not as strong as those noted for the IRI method. The Spearman's rank correlation coefficients between the GS and field methods (FS and SS), however, were similar to those of IRI. The two laboratory methods, IRI and GS, were negatively and moderately correlated.
|
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
For field method FS, obviously some kernels would come out of the plot combine during harvest, which might cause confounding effects on the counting after harvest. To determine whether this had a significant confounding effect on the FS method, notes on KS before and after harvest were taken in 2003. The results showed strong correlations between these two counts (r = 0.86 and 0.96 at Carrington and Prosper, respectively, P < 0.001). Additionally, it is hard to randomly select a spot before harvest because of the standing plants. Further, more KS would be created by moving plants before harvest. Therefore, KS data reported in this study were based on counts after harvest.
Kernel shattering in the field can be affected by various factors related to the environment and genotypes. Therefore, it is not surprising that data generated by the field methods FS and SS varied across locations and years in this study. Poehlman (1987) stated that large yield loss might occur under favorable conditions for KS. Platt and Wells (1949) proposed that crop conditions leading to large and plump kernels and weather conditions, particularly winds at maturity, were two primary factors contributing to shattering in the field. Later, Clarke and DePauw (1983) suggested that irrigation could increase kernel size, thereby enhancing KS and providing optimal conditions for differentiating between genotypes. Contrary to Clarke and DePauw (1983), this study showed that less KS was generally observed at Carrington, the irrigated site. In 2003, although large kernels were produced at Carrington, the occurrence of severe lodging might have protected the spikes from shattering. Vogel (1941) had experienced similar findings regarding the effect of lodging on KS. Additionally, greater values for glume strength were observed in spikes from Carrington in 2003. Therefore, both lodging and stronger glume strength could have contributed to lowering KS at Carrington in 2003. In 2004, severe FHB infection occurred at Carrington, which caused shriveled seed size. Hence, the infected, shriveled kernels did not shatter, which resulted in low KS. Therefore, irrigation alone probably does not cause an increase in KS in the field. Other factors such as lodging and disease, as this study showed, are critical to induce KS under field conditions.
Due to environmental and growing conditions, field methods FS and SS were not very consistent in our study. The results from Carrington 2003, an environment with unfavorable conditions for KS, did not strongly correlate with the results from Prosper 2005, which was an environment with favorable conditions for KS for both FS and SS methods. This inconsistency raises a question on repeatability of the field methods. The results produced with field methods in environments with low KS may not accurately predict the shattering potential of particular genotypes. Platt and Wells (1949) concluded that one field test might not be sufficient for evaluating KS due to the pronounced effect of the environment. Therefore, multiple environments might be required to evaluate KS with field methods. Under a favorable environment for KS, however, such as Prosper 2005, a wide range of KS was observed among genotypes assessed in our study. The values of range/LSD0.05 were high for both field methods FS and SS. These results suggest that the FS and SS methods can work very well to characterize genotypes under favorable conditions for KS.
High CV values for the field methods FS and SS might be due to the large error variance observed in this study. The error variance increased with the amount of KS in this study. The relatively low CV values at Prosper in 2005 might be due to their high means. Clarke and DePauw (1983) also reported high error variance for KS in their study, which they attributed to microenvironment changes. They suggested that a large number of replications might be required for evaluating KS in the field. Therefore, the number of replications needs to be considered, particularly when using these two field methods.
The KS levels with the SS method were generally higher than with FS in 2004 and 2005. The difference might be caused by their different estimation methods. The FS method is based on counting kernels on the ground surface, some of which might be buried in the soil or eaten by insects or birds. Hence, the FS method may underestimate KS. The SS method is based on lost kernels from the spikes of the main stem or primary tillers, which generally shatter more than the other smaller tillers since these spikes mature earlier and would experience more wind forces. Therefore, the SS method may overestimate KS. This bias, however, was larger for susceptible genotypes than for resistant genotypes, which resulted in a larger range, and thereby higher values of range/LSD0.05 for the SS method. This allowed the SS method to better differentiate between genotypes for KS.
In this study, 20 spikes per plot were sampled for the SS method, which indicates that one or two rows could be enough to get a sample. The FS method, however, requires large plots to allow kernels to fall within their own plot area and avoid confusion with adjacent plots. This might cause practical limitations in the use of the FS method in early generations of testing. Therefore, the SS method might be more efficient and more appropriate than FS for testing a large number of genotypes.
Field method DYL was designed to directly measure yield loss due to KS; however, this method did not correlate well with the other field methods. This might be due to the fact that its yield loss was partly attributed to the whole spike loss, which was observed for most genotypes after harvesting late. Based on this study, even the resistant genotypes showed >12% yield loss using the DYL method. Considering the low value of range/LSD0.05 for the DYL method and the high resource requirement, the DYL method may be the least desirable method, although it can be used to identify susceptible genotypes for KS under favorable environments.
The VS method is currently the most common method used in breeding programs. The strong correlation between VS and the other field methods in 2005 indicates that it is useful in screening for KS under favorable environments since it is cost efficient and quick. This method requires more experience, however, to accurately score genotypes. Additionally, the discrimination ability of the VS method was low (the lowest value of range/LSD0.05 in 2005), indicating that VS is not as precise as other field methods. A modified scale of 1 to 9 instead of 1 to 5 might improve the discrimination ability; however, this would require more experience for evaluating. In disease evaluation, several reports have suggested that a percentage scale could be more precise than a five- or 10-point scale (James, 1974; Christ, 1991; Danielsen and Munk, 2004). This may also be true for the VS method in KS evaluation. Therefore, usefulness of the VS method may depend on further modification as suggested above.
All field methods appear to be influenced by environmental conditions. These conditions were sporadic, unpredictable, and therefore unreliable. In this study, Prosper 2005 was the only environment conducive for KS and the one where the best differentiation between genotypes was observed with field methods. Unfavorable environmental conditions usually result in less KS, which may cause difficulties in discriminating between genotypes. The KS performance of a genotype in such an environment may be misleading. Therefore, evaluating KS using field methods will require several site-year testing environments to be effective. In addition, delaying harvest is often required for KS to occur under field conditions. To get sufficient shattering to occur, Porter (1959) allowed plant materials to stand in the field for 45 d after maturity. Therefore, laboratory methods might be advantageous since they are less dependent on the environment than field methods. In the past, various laboratory methods have been tried; however, most have proven to be laborious or unreliable for practical use (Harrington and Waywell, 1950).
The IRI protocol was developed as a new method. Strong correlations (both Pearson and Spearman's rank) between IRI and field scores in 2005 suggest that IRI was successful in simulating favorable conditions for KS, which demonstrates the effectiveness of this method. Consistent strong correlations between environments for the IRI method, especially between the 2003 and 2005 trials, which represented totally different environmental conditions for KS, suggest that IRI is less dependent on environmental conditions for KS than the field methods. Therefore, using the IRI method, the samples in environments with low KS may have the same reliability as those in environments with high KS to predict the KS potential of genotypes. Additionally, the IRI method needs only a few spikes and less time, which indicates that it might be useful in evaluating large breeding populations in early generations; however, the quality of samples to determine the IRI score is crucial for the accuracy of this method. In the 2004 trials, for instance, the IRI method appeared to have high error variance and CV. These results might be explained by the presence of nonuniform samples, which were caused by uneven germination due to climatic and soil conditions at Prosper and severe FHB incidence at Carrington. Consequently, the values of range/LSD0.05 were very low and the correlations between 2004 and other years were also relatively low compared with those between the other years. This indicates that sample uniformity is essential for using the IRI method accurately.
Glume strength is an important factor in preventing KS and can be used as a good indicator for KS (Vogel, 1941; Lebsock, 1950; Longwell, 1952); however, other morphological and agronomic traits, such as spike compactness, awns, and kernel size, may also significantly affect KS (Vogel, 1941; Chang 1943). In this study, it was found that under field conditions, the genotypes with strong GS had consistently low KS while the genotypes with low GS had varying levels of KS. This may explain the relatively low Pearson correlations between GS and the field methods. Some genotypes with low GS had low KS, which may be explained by other factors. Similar results were reported by Vogel (1941) and Harrington and Waywell (1950), who concluded that GS could be an important component, but that it is not the only factor influencing KS. Glume strength was also found to be generally consistent across environments and had low CVs, a good indication that GS was less affected by environmental conditions than were the field methods. Additionally, the rankings in 2003 were strongly correlated with those in 2004 and 2005, where a smaller sample size was used. Hence, few environments and a small sample size may be sufficient to obtain reliable data using the GS method. Very strong GS could cause problems in threshing (Platt and Wells, 1949), however, as was observed in this study. Therefore, breeding cultivars with resistance to KS should avoid selecting genotypes with very strong glume strength.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The laboratory methods, GS and IRI, were more consistent across environments than the field methods and only small sample sizes were required for evaluation. The laboratory methods were moderately to strongly correlated with the field methods. The laboratory methods appear to be less dependent on environmental conditions and could be more efficient than field methods in environments where the incidence of KS is limited. Therefore, they can be used successfully to indirectly screen wheat genotypes for KS under conditions that are not conducive to KS. The IRI method developed in this study is very promising because it is fast and was more strongly correlated with the field methods than GS, but it may need to be optimized. Of the genotypes included in this study, all advanced breeding lines had observable amounts of KS. Most cultivars had low KS. Based on both field methods and laboratory methods conducted in this study, four cultivars, Parshall, Glenn, Amidon, and Butte86, consistently showed very low KS and had very good KS resistance. These cultivars can be utilized in breeding programs to improve KS resistance in spring wheat.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication December 18, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Zhang, M. Mergoum, S. Kianian, D. W. Meyer, S. Simsek, and P. K. Singh Genetic Relationship and QTL Association between Kernel Shattering and Agronomic Traits in Wheat Crop Sci., March 17, 2009; 49(2): 451 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Mergoum, R. C. Frohberg, R. W. Stack, P. K. Singh, S. Ali, J. W. Rasmussen, T. L. Friesen, and T. B. Adhikari Registration of Spring Wheat Germplasm ND 756 Combining Resistances to Foliar Diseases and Fusarium Head Blight Journal of Plant Registrations, January 1, 2008; 2(1): 61 - 64. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
for the 24 hard red spring wheat genotypes included at Carrington, ND, in 2003 and 2004.