| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop Science and Dep. of Plant Pathology, North Carolina State Univ., Raleigh, NC 27695-7620
b Dep. of Crop Sciences, Univ. of Illinois, Urbana-Champaign, IL 61801
c Dep. of Plant Pathology, North Carolina State Univ., Raleigh, NC 27695-7567
d USDA-ARS National Center for Agricultural Utilization Research, 1815 N. University St., Peoria, IL 61604
e USDA-ARS, Plant Science Research Unit, Dep. of Crop Science, North Carolina State Univ., Raleigh, NC 27695-7620
* Corresponding author (james_holland{at}ncsu.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: BIC, Bayesian Information Criterion • CIM, composite interval mapping • MIM, multiple interval mapping • PCR, polymerase chain reaction • QTL, quantitative trait locus (loci) • SSR, simple sequence repeat
a Dep. of Crop Science and Dep. of Plant Pathology, North Carolina State Univ., Raleigh, NC 27695-7620
b Dep. of Crop Sciences, Univ. of Illinois, Urbana-Champaign, IL 61801
c Dep. of Plant Pathology, North Carolina State Univ., Raleigh, NC 27695-7567
d USDA-ARS National Center for Agricultural Utilization Research, 1815 N. University St., Peoria, IL 61604
e USDA-ARS, Plant Science Research Unit, Dep. of Crop Science, North Carolina State Univ., Raleigh, NC 27695-7620
* Corresponding author (james_holland{at}ncsu.edu).
Resistance to Fusarium ear rot [caused by Fusarium verticillioides (Sacc.) Nirenberg (synonym F. moniliforme Sheldon) (teleomorph: Gibberella moniliformis) and F. proliferatum (Matsushima) Nirenberg (teleomorph: G. intermedia)] and fumonisin contamination is heritable and controlled by at least 11 gene regions in a maize (Zea mays L.) population created by backcrossing the highly resistant donor line, GE440, to the susceptible but commercially successful recurrent parent line, FR1064. The relationship between resistances to Fusarium ear rot and fumonisin contamination and agronomic performance has not been reported. Therefore, the objective of this study was to examine the relationship between disease resistance and agronomic utility in this population by measuring resistances to Fusarium ear rot and fumonisin contamination in BC1F1:2 lines, and yield and agronomic performance in topcrosses of these lines. Fumonisin contamination was not correlated with yield, but two fumonisin quantitative trait loci (QTL) mapped to similar positions as yield QTL. Fusarium ear rot had a small positive correlation with topcross yield (r = 0.29), but QTL for the two traits mapped to distinct genomic positions. Similar results for other traits indicate that QTL can contribute in opposite directions to the overall genetic correlations between traits and that some trait correlations arise in the absence of detectable QTL effects on both traits. In general, no strong relationships were observed between disease resistance traits and agronomic traits, thus selection for increased resistance should not unduly affect agronomic performance.
Abbreviations: BIC, Bayesian Information Criterion • CIM, composite interval mapping • MIM, multiple interval mapping • PCR, polymerase chain reaction • QTL, quantitative trait locus (loci) • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To limit the negative impacts of fumonisin contamination and Fusarium ear rot, selection for resistance to these two aspects of the disease is an important objective. Robertson et al. (2006) reported high family mean heritabilities and high positive genotypic correlations between fumonisin contamination and Fusarium ear rot, suggesting that phenotypic selection should be effective at improving resistance to these traits. The relationship between these disease resistance traits and other important agronomic traits also impacts the feasibility of developing cultivars with improved resistance, especially if the sources of disease resistance are unadapted exotic germplasm. Negative genetic relationships between disease resistance and other agronomic traits may occur due to negative pleiotropy or to repulsion-phase linkages between favorable alleles at disease resistance loci and at genes affecting agronomic traits. Conversely, positive genetic relationships between disease resistance and yield may occur due to the protection afforded by resistance genes against the yield-reducing effects of ear rot.
Robertson et al. (2006) studied the inheritance of resistance to Fusarium ear rot and fumonisin contamination in a population derived from the first backcross of GE440 to FR1064. GE440 has high levels of resistance to both Fusarium ear rot and fumonisin contamination (Robertson et al., 2006) but is an older inbred line developed from the Hastings Prolific open-pollinated variety (Jenkins and Robert, 1959) that is not acceptable for commercial hybrid production. The earliest published record of GE440 indicates that it was used in field studies in 1951 (Jenkins et al., 1957). In contrast, FR1064 is a widely used B73-type commercial inbred developed by Illinois Foundation Seeds (Clements et al., 2004). Therefore, further backcrossing of resistance alleles from GE440 into the FR1064 background via phenotypic selection and marker-assisted selection will be conducted with this population. However, the question remains regarding what the effects of selection for increased disease resistance on other agronomic traits will be in this backcrossing program.
The relationship between resistances to Fusarium ear rot and fumonisin contamination and agronomic performance has not been previously reported. The objectives of this study were to investigate the relationships between disease resistance traits measured in partially inbred lines and agronomically important traits (grain yield, grain moisture, lodging, plant and ear height, and silk and anthesis dates) measured in topcrosses of these lines to a common tester. Relationships among traits were studied with correlation analysis and comparison of quantitative trait loci (QTL) mapped for the different traits on a common linkage map. These two analyses provided complementary approaches to study the genetic relationship between Fusarium ear rot and fumonisin contamination resistance and other phenotypic traits.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Field Evaluation
In 2003 and 2004, an experiment was planted in four locations per year in North Carolina: the Central Crops Research Station at Clayton, the Tidewater Research Station at Plymouth, the Peanut Belt Research Station at Lewiston, and the Sand Hills Research Station at Jackson Springs. Soils at the experiment sites are classified as Marlboro loamy sand (fine, kaolinitic, thermic Typic Paleudult) at Clayton, Portsmouth fine sandy loam (fine-loamy over sandy or sandy-skeletal, mixed, thermic, Typic Umbraquult) at Plymouth, Norfolk sandy loam (fine-loamy, kaolinitic, thermic Typic Kandiudult) at Lewiston, and Candor sand (sandy, siliceous, thermic Grossarenic Kandiudult) at Jackson Springs. Experimental entries included the 213 BC1–derived family topcrosses, the two parental line topcrosses [FR1064 x (FR615 x FR697) and GE440 x (FR615 x FR697)], and seven check cultivars (B73 x Mo17; Pioneer brand hybrids 31G66, 31G98, 32R25, and 33M54; and Dekalb brand hybrids DK697 and DK743). Because topcross seed was limited for two BC1–derived family topcrosses, those entries were replaced by a check hybrid (Pioneer brand hybrid 32K61) in some plots.
The treatment design was a sets design, wherein each BC1F1:2 family topcross hybrid was randomly assigned to one of three sets. Each set contained 71 BC1F1:2 topcrosses, the two parent line topcross hybrids, and seven check hybrids. The experimental design was a replication-within-sets design, and each set was arranged as an 8 by 10 
-lattice with two replications at each location. A sets design was used because the size of fields at some research stations prohibited use of a single large incomplete block design. Sets were planted in adjacent field blocks at each location. Experimental units were two-row plots 3.05 m in length, separated by 1.22-m alleys. Interrow spacing was 0.914 m in Lewiston and 0.9652 m in Clayton, Plymouth, and Jackson Springs. Plots were overseeded and thinned to target population densities of 44 plants per plot (62,288 plants ha–1 in Clayton, Plymouth, and Jackson Springs, or 65,750 plants ha–1 in Lewiston), or less in some environments because of poor germination due to cold weather. For Clayton, the target stands for 2003 and 2004 were 31 and 44 plants per plot, respectively. For Plymouth, the target stands for 2003 and 2004 were 38 and 44 plants per plot, respectively. For Lewiston, the target stands for 2003 and 2004 were 28 and 44 plants per plot, respectively. For Jackson Spring, the target stands for 2003 and 2004 were 28 and 44 plants per plot, respectively. Stand counts were then taken approximately 4 wk after planting and plots were thinned manually to the target population density, if needed.
When all plants had reached physiological maturity, ear and plant heights were measured on six plants per plot. Plant height was measured from the ground to the terminal node; ear height was measured from the ground to the node of the topmost ear. Counts were taken of root-lodged plants (leaning greater than 30% from vertical with intact stalks) and stalk-lodged plants (broken below the ear or plants with dropped ears). Plots were harvested mechanically, and grain yield and grain moisture were measured.
Genotyping and Linkage Map Construction
The procedure for the development of the linkage map is detailed in Robertson-Hoyt et al. (2006). Briefly, young leaves from eight plants per family and parent line were sampled for DNA extraction. DNA extractions were performed using the DNeasy Plant Mini Kit system (Qiagen, Inc., Valencia, CA). Simple sequence repeat (SSR) primers were selected for screening in both populations using consensus map information from the Maize Genetics and Genomics Database (www.maizegdb.org) to reference locations of marker loci. Parental lines were screened with SSR primers to identify polymorphisms (Senior et al., 1998). Polymerase chain reaction (PCR) products were separated by electrophoresis in 4% SFR agarose gels (Amresco, Solon, OH) and visualized by staining gels with 0.05% ethidium bromide and exposing them to ultraviolet light.
Locus orders and recombination frequencies were estimated using multipoint mapping in MapmakerEXP (Whitehead Institute, Cambridge, MA). Recombination frequencies were then transformed to centimorgans (cM) using Haldane's mapping function in MapmakerEXP. The resulting map consisted of 105 loci distributed across the 10 chromosomes and was consistent with consensus genetic maps of maize (www.maizegdb.org). Average intermarker distance was 20.7 cM and the largest gap between markers was 57.6 cM (Robertson-Hoyt et al., 2006).
Statistical Analyses
Estimation of Family Means
Yields were adjusted to 155 g kg–1 grain moisture. Percent stand was computed as the plot stand count divided by the target stand for the given environment. Percent lodging was computed for each plot as number of root- or stalk-lodged plants divided by plot stand. Traditional analyses of variance and spatial analyses were performed on the data for each set within each environment separately to estimate genotypic least square means for each trait within each set and environment (Brownie et al., 1993). Models with up to fourth-order polynomial effects of row and columns in the field layout were tested. Trend effects were maintained if significant in the model at P < 0.01 (Brownie et al., 1993). The following models were compared using PROC MIXED in SAS version 8.2 (Littell et al., 1996; SAS Institute, 1999): a model including complete and incomplete block effects, a model with significant row and column trend effects, a model with correlated errors, and a model with both significant trend effects and correlated errors (Brownie et al., 1993). Percent stand was included as a covariate if significant at P = 0.01. For each set within each environment, the model that minimized Akaike's Information Criterion (Lynch and Walsh, 1997) was chosen. Entry least square means from each set within each environment were then calculated using the most appropriate statistical model. These least square means were then used as the basis for a combined analysis across all experiments and environments. Before conducting a combined analysis across environments, we checked for heteroscedasticity for grain yield using Bartlett's test (Milliken and Johnson, 1992). The combined analysis across environments and sets was based on the model:
 |
ijk = residual error. In this analysis, the residual variation is based on the means of check cultivars repeated across different sets (the interaction between set and repeated check genotypes). For the purposes of comparing the experimental entries to the parental lines (in hybrid combination) and hybrid checks, genotypes were considered to be fixed effects (Holland et al., 2003) and all other effects were considered to be random effects. This mixed model was analyzed using PROC MIXED in SAS and the least-square means of parents and hybrid checks adjusted for set effects were obtained and the average standard error of genotype mean comparisons was estimated for each trait. As part of this analysis, contrasts were used to test the null hypothesis that the 10 BC1–derived families with lowest mean fumonisin concentration as lines per se in the experiment by Robertson et al. (2006) were not significantly different for agronomic traits in topcrosses than the 10 BC1–derived families with highest mean fumonisin concentration in the same study. This two-step analysis based on combining results of individual environment analyses is not maximally efficient relative to a single combined analysis, but the use of different spatial analysis models in different environments prevented simple application of a single mixed model for this data set. For the purposes of estimating genetic variance components, parental topcrosses and hybrid checks were dropped from the data set and genotypes were considered random effects. Heritabilities for each trait on an entry mean basis were estimated from the univariate mixed model analyses using PROC MIXED of SAS (Holland et al., 2003). Approximate standard errors for heritability estimates were estimated with the delta method as described by Holland et al. (2003). PROC CORR of SAS (SAS Institute, 1999) was used to estimate the correlations between the genotypic means of the different traits measured in the hybrid trials with the means of fumonisin contamination and Fusarium ear rot measured in an inbred study (Robertson et al., 2006). This was performed to approximate a genotypic correlation between traits, because ear rot and fumonisin were measured mostly in different environments than the topcross agronomic traits. There is a small bias due to genotype-by-environment covariances in this estimator because two of the eight yield-trial environments coincided with two of the four disease-evaluation environments (Casler, 1982).
QTL Detection and Estimation
Quantitative trait loci were identified using the procedure described in Robertson-Hoyt et al. (2006) based on the genotypic least square means across environments. Briefly, composite interval mapping (CIM) was implemented in Windows QTL Cartographer 2.5 only to provide initial models for further analysis by the multiple interval mapping (MIM) method. QTL peaks from the CIM analysis, with a LOD threshold value of 2.0 and minimum 5 cM between QTL, were used as the initial models for MIM in Windows QTL Cartographer 2.5. Models were then created and tested in an iterative, stepwise fashion, searching for new QTL to add to the model, and testing the significance of the QTL after each search cycle. New models were accepted if they decreased the Bayesian Information Criterion (BIC) (Piepho and Gauch, 2001). After no additional QTL could be added to the model according to the BIC, each pair of QTL in the model was then tested for epistatic interactions. Epistatic interactions were kept in the model if they decreased the BIC. While developing the model using multiple interval mapping, we were aware that overfitting the model was possible. Bogdan et al. (2004) proposed a modification of the BIC with improved model selection properties, but not all statistics required to calculate this modified BIC are available from QTL Cartographer. Therefore, to avoid model overfitting, we checked if the proportion of total variation due to QTL exceeded the trait heritability. If this occurred, the QTL with smallest effect was dropped from the model, and the model was retested. If the model still explained a higher proportion of the phenotypic variance than the heritable proportion, the QTL with smallest effect in this reduced model was dropped. This process was continued until we obtained the best model according to the BIC that did not explain a greater proportion of the variation than the heritability of the trait.
When no additional QTL main effects could be added to the model, the model with minimum BIC was chosen and QTL effects were simultaneously estimated using the "summary" option of QTL Cartographer. Genetic variability explained by QTL for each trait was calculated as the total phenotypic variation explained by QTL divided by the entry mean heritability of the trait. When comparing QTL for two or more traits, QTL positions were considered to be overlapping when they were within 20 cM (Visscher et al., 1996; Melchinger et al., 1998; Cardinal and Lee, 2005).
To test the null hypothesis that the proportion of loci at which BC1–derived lines possessed GE440 alleles had no effect on topcross agronomic performance, we estimated correlations between proportion of segregating loci in the BC1–derived lines and topcross trait means of each line.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
BC1F1:2 topcrosses differed significantly (P < 0.0001) for all traits except for root lodging, for which entries did not significantly differ. Because root lodging values were not significantly different, root lodging was not included in any further analyses. The hybrids created from the two parental lines did not differ significantly for days to silking or anthesis. However, the GE440 topcross hybrid had significantly higher values than the FR1064 topcross hybrid for grain yield, grain moisture, plant height, ear height, percent stalk lodging, and percent target stand.
We found evidence of transgressive segregation for three of the traits measured. No family hybrids had significantly lower mean yield than the FR1064 topcross, but one family hybrid had significantly higher mean yield than the GE440 topcross across environments (Table 1 ). Two family hybrids had significantly lower mean grain moisture than the FR1064 topcross, but no family hybrid had significantly higher mean grain moisture than the GE440 topcross. Two family hybrids mean silking dates were significantly earlier than the FR1064 topcross, but no family hybrid mean silking date was later than the GE440 topcross. No transgressive segregation was detected for lodging, plant height, ear height, anthesis date, or target stand (Table 1).
|
Heritabilities
Heritabilities on an entry mean basis for traits measured ranged from moderately low to high: grain yield (h2 = 0.54, SE = 0.05), grain moisture (h2 = 0.65, SE = 0.04), stalk lodging (h2 = 0.30, SE = 0.07), plant height (h2 = 0.80, SE = 0.02), ear height (h2 = 0.84, SE = 0.02), days to silking (h2 = 0.37, SE = 0.09), days to anthesis (h2 = 0.44, SE = 0.08), and percent target stand (h2 = 0.84, SE = 0.02).
Correlations
Many of the correlations between Fusarium ear rot or fumonisin concentration of BC1F1:2 families per se and agronomic traits measured in the hybrid study were significant. Significant correlations ranged from r = –0.30 (P < 0.0001, between Fusarium ear rot and moisture content) to r = 0.29 (P < 0.0001 between Fusarium ear rot and yield; Table 2
). The correlation between the proportion of 105 marker loci in BC1F1:2 lines that had GE440 alleles and topcross yield was positive (r = 0.17, P = 0.016). This indicates that, in general, having more GE440 alleles improved topcross yield, although the correlation was low. The proportion of loci in BC1F1:2 lines that had GE440 alleles was also correlated positively with topcross grain moisture (r = 0.34, P < 0.0001), plant height (r = 0.30, P < 0.0001), and ear height (r = 0.34, P < 0.0001), indicating that, in general, the GE440 genetic background is associated with increased grain moisture content as well as increased plant and ear heights.
|
|
Comparison of Disease Resistance QTL and Agronomic Performance QTL
Fusarium ear rot was positively correlated with topcross yield (Table 2), but no Fusarium ear rot QTL (reported in Robertson-Hoyt et al., 2006) mapped to the same positions as yield QTL. In contrast, despite the lack of significant correlation between fumonisin content and topcross yield, two fumonisin contamination QTL mapped to similar positions as yield QTL. In the region of 95.00 to 109.8 cM on chromosome 1, the fumonisin QTL (which accounted for 10% of the phenotypic variation) overlapped the yield QTL with the largest effect. On chromosome 7, in the region between 24 and 30 cM, a fumonisin-contamination QTL (which accounted for 7% of the phenotypic variation) overlapped a yield QTL. The GE440 allele contributed to reduced fumonisin content at all QTL, but the GE440 allele was associated with reduced yield at the overlapping yield QTL on chromosome 1, whereas the GE440 allele was associated with increased yield at the overlapping QTL on chromosome 7. The fumonisin QTL on chromosome 1 is also overlapping with a plant height QTL (FR1064 beneficial allele) and an ear height QTL (FR1064 beneficial allele) (Table 3; Robertson-Hoyt et al., 2006).
Fumonisin contamination and Fusarium ear rot were both positively correlated with stalk lodging and negatively correlated with grain moisture (Table 2). However, there were no lodging QTL identified, and Fusarium ear rot and fumonisin contamination QTL did not overlap with grain moisture QTL (Table 3; Robertson-Hoyt et al., 2006).
Fusarium ear rot and plant height were positively correlated, and one region on chromosome 5 (83–100 cM) contained QTL for both traits. The favorable alleles for both traits (shorter plants and lower ear rot) in this region are from GE440. There are three QTL overlapping between fumonisin concentration and plant height, even though these two traits were not significantly correlated (chromosome 1, 95–104 cM; chromosome 4, 135–145 cM; and chromosome 5, 83–97 cM).
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The FR1064 topcross hybrid's poor yield was also likely associated with its poor stand (61% of target stand compared with 98% for GE440 topcross) (Table 1), despite our attempt to adjust yield for differences in stand by using stand as a covariate in the analysis. A similar significant difference was observed between the high fumonisin progeny mean stands (88%) and the low fumonisin group mean stands (97%), which also had higher and lower mean ear rot, respectively (Robertson et al., 2006). The topcross seeds were produced using GE440, FR1064, or the BC1–derived family as the female parent in an environment prone to Fusarium ear rot. Therefore, it is likely that the FR1064 topcross and the more susceptible family topcrosses had higher proportions of seed that was damaged by Fusarium spp., leading to lower germination and lower stands. However, although there was a difference between the yield of the two parental topcross hybrids, there was no significant yield difference between the two BC1–derived family groups (Table 1). This suggests that using stand as a covariate in our analyses corrected for differences in stand within part of the range of variability, but perhaps was not adequate to adjust for the extremely low stands of the FR1064 topcross. These results further suggest that backcrossing Fusarium ear rot resistance alleles from GE440 into the FR1064 background could broaden that range of adaptation of FR1064 to include environments that are conducive to Fusarium ear rot, such as North Carolina (Shelby et al., 1994).
Based on the percentage of genotypic variation explained by QTL (Table 3), we were able to identify QTL associated with the majority of genetic variation for yield, plant height, and ear height (in each case, the combined effects of mapped QTL explained greater than 70% of the genotypic variation). Many QTL influencing moisture were identified. However, most of the QTL identified for moisture had small effects, suggesting that there may be many very small effect QTL that were not identified. For silk date and anthesis date, however, only a small number of QTL were identified (for silk and anthesis date, QTL explained only 39% of the genotypic variation) and for lodging no QTL were identified. For silk and anthesis date, the low number of QTL identified could be due to low power of detection, because only two environments were used to collect data on flowering time. Flowering time in hybrids, however, was not correlated with resistance to either fumonisin or ear rot, which confirms the results measured in lines per se by Robertson et al. (2006) and demonstrates that resistance is not due to disease escape due to flowering-time effects. We were unable to map QTL with effects on mean lodging ratings because a large proportion of the variation in stalk lodging was due to genotype by environment interaction.
Grain moisture was negatively correlated with Fusarium ear rot and fumonisin contamination. Bush et al. (2004) reported that early harvest (to force dry down as fast as possible) is a good control strategy for reducing fumonisin contamination, by escaping the disease pressure that would lead to high levels of fumonisin. The correlation we observed in this study between lines with low mean fumonisin and topcrosses with higher moisture concentration is likely due to linkage between high moisture alleles and disease resistance alleles in GE440. This illustrates the commonly observed phenomenon that introgression of unadapted germplasm is often accompanied by unexpected and possibly unwanted phenotype change along with favorable change (Goodman, 1985).
The results of the parental line comparisons, correlation analysis, and QTL analysis were not always similar. For example, there was a positive, but small (r = 0.15), correlation between Fusarium ear rot and plant height. This was surprising because GE440 had significantly lower ear rot than FR1064 and its topcross was significantly taller than the FR1064 topcross. However, at the one QTL region that affected both Fusarium ear rot and plant height, the GE440 allele was associated with shorter plants. Therefore, this QTL region contributes to the positive correlation between these two traits. Another example was that the trait pair that had the greatest number (three) of overlapping QTL (fumonisin content and plant height) was not significantly correlated. Finally, two overlapping QTL for fumonisin concentration and grain yield were discovered, despite the lack of correlation between these two traits. Careful consideration of these results, however, reveals that different QTL often contributed in opposite directions to the total genetic covariance between the two traits. For example, of the three QTL regions that affected both fumonisin concentration and plant height, the GE440 allele of the two most important QTL had nearly identical effect magnitudes but opposite signs on plant height (–2.7 cm on chromosome 1 and +2.8 cm on chromosome 4) (Table 3). Similarly, for grain yield and fumonisin concentration, the overlapping QTL region on chromosome 1 would have contributed a positive genetic covariance, whereas the region on chromosome 7 would have contributed a negative genetic covariance. Because the contributions of these unlinked gene regions would be summed in their respective total genetic covariances, they would have partly canceled each other out in both cases.
In general, marker-assisted selection could prove useful in helping to break linkages between resistance QTL and other important agronomic traits if backcrossing resistance QTL is the breeding goal. However, the QTL identified in this study explain such a small percentage of the variation that marker-assisted selection for improving topcross agronomic traits cannot be recommended. In summary, the correlation results suggest that if ear rot resistance alleles were backcrossed into FR1064, yield, stalk lodging, and plant height would decrease, while grain moisture content would increase. If fumonisin contamination resistance alleles were backcrossed into FR1064, stalk lodging would decrease, while grain moisture content would increase. In both cases, a backcross breeding program to introgress resistance alleles from GE440 into FR1064 appears promising because the results of this study predict only minor negative impacts on testcross agronomic performance of such introgression.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication October 20, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
dms/fumonbg3.html (verified 18 June 2007). U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition and the Center for Veterinary Medicine, College Park, MD.dms/fumongu2.html (verified 18 June 2007). U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition and the Center for Veterinary Medicine, College Park, MD.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
