| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Crop Science Dep., Williams Hall, Box 7620 North Carolina State Univ., Raleigh, NC 27695-7620
b Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
c Curtis Hall, Univ. of Missouri, Columbia, MO 65211
* Corresponding author (andrea_cardinal{at}ncsu.edu)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AIC, Akaike's information criterion • CIM, composite interval mapping • cM, centimorgan • ECB, European corn borer • LG, linkage group • QTL, quantitative trait loci • RFLP, restriction fragment length polymorphism • RIL, recombinant inbred lines • SCB, sugarcane corn borer • SSR, simple sequence repeats • SWCB, Southwestern corn borer
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Breeding for enhanced resistance to stalk tunneling has been an effective means of reducing the negative effects of ECB on grain production (Duvick, 1992). However, conventional selection using artificial infestations is laborious, complicated by unfavorable correlations with other traits, and potentially limited by a lack of information on genetic and biochemical factors associated with resistance (Klenke et al., 1986b; Beeghly et al., 1997). Some of these difficulties could be addressed and resolved through genetic analysis facilitated by molecular genetic maps (Paterson et al., 1991). Information from such analysis could enable marker-assisted selection, identify candidate genes, and define the biological basis of resistance mechanisms (McMullen et al., 1998).
Genetic factors for resistance to stalk tunneling by ECB in maize have been mapped using whole-arm chromosomal translocations and loci detected by DNA polymorphism. Genetic factors for resistance from inbred line B52 were identified with translocations on chromosomes 1, 2, 3, and 4 (Onukogu et al., 1978). Schön et al. (1993) reported QTL on chromosomes 1,2,3,7, and 10 in a B73xB52 F2:3 population. Jampantong (1999) detected QTL associated with resistance to ECB tunneling on chromosomes 2, 5, 6, 8, and 9 of a B73xMo47 F2:3 population (Fig. 1a,b).
|
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Phenotypic Data
Two hundred F6:8 lines and the parental lines, B52 and B73, were planted in 2.81 m long single-row plots with 0.76-m spacing between rows, at a rate of 17 kernels/plot in 14 by 15 simple lattice designs at the ISU Agronomy and Agricultural Engineering Research Center (AAERC) and the Hinds Farm, Ames, IA, on 6 May and 28 April 1997, respectively. Five entries each of B73 and B52 were included in all replications. The experiment was repeated in 1998, but 14 RILs were omitted from the 1998 experiments because the DNA marker data revealed that they were contaminated. In 1998, the experimental design was a 14 by 14 simple lattice planted at Hinds Farm on 29 April (Replication 1) and 1 May (Replication 2) and at AAERC on 5 May with the same plot dimensions and seeding rates as in 1997.
The soil characteristics of the research plots differed greatly. The Hinds Farm is located on a floodplain and has sandy soil classified as Hanlon-Spillville complex. The soil at the AAERC has been classified as Nicollet and Webster clay loam.
All plots were fertilized before planting with 134.5 kg ha-1 of actual N in the spring of both years. The herbicide metalochlor [2-chloro-6'-ethyl-N-(2-methoxy-1-methylethyl)acet-o-toluidide] (3.4 L ha-1) was applied pre-planting and bromoxynil (3,5-dibromo-4-hydroxybenzonitrile) (1.7 L ha-1) was applied at post-emergence at Hinds Farm in 1997. The herbicide propachlor (2-chloro-N-isopropylacetanilide) (28 kg ha-1) was applied preplanting and bromoxynil-atrazine (2-chloro-4-ethylamine-6-isopropylamino-S-triazine) (2.7 L ha-1) was applied at post-emergence at Hinds Farm in 1998. Cyanazine (2.4 kg ha-1) cyanazine [2-(4-chloro-6-ethylamino-1,3,5-triazin-2-ylamino)-2-methylpropionitrile] and alachlor (2.8 kg ha-1) [2-chloro-2',6'-diethyl-N-(methoxymethyl) acetanilide] were incorporated at preplanting at AAERC Farm in 1997 and 1998. Manual weeding was necessary at both locations in both years.
Freshly hatched ECB larvae were applied in the leaf axils of the top ear shoot and of the two adjacent nodes of the first six plants in each plot when approximately 50% of the RILs reached anthesis. In 1997, the application dates were 22 to 24 July at Hinds Farm and 5 to 8 August at the AAERC. In 1998, the dates were 20 to 22 July and 22 to 27 July at Hinds Farm and the AAERC. Overall, 400 to 500 larvae plant-1 were applied in 1997 and 300 to 400 larvae plant-1 were applied in 1998.
The date of anthesis for each plot was recorded in all environments as the date when 50% of the plants in a plot exerted anthers. The trait anthesis was defined as the number of days from the date of planting until the date of anthesis. After anthesis, plant heights of the six infested plants were measured to the nearest 5 cm from the soil surface to the node of the uppermost leaf. The trait plant height was defined as the average height in cm of the six plants.
Previous studies have reported ECB tunneling as a ratio of tunnel length and plant height because large differences in plant height among genotypes could have a confounding effect on the assessment of resistance (Coors, 1987). Herein, ECB tunneling is reported in centimeters and not as a ratio of tunnel length and plant height. The ECB tunneling was measured on the infested plants approximately 45 to 50 d after application of the larvae. Stalks were split longitudinally from the soil surface to the second node below the tassel. All tunnels in the stalks were measured. When there were parallel tunnels, only the longest tunnel was recorded. Holes in the stalk were counted as 1-cm tunnels. Two cm were added to the total length of a tunnel when the top of the plant was missing, and the stalk was terminated by ECB tunneling. In 1997, tunneling was measured on 13 and 14 September at Hinds Farm and from 20 to 27 September at the AAERC. In 1998, measurements were taken on 11 to 15 September at Hinds Farm and on 17 to 19 September at the AAERC. The trait ECB tunneling was defined as the average total tunnel length in cm of the infested plants in a plot. For QTL analysis, the phenotypic values for plant height and anthesis date were the entry means for all environments (i.e., across locations). For QTL analysis, the phenotypic values for ECB tunneling were the entry means at a given location (e.g., Hinds Farm) or the entry means for all environments (i.e., across locations), depending upon the analysis performed.
Analysis of Phenotypic Data
The error variances for ECB tunneling at the AAERC and Hinds Farm were tested for homogeneity by Box's test (Milliken and Johnson, 1992) to determine if a combined analysis was justified.
The data from each environment (e.g., Hinds Farm 1997, AAERC 1997, Hinds Farm 1998, and AAERC 1998) were analyzed by Proc Mixed (SAS Institute Inc., 1997). Complete and incomplete blocks were considered to be random effects. Lines were fixed effects. The least squares means for each RIL from each environment were used for the overall analysis including environments and genotypes as factors. The combined analysis was performed by Proc GLM (SAS Institute Inc., 1990) with environment as a random effect. The least squares means (across all environments) for each RIL were used in QTL analysis.
To estimate genetic variance components, heritabilities, and genetic and phenotypic correlations (rg and rp, respectively) Proc GLM was used to conduct an analysis of variance with RILs and environments as random effects. Data collected from B52, B73, and the contaminated RILs were not included in the estimation of correlations, heritabilities, and variance components. Approximate standard errors of the genetic correlations were estimated according to Mode and Robinson (1959). Exact confidence intervals for heritability (H2) estimates on an entry-mean basis were calculated according to Knapp (1985). Approximate standard errors (SE) of heritability estimates on a plot basis were calculated according to Hallauer and Miranda (1981) and Mode and Robinson (1959).
Genotypic Data
DNA was extracted from lyophilized leaf tissue (Saghai-Maroof et al., 1984) collected from 10 F6:7 plants of each RIL. Segregation data at RFLP loci were produced according to published procedures (Veldboom et al., 1994) with a set of 120 DNA clones available from the University of Missouri, Columbia (MaizeDB), Iowa State University, and Brookhaven National Laboratory. Segregation data at 65 loci defined by simple sequence repeats (SSR) were produced according to the protocol described by Senior et al. (1996) with primer sequences deposited in MaizeDB (http://nucleus.agron.missouri.edu/ssr.html; verified 24 Jan. 2001).
One hundred eighty-three RILs were used for linkage mapping and QTL analysis. Seventeen RILs (three RILs were discarded after the 1998 experiments) were discarded from the original set of 200 because they were contaminated or had more than 10% of loci with non-parental alleles. A Chi-square test was performed for each locus to test for segregation distortion.
Linkage analysis was performed with MAPMAKER/EXP version 3.0 (Lander et al., 1987). Loci were assigned to linkage groups (LG) with a minimum LOD score of 3.0 and a maximum Haldane distance of 40 centimorgan (cM). Three-point linkage analysis was performed for each linkage group. The "order" command was used at least 16 times per linkage group with an informativeness criteria of 160 individuals and a minimum distance between loci of 5 to 8 cM (depending on the linkage group). All combinations of "triple error detection" on or off and "use three point" on or off were used several times. The order that was found to be "best" in the majority of the runs was used as the initial map for adding new loci to the LG. The "compare" command was used to check if that order was indeed best. One order was declared to be better than another when they differed by a log likelihood of 2.0. Then, the "try" command was used to map the remaining loci. Finally, the "ripple" command was used to verify local locus orders. Unmapped loci were placed with the "place" command.
QTL Mapping
Composite interval mapping (CIM; Zeng, 1994) was conducted with PLABQTL version 1.1 (Utz and Melchinger, 1996) and QTL Cartographer version 1.13 (Basten et al., 1999). The programs were used as searching tools for QTL that would be subsequently evaluated in a multiple regression model. Mapped marker loci and a single locus from each tightly-linked (0.0–0.1 cM) pair or group of loci were used for QTL mapping with PLABQTL. Cofactors were chosen by stepwise regression and Akaike's information criterion, AIC, of 3.0 with the "cov" statement in PLABQTL. With QTL Cartographer, cofactors were chosen with a combination of forward and backward stepwise regression using the FB option in SRmapqtl. Cofactors with an F-statistic higher than 3.5 in the multiple regression were used in CIM.
A LOD threshold of 2.50 was used to declare the presence of a QTL with PLABQTL. To declare the presence of a QTL with QTL cartographer, 500 permutations were performed for each trait to determine the genome-wise significance level at 
= 0.05 (Churchill and Doerge, 1994). Then, QTL detected by either program were integrated in a single multiple regression model by PLABQTL. Model selection was performed by backward and forward stepwise regression using the AIC to choose the best model (Jansen, 1993). Two models were considered to be significantly different if their AIC values differed by more than 2.0. If two models were not significantly different according to their AIC values, the model with the fewest parameters and highest R2 was chosen.
Digenic epistasis for ECB tunneling was tested between all pairs of loci with two-locus analyses of variance using the SAS routine Epistacy (Holland, 1998) Interactions with P-values less than 0.00026 were declared significant. This threshold was chosen based on a conservative estimate of the minimum number of independent tests among 20 chromosome arms of maize (Holland et al., 1997). Interaction terms were sequentially added to a multiple regression model that included the marker loci closest to each QTL estimated from CIM. Interactions significant at a P < 0.05 level in the multiple regression model were maintained in the final model. Models that contained the QTL main effects and no more than two digenic epistatic interactions were developed. The final "best model" explained the greatest proportion of the phenotypic variation and had significant (P < 0.05) main effects of the initial QTL and interaction terms.
Comparison of QTL Detected in Different Studies
To determine if common QTL were detected in this experiment and other ECB tunneling experiments, we compared map positions of QTL in the different studies. Recombination distances among genetic markers are estimates that can vary among repeated experiments due to genotypic sampling and environmental factors such as temperature that affect recombination per se. Estimates of QTL map positions are affected by the same factors and also by imprecision in evaluating quantitative phenotypes. The inability to define QTL positions precisely makes comparison of QTL detected in different experiments difficult. Additionally, congruency of QTL positions detected in different studies could be affected by the usage of different sets of genetic markers (Beavis, 1994). Even under ideal conditions, computing confidence intervals for QTL positions in CIM analysis is an unresolved problem (Visscher et al., 1996). QTL positions estimated by other mapping algorithms have confidence intervals that typically span about 20 cM (Lee, 1995; Kearsey and Farquhar, 1998). Therefore, we followed Melchinger et al. (1998) and Groh et al. (1998) by declaring QTL identified in different experiments that differed in their estimated positions by less than 20 cM to be common.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In 1997, there was a significant natural infestation of ECB prior to anthesis at both locations, and the larvae initially fed on the newly emerging leaf-blade tissue. The final larval instars of such infestations can bore into stalks. The resulting tunnels cannot be distinguished from those caused by the artificially infested larvae. Presumably, more larvae can survive to cause stalk tunneling on those RILs with higher levels of leaf-blade damage incurred during the initial infestation. Therefore, to minimize the possible confounding effect of extra tunneling due to the natural infestation, each plot was rated visually for leaf-blade damage. The leaf-blade damage was used as a covariate in the analysis of ECB tunneling in 1997. Ratings were taken before application of ECB larvae and were based on a nine-point scale of leaf damage (Guthrie et al., 1960). The leaf-damage score had a significant effect as a covariate in the statistical analysis of ECB tunneling, indicating that it was effective in reducing the confounding effects of feeding damage from the natural infestation.
The phenotypic evaluation of the parents and the RILs is summarized in Tables 1 and 2. The error variances for ECB tunneling at the AAERC and Hinds Farm locations were significantly different (P < 0.01; Milliken and Johnson, 1992). Therefore, a separate analysis was performed for each location with the tunneling data collected in both years at a given location (Table 1).
|
|
2g) and phenotypic variances (2p) were both 4.6 times larger at the AAERC than at Hinds Farm (Table 2). However, B73 and B52 differed significantly from each other and from the mean of the RILs in all analyses of ECB tunneling. Significant transgressive segregation for higher ECB tunneling was observed in all the analyses, and there was significant genetic variation for tunneling in all environments (Table 1). The H2 on an entry-mean basis were between 0.71 and 0.78 (Table 2). The H2 estimate on a plot basis (0.36) was less than half the value of the estimate on an entry-mean basis, demonstrating the importance of replication over locations and years for obtaining good estimates of ECB tunneling means (Table 2). The heritability estimates were similar to those reported previously for ECB tunneling, plant height and days to anthesis (Schön et al., 1993; Lee, 1993; Sadehdel-Moghaddam et al., 1983). B73 and B52 differed significantly for anthesis date but not for plant height. The population of RILs varied significantly for both traits. The mean of the RILs was equal to B73 for anthesis date but less than both parents for plant height (Table 1). Significant transgressive segregation was observed for both traits. Anthesis date and plant height had very high heritabilities on an entry-mean basis, 0.95 and 0.97 (Table 2).
Anthesis date was positively correlated with plant height (entry-mean basis rg = 0.416, rp = 0.402). ECB tunneling was not significantly correlated with anthesis date (entry mean basis rg = -0.138, rp = -0.122).
Segregation and Linkage Analysis
The genetic map used for QTL mapping consisted of 1667 cM defined by 161 loci that mapped to unique positions. Overall, 185 loci were mapped, but 24 of them did not define unique positions on the map because of tight linkage and were not used for QTL mapping. In general, the order of loci in each LG coincided with the publicly available genetic maps (Davis et al., 1999). Fifty-one of the 185 loci exhibited significant (P < 0.001) deviation from the expected genotypic ratios of an F6:7 population. Twenty-six of 51 loci had an excess of B73 alleles, and six loci (phi096, phi026, phi079, BNL15.45, BNL13.05, and BNL9.11) had an excess of B52 alleles. The six loci with an excess of B52 alleles mapped to two clusters on chromosomes 4 and 6. Nineteen loci had excess of heterozygotes. Loci were assigned to 11 linkage groups. Chromosome 3 was divided in two linkage groups.
QTL Analysis
QTL were detected on five chromosomes (2, 3, 5, 7, and 9) in all analyses of ECB tunneling (Table 3). Chromosome eight contained a QTL only in the combined analysis. There was evidence of more than one QTL on chromosomes 2, 3, 5, and 9. In the combined analysis, nine QTL were detected, and they were associated with 46.2 and 59.2% of the phenotypic and genotypic variances, respectively. The alleles for decreased tunneling (resistance) were inherited from B52 at most of the QTL in all analyses and environments.
|
Epistatic interactions for ECB tunneling were detected in the combined analysis for two pairs of loci, NPI429 and NPI104 (chromosomes 1 and 5), and bnlg128 and UMC147 (chromosomes 3b and 5). Genotypes that were homozygous for the B52 allele at bnlg128 and homozygous for the B73 allele at UMC147 had 13.91 cm of ECB tunneling. Similarly, genotypes homozygous for the other parental allele at those loci had 14.11 cm of tunneling. Genotypes homozygous for B73 alleles and for B52 alleles at both loci had 17.55 and 15.60 cm of tunneling, respectively. Either pair of loci could be included as a significant interaction term in a multiple regression model containing the main effects at the other QTL. However, the interaction terms were not significant when both were included in the full model. The model with the interaction term for bnlg128 and UMC147 explained 2.4% more of the phenotypic variation. The regression model including the bnlg128*UMC147 interaction along with main effects of nine QTL explained 52.1% of the phenotypic variation (Table 4). The full regression model with main effects of the nine QTL without the interaction term explained 47.4% of the phenotypic variation. The R2 of this model (Proc GLM, SAS) was 1% higher than the model fit with PLABQTL (Table 3) because the QTL were fit at the marker position in the regression analysis in Table 4, whereas the QTL were fit at their best estimated positions by PLABQTL.
|
|
Comparison of QTL Detected in Different Studies
Two QTL for ECB tunneling were detected herein and by Schön et al. (1993) (UMC4 on chromosome 2 and dupssr5 on chromosome 3; Table 3). The B52 alleles at these QTL decreased tunnel length. Additionally, two QTL detected at Hinds Farm only (Table 3), one near bnlg108 and one near UMC59, coincided in their location and sign of their effects with QTL detected by Schön et al. (1993).
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our results are consistent with other reports in which the majority of the QTL for resistance were contributed by the resistant parent, but additional resistance factors were derived from the susceptible parent (Bohn et al., 1997; Cardinal et al., 1998; Jampantong, 1999).
Computing appropriate confidence intervals for QTL position in CIM analysis is an unresolved problem (Visscher et al., 1996). The precision of estimating QTL positions declines when the marker loci exhibit severe segregation distortion as observed on chromosome 9 in the region of the QTL for ECB tunneling QTL (UMC81-UMC114). The decline in precision occurs because segregation distortion can result in incorrect estimates of locus order and biased estimates of recombination frequency when using the mapping algorithms of MAPMAKER/EXP (Lorieux et al., 1995a; Lorieux et al., 1995b). Therefore, QTL that differed in their estimated positions by less than 20 cM were considered to be common for the purpose of comparing across different experiments.
Few QTL for ECB tunneling coincided between our study and Schön et al. (1993). Several studies of traits such as plant height and grain yield, have reported similar results in which few QTL were validated in two different generations of the same population (with or without resampling) or in different environments (Austin and Lee, 1996; Beavis, 1994; Melchinger et al., 1998). Many factors could have caused the difference in observations presented here and by Schön et al. (1993). Different methods of QTL analysis, sample sizes, environmental conditions, and inbreeding levels were used in this study. These factors have been reported previously to affect QTL results (Beavis, 1994; Melchinger et al., 1998; Zeng, 1994; Jansen and Stam, 1994).
QTL for ECB tunneling have been mapped in three other single-cross maize populations in the F3 generation; B73xDe811, Mo17xB52 and B73xMo47 (Lee, 1993; Jampantong, 1999). Two QTL (on chromosomes 3 near dupssr5, and 5 near UMC54) were common between this study and the B73xDe811 population. The additive effect at both QTL showed that resistance factors were contributed by the resistant parent (B52 and De811 in each experiment, respectively). Two QTL (chromosome 2 near dupssr21 and chromosome 9 near UMC81) were detected both in this study and in the Mo17*B52 population. In both populations the B52 allele was associated with a decrease in tunnel length. QTL on chromosomes 3 (dupssr5-UMC165), 5 (UMC54), and 9 (UMC81) were also reported in the B73xMo47 population (Jampantong, 1999). The QTL on chromosome 3 detected in the B73xMo47 population coincided with two QTL in this study; one QTL (dupssr5) was detected in the combined and AAERC analyses (Table 3) and the other QTL (UMC165) was detected at Hinds Farm. In both studies, the non-B73 alleles (B52 or Mo47) were associated with the resistant phenotype. Similarly, the non-B73 alleles contributed to the reduced tunnel length at the other two common QTL on chromosomes 5 and 9. Summarizing the above comparisons, four genomic regions have been consistently associated with resistance to ECB tunneling in the stalk: chromosome 2 near bnlg108-dupssr21, chromosome 3 near dupssr5-UMC165, chromosome 5 near BNL10.12-UMC54, and chromosome 9 near UMC81-UMC153.
ECB tunneling and anthesis date were not significantly correlated, but some QTL for both traits mapped to similar positions. Nonsignificant genotypic correlation between two traits implies that the sum of the correlations of individual locus effects across the genome is close to zero. However, such a result does not necessarily mean that two traits do not share genetic factors. Pleiotropy or linkage of QTL affecting different traits can contribute to a genetic correlation, but pleiotropic or linkage relationships at different QTL of opposite effect may result in a nonsignificant whole-genome correlation. This was observed for common QTL affecting both ECB tunneling and anthesis date since two pairs of QTL had opposite additive signs and one pair of QTL had additive effects of the same sign.
Comparison of QTL affecting resistance to leaf-feeding and stalk-tunneling stages of ECB suggests that the genetic controls of resistance to these different stages of ECB feeding have little in common. Only one QTL on the long arm of chromosome five has been associated with resistance to both leaf-feeding and stalk tunneling by ECB in a common population (Jampantong, 1999). Furthermore, comparing the results presented here with those from a QTL mapping study for resistance to ECB leaf feeding in the Mo17*H99 population (Cardinal et al., 1998) reveals that only two genomic regions (7L and near the centromere on chromosome 9) could potentially have genetic factors for resistance to both leaf feeding and stalk tunneling by ECB. Similarly, only one QTL (chromosome 9 near UMC114-bnl8.17) for resistance to ECB stalk tunneling presented in Lee (1993) had a similar genomic location to a QTL for resistance to ECB leaf feeding (Cardinal et al., 1998). Biometrical investigations of the relationship between resistance to leaf feeding and stalk tunneling indicate that there is no genotypic or phenotypic correlation between the two traits. This result was found not only for ECB, but also for southwestern corn borer (SWCB) [Diatraea grandiosella (Dyar)], and sugar cane borer (SCB) [Diatraea saccharalis (Fabricius)] (Bergvinson, 1999, personal communication; Russell et al., 1974; Sadehdel-Moghaddam et al., 1983; Klenke et al., 1986a; Coors, 1987).
Surprisingly, however, there was greater concordance of map positions of QTL for ECB tunneling resistance and QTL affecting resistance to leaf feeding by SWCB and SCB. Three QTL (on chromosome 2 near UMC4, on 5 near NPI104, and on 7 near bnlg657) were detected both for resistance to Sugarcane corn borer (SCB) by Bohn et al. (1996) and for ECB tunneling herein. Five QTL on chromosomes 2 (UMC4), 3 (dupssr5), 5 (UMC54), 7 (bnlg657), and 9 (UMC81) were detected both for resistance to SWCB or SCB by Bohn et al. (1997) and herein for ECB. Finally, QTL on chromosomes 5 (UMC54), 7 (bnlg657), 8 (NPI268), and 9 (UMC81) were detected for resistance to SWCB or SCB by Groh et al. (1998) and for resistance to ECB tunneling herein.
These comparative QTL results suggest that resistance to ECB tunneling in temperate maize and resistance to leaf feeding by insect pests of tropical maize may be conferred in part by common mechanisms, while resistances to stalk tunneling and leaf-blade feeding by ECB in temperate maize are mediated largely through different mechanisms. Also, 4-dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA) is involved with resistance to leaf-blade feeding by ECB in temperate but not tropical maize germplasm (Klun et al., 1967; Frey et al., 1997; Abel et al., 1995). In tropical maize, DIMBOA has not been associated with resistance to leaf-blade feeding by SWCB (Hedin et al., 1984). In addition, in temperate maize, DIMBOA has been suspected to have only a minor influence on resistance to stalk tunneling by ECB (Klun and Robinson, 1969). Comparative genetic mapping provides further evidence for the involvement of DIMBOA in resistance to leaf-blade feeding but not stalk tunneling by ECB in temperate maize. Genes coding for enzymes involved in the synthesis of precursors of DIMBOA have been cloned and genetically mapped (Frey et al., 1997) to a region of chromosome 4 which was also associated with QTL of very large effect on leaf-blade feeding by ECB in temperate maize (Veldboom and Lee 1993; Cardinal et al., 1998; and Jampantong, 1999). However, that region was not associated with stalk tunneling by ECB in this experiment.
Cell wall composition in the leaf blade or in the leaf sheath could be an important common mechanism of resistance against different borers (ECB, SCB, and SWCB) in maize. Bohn et al. (1996) suggested that fiber and cell wall phenolic acid contents were involved in leaf feeding resistance to SCB. Similarly, resistance to stalk tunneling by ECB was explained in large part by cell wall components indicated by acid-detergent fiber, neutral-detergent-fiber, and lignin and silica content of leaf sheaths in temperate maize (Coors, 1988; Beeghly et al., 1997). Chromosomal regions containing genes involved in the synthesis of cell wall components could be associated with resistance to different species and different feeding patterns of corn borers in maize. To test these ideas, it would be important to map QTL for cell wall components and resistance to insect tunneling in the same maize population.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication May 9, 2000.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Bachlava, R. E. Dewey, J. Auclair, S. Wang, J. W. Burton, and A. J. Cardinal Mapping Genes Encoding Microsomal {omega}-6 Desaturase Enzymes and Their Cosegregation with QTL Affecting Oleate Content in Soybean Crop Sci., March 19, 2008; 48(2): 640 - 650. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Robertson-Hoyt, M. P. Jines, P. J. Balint-Kurti, C. E. Kleinschmidt, D. G. White, G. A. Payne, C. M. Maragos, T. L. Molnar, and J. B. Holland QTL Mapping for Fusarium Ear Rot and Fumonisin Contamination Resistance in Two Maize Populations Crop Sci., June 20, 2006; 46(4): 1734 - 1743. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Chardon, B. Virlon, L. Moreau, M. Falque, J. Joets, L. Decousset, A. Murigneux, and A. Charcosset Genetic Architecture of Flowering Time in Maize As Inferred From Quantitative Trait Loci Meta-analysis and Synteny Conservation With the Rice Genome Genetics, December 1, 2004; 168(4): 2169 - 2185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Krakowsky, M. Lee, W. L. Woodman-Clikeman, M. J. Long, and N. Sharopova QTL Mapping of Resistance to Stalk Tunneling by the European Corn Borer in RILs of Maize Population B73 x De811 Crop Sci., January 1, 2004; 44(1): 274 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Krakowsky, M. J. Brinkman, W. L. Woodman-Clikeman, and M. Lee Genetic Components of Resistance to Stalk Tunneling by the European Corn Borer in Maize Crop Sci., July 1, 2002; 42(4): 1309 - 1315. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Jampatong, M. D. McMullen, B. D. Barry, L. L. Darrah, P. F. Byrne, and H. Kross Quantitative Trait Loci for First- and Second-Generation European Corn Borer Resistance Derived from the Maize Inbred Mo47 Crop Sci., March 1, 2002; 42(2): 584 - 593. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
