| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Horticulture and Crop Science, The Ohio State University/Ohio Agricultural Research & Development Center, 1680 Madison Ave., Wooster, OH 44691
b Department of Plant Pathology, The Ohio State University/OARDC, 1680 Madison Ave., Wooster, OH 44691
c University of Hohenheim, 350 Institute of Plant Breeding, Seed Science, and Population Genetics, 70593 Stuttgart, Germany
d Quality Seed CC, P.O. Box 100881, Scottsville 3209, KZN, Republic of South Africa
* Corresponding author (pratt.3{at}osu.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: GLS, gray leaf spot • MAS, marker assisted selection • PLAA, percent leaf area affected • QTL, quantitative trait locus
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cercospora zeae-maydis overwinters on crop residue left in the field (de Nazareno et al., 1993). The widespread adoption of conservation tillage practices, with the associated higher levels of residue remaining on the soil surface, enabled the spread of C. zeae-maydis and increased the incidence of GLS in the U.S. Corn Belt (Latterell and Rossi, 1983). Warm, humid conditions favor the spread of the disease, and fields planted along riverbeds or other low-lying areas are most likely to experience severe epidemics (Lipps, 1987, Rupe et al., 1982). In sub-Saharan Africa, small-scale farming systems are heavily affected because of widespread maize cultivation and favorable agroecological conditions (Bigirwa et al., 2001). A survey of maize production and disease severity revealed GLS is endemic throughout diverse agroecological zones of Uganda (Bigirwa et al., 2001).
Methods to manage GLS epidemics include conventional tillage that buries crop residue, fungicide application, and utilization of resistant hybrids (Ward et al., 1997). The desirable attributes of conservation tillage make it unlikely that growers who have adopted it will resume conventional tillage practices. Cercospora zeae-maydis is endemic in much of the Corn Belt; a reduction in conservation tillage would have to be universally adopted to have an economic impact on GLS epidemics (Lipps et al., 1996). Fungicide application is costly and not practical in most operations, but is a method of last resort used in South Africa. Most hybrids currently in production are moderately to highly susceptible to C. zeae-maydis (Thomison and Lipps, 1997). Availability and adoption of resistant hybrids would provide a cost-effective means of controlling GLS while allowing growers to continue to reap the benefits of conservation tillage. Development of improved germplasm with resistance to multiple foliar pathogens (including C. zeae-maydis) has been identified as one of the top priorities for research and development of maize in sub-Saharan Africa (DeVries and Toenniessen, 2001).
The development of resistant germplasm can be accomplished by introgressing resistance factors from donors into elite maize germplasm. The identification of molecular markers linked to disease resistance loci would aid breeding efforts by complementing traditional phenotypic selection with marker assisted selection (MAS) and assist in strategic deployment of resistance factors in a manner that could prolong their effectiveness (Simmonds, 1985).
Early studies of temperate adapted germplasm addressed the genetic basis of resistance to C. zeae-maydis by diallel and generation means analyses and concluded that resistance is under additive genetic control, with some dominance effects (Thompson et al., 1987; Ulrich et al., 1990; Gevers et al., 1994; Coates and White, 1998). Genotypes were selected in these studies, resulting in a fixed model (Griffing, 1956). The conclusions reached were, therefore, applicable only for the selected inbred lines. Gevers and Lake (1994) suggested that a single gene in South African germplasm conferred resistance to C. zeae-maydis, but their results have not been confirmed.
QTL mapping studies have made limited progress in identifying consensus QTL for resistance to C. zeae-maydis (Bubeck et al., 1993; Saghai-Maroof et al., 1996; Clements et al., 2000; Lehmensiek et al., 2001). Bubeck et al. (1993) used two different inbreds, NC250A and ADENT, as sources of resistance in three F2:3 mapping populations. They identified QTL on five different chromosomes, but only one of these, on the short arm of chromosome two, was consistent over three environments. Saghai-Maroof et al. (1996) employed selective genotyping to identify three QTL on chromosomes 1, 4, and 8 that collectively explained 44 to 64% of the variation across two generations, F2 and F2:3, across two seasons in one location. The susceptible parent, B73, contributed the QTL on chromosome 4 and the other two were from the resistant parent, Va14. In tests for epistatic interactions, they demonstrated that the QTL on chromosome 4 had little or no effect when the QTL on chromosome 1 was homozygous for the Va14 allele. In addition, the QTL on chromosome 8 displayed recessive gene action. Using the inbred O61 as a resistance source, Clements et al. (2000) evaluated a BC1S1 population for 2 yr at one site and 1 yr at a second site. They found five QTL, all from the resistant parent, which were significantly associated with resistance to C. zeae-maydis in both years and locations (Table 1). Lehmensiek et al. (2001) used bulked segregant analysis to identify QTL on chromosomes 1, 3, and 5 associated with resistance in an F2 population derived from proprietary parental lines. They confirmed these QTL in sister F2 populations evaluated in two environments. These four studies collectively utilized five resistant inbreds, and chromosome 1, bin 1.05/1.06 is the nearest to a consensus QTL identified, with three of the five inbreds contributing resistance from this region (Table 1). To date, all studies have been conducted in the USA or RSA, but not in both environments.
|
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
F2:3 progeny lines were derived from single F2 plants and F2:4 progeny lines were derived from single ears obtained through self-pollination of one plant within each of the 144 F2:3 lines planted at Wooster in 1998. Controlled pollination techniques were used and no seed were bulked at any time. Thus, all progenies shared identity through descent within pedigreed family lines.
Disease Evaluations
F2:3 Population
A population of 144 F2:3 lines was evaluated at Wooster and Cedara in 1998. Plots were planted at Wooster in a simple 12 x 12 lattice design (two replicates) in Typic Fragiudalfs (Wooster silt loam) soil. Each plot was a single row into which 25 seeds were sown. Parental check plots were incorporated into guard rows. All plots grown in Ohio were 5 m long with 0.75-m row spacing, and were planted and maintained according to standard agronomic practices (Ohio State University Extension, 1995) with the following exceptions: nitrogen was applied at a rate of 27.5 kg/ha and chlorpyrifos (O,O-diethyl O-3,5,6-trichloro-2-pyridyl phosphorothioate) insecticide was applied in the furrow with the seed at planting. Weeds were controlled with herbicides, and hand weeding was performed as necessary in all plots.
The 1998 Wooster plots were inoculated four times at 7-d intervals starting the last week of July, during the V9 to V15 growth stages (Ritchie et al., 1989), with inoculum grown on sorghum [Sorghum bicolor (L.) Moench] kernels. Five to 10 sorghum kernels were dropped into the leaf whorl of every plant in each plot. Low volume overhead sprinkler irrigation was applied at dawn and dusk each day to extend the leaf wetness period. Ear leaves were directly exposed to inoculum. Five plants per plot were selected for rating, and these plants were tagged with surveyor's tape before evaluations to ensure that the same plants were rated each time. The end plants of each plot were not rated to avoid possible border effects.
We assessed GLS severity by rating leaves at midplant height (usually the ear leaf) using the percent leaf area affected (PLAA) scale developed by Smith (1989). This scale assigns a PLAA score on the basis of visual estimates of the percentage of leaf surface area covered by lesions. We chose midplant leaves because leaves at that position and above contribute most to grain yield (Ritchie et al., 1989; Solomonovitz et al., 1992). Given the high degree of resistance of some lines, plants with no lesions on leaves at mid-plant height received a score of 0%, even if not completely devoid of lesions on lower leaves. GLS severity assessments were performed five times in 1998 with approximately 7 d between assessments, beginning the third week of August.
Cedara plots 4.5 m long were planted in the same design with 0.8 m between plots. The fertilizer regime was 450 kg/ha (30N, 45P, 60K), followed by a top dress of 200 kg/ha of 28% N. Herbicide spray regime was EPTC (S-ethyl-N,N-dipropylthiocarbonate), (4.0 L/ha, pre-plant), atrazine (6-chloro-N 2-ethyl-N 4-isopropyl-1,3,5-triazine-2,4-diamine), (3.5 L/ha), and metolachlor [2-choloro-6'-ethyl-N-(2-methoxy-1-methylethyl)acet-o-toluidide], (0.65 L/ha, post-plant). The insecticides deltamethrin [(S)-
-cyano-3-phenoxybenzyl(1R,3R)-3-(2,2-dibromovinyl)-2,2 dimethylcyclopropanecarboxylate] and monocrotophos [dimethyl (E)-1-methyl-2-(methylcarbamoyl)-vinyl phosphate] were applied after planting at 0.2 and 0.75 L/ha, respectively. The soil type at Cedara is Clay Plinthusalf (kaolinitic thermic). The rating methodology in Cedara was the same as at Wooster, but natural inoculum was relied on.
F2:4 Population
For trait-based selection of the phenotypic classes (Lebowitz et al., 1987), we chose lines that fell inside, or outside, approximately one standard deviation (±0.1) from the mean based on maximum PLAA data collected at Wooster and Cedara in 1998. Those that fell outside were considered the resistant and susceptible tails (23 lines in each tail) and those that fell inside were considered the intermediate class. The additional generation of self-pollination allowed confirmation of the previous phenotypic classification and served to increase the proportion of additive variation.
The 46 F2:4 lines derived from the F2:3 tails were evaluated at Wooster in 1999, Apple Creek, OH, in 2000 at a site that has been under conservation tillage maize production for at least 10 yr, and at Cedara in 2001. In the 2001 Cedara experiment, the selected 46 F3:4 lines plus 54 randomly selected lines representing the intermediate class, were grown. The same agronomic practices and disease evaluation methodology used in 1998 were again employed. A randomized complete block design with two replicates was used at each location. In addition to the F2:4 progeny lines and parents evaluated, the inbreds NC250A (resistant) (Freppon et al., 1994), Va14 (resistant) (Saghai-Maroof et al., 1996), Pa875 (resistant) (Elwinger et al., 1990), B103 (susceptible) (personal observation), and B73 (susceptible) (Saghai-Maroof et al., 1996) were evaluated as controls at Apple Creek.
We relied on natural inoculum for disease development at the Apple Creek and Cedara sites. Inoculum for the Wooster site was prepared as previously described except that inoculum was grown on oat (Avena sativa L.) kernels (Freppon et al., 1994). Disease ratings of the F2:4 lines were performed by the same procedures as in 1998.
Genotypic Analysis
The selected 46 F2:4 lines were genotyped with simple sequence repeat (SSR) and restriction fragment length polymorphism (RFLP) molecular markers for QTL detection by Kruskal–Wallis analysis. For SSR marker analysis, tissue from five plants per line was harvested from field-grown plants at Wooster in 1999 and bulked for DNA extraction. DNA was extracted by grinding fresh tissue in 30 µL 0.5 M NaOH in a 1.5-mL Eppendorf tube. The mixture was briefly centrifuged and 10 µL of DNA-containing supernatant was diluted in 490 µL of 100 mM Tris. Polymerase chain reaction (PCR) primers were purchased from Research Genetics (Huntsville, AL), and PCR was performed according to the manufacturer's protocol. The PCR products were resolved on a 4% (w/v) high-resolution agarose gel (Amresco, Solon, OH). Over 300 SSR primers were screened for useful polymorphisms between the two parents.
DNA was extracted by a modified CTAB method (Gardiner, 1998) from F3:4 plant tissue grown in the greenhouse and bulked from five plants to perform RFLP analysis. The DNA samples were digested with restriction enzymes (BamHI, DraI, EcoRI, and EcoRV) according to the manufacturer's instructions, electrophoresed on 0.8% (w/v) agarose gels, and transferred to nylon membranes. Southern hybridizations were performed by standard procedures (Gardiner, 1998). After washing, membranes were wrapped in cellophane and placed on molecular imaging screens for 48 to 96h before scanning. Hybridization signals were detected with a Storm 840 PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
A total of 97 individual markers, including 50 SSRs and 47 RFLPs, were available for initial QTL discovery. Genome coverage ranged from 6 to 13 markers per chromosome. The markers utilized for initial QTL discovery were as follows: Chromosome 1; bnlg1832, bnlg1811, bnlg1614, bnlg1112, bnlg147, bnlg2238, bnlg1884, bnlg182, umc58, umc84, bnlg131, bnl5.59, bnlg149, phi095, umc72, umc133, Chromosome 2; bnlg371, bnlg1045, bnlg1520, umc255, asg20, umc36, umc137, csu6, Chromosome 3; phi073, bnlg1523, bnlg1144, bnlg2136, bnlg1449, umc102, umc32, bnl8.35, umc63, bnlg1496, bnlg1160, phi046, Chromosome 4; nc005, phi072, umc1067, umc169 bnlg15.07, umc127, umc42, php20608, bnlg589, Chromosome 5; mmc282, phi100, npi409, umc43, umc108, phi085, php1017, mac.BO3, Chromosome 6; npi373, umc59, umc65, umc21 umc134, umc1023, phi089, phi077, bnlg1371, bnlg345, bnlg1740, Chromosome 7; bnlg1200, bnlg155, umc168, asg49, umc35, umc245, umc254, Chromosome 8; umc124, bnlg162, agrr21, umc117, bnlg119, Chromosome 9; bnlg1525, phi065, bnlg1714, bnlg1724, umc113, umc109, asg12, umc25, bnlg127, csu54, umc94, Chromosome 10; npi232, npi285, phi063, umc130, umc259, npi287, bnlg1839, bnlg1526, umc44.
To construct linkage maps of chromosomal regions with putative resistance QTL detected by Kruskal–Wallis analysis, 20 additional SSR markers that mapped to chromosomes 2 and 4 were screened. Four more polymorphic SSR markers on chromosome arm 2L, and one more on 4L were added for construction of linkage groups. Ten previously described resistance gene analog (RGA) probes (Collins et al., 1998) were obtained from Scot Hulbert, Kansas State University, and the parents plus the F1 were screened for RFLPs. Five of the RGA probes provided useful polymorphisms and were mapped. Five of the probes were either not polymorphic, or yielded complex banding patterns that were difficult to interpret, and were not used for further analysis. Of the 10 RGAs screened, one mapped to chromosome 2 and two mapped to chromosome 4 and were added to those linkage groups. Composite interval mapping was performed with these linkage groups. The additional 54 F2:4 lines (intermediate class) evaluated at Cedara also were genotyped on chromosomes 2 and 4 (see Composite Interval Mapping below).
Statistical Analysis
The maximum PLAA values of five plants per replicate plot were averaged to yield a mean score, and the overall mean value of the replicates was then determined. Area under the disease progress curve (AUDPC) values were also calculated (Campbell and Madden, 1990). The F2:3 and F2:4 data were analyzed for normality by plotting the residuals versus their expected values (Sabin and Stafford, 1990) with SAS Proc Univariate (SAS Institute, 1985). The residuals of the mean PLAA data were not normally distributed, so the data were transformed by a log10 (datum +1) transformation (Sabin and Stafford, 1990) because some plot means were zero.
Spearman rank correlations were performed with SAS using the maximum PLAA data to determine how the relative ranking of lines in each generation varied across environments. Genotype x environment (G x E) interactions in the F2:3 and F2:4 generations were tested by analysis of variance (ANOVA) with PLAA as dependent variable and environment, replication within environment, and genotype by environment as random variables. ANOVA was performed using SAS PROC GLM (SAS, 1985). Homogeneity of error variance was tested across locations using Cochran's test (Winer, 1971).
QTL Analysis
Selective Genotyping
A selective genotyping strategy was adopted because the individuals in the extreme phenotypic classes contain the largest amount of linkage information and the amount of genotyping needed is reduced (Darvasi and Soller, 1992). Progeny with phenotypes more than one standard deviation from the population mean comprise about 33% of the total population but contribute about 81% of the total linkage information (Lander and Botstein, 1989). We took such a sample from the tails of the distribution (32% of total lines) to realize the degree of power it afforded and also because less restrictive sampling of the tails (in contrast with e.g., 5–10% selection intensity) would assist in alleviating the tendency of selective genotyping toward upward bias in estimation of QTL effects. To minimize the influence of random error in classification of resistant and susceptible tails we used mean phenotypic values arising from multiple environments (Wooster and Cedara, 1998). Weller and Wyler (1992) proposed genotyping a sample of the individuals with phenotypes close to the mean to more accurately detect a marker-linked variance effect when conducting selective genotyping. We also included a random sample of 54 F2:4 lines representing the intermediate phenotype in the study conducted at Cedara in 2001 and genotyped them on chromosomes bearing significant markers identified by the initial QTL detection using Kruskall-Wallis analysis.
Kruskal-Wallis Analysis
The Kruskal–Wallis rank sum test, which makes no assumptions about the probability distributions of the phenotypic data, was performed. This analysis was performed with maximum PLAA values in the MapQTL program (Van Ooijen and Voorrips, 2001). The G x E term for PLAA was highly significant in the F2:4 generation. Heterogeneity of error was not significant in the Wooster, 1999, and Apple Creek, 2000, locations, so these respective data sets were combined for QTL analysis. Data were analyzed separately for Cedara 2001 because disease severity was much higher and apparently contributed to substantial magnitude effects that resulted in significant heterogeneity of error when combined with other F2:4 data sets.
Composite Interval Mapping
A more robust population size (100 lines) for linkage map construction on chromosomes 2 and 4 was created with the 54 F2:4 lines displaying intermediate phenotypic values in addition to the 46 lines representing the extremes. Ten markers on chromosome 2 (5 SSRs, 4 RFLPs, and 1 RGA) and 8 markers on chromosome 4 (1 SSR, 5 RFLPs, and 2 RGAS) were used for genetic map construction in a composite interval mapping approach to identify more precisely the position of the QTL. These markers were used to genotype all 100 lines. A genetic linkage map of the long arms of chromosomes 2 and 4 was constructed with Joinmap 3.0 linkage analysis software (Van Ooijen and Voorrips, 2001). Linkage groups were determined by a log-likelihood (LOD) threshold of 3.0. The calculation of linkage maps was performed with all pairwise recombination estimates smaller than 0.45 and a LOD score larger than 0.05. Kosambi's mapping function was used. Two markers, bnlg1045 and umc255, both on chromosome 2, were excluded from the analysis because they displayed significant (P > 0.01) segregation distortion by chi-square analysis. The final map positions of markers were compared to the published maize SSR map (Maize Mapping Project, 2002, www.maizemap.org; verified 13 November 2003). Composite interval mapping was performed with this marker set and the F2:4 phenotypic data from Ohio in 1999 and 2000 and Cedara in 2001.
Maximum transformed PLAA values were used for composite interval mapping analysis. Linkage groups were scanned for QTLs at 5-cM intervals with LOD thresholds corresponding to a genome-wide error rate of 5% calculated by 1000 permutations of the data (Churchill and Doerge, 1994) in the QTL analysis program MapQTL 4.0 (Van Ooijen et al., 2002). Marker cofactors were selected by the Automatic Cofactor Selection option in MapQTL and then modified according to the program instructions (Van Ooijen et al., 2002) to finish with a set of cofactor loci closest to the significant maxima in the QTL likelihood map (see Results). The percent phenotypic variation explained by the significant intervals and estimates of their additive genetic effects were also calculated in MapQTL.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Significant location and G x E interactions were detected across F2:4 evaluation sites (data not presented). Homogeneity of variance tests of the F2:4 data revealed the two Ohio locations to be homogeneous so they were combined for QTL analysis. The combined Cedara and Ohio data were not homogeneous. Heterogeneity between these data was attributed to the high PLAA values at Cedara that were not observed at either Ohio location. Analysis of combined Ohio data sets yielded no significant G x E interaction (Table 2). Genotypic rank changes appeared to be minimal based on Spearman rank correlations of PLAA means of the F2:4 progenies (Table 3).
|
|
|
Composite Interval Mapping
Markers on chromosome arms 2L and 4L used to construct genetic linkage maps are presented in Fig. 2
. On chromosome 2, 10 markers covered a linkage group of 121 cM, for 12.1-cM average distance between markers. For chromosome 4, a linkage group of 73 cM was constructed with six markers giving a 12.2-cM average interval between markers. These linkage groups contained the markers bnlg1520 (2L) and umc127 (4L) that were identified as linked to resistance loci by Kruskal–Wallis analysis. The order of markers within linkage groups was in general agreement with the B73 x Mo17 maize map (Maize Mapping Project, 2002, www.maizemap.org).
|
|
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A resistance interval on chromosome 2 was proximal to bnlg1520, but linked to this SSR locus across environments by composite interval analysis. The bnlg1520 marker was significant in Ohio where disease severity was moderate but it did not display significance at Cedara in 2001 where disease severity was high. It also displayed recessive inheritance, making it less desirable for breeding purposes. Saghai-Maroof et al. (1996) reported that a QTL on chromosome 8 was also recessive. The marker interval on chromosome 4L containing umc127 was consistent across all environments and explained up to 24% of the variation. The intervals on chromosomes 2 and 4 explained 40 to 47% of the variation across both test locations representing multiple years and macroenvironments.
Previous studies identified C. zeae-maydis resistance QTL on all 10 maize chromosomes (Bubeck et al., 1993; Saghai-Maroof et al., 1996; Clements et al., 2000; Lehmensiek et al., 2001). Of these only chromosomes 9 and 10 contained loci that were not significant across different environments (see Clements et al., 2000). Since different sources of resistance were used in each previous mapping study, and most others used B73 as the susceptible parent, it is not surprising that we identified only one region in common with their findings. The marker umc127 maps to chromosome 4, bin 4.08, a region also identified by Saghai-Maroof et al. (1996) as associated with resistance. In their study, the resistance source was the susceptible parent, B73, and the major QTL they identified on chromosome 1 had an epistatic effect on this locus. Our results show the resistance locus around bin 4.08 is contributed by the resistant parent, VO613Y, and displays additive gene action, a result that is consistent with previous reports of the inheritance of C. zeae-maydis resistance (Thompson et al., 1987; Ulrich et al., 1990; Coates and White, 1998). Our results were not in complete agreement with Gevers and Lake (1994) who proposed single gene inheritance of resistance in maize inbred VO613Y. We conclude that resistance in VO613Y may be unique because it apparently displays rather simple inheritance—we identified only two significant resistance QTL.
We mapped the RGAs to genomic locations generally consistent with those reported by Collins et al. (1998). They mapped PIC21 to chromosome three, while our map places it on chromosome four, linked to PIC14. This discrepancy is probably due to differences in genetic material used in the two studies and limitation in our population size (46) and theirs (70). Collins et al., (1998) demonstrated cosegregation between rps1 and an RGA. We could not identify any candidate C. zeae-maydis resistance genes with our data, but the association of RGAs with both QTL is worthy of further investigation.
Another of our objectives was to evaluate germplasm resistant to C. zeae-maydis across contrasting environments and determine if this resistance source could be deployed in diverse environments. Two taxonomically identical, but genetically distinct, sibling species of C. zeae-maydis (type I and type II) have been characterized according to DNA molecular marker profiles (Wang et al., 1998). No difference in pathogenic ability of the two types has been reported in the literature, and the observed reaction of U.S. maize hybrids to the two types has been consistent (Lipps et al., 1996). It is reported that the U.S. Corn Belt contains primarily type I, and maize producing regions of sub-Saharan Africa contain type II (Dunkle and Levy, 2000). A differential response of maize genotypes to types I and II of C. zeae-maydis would indicate a difference in virulence between the two types. While there was slight variation in disease pressure, and possibly different C. zeae-maydis types present at Wooster and Cedara in 1998, the F2:3 lines maintained their relative rankings across these environments (Table 3). Dunkle and Levy (2000) confirmed that both types of C. zeae-maydis are present at the Apple Creek, OH, location where we grew the F2:4 population during the 2000 season. The inbred VO613Y, and certain of its progeny lines, had a high level of resistance in all environments tested, with PLAA values at or near zero in all locations. This is further evidence for the lack of physiological races of the pathogen in the locations we utilized. Given the results of our disease assessments on the two continents, the prospects are promising for deployment of VO613Y-derived resistance to aid in management of GLS in both the U.S. Corn Belt and RSA.
Recent studies in maize demonstrated that MAS could be more economical and yield greater gains than phenotypic selection, especially for traits that were difficult or costly to measure (Dreher et al., 2000; Yousef and Juvik, 2001). Because of the lack of an effective method for evaluating resistance to C. zeae-maydis in the greenhouse, U.S. breeders essentially can evaluate only one generation per year. A MAS strategy would accelerate the breeding process by allowing indirect selection of resistant individuals based on marker genotype. Selection based on marker genotype in a backcross strategy would increase efficiency by allowing selection for the resistance locus from the donor parent and for the recurrent parent genotype at all other loci. This increased efficiency should offset the higher financial cost associated with use of molecular markers (Dreher et al., 2000). MAS for recessive alleles, such as the QTL on chromosome 2L, is especially advantageous over conventional means. Coupling MAS with phenotypic selection may give the greatest gain since our resistance QTL explained about 40% of the phenotypic variation.
Developing maize lines with multiple disease resistance is a high priority in many breeding programs, especially in sub-Saharan Africa, where increasing intensity of maize production has resulted in maize being produced essentially year-round in many areas with environments that are favorable to disease development. Because of the nature of these farming systems, and the potential catastrophic consequences of crop failure caused by disease epidemics (Allard, 1999), the development of durable resistance (Simmonds, 1985) should be the top priority. Given the large environmental component characteristic of GLS epidemics (Bubeck et al., 1993; Saghai-Maroof et al., 1996), and the assumption that only moderately resistant hybrids are necessary to manage GLS epidemics in many locations in the USA (Pratt et al., 1997), the markers we have identified should be useful for introgressing resistance loci into elite maize germplasm via a MAS strategy. We conclude that the most straightforward strategy would be to backcross the resistance donor with an agronomically elite line as the recurrent parent.
Identification of mapped C. zeae-maydis resistance loci from different resistance sources now provides the opportunity for combining resistance loci. Such resistance QTL combinations could provide more durable protection (Simmonds, 1985) against GLS than might be obtained with resistance conferred by only a single source.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication August 5, 2002.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
