Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 21 November 2006
Published in Crop Sci 46:2590-2597 (2006)
© 2006 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Agricola
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Related Collections
Right arrow Wheat
Right arrow Plant Disease
Right arrow Crop Genetics

CROP BREEDING & GENETICS

Characterization of Resistance to Fusarium graminearum in a Recombinant Inbred Line Population of Wheat

Resistance to Fungal Spread, Mycotoxin Accumulation, and Grain Yield Loss, and Trait Relationships

Guo-Liang Jianga,*, Yanhong Dongb, Janet M. Lewisa, Lee Silera and Richard W. Ward{dagger},*,a

a Dep. of Crop and Soil Sciences, Michigan State Univ., East Lansing, MI 48824
b Dep. of Plant Pathology, Univ. of Minnesota, 1991 Upper Buford Cir., St. Paul, MN 55108

* Corresponding authors (gljiang{at}msu.edu; r.w.ward{at}cgiar.org)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Fusarium head blight (FHB or scab) caused by Fusarium species can lead to tremendous loss of grain yield and quality in wheat (Triticum aestivum L.) as well as severe mycotoxin contamination. Characterization of different types of resistance is of great significance for genetics and breeding studies. One hundred fifty-two F6:7 recombinant inbred lines (RILs) derived from a cross CJ 9306/Veery were assessed in the greenhouse by single-floret inoculation to analyze three types of FHB resistance, i.e., resistance to fungal spread within spikes, deoxynivalenol (DON) accumulation, and grain yield loss. Each of the investigated resistance types exhibited a large and continuous variation in the RIL population, and thus was inherited quantitatively. Estimates of broad-sense heritability varied with resistance types, higher and less variable for spread resistance (0.85–0.93), but lower and more variable for resistance to DON accumulation (0.64–0.92) and/or resistance to grain yield loss (0.61–0.83). For spread resistance parameters (the number and percentage of scabby spikelets) and DON content, frequency distributions exhibited a few peaks and/or were skewed toward resistance, suggesting the existence of major resistance genes. Genetic correlation coefficients of spread resistance parameters with DON content and yield loss parameters were 0.85 to 0.92 and 0.78 to 0.94, respectively. The genetic correlations ranged from 0.63 to 0.77 between DON content and yield loss parameters. There were no noticeable genetic associations between FHB resistance and agronomic traits such as plant height, spike length, number of spikelets and grains per spike, grain weight, and heading date.

Abbreviations: AUDPC, area under disease progress curve • CoV, co-variance • DON, deoxynivalenol • dpi, days postinoculation • FHB, Fusarium head blight • MP, mean product • MS, mean square • NIRS, number of infected rachis sections • NSS, number of scabby spikelets • PSS, percentage of scabby spikelets • QTL, quantitative trait locus/loci • RIL, recombinant inbred line


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
FUSARIUM HEAD BLIGHT is a devastating disease in wheat and barley (Hordeum vulgare L.) worldwide. FHB epidemics resulted in economic losses of more than U.S. $3000 million estimated during the last decade in the USA alone (Windels, 2000). Besides the decrease of grain yield and quality, it also causes severe mycotoxin (deoxynivalenol or DON) contamination in the infected grains, which is detrimental to the health of livestock and human beings. Lack of resistant cultivars and other adequately effective control methods hinders crop production. From the economic and environmental points of view, development of resistant cultivars is the best approach to control this frustrating disease.

The expression of resistance to FHB in wheat and barley is complex (Kolb et al., 2001), and there is no complete resistance to FHB in wheat, although sources of partial resistance have been identified through extensive surveys of germplasm (Browne et al., 2005; King, 1996; McKendry et al., 2004). Schroeder and Christensen (1963) first defined two types of FHB resistance in wheat: resistance to initial infection and resistance to the spread of the pathogen within the tissue or spike, which were subsequently designated respectively as Type I and Type II resistance. On the basis of these definitions, additional types of resistance have also been proposed: such as resistance to mycotoxin accumulation (Miller et al., 1985), resistance to kernel infection, and tolerance (Mesterhazy, 1995; Mesterhazy et al., 1999). So far, however, very few sources of Type I resistance have been identified, and almost all the germplasms with other types of resistance have proved to possess Type II resistance to some extent. Therefore, Type II resistance is predominant in wheat breeding and related research. Knowledge of other types of resistance is insufficient. Recent studies suggested the possibility of independent quantitative trait loci (QTL) or genes for resistance to DON accumulation and kernel infection different from Type II resistance (Somers et al., 2003; Yang et al., 2005). It is clear that characterization of the genetic variability of the different types of resistance and their association is of significance and interest for the development of resistant cultivars.

The objectives of our study were to: (i) characterize the genetic variation and heritabilities of resistance to fungal spread, DON accumulation, and grain yield loss in a RIL population; (ii) measure the relationships among these three types of resistance; and (iii) characterize the relationship between resistance and agronomic traits.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Materials and Experimental Design
A set of 152 F6:7 RILs derived from a cross Veery/CJ 9306 and two parents were used to evaluate FHB resistance. CJ 9306 is a novel germplasm line with a very high level of FHB resistance and good agronomic traits. This germplasm was developed through multiple-parent crossing and recurrent selection combined with modified pedigree methods with the aid of a dominant male-sterile gene Ta1 in China (Jiang, 1997). ‘Veery’ [Kavkaz/Buho//KAL/BB (CM33027)] is highly susceptible to FHB and was developed at CIMMYT, Mexico. The cross (Veery/CJ 9306) was made at Nanjing Agricultural University, China, in the spring of 1998. The hybrid (F1) was planted in the autumn of the same year, and F2 seeds were randomly harvested in the summer of 1999. Subsequently, the RILs were developed in the greenhouse at Michigan State University by single seed decent.

In December of 2001 and November of 2003, 152 F6:7 RILs were grown in the greenhouse at Michigan State University in a completely randomized design with two replications. For each line, six plants were planted in two pots (10.5 x 10.5 x 12.4 cm), each having three plants per replication. The two parental lines were planted as the controls many times at an interval of 1 wk so that they could be included in each inoculation to estimate the possible differences in the infection rate among inoculation dates. The plants were grouped and inoculated according to their heading and flowering stages in March of 2002 and January of 2004, respectively.

Inoculation and Assessment of Resistance to Fungal Spread
Single-floret inoculation was conducted immediately before or after initial anthesis (Jiang et al., 2001). The inoculum was F. graminearum isolate PH-1 (NRRL 31084) for 2002 and a mixture of two isolates PH-1 and WF-1 for 2004. The isolates PH-1 and WF-1 were separately isolated from scabby grain in a Michigan wheat field in 1996 and 2002, respectively. Twelve to 15 µL of conidospore suspension at 5 x 104 spores/mL, produced by CMC liquid culture (Cappellini and Peterson, 1965), was pipetted into a central basal floret of the spike with a self-refilling syringe. Six to eight spikes of each RIL were inoculated per replication. For each single batch of inoculation, all the plants with similar growth stage (just before or after anthesis of central spikelets) were inoculated on the same day, including the checks. In most cases, the primary spikes or the first tiller spikes were sampled for inoculation. The inoculated spikes were tagged to indicate the date of inoculation and record the symptoms of disease. The inoculated plants/pots were placed in a misting chamber equipped with an auto-mist-irrigation system programmed to deliver 20 s of mist at intervals of 6 min and a temperature controlling system set at 22 to 26°C. After 3 d of mist-irrigation, the pots were transferred to another greenhouse compartment. The number of scabby spikelets (NSS) on the inoculated spikes was visually counted at 5, 9, 13, 17, 21, and 25 d postinoculation (dpi), respectively. Scabby symptoms varied from light or dark brown glumes, water-soaked spots on the glumes, brown main rachis, and bleached spikelets to pink seeds. The criteria of disease scoring (Jiang et al., 1995) were as follows: 0.5–symptoms limited to the inoculated floret; 1.0–symptoms limited to the inoculated spikelet; 1.8–symptoms limited to the inoculated spikelet and the rachis section to which that spikelet was attached; 2.0 or greater–symptoms observed in one or more spikelets in addition to the inoculated spikelet (i.e., the total number of scabby spikelets).

At 25 dpi, the number of infected rachis sections (NIRS) and total spikelets (NTS) were also estimated. NIRS was defined as the number of rachis nodes and just below adjacent internodes both with visible infected symptoms. The percentage of scabby spikelets (PSS), a proportion of scabby spikelets to total spikelets, was calculated for each observation. On the basis of PSS data, the area under disease progress curve (AUDPC) was computed. Since 2002 data suggested extremely high correlations between NSS or PSS and AUDPC or NIRS, only NSS and PSS were determined at 21 and 25 dpi in 2004.

Deoxynivalenol Test and Estimation of Grain Yield Loss
After all the plants had matured, inoculated spikes and noninoculated spikes for each replication were harvested separately and threshed carefully with a head thresher at a lower speed to avoid the loss of the scabby or shriveled kernels. In 2004, the total grains from inoculated spikes and noninoculated spikes were counted separately for each pot. Grain weight per spike and 1000-grain weight were determined as well. Then, to estimate the relative loss of grain yield caused by FHB, the reductions of number of grains per spike, grain weight per spike, and 1000-grain weight were calculated, relative to noninoculated spikes (CK), i.e., the relative decrease, by the following formula: RD = (MCK – MIS)/MCK x 100, where RD = relative decrease, MCK = the mean of the check spikes (i.e., noninoculated spikes), and MIS = the mean of inoculated spikes.

Ten to 12 scabby kernels were randomly taken from the inoculated spikes to serve as a sample for DON test. DON extraction and analysis were based on a modified method of Mirocha et al. (1998). Briefly, seeds were weighed and placed into a 3.7 mL (1-dram) glass vial capped with a screw cap and extracted by soaking and shaking with 2 mL of acetonitrile/water (84/16 v/v) for 24 h. The extract was passed through a minicolumn packed with C18 and aluminum oxide. One and a half milliliters of the filtrate were placed into a 0.5-dram glass vial and evaporated to dryness under nitrogen. Twenty-five microliters of TMS reagent (TMSI/TMCS 100:1) were added, and the vial was rotated so that the reagent contacted with all residue in the vial. The vial was placed on a shaker for 10 min, and then 200 µL of isooctane were added followed by 200 µL of HPLC water to quench the reaction. After vortex, the upper layer was transferred to a GC vial. Selected ion monitoring (SIM) was used for GC/MS analysis (Shimadzu GC–MS-QP2010, Shimadzu Corporation, Kyoto, Japan), with fragment ion (m/z value) of 235.10 as target ion and 259.10 and 422.10 as reference ions.

Determination of Agronomic Traits
After physiological maturity, the plant height of inoculated plants was recorded as the distance (cm) from the soil surface to the tip of heads, excluding awns, and spike length was recorded as the distance (cm) from the bottom rachis node to the tip of heads. The number of total spikelets, number of grains per spike, grain weight per spike (g), and 1000-kernel weight (g) were estimated as described above. Heading date, based on RILs, was recorded as the days from the emergence of seedlings to the date when 50% of heads had emerged from the boot. For the controls, heading date was recorded only for the plants with the same sowing date as the RILs.

Statistical Analysis
Statistical analysis was based on replication means for all the inoculated spikes within a replication. A one-way ANOVA was computed first for single year data, and then two-way ANOVA based on 2-yr combined data was conducted to estimate the year effect and genotype x year interaction. For NSS and PSS, only the data at 25 dpi are presented in this paper. For PSS, because of a high consistency between the results of original observed values and arc-sin transformed values, the results based on original data are presented. Since AUDPC and NIRS were determined in only 1 yr (2002) and exhibited extremely high correlations with NSS or PSS, the data for AUDPC and NIRS are not presented but the results are discussed. Broad-sense heritability (hB2) on a RIL mean basis was estimated on the basis of ANOVA results (Fehr, 1987). For single year data, hB2 = {sigma}g2/({sigma}g2 + {sigma}e2/r), in which {sigma}g2 = (MSRIL – MSe)/r and {sigma}e2 = MSe. For 2-yr data, hB2 = {sigma}g2/[{sigma}g2 + {sigma}gy2/y + {sigma}e2/(yr)], in which {sigma}g2 = (MSRIL – MSgy)/(yr), {sigma}gy2 = (MSgy – MSe)/r, and {sigma}e2 = MSe. Here r and y represent the number of replications and number of years, respectively. The exact confidence intervals for heritability were calculated according to Knapp et al. (1985).

Simple correlation between traits was analyzed on the basis of RIL means. In the case that the differences among the RILs for agronomic traits observed were significant, phenotypic and genotypic coefficients of correlation between resistance and agronomic traits were computed on the basis of ANOVA and analysis of covariance results in a similar way for estimation of heritability described above (Fehr, 1987), i.e., rp12 = CoVp12/{surd}({sigma}p12·{sigma}p22) and rg12 = CoVg12/{surd}({sigma}g12·{sigma}g22), where the subscripts 1 and 2 represent traits 1 and 2, respectively. For single-year data, CoVg12 = (MPRIL – MPe)/r, CoVp12 = CoVg12 + CoVe12, and CoVe12 = MPe. For 2-yr combined analysis, CoVg12 = (MPRIL – MPgy)/(yr), CoVp12 = CoVg12 + CoVgy12 + CoVe12, CoVgy12 = (MPgy – MPe)/r, and CoVe12 = MPe.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Resistance to Fungal Spread
All the indications of resistance to fungal spread (NSS, PSS, AUDPC, and NIRS) for the controls or parents were not significantly different for the date of inoculation within each year (F = 0.003–0.034 and P > 0.99 for 2002, and F = 0.052–0.054 and P > 0.99 for 2004). CJ 9306 showed significantly lower disease severity than Veery for all the FHB related parameters analyzed (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Means, coefficients of variation, F values, estimates, and exact 90% confidence intervals of broad-sense heritability of Fusarium head blight scores for resistance to fungal spread in a 152 recombinant inbred line (RIL) population of wheat by greenhouse-based, single-floret inoculation in 2 yr (2002 and 2004).{dagger}

 
One-year and/or 2-yr ANOVA showed that the differences among RILs were highly significant for all the resistance parameters (Table 1). For NSS and PSS, 2-yr ANOVA also suggested a significant difference between years (F = 7.54–13.52, P < 0.01) and a significant genotype x year interaction (F = 6.63–7.12, P < 0.01), which might be partly attributed to different isolates used for inoculation in the 2 yr. In most cases, the averages of RIL population, with large variability, were around the mid-parent values. The coefficients of variation were around 56 to 63% for NSS, PSS, and AUDPC. Comparatively speaking, the coefficient of variation and F value for NIRS (data not shown) were smaller than those of other measures. Frequency distributions were continuous and exhibited a few peaks except for NIRS (Fig. 1 ). Transgressive segregation was evident for all the measures of disease, especially toward susceptibility. Some lines exhibited smaller values for NIRS and AUDPC than CJ 9306. According to the 2-yr data of NSS and PSS, however, very few RILs were consistently superior to CJ 9306 in the resistance to fungal spread (referred to NSS and/or PSS).


Figure 1
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Frequency distribution of 152 wheat RILs derived from the cross Veery/CJ 9306 for resistance to fugal spread of Fusarium graminearum by single-floret inoculation in greenhouse.

 
Resistance to Deoxynivalenol Accumulation
The DON test results showed that CJ 9306 was highly resistant to DON accumulation within visually scabby seeds. On average, the DON content for CJ 9306 (14.1 µg g–1) was approximately one tenth of Veery (134.2 µg g–1) (Table 2). One-year and 2-yr ANOVA results suggested highly significant differences among RILs for DON concentration (Table 2). Similarly to resistance to fungal spread (NSS and PSS), year difference and genotype x year interaction for resistance to DON accumulation were also significant (F = 6.21 and 2.79, respectively, P < 0.01) (Table 2). The mean of the RIL population for DON concentration was lower than the mid-parent value. The coefficient of variation based on 2-yr combined analysis was 75.6%. Frequency distribution (Fig. 2 ) showed a wide range of variation for the resistance within the population, 0.1 to 235.6 µg g–1 in DON concentration and was obviously skewed toward the resistant parent. There were 20 RILs showing lower DON contents than CJ 9306 for either 2002 or 2004. However, the differences between these lines and CJ 9306 were not statistically significant.


View this table:
[in this window]
[in a new window]
  		Table 2. Means, coefficients of variation, F values, estimates, and exact 90% confidence intervals of broad-sense heritability of deoxynivalenol concentration (µg g–1) for resistance to mycotoxin accumulation by Fusarium graminearum in a 152 recombinant inbred line (RIL) population of wheat over 2 yr (2002 and 2004) under single-floret inoculation.

 

Figure 2
View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2. Frequency distribution of 152 wheat RILs derived from the cross Veery/CJ 9306 for resistance to mycotoxin accumulation (deoxynivalenol concentration) by Fusarium graminearum under single-floret inoculation in greenhouse.

 
Resistance to Grain Yield Loss
Data in Table 3 indicated that, for the losses of the grain yield components, CJ 9306 and Veery had distinctly different responses to the disease. On average, there was no significant loss of grain yield for CJ 9306. However, the relative decrease of the number of grains per spike, grain weight per spike and 1000-kernel weight in Veery was 18.8, 48.2, and 33.5%, respectively. Within the RIL population, there was a large variation for the relative loss of grain yield caused by FHB. ANOVA showed that the differences in the relative decrease of three yield components were significant among the 152 RILs (Table 3). The average of the RIL population for the number of grains per spike was reduced by 10.1%, close to a half of the average decrease of susceptible parent Veery. For grain weight per spike and 1000-kernel weight, however, the average decrease of the RIL population was 37.1 and 32.5%, respectively, obviously larger than the average of the two parents or the half value of Veery.


View this table:
[in this window]
[in a new window]
  		Table 3. Means, coefficients of variation, F values, estimates, and exact 90% confidence intervals of broad-sense heritability of relative decrease of number of grains per spike, grain weight per spike (g), and 1000-kernel weight (g) for resistance to grain yield loss by Fusarium graminearum in a 152 recombinant inbred line (RIL) population of wheat in 2004 under single-floret inoculation.

 
The frequency distributions indicated that the components of grain yield were significantly reduced because of the infection of FHB for most RILs (Fig. 3 ). In some cases, the values of decrease in grains and grain weight per spike were negative. This phenomenon was most likely attributable to larger inoculated spikes, compared with the noninoculated spikes, because most of the inoculated spikes were the primary spikes or the first tiller spikes. For 1000-kernel weight, however, only a few lines exhibited a negative decrease.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Frequency distribution of 152 wheat RILs derived from the cross Veery/CJ 9306 for resistance to grain yield loss by Fusarium graminearum under single-floret inoculation in greenhouse.

 
Heritability and Association of Resistance
The estimates of heritability suggested a higher broad-sense heritability for the resistance to FHB spread within the spikes (Table 1). For single year data, the estimates for NSS, PSS, and AUDPC were 0.88 to 0.93 but 0.82 for NIRS. Comparatively speaking, the estimates of broad-sense heritability for 2004 were larger than those for 2002. Heritabilities based on 2-yr combined analysis were reduced because of elimination of genotype x year variance but still were 0.85 to 0.86. Heritability for DON concentration varied largely with years (Table 2), and the average estimate (0.64) based on 2-yr data was lower than those of the parameters of resistance to spread NSS and PSS. For the resistance to grain yield loss, grain weight per spike, and 1000-kernel weight showed a higher heritability than the number of grains per spike (Table 3). Under the same conditions (for 2004 data), the estimates of heritability for the parameters of resistance to grain yield loss were smaller than those of the spread resistance parameters NSS and PSS.

Correlation analysis of either single-year or 2-yr data showed that there were extremely high correlations between different indications of the resistance to fungal spread (rp = 0.88–0.98, P < 0.01, and rg = 0.91–0.99). Both phenotypic and genotypic correlation coefficients between NSS and PSS were 0.98 (Table 4). Phenotypic correlation between DON content and spread resistance parameters or yield loss resistance parameters was low to moderate (rp = 0.36–0.60, P < 0.01). However, the genetic correlation coefficients were high or higher (rg = 0.63–0.92). Among the three indications of resistance to grain yield loss, the phenotypic correlation varied with a range of 0.35 to 0.88 (P < 0.01), but the genetic correlation was higher (rg = 0.73–0.96). In comparison, relative decrease of grain weight per spike or 1000-grain weight showed a higher correlation with resistance to fungal spread (rp = 0.79–0.84, P < 0.01, and rg = 0.93–0.94) or resistance to DON accumulation (rp = 0.53 or 0.56, P < 0.01, and rg = 0.75 or 0.77).


View this table:
[in this window]
[in a new window]
  		Table 4. Coefficients of phenotypic (above) and genotypic (below) correlation between different indications of Fusarium head blight resistance in a 152 recombinant inbred line population of wheat by single-floret inoculation.{dagger}

 
Correlation between Resistance and Agronomic Traits
ANOVA indicated that the differences among RILs were highly significant for all the investigated agronomic traits (F = 3.83–12.71 and P < 0.01), although the differences between two parents were not significant for most of the agronomic traits (data not shown). The range of variation in the RIL population was far beyond the range of two parents, i.e., transgressive segregation for agronomic performance was very evident in this RIL population.

Phenotypic and genotypic correlation coefficients (data not shown) between resistance parameters and agronomic traits indicated that, in most cases, no noticeable genetic correlations existed between the resistance parameters and agronomic traits. No significant correlations were detected between any of the resistance parameters and plant height, number of total spikelets per spike, and/or 1000-kernel weight. Likewise, no significant correlations were observed between any of the agronomic traits and NSS, DON, and/or relative decrease of grains per spike. In a few cases that the phenotypic correlation coefficients were statistically significant, the magnitudes of the correlations were low, and the coefficients of determination were smaller than 0.06. Most of the genetic coefficients of determination were smaller than 0.04 (i.e., |rg| < 0. 20), and none were over 0.13.


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The results of this study indicate that there is a wide range of variability for FHB resistance in the RIL population derived from the cross Veery/CJ 9306. For resistance to fungal spread within the spike, the expression of resistance ranged between almost no spread of disease to entirely diseased spikes. Since CJ 9306 is extremely resistant, few lines in its derived RIL population consistently exhibited superiority over CJ 9306 in resistance to spread.

The four parameters of resistance to fungal spread were highly correlated with each other, indicating that any of them could effectively reflect the difference in resistance observed among RILs. In comparison, NSS is the simplest and easiest measure, but this measure does not refer to the size of the spike or the number of spikelets per spike. PSS provides such a complement to NSS. AUDPC is a comprehensive indication of dynamic development of disease and thus is especially of interest for basic research (Jeger and Viljanen-Rollinson, 2001). However, the AUDPC measure is time-consuming and laborious since it requires multiple observations, and thus it would not be practical in large-scale germplasm screening and breeding programs. NIRS is as simple as NSS, but it is of more biological and less economic interest. The results of our study showed that the average NIRS (data not shown) was greater than NSS, 2.4 higher for RIL average, and 2.9 higher for parental mean. This is consistent with other research (Argyris et al., 2005; Lewis et al., 2003; TeKrony et al., 2000). Under the same conditions, the estimate of broad-sense heritability for NSS and/or PSS was close to that of AUDPC, while NIRS exhibited a lower estimate. In addition, NSS or PSS at 21 dpi (data not shown) exhibited a very high correlation with NSS or PSS at 25 dpi (rp = 0.94–0.97, rg = 0.97–0.99) and with AUDPC (rp = 0.97–0.98, rg = 0.98–0.99). A similar result was reported by Buerstmayr et al. (2000). Therefore, a single FHB scoring (NSS or/and PSS) at 21 to 25 dpi is practical and sufficient for large-scale evaluation of FHB resistance and even for basic research.

Since Miller et al. (1985) analyzed the relationship between FHB resistance and DON concentration, resistance to mycotoxin accumulation in wheat has received increasing attention. However, this type of resistance is generally limited to basic research because of the technological requirements for DON testing. A simple and quick assessment would make it applicable in breeding, and small sample sizes are especially important for early-generation selection. Most of the previous studies have reported that at least 5 g of kernels were required, and both infected and uninfected grains from the inoculated spikes were mixed to serve as the sample for DON test (Browne et al., 2005; Mesterhazy et al., 1999; Miedaner et al., 2003; Somers et al., 2003; Zhou et al., 2002). However, the DON concentration determined in this way did not exactly reflect the resistance to toxin accumulation because it was largely influenced by the proportion of uninfected kernels. A measurement separating DON from kernel damage is of great interest for understanding of resistance to DON accumulation different from other types of resistance, in which independent genes or QTL might be involved (Somers et al., 2003; Yang et al., 2005).

In our preliminary experiment (data not shown) with 40 pairs of infected and uninfected samples taken from inoculated spikes, no DON was traced in most of the uninfected samples, and even in the samples with DON detected, the DON contents were very low, only 0.26–4.0 µg g–1. However, the DON contents in the corresponding infected samples were higher than 100 µg g–1 except for two lines. It is clear that if both infected and uninfected seeds were used in DON testing, the results would be significantly biased. In this study, resistance to DON accumulation was defined as the ability to reduce or limit DON accumulation within the visually scabby grains. Therefore, we only used the infected and/or shriveled seeds as the samples of DON testing, to eliminate the affects resulted from the differences in severity or percentage of kernel infection on the spikes. The results indicated that the variability of resistance to DON accumulation in the RIL population was very large, ranging from 0.1 to 235.6 µg g–1. The variable heritability with lower average estimates indicated that resistance to DON accumulation was greatly influenced by the environment, in comparison with resistance to fungal spread. However, the frequency distribution skewing toward resistance implied the existence of major genes. Evident transgressive segregation suggested a possibility to develop new cultivars or lines that have higher resistance to DON accumulation than CJ 9306. The identification of QTL and marker-assisted selection would speed up the realization of this goal (Somers et al., 2003). In addition, frequency distributions (Fig. 2 and 3) and correlation analysis suggested some differences between resistance to DON accumulation and resistance to grain yield loss, although the phenotypic correlations were significant. It seems possible that there were different genes or QTL involved in these two types of resistance. Further investigations are needed.

Correlation between DON accumulation and FHB resistance ranging from 0.37 to 0.89 have been reported in previous studies (Bai et al., 2001; Mesterhazy et al., 1999; Miedaner et al., 2003; Somers et al., 2003; Zhou et al., 2002). We suppose that the inconsistencies were attributed to DON sampling and analysis methods as well as the types and numbers of experimental materials. The present study shows that there was a high genotypic correlation (rg = 0.92) between NSS/PSS and DON content, suggesting that the resistance to DON accumulation could be achieved by indirect selection for spread resistance to FHB within spikes (Miedaner et al., 2003). This is reflected by the fact that DON accumulation was not taken into consideration in the development of resistant germplasm CJ 9306. Of course, genes or QTL for DON accumulation or degradation independent of spread resistance would be valuable for pyramiding FHB resistance (Somers et al., 2003).

In general, yield loss caused by FHB is defined as the amount of grain yield reduction in the field under disease conditions when compared with yield under conditions without disease. Clearly, it is difficult to employ this measure in genetics and breeding research because it depends on many factors such as disease incidence, disease severity, and kernel infection. In genetic studies of yield loss due to FHB, kernel infection severity (i.e., percentage of Fusarium-damaged kernels) was usually determined (Browne et al., 2005; Mesterhazy, 1995; Yang et al., 2005). However, 1000-kernel weight or grain weight per spike was rarely measured (Mesterhazy et al., 1999). In this study, we measured three traits (relative reduction of number of grains per spike, grain weight per spike, and 1000-kernel weight) to estimate the resistance to grain yield loss, which reflects the property of plants to resist grain yield loss after the fungus had successfully infected and produced disease symptoms on the plants/spikes. The estimates of broad-sense heritability for the relative decrease of grain weight per spike (0.83) and 1000-grain weight (0.76) were distinctly higher than heritability of relative decrease of grains per spike (0.61). It implied that the reduction of grains per spike by FHB was more inconstant and easily influenced by the environments. Moreover, the relative loss of either grain weight per spike or 1000-kernel weight after FHB infection was larger than the decrease of the number of grains per spike. In other words, the total grain yield loss was predominantly due to the decrease of grain weight rather than grain number. Therefore, the relative decrease of grain weight per spike and 1000-kernel weight would be more effective in evaluation of resistance to grain yield loss caused by FHB.

Clarification of the relationship between FHB resistance and agronomic traits is of great significance for development of resistant cultivars acceptable in crop production. Earlier breeding experience indicated that it is difficult to combine the resistance with excellent agronomic performance; studies also suggested significant correlations between FHB resistance and some agronomic traits (Chen, 1983; He, 1983; King, 1996; Liao and Yu, 1985). For instance, significant positive correlation between FHB resistance and plant height was reported in some investigations (Buerstmayr et al., 2000; Liu et al., 1998; Somers et al., 2003). However, it was not confirmed in other studies (Gocho et al., 1983; Jiang and Wu, 1989; Zhou et al., 1988). The present study further demonstrated that there were no noticeable genetic associations between FHB resistance and agronomic traits such as plant height, spike length, number of spikelets and grains per spike, grain weight, and heading date. Although the phenotypic correlation coefficients were significant in a few cases, the coefficients of determination were very small (rp2 < 0.06 and rg2 < 0.13). In breeding practice, such correlations seem to be negligible. Thus, development of FHB resistant and high-yielding cultivars would be possible (Jiang and Wu, 1996).

In conclusion, the investigated resistance types (resistance to fungal spread, mycotoxin accumulation, and grain yield loss) exhibited a large and continuous variation in the RIL population and thus were inherited quantitatively. Estimates of broad-sense heritability varied with the resistance types. Heritability of resistance to fungal spread was higher than those of resistance to DON accumulation and/or resistance to grain yield loss. Compared with resistance to fungal spread, resistance to DON accumulation or yield loss was more unstable and easily influenced by environments. Frequency distributions for spread resistance parameters (except for NIRS) and DON concentration exhibited a few peaks or were distinctly skewed toward resistance, suggesting the existence of major genes. A preliminary QTL analysis has detected one major QTL on 3BS and two minor QTL on 2DL and 5BL for the NSS (Jiang et al., 2005). Therefore, the FHB resistance in CJ 9306 was inherited as a quantitative trait with both major genes and minor genes. We expect that further QTL analysis in progress will provide detailed elaborations.


   ACKNOWLEDGMENTS
 
This research was partly supported by the grants from US Wheat and Barley Scab Initiative (via USDA-ARS).


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This research was partly supported by the grants from U.S. Wheat and Barley Scab Initiative (via USDA-ARS).

{dagger} Current address of R.W. Ward: CIMMYT Wheat Program, Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico. Back

Received for publication January 11, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Agricola
Right arrow Articles by Jiang, G.-L.
Right arrow Articles by Ward, R. W.
Related Collections
Right arrow Wheat
Right arrow Plant Disease
Right arrow Crop Genetics

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