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


     

Published online 2 October 2006
Published in Crop Sci 46:2387-2395 (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 Parh, D. K.
Right arrow Articles by Godwin, I. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Parh, D. K.
Right arrow Articles by Godwin, I. D.
Agricola
Right arrow Articles by Parh, D. K.
Right arrow Articles by Godwin, I. D.
Related Collections
Right arrow Crop Genetics
Right arrow Plant Disease

CROP BREEDING & GENETICS

Genetic Components of Variance and the Role of Pollen Traits in Sorghum Ergot Resistance

D. K. Parha,*, D. R. Jordana, E. A. B. Aitkenb, B. J. Gogelc, C. L. McIntyred and I. D. Godwine

a Queensland Dep. of Primary Industries and Fisheries, Hermitage Research Station, Warwick, QLD 4370, Australia
b School of Integrative Biology, Univ. of Queensland, Brisbane, QLD 4072, Australia
c Univ. of Adelaide, PMB1, Glen Osmond, SA 5064, Australia
d CSIRO Plant Industry, Queensland Bioscience Precinct, 306 Carmody Rd., St Lucia, QLD 4067, Australia
e School of Land and Food Sciences, Univ. of Queensland, Brisbane, QLD 4072, Australia

* Corresponding author (dipal.parh{at}dpi.qld.gov.au)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Ergot (caused predominantly by Claviceps africana Freder., Mantle & De Milliano) is a disease of sorghum (Sorghum bicolor L. Moench), with pollen traits reported to be associated with resistance. This study investigated the genetic architecture and the role of pollen quantity (PQ) and pollen viability (PV) in ergot resistance in an F5 recombinant inbred line (RIL) sorghum population developed from the cross between an elite germplasm line, 31945-2-2 from Australia, and a recently reported putatively resistance source line, IS8525. Percentage ergot infection (PCERGOT) in IS8525 was very low, while in contrast, 31945-2-2 was heavily infected at all inoculation dates of the two field trials conducted under artificial epiphytotic conditions during 2001 and 2002. The distribution of the predicted means for PCERGOT, PQ, and PV was normal, suggesting that the traits are polygenic in nature. Genetic correlations between the two pollen traits and PCERGOT were moderately negative, indicating that there might be some common genetic factors controlling these traits. However, low R2 values of PQ (11%) and PV (9%) suggest that only a small part of the total variability in ergot resistance in this population was due to the variability in PQ and PV. Correlations between ergot scores for different sampling dates (SDATEs) and years were moderate to high, indicating relatively low levels of genotype x environment interaction. Genetic variance and broad sense heritability were also high, indicating that ergot resistance is likely to respond well to conventional selection.

Abbreviations: FA(2), two-factor analytic model • PCERGOT, percentage ergot infection • PDATE, planting date • PQ, pollen quality • PV, pollen viability • QDPI&F, Queensland Department of Primary Industries and Fisheries • RH, relative humidity • RIL, recombinant inbred line • SDATE, sampling date


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SORGHUM ERGOT is one of the major diseases of sorghum. The disease is endemic in Asia and Africa where it was first recorded more than 80 yr ago. A pandemic outbreak of the disease occurred during the mid 1990s when the disease was simultaneously reported in Brazil, South Africa, Australia, and the USA. This caused great concern in the sorghum industry in Australia and worldwide. The infection process mimics fertilization, and therefore fertilized ovaries resist infection. The disease is confined to the sorghum inflorescence and instead of normal pollination, fertilization, and production of seeds, the ovaries of the individual spikelets are colonized by fungal hyphae that develop into spore-bearing fungal masses called sphacelia. Factors such as low temperatures, cloudiness, wind, and rain during flowering that reduce PV and PQ (Brooking, 1976; Bandyopadhyay et al., 1998; Wang et al., 2000) and consequently affect pollination and fertilization, increase the risk of ergot infection. Male-sterile plants in hybrid seed production blocks are particularly vulnerable to ergot damage and losses up to 80% have been reported in India and Zimbabwe (Bandyopadhyay et al., 1998). In Australia the major impact of the disease has come through its significant impact on animal health. Sorghum grain contaminated with even low levels of sclerotia (fungal bodies) can cause toxicity when fed to livestock, particularly sows, dairy cattle, and beef cattle in feedlots (Blaney et al., 2000).

Environmental variables play significant roles in both the infection process and severity of the disease. A number of studies (McLaren and Flett, 1998; Wang et al., 2000; McLaren, 2002) have extensively examined the role of weather variables on ergot incidence, and models were developed to predict ergot severity based on weather variables associated with the critical periods of host development. These models were shown to accurately predict ergot severity of a genotype assuming the presence of viable inoculum, but may not be applicable under conditions of natural infection (Wang et al., 2000). The use of resistant cultivars is seen as being the most promising and most economically viable means of controlling ergot, but has met with limited success to date. Resistance to ergot has been reported as pollen-mediated (i.e., efficient pollination and fertilization restrict infection) in different crops including sorghum (Watkins and Littlefield, 1976; Willingale et al., 1986; McLaren, 1997). In the past, a number of resistance sources have been reported (Sundaram, 1971; Khadke et al., 1978; McLaren, 1992; Tegegne et al., 1994), but it appears that resistance in the reported lines could be a pollen-mediated disease escape or environmental effects overshadowed genetic effects on the phenotype. Recently, several new putative sources of resistance have been identified and found to be effective in male-sterile backgrounds (Dahlberg et al., 2001; Reed et al., 2002), suggesting that resistance to ergot is not entirely pollen-mediated.

The objective of this research was to characterize the genetic basis of sorghum ergot resistance and its relationship with PV and PQ in a RIL population derived from a cross between a susceptible line and the resistance source line IS8525.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Plant Material
This study utilized a population of 290 F5 RILs and their inbred parent lines. The RILs were derived by single seed descent from a cross between two sorghum inbred lines IS8525 and 31945-2-2. Germplasm 31945-2-2 is an elite line developed by the Queensland Department of Primary Industries and Fisheries (QDPI&F), Australia. It produces a good quantity of viable pollen under favorable environmental conditions, but pollen production is severely hindered as the temperature falls below 13°C. A resistance source, IS8525, has been reported to possess the highest known level of ergot resistance from different national and international trials (Dahlberg et al., 2001; Reed et al., 2002; Ryley et al., 2002). It produces a moderate quantity of highly viable pollen under both favorable and adverse environmental conditions.

Field Trials
Two field trials were conducted at the QDPI&F, Hermitage Research Station (altitude 480 m, 280°10' S, 152°02' E), Warwick, QLD, Australia, during the 2001 and 2002 growing seasons. In each year, there were two planting dates (PDATEs) (10 and 19 Feb. 2001 and 30 Jan. and 11 Feb. 2002) to ensure that sufficient variation in flowering date occurred for each RIL to provide a range of inoculation dates for ergot infection. These sowing dates were expected to coincide with the onset of weather conditions conducive to ergot development (Wang et al., 2000). The trial was laid out as a 25 row by 48 column rectangular array of plots with columns 1 to 24 and 25 to 48 representing the PDATE 1 and PDATE 2 plots, respectively. Each block of 12 columns constituted a complete replicate. Consequently, the RILs and parental checks, 31945-2-2 and IS8525, were replicated twice in each PDATE. The trial consisted of single row plots of dimension 3.0 m long and 76 cm wide and containing 20 to 25 plants. The trials were conducted in Talgai shallow phase and Ellinthorp clay soil series in 2001 and 2002, respectively. The surface of these soils are cracking and self-mulching gray clay with abundant CaCO3 concretions (McKeown, 1978). In both years, the experimental plot was fertilized with 130 kg N ha–1 as urea {approx}1 mo before sowing. A starter dose of 12.0, 22.5, 2.5, and 3.0 kg N, P, S and Zn ha–1 was applied immediately before planting. Thinning, weeding, insect control and postsowing irrigation were performed as required following standard practices.

Inoculum and Inoculation
A conidial suspension was prepared by washing infected sorghum heads containing fresh honeydew collected from a nearby ergot-infected disease plot. The resultant suspension was filtered through a fine mesh cloth and the spore concentration was determined using percentage transmittance by a colorimeter (Herde et al., 2006). The concentrated spore suspension was then diluted with water in a 6.0-L sprayer to contain {approx}1 x 106 conidia mL–1. Inoculation was performed on six different dates from 27 April to 15 May 2001 and on 10 different dates from 4 April to 1 May 2002. On each date, on the basis of flowering, a single plant (marked by colored paint on the flag leaf) of only those RILs that had flowered in the middle third of the panicle was selected for inoculation. Florets were removed from the top and the bottom third of the selected plant and the remaining middle third was sprayed with inoculum until it ran off the head. The PCERGOT of the inoculated plant was recorded at the soft dough stage of grain development ({approx}3–4 wk after inoculation).

Pollen Quantity and Viability
Pollen quantity was measured on fine sunny days. Windy, cloudy, and foggy days were avoided as these conditions affect pollen production and PQ measurement (Artschwager and McGuire, 1949; Quinby, 1958; Pepper and Prine, 1972; Ryley et al., 2002). As each individual RIL flowered, 1 to 3 heads were flicked once and rated 1 to 10 (1 for no visible pollen, and 10 for copious quantity of visible pollen) by observing the density of the resultant clouds of pollen. The pollen of each individual RIL was collected on a 0.8% water–agar plate and sprayed with iodine solution (2.5 g L–1 of iodine in 7:3 water–ethanol) immediately after collection. Approximately 100 random pollen grains per plate were counted under a compound microscope to determine PV. Only those pollen grains that stained dark brown to black were considered viable. Pollen grains that were empty, pale, and disrupted were considered nonviable (McLaren and Wehner, 1992).

Weather Data
Weather data including maximum and minimum temperatures and relative humidity (RH) were collected from the weather stations on the research station located in close proximity to the trials. Mean maximum and minimum temperatures and RH were calculated based on 1- to 5-d postinoculation (for PCERGOT), 21- to 28-d preinoculation (for PV), and daily values (for PQ) to examine the effect of these weather variables on the respective traits. These periods were reported as critical for ergot development in the published literature (Bandyopadhyay et al., 1998; Wang et al., 2000).

Data Analyses
The PCERGOT, PQ, and PV data were analyzed using linear mixed models including terms to reflect the sources of variation in the data, which are year, PDATE, replicate, plot, genotype (G), SDATE, and the appropriate interactions between these terms. The plot errors within years were modeled spatially using the approach of Gilmour et al. (1997) and an autoregressive correlation structure was specified to model the correlation between the plot errors for individual plots across SDATEs within years. Factor analytic models (Smith et al., 2001) were used to model the genotype x SDATE interactions. The factor analytic model accommodates genetic variance heterogeneity between SDATEs and heterogeneity of genetic correlation between pairs of SDATEs. In addition, a plot of the estimated loadings from a factor analytic model offers a mechanism for clustering (grouping) SDATEs in terms of genetic correlations [see Smith et al. (2001) for details]. In these plots, the squared length of the vector for a SDATE is the genetic variance explained by the two factors, and the cosine of the angle between the vectors for two SDATEs is the genetic correlation due to the two factors. All mixed model analyses in this paper were conducted using the SAMM (Butler et al., 2003) suite of functions (written for S-language environments; Becker et al., 1988) for mixed model analysis. Additional histograms, linear regression analyses between the two pollen traits and PCERGOT, and associated statistics were produced using the Analyse-it statistical package (v. 1.68, Analyse-It Software Ltd., England).

Heritability Estimates
A linear mixed model containing the terms outlined in Table 1 plus a random genotype main effect and for the grouping factor, that is, group replacing SDATE, was fitted to produce overall heritability estimates for the three traits according to Cullis et al. (2006). These authors propose a generalised heritability estimate for more complex variance structures using the following formula:

Formula 1[1]
where AVESED is the prediction standard error of difference averaged for all pairs of genotypes, and {sigma}g2 is the genotypic variance. In the case of balanced data and a simple model including an overall mean and random variety and residual terms only, this function returns the same result as the standard heritability measure (Cullis et al., 2006). To produce heritability estimates within each group, the same formula was used at the group level for a model specifying genotypic variance heterogeneity between groups.


View this table:
[in this window]
[in a new window]
  		Table 1. Terms included in linear mixed model analysis of percentage ergot (PCERGOT), pollen quantity (PQ), and pollen viability (PV). A tick ({checkmark}) indicates terms included in the final model; a dash (–) indicates nonsignificant terms that were omitted from the final model.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Statistical Analyses and Genotype x Sampling Date Interactions
The terms included in the linear mixed model analyses of PCERGOT, PQ, and PV are listed in Table 1 together with their assigned fixed (F) or random (R) status. The significance of the random terms was assessed using maximum likelihood ratio tests. A two-factor analytic model [FA(2)] was used to model the genotype x SDATE interactions for all traits. Figure 1 is a plot of the loadings for the first factor against the loadings for the second factor for PCERGOT, PQ, and PV, respectively. The plot for PCERGOT shows three clusters (groups) of SDATEs as indicated in the corresponding table. Similarly there were two clusters for PQ. Genotypes appeared to behave reasonably dissimilarly across SDATEs for PV with seven clusters, only three of which include multiple SDATEs. Table 2 is a summary of the genetic correlations for the FA(2) analysis of each trait. With the exception of Group 1/Group 3 (G1/G3), the correlations within and between groups were high for PCERGOT.


Figure 1
View larger version (27K):
[in this window]
[in a new window]
  		Fig. 1. Plot of sampling date loadings from Factor Analytic Model 2 showing groupings for percentage ergot (PCERGOT), pollen quantity, and pollen viability. The sampling dates included in each group for the respective traits are indicated in the table alongside each plot.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Genetic correlations within and between groups of sampling dates (SDATEs) identified by Factor Analytic Model 2 for percentage ergot (PCERGOT), pollen quantity (PQ), and pollen viability (PV).

 
A fixed effect analysis of the three traits for year, genotype, cluster grouping, and genotype x group interaction terms is presented in Table 3. The year effect was not significant, but genotype main effect was significant for all traits (P < 0.05). The group main effect was only significant for PCERGOT (P < 0.05), and the genotype x group interaction term was significant for PCERGOT and PQ (P < 0.05), but not for PV.


View this table:
[in this window]
[in a new window]
  		Table 3. Sources of variation and level of significance (P value for {chi}2) from the fixed effect analysis for percentage ergot (PCERGOT), pollen quantity (PQ), and pollen viability (PV) conducted at the Hermitage Research Station, Warwick, Australia, during 2001 and 2002.

 
Genotypic Performances for PCERGOT, PQ, and PV
The frequency histograms of the predicted genotype means (overall and group data), as illustrated in Fig. 2, 3, and 4 for PCERGOT, PQ, and PV respectively, showed continuous variation indicating the polygenic nature of the traits. Although there is a little skewness and kurtosis in each histogram, the square-root transformed data appeared to follow the assumption of normality (data not shown). The mean PCERGOT of the RILs ranged from 3.60 to 44.25 for the overall data. The PCERGOT was low for the resistant parent IS8525 (4.08) and only three RILs, namely RIL-227 (3.60), RIL-206 (3.79), and RIL-44 (4.04), showed a level of resistance comparable with IS8525. The PCERGOT was 30.18 for 31945-2-2, the other parent of the cross. The overall predicted genotypic mean for PCERGOT was 15.63. The mean PCERGOT was significantly higher for G1 (19.81) than that of G2 (13.47) and G3 (12.43) (Fig. 2), with the lowest mean maximum (21.59°C) and minimum temperatures (7.56°C) and highest RH (80.97%) averaged across 1 to 5 d from the date of inoculation.


Figure 2
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 2. Frequency histogram of percentage ergot (PCERGOT) using overall, Group 1 [G1, Sampling Dates (SDATEs) 3–6 of 2001 and SDATE 10 of 2002], Group 2 (G2, SDATEs 1–2 of 2001 and SDATEs 6–9 of 2002), and Group 3 (G3, SDATEs 1–5 of 2002) data in the recombinant inbred line (RIL) population 31945-2-2 x IS8525 (n = 292) conducted at the Hermitage Research Station, Warwick, QLD, Australia, during 2001 and 2002. MP = midparental value.

 

Figure 3
View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3. Frequency histogram of pollen quantity (PQ) using overall, Group 1 [G1, Sampling Dates (SDATEs) 1–4 of 2002] and Group 2 (G2, SDATEs 1–6 of 2001 and 5 of 2002) data in the recombinant inbred line (RIL) population 31945-2-2 x IS8525 (n = 292) conducted at the Hermitage Research Station, Warwick, QLD, Australia, during 2001 and 2002. (PQ 1 indicates no visible pollen and 10 indicates copious quantity of visible pollen on a 1–10 scale).

 

Figure 4
View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. Frequency histogram of pollen viability (PV) using overall and Group 3 (Sampling Dates 3 and 4 of 2001 and 3 of 2002) data in the recombinant inbred line (RIL) population 31945-2-2 x IS8525 (n = 292) conducted at the Hermitage Research Station, Warwick, QLD, Australia, during 2001 and 2002.

 
The PQ of the RILs ranged from a score of 2.15 to 6.39 based on a 1-to-10 scale. Pollen production was mostly consistent across the SDATEs in IS8525 (4.78) but was reduced significantly in 31945-2-2 with cooler temperature (7.50°C) on the SDATE. The average PQ of 31945-2-2 was 4.49 for G1 but was only 2.77 for G2 (Fig. 3).

The PV of the RILs ranged from 72.40 to 94.75. The PV was consistently high for IS8525 (91.95 and 92.63 for overall and G3, respectively) and was moderately high for 31945-2-2 (Fig. 4).

Heritability Estimates of the Traits
The overall heritability (broad-sense) estimates were moderate to high for all traits, 0.842 (PCERGOT), 0.497 (PQ), and 0.664 (PV). However, the estimate was moderately low (0.339) for PV with G3 data (Table 4).


View this table:
[in this window]
[in a new window]
  		Table 4. Estimates of heritability (broad-sense) for percentage ergot (PCERGOT), pollen quantity (PQ), and pollen viability (PV) using overall and group data in the recombinant inbred line (RIL) population 31945-2-2 x IS8525.

 
Relationship between PCERGOT and the Pollen Traits
The relationships between PCERGOT and the two pollen traits are displayed in Fig. 5. Although the relationships were significant (P < 0.0001 for both PQ and PV), the R2 values indicate that only a limited amount of the variation in PCERGOT was explained due to the variation in PQ and PV (R2 = 11 and 9% of the total variation, respectively). The inclusion of interaction term (i.e., PQ x PV) in the analysis had no appreciable change on the relationship between PCERGOT and pollen traits which appeared to act in an additive fashion.


Figure 5
View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between two pollen traits, pollen quantity (PQ) and pollen viability (PV), and percentage ergot (PCERGOT) in the recombinant inbred line population 31945-2-2 x IS8525 (n = 292) conducted at the Hermitage Research Station, Warwick, QLD, Australia, during 2001 and 2002.

 
A linear mixed multitrait analysis of the PCERGOT, PQ, and PV data was undertaken. The estimated genetic correlations (Table 5) between the two pollen traits and PCERGOT were negative and moderate indicating that a decrease in PQ and PV concomitantly increased PCERGOT and there might be some common genetic factors involved in controlling these traits. The relationship was positive and moderate for PQ and PV.


View this table:
[in this window]
[in a new window]
  		Table 5. Genotypic correlations between percentage ergot (PCERGOT), pollen quantity (PQ), and pollen viability (PV) in the recombinant inbred line population 31945-2-2 x IS8525 (n = 292).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
For the first time, genetic components of ergot resistance including two pollen traits reported previously as being implicated in ergot resistance, were investigated in sorghum. Our study suggests that the genetic variance for ergot resistance in this population of sorghum is moderately high and this coupled with high heritability indicate that the trait could be amenable to breeding programs through selection. The factor analytic model for genotype x environment interaction identified three groups among the SDATEs of PCERGOT, but observed genetic correlations between groups of SDATEs were moderate to high. This, along with a nonsignificant year effect on ergot severity and high Spearman rank correlation (r = 0.76, P < 0.0001) between 2 yr of data (data not shown) suggest that relative rankings of the RILs developed from the cross 31945-2-2 x IS8525 did not vary greatly, and that predictions of genotypic performance would be feasible across environments. This is consistent with the observations of Dahlberg et al. (2001) and Reed et al. (2002), who reported that IS8525 possessed the highest level of ergot resistance among various sorghum accessions tested, and also this source of resistance was more stable across diverse environments. Relative ranking of other sources of resistance have been reported to be unpredictable across genotypes. Our study confirms that IS8525 is an extremely stable source of ergot resistance. This could be due to the limited contribution of pollen traits which are influenced by the environment to the resistance of IS8525.

The PCERGOT rating of IS8525 was very low as was observed earlier by different studies (Dahlberg et al., 2001; Ryley et al., 2002; Reed et al., 2002). Using overall predicted genotypic means, only three RILs were found comparable with the level of resistance to IS8525. The PQ was moderate but consistent and the PV was very high for IS8525 across the SDATEs. On the other hand, PCERGOT was high for the susceptible parent, 31945-2-2, throughout the SDATEs, indicating that ergot-favorable weather persisted during both seasons of the field trials. The PV of 31945-2-2 was moderately high but the pollen production was reduced substantially with cooler temperatures, as is evident from the histogram of G2 data for PQ (Fig. 3).

An FA(2) model (Smith et al., 2001) was used to model genotype x SDATE interactions. This analysis results in more accurate prediction of overall genotypic means by accommodating genotype x SDATE variance heterogeneity and heterogeneity of genetic correlation between pairs of SDATEs. A substantial reduction in error variance could be attributed to large gains in heritabilities obtained for the overall and group data sets for PCERGOT, as well as moderate to high heritabilities for the two pollen traits (Table 4) even though these traits were reported to be highly influenced by environment. The within- and between-group genetic correlations also support these heritability estimates except for the G3 data set for PV. The moderately low heritability estimate for G3 data set for PV could be attributed to less genetic variance due to 10 missing values (data not shown) for RILs, as well as high error group variance compared with the overall data set.

The importance of weather variables such as pre- and postflowering temperatures, RH, and so forth on ergot infection has been well documented in a number of studies. Although this study did not aim to elucidate the effect of weather variables on ergot development, a comparison of the mean PCERGOT of the three groups identified by the FA(2) indicated a strong influence of mean maximum and minimum temperatures and RH at 1 to 5 d after inoculation. This agrees with earlier studies. The mean PCERGOT of the RILs of G1 was significantly higher, with the lowest maximum (21.59°C) and minimum temperatures (7.56°C) and highest RH (80.97%). As the temperature increased and the RH decreased as in G2 and G3, the relative ergot severity on RILs also decreased (Fig. 2). For PQ and PV, although the group mean was not significant, a negative trend in pollen production and PV was observed for many RILs, including parental line 31945-2-2 with decreasing mean minimum temperature on the SDATE and 3 to 4 wk of preinoculation, respectively (Fig. 3, 4).

The two pollen traits, PQ and PV, have previously been reported to be associated with resistance (escape) to the disease (Bandyopadhyay et al., 1998; McLaren, 1997). The relationship between PV and ergot severity was reported to be linear (McLaren and Flett, 1998), with severity increasing as PV decreased. Genotypes also differ in their ability to produce pollen (Pérez et al., 1999). Though there is no published method for the quantitative measurement of PQ, a qualitative measurement indicated that ergot severity was negligible when the product of PV (0–100%) and PQ (0–10) (PV x PQ) exceeded 100; the disease severity increased in a linear fashion as PV x PQ decreased below 100 (Ryley et al., 2002). Thus, genotypes that produce copious quantities of viable pollen should be more resistant. Consequently, male-sterile lines that do not produce pollen are very susceptible to ergot infection. In this study, the genetic correlations observed between PCERGOT and the two pollen traits were moderate and negative, indicating that some genetic factors (pleiotropy or close linkage) may be common in controlling these traits and that the decrease in PQ and PV are associated with increase in PCERGOT. However, the small R2 values suggest that only a small part of the total variability in ergot infection can be explained by the variability in the pollen traits. This suggests that another mechanism of resistance may be operating in controlling ergot resistance in this population of sorghum.

Dahlberg et al. (2001) indicated that the resistance mechanism in IS8525 was not pollen mediated but physiological. A combination of flower characteristics including least exposure time of stigma to inoculum before pollination, rapid stigma drying after pollination, and small stigma were reported as being responsible for ergot resistance in this line. Their study also reported another restorer line, Tx2737, which was an excellent pollen shedder, a desirable pollen characteristic for resistance to ergot, was susceptible to ergot because of the protogynous nature of the flower. Reed et al. (2002) confirmed the genetic resistance of IS8525 to ergot in their study as well as observing moderate resistance in male-sterile hybrids produced using this line. However, unlike the present study, ergot resistance in those lines was not stable across diverse environments.

Ergot resistance in our study appears polygenic. Lynch and Walsh (1998) suggested that the difference between the mean of a RIL population and its midparental value can be used as a measure of epistasis. By this measure, epistasis for ergot infection seems to be relatively unimportant in this population of sorghum (Fig. 2).

For a trait showing continuous variation, it is assumed that apart from polygenes, major genes are also involved in the inheritance system (Wang et al., 2001). The high heritability found for PCERGOT indicates that this may well be the case for this trait in sorghum. Mendelian approaches can be utilized to analyze the inheritance of major genes with high heritability, but a QTL mapping approach is more useful to uncover the genetic structure of variation of traits controlled by polygenes. We are currently developing a genetic map of this population of sorghum for QTL analysis to unravel the genetic basis of ergot resistance.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
From this study we conclude that ergot resistance derived from IS8525 in this RIL population is likely to be responsive to conventional selection given the apparent absence of epistasis, low genotype x environment variation, moderate levels of genetic variation, and high heritability. At least three of the RILs in this population have a greater level of ergot resistance than that of IS8525. The two pollen traits are also amenable to selection and show moderate genotypic correlations with ergot resistance indicating that there might be some common genetic factors involved in controlling these characters. The modest amount of variation in ergot infection explained by these traits supports the findings of Dahlberg et al. (2001) that pollen traits are of less importance in the resistance of IS8525 than in other resistance sources. These findings, along with the lower genetic variances of the pollen traits, indicate that while indirect selection for ergot resistance using these traits would result in some increase in resistance, direct selection for ergot resistance would still be required.


   ACKNOWLEDGMENTS
 
This research was supported by the financial help of the Cooperative Research Centre for Tropical Plant Protection, University of Queensland, Australia; the Queensland Department of Primary Industries and Fisheries, Australia; and the Commonwealth Scientific and Industrial Research Organization, Australia.

Received for publication December 15, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 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 Parh, D. K.
Right arrow Articles by Godwin, I. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Parh, D. K.
Right arrow Articles by Godwin, I. D.
Agricola
Right arrow Articles by Parh, D. K.
Right arrow Articles by Godwin, I. D.
Related Collections
Right arrow Crop Genetics
Right arrow Plant Disease

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