HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Harder, D.E. Articles by Legge, W.G. Search for Related Content PubMed Articles by Harder, D.E. Articles by Legge, W.G. Agricola Articles by Harder, D.E. Articles by Legge, W.G.

PLANT GENETIC RESOURCES

Effectiveness of Different Sources of Stem Rust Resistance in Barley

^a Cereal Research Centre, Agriculture and Agri-Food Canada, 195 Dafoe Road, Winnipeg, MB, R3T 2M9 Canada
^b Research Centre, Agriculture and Agri-Food Canada, P.O. Box 1000A, R.R.#3, Brandon, MB, R7A 5Y3, Canada

dharder{at}em.agr.ca

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
Stem rust (Puccinia graminis Pers. f. sp. tritici Eriks. & Henn.) pathotype QCCJ is a potential threat to barley (Hordeum vulgare L.) production in North America. A field test was conducted for 3 yr to evaluate the effectiveness of several new sources of stem rust resistance to reduce losses. A randomized complete block design was used, with a split-plot arrangement of treated (propiconizole fungicide) and untreated plots. Nine lines used in the test were classed as resistant (R), moderately resistant (MR), moderately susceptible (MS), and susceptible (S). The yield losses in each group were: R—Q21861 (12%), QSM-041 (12%), BM8923-46 (12%); MR— `Diamond' (21%), SB90585 (26%); MS—`Robust' (30%), `Bonanza' (33%), `Harrington' (37%); S—`Klages' (53%). There were highly significant effects on 1000-kernel weight, test weight, and percentage kernel plumpness for all entries. The losses in kernel weight ranged from 7% (Q21861) to 43% (Klages), in test weight from 3% (Q21861) to 26% (Klages), and kernel plumpness from 5% (Q21861) to 95% (Klages). There were no significant effects of rust infection on plant height and days to heading. Effects on lodging were variable, and reduced days to maturity was highly significant for all entries. Lines with combinations of genes Rpg1/rpg4 (Q21861, Q/SM-041) and Rpg1/Rpg3 (BM8923-46) provided the highest levels of protection, with a somewhat lesser level provided by Rpg1/RpgU (Diamond). Combinations of these genes should provide effective stem rust resistance in barley breeding.

Abbreviations: AAFC, Agriculture & Agri-Food Canada • kw, 1000-kernel weight • MR, moderately resistant • MS, moderately susceptible • PGR, Plant Gene Resources of Canada • plump, percentage kernel plumpness • R, resistant • S, susceptible • tw, test weight

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
THE NORTH CENTRAL GREAT PLAINS of the USA and the eastern prairie region of Canada are important barley-growing regions under risk of crop loss due to stem rust. Before the introduction in North America of barley cultivars with gene Rpg1 for stem rust resistance, yield losses to natural epidemics of stem rust were as high as 12 to 15% in the north-central USA (Roelfs, 1978) and in Canada (McDonald, 1970). Extensive losses to naturally occurring stem rust were reported in susceptible cultivars in Australia in 1982 and 1983 (Dill-Macky et al., 1990). Although P. g. tritici occurs on barley in many parts of the world, there is little other data on estimated losses in barley due to this disease.

Gene Rpg1 has provided effective resistance to stem rust since the introduction of barley cultivars with this gene in 1938, and it has remained as a principal source of resistance (Steffenson, 1992). Other genes for stem rust resistance known in barley are Rpg2 from `Hietpas-5' (Patterson, 1951), Rpg3 from PI 382313 (Jedel et al., 1989), rpg4 from Q21861/PI 584766 (Jin et al., 1994), RpgU occurring in several barley cultivars (Fox and Harder, 1995), and a second recessive gene in Q21861 (Jin et al., 1994; Fox and Harder, 1995). Factors modifying resistance to stem rust were observed in the cross OAC 21/`Chevron' (Lejeune, 1946) and in the crosses Minn. 615/`Kindred' and Minn. 615/`Montcalm' (Miller and Lambert, 1955). The expression of Rpg1 varies to some extent with genetic background (Steffenson, 1992). The effectiveness of the Rpg1 resistance appears to be enhanced by the presence of gene Rpg3 (Jedel et al., 1989). Experience with P. g. tritici in inoculated field nurseries has shown a wide range of reactions to stem rust by various barley lines and cultivars (D.E. Harder, 1990–1998, unpublished data), further indicating interactions of genes for resistance with a number of other genetic factors to confer variable resistance levels.

In 1988 a new pathotype of P. g. tritici was first observed (Martens et al., 1989) that showed higher levels of virulence to barley cultivars with Rpg1. This pathotype was designated as QCC according to the original nomenclature of Roelfs and Martens (1988) and subsequently identified as QCCJ to differentiate it from other isolates of QCC that are avirulent to Rpg1 (Fox et al., 1995). In Manitoba field nurseries inoculated with QCCJ, various barley lines or cultivars, with or without Rpg1, have shown very high levels of infection, typically 80 to 90%, with susceptible-type reactions (D.E. Harder, 1994–1998, unpublished data). Late-sown barley fields in Manitoba's Red River Valley have shown up to 65% infection levels (Harder, 1997). A general epidemic of stem rust in barley involving QCCJ has yet to occur, although the observations suggest that a potential for an epidemic exists. Western Canadian breeding programs are actively incorporating stem rust resistance into breeding lines. The known genes for resistance other than Rpg1 confer varying intermediate levels of resistance to QCCJ (Liu and Harder, 1996). The level of protection from QCCJ under field conditions afforded by alternative resistance sources is not documented. The objective of this study was to determine the effectiveness of various sources of stem rust resistance in barley to prevent yield and quality losses.

	Materials and methods

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
Lines and Cultivars
Nine barley lines or cultivars were selected for measurement of responses to stem rust infection. These were selected on the basis of previously observed field reactions in nurseries inoculated with P. g. tritici pathotype QCCJ, their known stem rust resistance genotype, agronomic type, and popularity with growers. The lines, their accession numbers, known resistance genotypes, and range of reactions observed (infection levels and reaction) in nurseries inoculated with pathotype QCCJ were:

BM8923-46 is a breeding line from the Agriculture & Agri-Food Canada (AAFC) Research Centre in Brandon, Manitoba, derived from the cross `Manley'/WB8738-1-1-7 (a selection from Manley/PI 382313). Manley shows resistance to non-QCCJ pathotypes that is characteristic of lines with gene Rpg1, and PI 382313 is a source of gene Rpg3 (Jedel et al., 1989). Field reaction is 30 MR to MS.

Bonanza (CI 14003) is a six-row malting-type barley that was extensively grown in western Canada from 1975 to 1986. Bonanza carries Rpg1 (Steffenson et al., 1993) and typically shows a 70 to 80 MS to S field reaction.

Diamond (Plant Gene Resources of Canada [PGR] 12328) (Kaufmann and Kibite, 1985) is a six-row feed barley with relatively good resistance to pathotype QCCJ in field tests. The ancestry of Diamond (`Galt'/`Unitan') includes Galt (`Glacier'/`Newal'/`Husky'), and `Peatland'. Galt, Husky, and Peatland carry Rpg1. In addition to Rpg1, Diamond also carries gene RpgU (Fox and Harder, 1995). Field reaction is 50 MR to MS.

Harrington (PGR 12181): is a two-row barley that has been the dominant cultivar grown for malting purposes in western Canada. Although Harrington has no known genes for stem rust resistance, it has consistently demonstrated a moderate to low level of resistance. Field reaction is 50 to 60 MS to S.

Klages (CI 15478) is a two-row malting barley originating from the USDA-ARS and registered in Canada through the University of Saskatchewan. Klages is a standard for two-row malting quality, but production was limited to the western prairie region due to disease susceptibility. Klages does not carry gene Rpg1 (Steffenson et al., 1993). Field reaction is 90-100 S.

Q21861 (PI584766) originated from the International Wheat and Maize Improvement Centre, Mexico and subsequently was selected in Queensland, Australia. Q21861 has genes Rpg1, rpg4, and a probable third recessive gene (Jin et al., 1994; Fox and Harder, 1995) for stem rust resistance. Field reaction is 20 MR-MS.

Q/SM-041 is a breeding line from the Crop Development Centre, University of Saskatchewan, derived from the cross Q21861/SM89010 (courtesy of Dr. B.G. Rossnagel, University of Saskatchewan). This is a doubled haploid line expected to have Rpg1 and rpg4. Field reaction is 20 to 30 MR to MS.

Robust (PI 476976) is a six-row malting barley from the University of Minnesota, which is a popular cultivar in Manitoba. In field tests using non-QCCJ pathotypes, Robust has resistance that is characteristic of the Rpg1 resistance (D.E. Harder, 1994–1998, unpublished data). Robust is moderately to highly susceptible to QCCJ. Field reaction: 80 to 90 MS to S.

SB 90585 is a breeding line from the Crop Development Centre, University of Saskatchewan (courtesy Dr. B.G. Rossnagel), derived from a cross between the Dutch line VD 40385 and SB 86154. SB 90585 appears to carry Rpg1 as determined from seedling tests with P. g. tritici pathotype MCC (D.E. Harder, 1995, unpublished data). It may also carry additional unknown resistance from VD 40385. Field reaction is 60 MR to MS.

Site and Experimental Design
In each of 1994, 1995, and 1996, the test was carried out at the AAFC Brandon Research Centre on Ramada orthotic black clay loam to silty clay loam soil. Previous use of the land was canola (Brassica napus L.) for the 1994 test, wheat (Triticum aestivum L.) for 1995, and fallow for 1996. Fertilizer was applied per soil test recommendations as follows: 56 kg ha^-1 actual N in 1994; 26 kg ha^-1 actual N in 1995; 66, 29, and 24 kg ha^-1, respectively, of actual N, P, and S in 1996.

Whole plots, in a randomized complete block design with four replicates, consisted of eight 5-m rows of each entry, 30 cm apart. A row of a mixture of susceptible wheat and barley lines (spreader row) split each whole plot into two four-row subplots. The spreader rows were sown about 10 d before the test plots in each year. The sowing dates for the test plots were 27 May 1994, 29 May 1995, and 3 June 1996.

Inoculation and Fungicide Treatments
The spreader rows were inoculated with P. g. tritici pathotype QCCJ at the early jointing stage by injecting an aqueous suspension of urediniospores into the hollow culm of plants chosen at random at 0.75-m intervals along each spreader row. At the first opportunity following injection when an overnight dew was expected, a follow-up inoculation was performed by dusting the spreader rows with urediniospore-talcum powder mixture (approximately 1:1000 by weight). Subsequent infections then spread naturally through the remainder of the nursery.

Stem rust was controlled in a randomized subplot of each whole plot with Tilt fungicide (propiconizole; Novartis Crop Protection, Canada), applied at the late boot stage at recommended rates with a backpack sprayer. Treatment dates were: 6 July, 27 July, 9 Aug. 1994; 19 July, 28 July, 15 Aug. 1995; 18 July, 26 July, 8 Aug., and 14 Aug. 1996. Each subplot was harvested with a small plot combine and analyzed separately.

Measurements
Data were taken on yield, 1000-kernel weight (kw), test weight (tw), percentage kernel plumpness (plump = percentage of kernels retained on a 2.4 by 19.1 mm slotted screen), time in days from sowing to heading, maturity, height, lodging, and stem rust severity. Other diseases, mainly spot blotch [Cochliobolus sativus (Ito and Kurib.) Drechs. ex Dast.] and net blotch (Pyrenophora teres Drechs.) that could affect yield or quality also were noted and measured as a single score on two replications each year. The progress of stem rust development was estimated by severity percentage and infection type at 3- to 7-d intervals through the growing season. Evaluations began 1 wk after the first infections appeared in the test plots. Infection types were assigned numerical values to facilitate digital processing as follows: 0 = R, or no visible infection; 1 = R to MR; 2 = MR; 3 = MR to MS; 4 = MR to MS to some S types; 5 = MS; 6 = MS to S; 7 = S. An infection index as a single value for purposes of analysis was derived by calculating (severity percentage rating x infection type)/2. The terminal infection indexes (last evaluation date) were used to compare the rust severities among the lines.

A supplementary experiment to evaluate the rate of stem rust development in each line was conducted at the AAFC Cereal Research Centre at Winnipeg in 1996. In this test each line was sown as randomized four-row, 2-m-long plots, replicated six times. To reduce interplot interference, each plot was surrounded by four rows of the wheat cultivar Katepwa, resistant to pathotype QCCJ. The entire experimental area was surrounded by eight rows of Katepwa. External to the eight-row wheat border, two rows of a mixture of susceptible wheat and barley cultivars were planted as rust spreaders, 10 d prior to planting the test plots. The spreader rows were inoculated by injection as described above. The spreaders were destroyed when the initial infections appeared in the plots to prevent further introduction of new inoculum. Rust progress observations were taken at 4-d intervals after the first appearance of infections. For each plot, 10 stems were chosen at random and infection percentage and infection-type ratings were taken. The data were analyzed as in the Brandon experiments.

Statistical Analysis
Preliminary analyses of variance were conducted for each trait each year individually using the GLM procedure (SAS Institute, 1985) for a randomized complete block design with split plots. Nonsignificant heterogeneity of variances justified pooling the analyses of variance across years. Year, genotype, and fungicide treatments were considered to be fixed effects, and replications were nested within years and used as the error term for years. The loss percentage in a trait due to stem rust was calculated as: 100 - (untreated subplot value/treated subplot value x 100). Using percentage of difference comparisons may exaggerate treatment differences because of inherent normal potentials and may increase coefficients of variation. However, this parameter is the most easily understood when considering disease effects. Since differences between groups of lines were large, conclusions regarding the effects of stem rust infection were not affected by considering only percentage of differences. Percentage reductions in traits were combined across years and subjected to analysis of variance with years and genotypes considered fixed effects and replication nested within years (error for year effects). Entry means were compared with the LSD value (P = 0.05). For the rust infection severity data, lines were compared by first analyzing for homogeneity of variance, then performing paired t tests adjusting for equality of variance when necessary.

	Results

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
Although statistical comparisons are possible considering only treated and untreated plots, the true effect of rust infections on the various lines may be somewhat obscured by partial effectiveness of fungicide treatment or entry x treatment interactions. Figures 1A and 1B show the progress of rust development in Brandon in 1996 with and without fungicide treatment. Despite four fungicide applications in 1996, high levels of stem rust developed on the more susceptible lines in the test. Klages, Bonanza, and Robust developed final infection indexes between 115 and 130 with fungicide treatment and between 288 and 310 without treatment. These results indicate that loss percentages caused by stem rust for the various components reported below, particularly for the more susceptible lines, may be conservative.

View larger version (32K): [in this window] [in a new window]   Fig. 1 Progress of stem rust development in plots of nine cultivars or lines of barley, inoculated with Puccinia graminis f. sp. tritici pathotype QCCJ, in Brandon, Manitoba, in 1996. (A) Plots not treated, (B) plots treated with propiconizole fungicide

Yield and Kernel Characteristics
Table 1 shows yield data for each of the 3 yr and for the years combined. Averaged across years, the yields of all lines were significantly reduced in the untreated plots compared with the treated plots. There were no differences in yield reduction between BM8923-46, Q/SM-041, and Q21861. The next most effective lines were Diamond and SB90585, with no significant differences between the two. The line SB90585 also was statistically similar to Robust and Bonanza, although there was an actual 7.1% difference between SB90585 and Bonanza (LSD = 7.78). Similarly, there were no statistical differences between Robust, Bonanza, and Harrington. Klages had a significantly higher loss than all the other lines.

Of all parameters measured, the proportion of plump kernels was the most severely affected by rust infection (Table 2). The exception was Q21861, where there was only a modest reduction. Lines BM8923-46 and Q/SM-041 had significantly greater reduction in plumpness than Q21861. The relative reductions in Diamond, SB90585, and Robust were similar, as were those for Robust, Harrington, and Bonanza. Klages showed very high reduction in plumpness.

Agronomic Characters
Effects of stem rust on days to heading and maturity, height, and lodging are shown in Table 3 . There was little effect on days to heading or on plant height. There was a highly significant acceleration in days to maturity without fungicide protection for all lines. Line Q21861, which consistently had shown the highest resistance in most parameters, matured only 1 d earlier, while Klages, the most susceptible line, matured 6.3 d earlier. For the remaining lines, the effect on maturity was inconsistent relative to yield responses. Bonanza matured at a rate similar to BM8923-46 and Q/SM-041. There was little or inconsistent effect on lodging. Lodging was significantly greater in untreated plots of Robust and Bonanza. However, for Klages, which had the highest yield loss, there was no significant effect on lodging. There was generally no effect on lodging among the lines that had shown the most resistance, with only Q/SM-041 showing a slight increase, similar to the moderately susceptible Harrington.

Evaluation of Stem Rust Infection Levels
The development of stem rust infections in the tests at Brandon in 1996 are shown in Fig. 1A and 1B, and at Winnipeg in 1996 in Fig. 2 . The terminal infection indexes are in Table 5 . The scores were higher in Brandon than in Winnipeg for all entries except Klages (Table 5). The more resistant lines appeared to show the greatest differences between the locations; however, the rankings for both locations were similar, with variations in rank occurring mainly within the resistance and susceptibility classes. At both locations lines Q21861, Q/SM-041, and BM8923-46 were grouped as the most resistant lines, followed by Diamond and SB90585. In Brandon, Bonanza, Klages, and Robust were similar in their terminal severities, but in Winnipeg these three lines were significantly different, with Klages the most susceptible.

View larger version (17K): [in this window] [in a new window]   Fig. 2 Progress of stem rust development in plots of nine cultivars or lines of barley, inoculated with Puccinia graminis f. sp. tritici pathotype QCCJ, in Winnipeg, Manitoba, in 1996

View this table: [in this window] [in a new window]   Table 5 Terminal rust infection indexes of nine barley lines or cultivars in plots inoculated with pathotype QCCJ of Puccinia graminis f. sp. tritici at Brandon and Winnipeg, Manitoba in 1996.

	Discussion

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
The recent threat to barley production in the Northern Plains of North America by the occurrence of P. g. tritici pathotype QCCJ and the fact that protection was provided mainly by a single gene (Rpg1) generated considerable interest in finding new sources of resistance. Genes such as Rpg3 (Jedel et al., 1989), rpg4 (Jin et al., 1994), and RpgU (Fox and Harder, 1995) have been identified and characterized based on the resistance phenotypes in segregating populations and in field nurseries. However, there was no information on the level of loss protection that these and possibly other sources of resistance may provide under stem rust epidemic conditions. To evaluate the loss avoidance against stem rust, the effects of other diseases such as net blotch and spot blotch also need to be considered. In terms of visual ratings of these diseases, the effect of fungicide treatment was highly significant, but relative to stem rust, the differences between the treated and untreated plots were not as large. This may be due to relatively less efficacy of propiconizole to control these diseases, or less favorable timing of treatment. Normally spot blotch and net blotch begin to show significant development in the field at the early tillering stage, whereas fungicide treatment was not begun until the boot stage, to coincide with stem rust development. Also, the relative scores for these diseases among the entries had little relationship to their respective rust scores or loss data. For example, Klages, Robust, and Bonanza, which had the highest rust scores and had among the highest losses, scored relatively low for net blotch and spot blotch. Therefore, although it is likely that these diseases contributed to some of the yield and agronomic differences between the treated and untreated plots, their effects generally were considered to be relatively neutral when comparing the responses of the various entries to stem rust. An exception may have been Harrington (see below).

For most of the components measured, the lines with Rpg1/rpg4 combined (Q21861 and Q/SM-041) and Rpg1/Rpg3 combined (BM8923-46) showed the most effective resistance. These lines fit into the resistant class. Genes Rpg3 or rpg4, when combined with Rpg1, appear to be equally effective.

During 1994 and 1995, Q21861 and Q/SM-041 had yield losses around 6%, but significantly higher losses in 1996 contributed to a somewhat higher 3-yr average loss. Gene rpg4 becomes relatively ineffective at higher temperatures (Jin et al., 1994). Observations of the amount of stem rust development on Q21861 in inoculated nurseries are highly variable, often depending on location in the nursery and direct exposure to the sun (D.E. Harder, 1994–1996, unpublished data). During August (the main grain-filling period) the average maximum and minimum temperatures, respectively, in 1996 were 27.1 and 10.3°C, as compared with 27.1 and 11.8°C in 1995, and 23.6 and 9.5°C, in 1994. Since temperatures in 1996 and 1995 were very similar, temperature does not explain the higher yield loss in the rpg4-containing lines in 1996. Other unknown environmental interactions may have been responsible.

Diamond (Rpg1, RpgU) and SB90585 (Rpg1, Rpg?) best fit a moderately resistant classification. From the similarity of their responses, it is possible that SB90585 also carries the RpgU gene. Gene RpgU could be present in breeding lines such as SB90585, since by pedigree analysis Fox and Harder (1995) suggested that this gene may occur in a number of North American barley cultivars.

Lines with apparently only Rpg1 (Robust and Bonanza) as well as Harrington (no known resistance genes) fit a MS category, with few differences between them. Fox and Harder (1995) had suggested that Bonanza may also carry RpgU, but from the comparison with Diamond in our study, Bonanza appears to lack this gene (unless Diamond has resistance in addition to Rpg1 and RpgU). Klages, with yield losses averaging 53%, was the most susceptible, and could be considered as a reliable susceptible control in experiments with stem rust on barley.

Dill-Macky et al. (1990) concluded that the yield losses in their study were mainly attributable to reductions in grain weight and size. Of the kernel characteristics, plump was the most severely affected by stem rust. This may be expected, since even a small amount of shriveling of most seeds could drastically reduce the number of plump seeds. Thus, this component may not indicate the actual amount of damage, although plumpness may have economic consequences because of its importance in malting barley quality. From the results of this study, the reduction in plump was relatively poor at discriminating the various levels of resistance. Within the resistant group, Q21861 appeared to be much less affected than Q/SM-041 and BM8923-46. There was little distinction in kernel plumpness among the lines within the MR and MS groups. The tw measure showed the lowest percentage reduction due to stem rust. This also could be expected, since with reduction in seed size, more seeds would fit into a given volume. This measure was a better indicator of the four classes of resistance than was plump. Like tw, kw reductions accurately discriminated the four different resistance and susceptibility classes, and best predicted the amount of reduction in yield. Considering all of the effects on seed quality, Q21861 appeared to be the least affected by stem rust, and Klages the most.

Several workers (Calpouzos et al., 1976; Dill-Macky et al., 1990; Fox and Harder, 1995) have concluded that terminal disease severities are a reliable indicator of resistance to stem rust in wheat or barley. Greaney (1933) developed linear regression models that predicted the amount of reduction in yield and test weight for each 10% increment in infection level of stem rust in wheat and oat (Avena sativa L.). When compared with yield loss data (Table 2), the terminal disease rankings (Table 4) at both Brandon and Winnipeg were similar, indicating that Q/SM-041, BM8923-46, and Q21861 were the most resistant, followed by Diamond and SB90585 being somewhat less resistant.

With increased apparent susceptibility, the relationship between yield losses and visual rust scores were less consistent. The rust disease index for Harrington ranked an intermediate sixth at both locations, but Harrington consistently ranked eighth for yield loss all 3 yr (Table 1). Robust ranked somewhat higher for yield loss protection than Harrington, but showed higher rust disease indexes, particularly at Brandon. Across a number of years in inoculated stem rust nurseries, Harrington has consistently given moderate to moderately susceptible rust disease scores (D.E. Harder, 1990–1998, unpublished). Although Harrington consistently showed a more resistant rust infection index response than Robust or Bonanza, the yield loss was as high as in the two more susceptible cultivars. Harrington is highly susceptible to diseases such as spot blotch and net blotch, which was reflected in the higher ratings for these diseases for this cultivar, and this may have contributed to the higher losses in Harrington relative to the rust infection index. At Brandon the rust epidemics were initiated more aggressively than at Winnipeg. As a result, at Brandon the terminal rust severities were obscured among the more susceptible lines. In contrast at Winnipeg they were clearly distinguishable, and conformed more closely to the agronomic data. Therefore, to use terminal rust severity data as an evaluation tool, it is important that the test nursery not be overwhelmed with an excessively severe epidemic. In climates where rust infections normally are readily obtained, a less aggressive initiation of the epidemic is recommended. If very high infection levels nevertheless should occur, it may be necessary to make evaluations earlier when checks are more clearly distinguishable.

Of the agronomic measures taken, days to heading and height were not affected, as may be expected. There was an inconsistent effect on lodging, and the lodging scores were poorly associated with losses in yield and seed quality. Past experience has shown heavily stem rusted fields of wheat to be badly lodged, as occurred with Robust and Bonanza. The time to maturity was significantly affected. Earlier than normal maturity is commonly observed in rust-affected cereals. In a detailed study McGrath and Pennypacker (1991) related reduced yield to accelerated senescence and shortened grain growth duration in wheat affected by stem rust. Under the conditions of our study this measure did not discriminate the resistant, moderately resistant, and moderately susceptible lines; however, it is possible that under less severe stem rust conditions these lines may have shown significant differences. Klages, the most susceptible line, showed the greatest acceleration to maturity. There is also a possibility that maturity was delayed in the control plots due to fungicide treatment. The times to maturity of the registered cultivars, however, were similar to those expected. The very early senescence of these lines due to stem rust would be an important factor in limiting potential yield and in reducing quality.

This study has shown that using known sources of stem rust resistance in barley is effective in reducing potential losses to this disease. This experiment involved more severe stem rust epidemic conditions than normally would occur in nature. With yield losses in the most resistant lines of 12% under these conditions, virtually no losses would be expected under natural field conditions. Field reactions of barley lines derived from similar crosses have varied considerably across a number of years of observation (D.E. Harder and W.G. Legge, 1990–1998, unpublished data), indicating considerable background genetic interactions with the stem rust resistance genes. Also, the stem rust resistance genes may interact positively, as was indicated for Rpg1 and Rpg3 (Jedel et al., 1989). Since gene Rpg1 remains effective against the bulk of the North American P. g. tritici population, it is important to retain this gene in breeding populations. Lines with this gene alone are ineffective against pathotype QCCJ, but in combination with genes rpg4 (line Q/SM-041) and Rpg3 (line BM8923-46), much improved resistance was attained. There were no differences between lines with either Rpg1/3 or Rpg1/rpg4. Barley lines with genes Rpg1, Rpg3 and rpg4 combined are yet to be developed, and it is possible that further improvements in resistance could result. There also appeared to be a contribution to Rpg1, although less strong, by RpgU (Diamond and possibly SB90585). In conclusion, breeding for stem rust resistance in barley using the sources of resistance used in this study would be a highly effective and economical way to control this disease.

	ACKNOWLEDGMENTS

 
Funding from the Western Grains Research Foundation, Cargill Research Fund, and the Brewing and Malting Barley Research Institute is gratefully acknowledged. The assistance of Dr. B.G. Rossnagel in planning the study and providing seed of SB90585 and Q/SM-041 is also gratefully acknowledged. Ken Dunsmore, Tom Henderson, and Bob Wilson provided technical assistance.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 
Contribution no. 1741, Cereal Research Centre, AAFC, Winnipeg.

Received for publication November 23, 1998.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results Discussion REFERENCES

 

• Calpouzos, L., A.P. Roelfs, M.E. Madson, F.B. Martin, J.R. Welsh, and R.D. Wilcoxson. 1976. A new model to measure yield losses caused by stem rust in spring wheat. Tech. Bull. Agric. Exp. Stn. Univ. Minnesota Bull. 307.
• Dill-Macky R., Rees R.G., Platz G.J. Stem rust epidemics and their effects on grain yield and quality in Australian barley cultivars. Austr. J. Agric. Res. 1990;41:1057-1063.
• Fox S.L., Harder D.E. Resistance to stem rust in selected barley lines and the inheritance of resistance to pathotype QCC. Can. J. Plant Sci. 1995;75:781-788.
• Fox S.L., Harder D.E., Kim W.K. Use of virulence and length variability within the rDNA repeat unit to distinguish isolates of Puccinia graminis f. sp. tritici pathotype QCC. Can. J. Plant Pathol. 1995;17:197-204.
• Greaney, F.J. 1933. Method of estimating losses from cereal rusts. Proc. World's Grain Exhib. & Conf. Regina, Canada. Contrib. no. 379. Div. of Botany, Exp. Farms Branch (now Research Branch), Agriculture & Agri-Food Canada, Ottawa, Canada. 2:224–235.
• Harder D.E. Stem rusts on wheat, barley, and oat in Canada in 1995. Can. J. Plant Pathol. 1997;19:171-175.
• Jedel P.E., Metcalfe D.R., Martens J.W. Assessment of barley accessions PI382313, PI382474, PI382915, and PI382976 for stem rust resistance. Crop Sci. 1989;29:1473-1477.[Abstract/Free Full Text]
• Jin Y., Steffenson B.J., Miller J.D. Inheritance of resistance to pathotypes QCC and MCC of Puccinia graminis f. sp. tritici in barley line Q21861 and temperature effects on the expression of resistance. Phytopathology 1994;84:452-455.
• Kaufmann M.L., Kibite S. Registration of `Diamond' barley. Crop Sci. 1985;25:706.[Free Full Text]
• Lejeune A.J. Correlated inheritance of stem rust reaction, nitrogen content of grain and kernel weight in a barley cross. Sci. Agric. 1946;26:198-211.
• Liu J.Q., Harder D.E. Components of resistance in barley to stem rust: receptivity, urediniospore production, latent period, and infection response in adult plants. Can. J. Bot. 1996;74:1298-1304.
• Martens J.W., Dunsmore K.M., Harder D.E. Incidence and virulence of Puccinia graminis in Canada on wheat and barley in 1988. Can. J. Plant Pathol. 1989;11:424-430.
• McDonald W.C. Diseases and other factors affecting average yields of barley in Manitoba, 1954–1968. Can. Plant Dis. Surv. 1970;50:113-117.
• McGrath M.T., Pennypacker S.P. Reduction in the rate and duration of grain growth in wheat due to stem rust and leaf rust. Phytopathology 1991;81:778-787.
• Miller J.D., Lambert J.W. Variability and inheritance of reaction of barley to pathotype 15b of stem rust. Agron. J. 1955;47:373-377.[Free Full Text]
• Patterson, F.L. 1951. Adult plant and seedling reaction of barley varieties and hybrids to 3 pathotypes of Puccinia graminis tritici. Ph.D. thesis. Dep. of Agronomy. Univ. of Wisconsin, Madison.
• Roelfs, A.P. 1978. Estimated losses caused by rust in small grain cereals in the United States — 1918 to 1976. USDA Misc. Publ. no. 1363. Washington, DC.
• Roelfs A.P., Martens J.W. An international system of nomenclature for Puccinia graminis f. sp. tritici. Phytopathology 1988;78:526-533.[ISI]
• SAS Institute. SAS user's guide: Statistics. Version 5 ed. Cary, NC: SAS Inst, 1985.
• Steffenson B.J. Analysis of durable resistance to stem rust in barley. Euphytica 1992;63:153-167.[ISI]
• Steffenson B.J., Miller J.D., Jin Y. Detection of the stem rust resistance gene Rpg1 in barley seedlings. Plant Dis. 1993;77:626-629.

This article has been cited by other articles:

				J.M. Nyachiro, J.H. Helm, M. Oro, P.E. Juskiw, and D.F. Salmon Registration of 'Sundre' Barley Journal of Plant Registrations, May 1, 2007; 1(1): 35 - 36. [Full Text] [PDF]

				J.M. Nyachiro, J.H. Helm, M. Cortez, M. Oro, P.E. Juskiw, and D.F. Salmon Registration of 'Manny' Barley Crop Sci., February 23, 2005; 45(2): 788 - 789. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Harder, D.E. Articles by Legge, W.G. Search for Related Content PubMed Articles by Harder, D.E. Articles by Legge, W.G. Agricola Articles by Harder, D.E. Articles by Legge, W.G.