Introduction

Oat (Avena sativa L.) is one of the important cereal crops grown worldwide (Gold et al. 2005), generally divided into hulled oat (with lemma and palea) and naked oat (bare grain/groat) (Zhao et al. 2007a, b). China is the site of origin of naked oat and its cultivation can be traced back to 5000 years ago (Qu et al. 2006). Naked oat is the popular type planted in China while hulled oat is grown in other countries (Guo et al. 2012; Zhen and Zhang 2012). According to statistical data in 2013, China is among the top ten world’s oat producers, namely EU, Russia, Canada, Australia, United States, Ukraine, Chile and China (USDA. Foreign Agricultural Service, World Agriculture Production, October 2014). Used as high-quality forage and healthy food, oat is considered as a high value crop. Oat cultivated acreage has rapidly increased to 1.0 million ha, mainly grown in Shanxi, Hebei, Inner-Mongolia, and Jilin as well as other provinces (Qu et al. 2006; Zhao et al. 2007a, b.

The productivity and quality of oat can be greatly affected by oat stem rust (Fig. 1) which is caused by the obligate biotrophic fungus P. graminis Per. f. sp. avenae Eriks. & E. Henn. (Pga) (Stakman et al. 1923). The first four races of Pga were identified using oat cultivar differentials: Victory, White Tartar and Monarch in North America in 1923 (Stakman et al. 1923) and the fifth race or race five was identified with differentials White Tartar, Richland and Jostrain in 1925 (Fetch and Dunsmore 2003). Since then, race identification for Pga has become routine in some countries (Harder 1994; Keiper et al. 2005). Severe epidemics of oat stem rust occurred frequently in North America in the early 1900s (Fetch 2003) and also in western Canada in the 1940s (Martens 1978). Recent epidemics in Manitoba and eastern Saskatchewan occurred in 1977 and 2002 and caused estimated yield losses of 35 and 5~10 %, respectively (Fetch 2003). In China, there were no published reports on oat stem rust before the present study. In 2008 and 2009, severe epidemics of oat stem rust occurred in the main oat producing areas in Baicheng, Jilin Province, and in 2012 and 2013, similar epidemics occurred in Zhangjiakou, Hebei province. The disease caused yield losses of approximately 10~15 %.

Fig. 1
figure 1

The natural occurrence of oat stem rust in a field in Baicheng, Jilin province, in the years 2008–2009 (provided by Leng Yanrui, Baicheng Academy of Agricultural Science)

Full size image

The recent severe epidemics of oat stem rust in China should be basically resulted from lack of the knowledge of the rust resistance of oat cultivars and the information on race or virulence of oat stem rust population. Therefore the objectives of this study were to characterize the races and virulence pathotypes of the Pga population using the oat differentials and to evaluate the resistance of the commonly planted oat cultivars.

Materials and methods

Collection of oat stem rust and oat cultivars

Of a total of 11 oat stem rust samples, four were collected in 2012 and provided by Hebei North University and seven were collected from production fields in September, 2013 also in Zhangjiakou (altitudes 1200–2000 m), Hebei Province. After naturally drying at room temperature, the samples were kept at 4 °C for later use.

Thirty-five oat cultivars were obtained with the help of local research institutes of plant protection (Baicheng Academy of Agricultural Sciences, Gansu Academy of Agricultural Sciences, Shanxi Academy of Agricultural Sciences, Inner Mongolia Academy of Agricultural Sciences). These oat cultivars included the major cultivars used in the major oat production areas in China.

Procedures of isolation and multiplication of single-uredium isolates

Single-uredium isolates were obtained using a previously described method (Huang et al. 1989) as it follows: Seeds of the universally susceptible variety “Marvellous” were sown in 12 cm diameter clay pots. When the primary leaves were fully expanded or when the seedlings were about 7 days old, the leaves were cut off from the seedlings and put onto two layers of filter paper wetted with 40 mg l−1 6-benzylaminopurine (BA) solution used to keep detached leaves green in 15 cm diameter petri dishes. Both ends of the detached leaves in the petri dishes were immobolized with glass stripes. Urediospores of single uredia on the hydrated samples were scratched and smeared onto the detached leaves using whittled-flat toothpicks. The inoculated leaves were then moistened with an atomizer containing 0.05 % Tween 20. The dishes were covered with the lids and placed in an incubation chamber for 20 h dark at 16~20 °C, transferred to a culture room with 14 h (light)/10 h (dark) at 18~20 ± 1 °C. Visible uredinia (diameter, 0.5~0.7 mm) appeared about 6 days after inoculation. The isolation or purification of single pustules was conducted using the above procedures, repeated three times to obtain a purified single pustule. In order to obtain sufficient urediospores for later race identification use, the multiplication of the purified single pustules also was done following the above method.

Differentials and designation of races

Twelve single gene lines Pg1, Pg2, Pg3, Pg4, Pg6, Pg8, Pg9 (Rodney-Pg9 background), Pg10, Pg12, Pg13, Pg15 and Pg16 were included in the differential host system, and Pg9-R (Pg9 in Rosen’s Mutant background), Pg-a, Rodney 0 and Marvelous (no known resistance genes) were used as supplementary differentials. These differentials, listed in Table 1, were provided by Dr. J. H. Yuan, Hebei North University (from Dr. Y. Jin, Cereal Disease Laboratory, USDA-ARS, St. Paul, MN).

Table 1 Background and origin of differentials and their low infection types caused by P. graminis f. sp. avenae
Full size table

The letter code nomenclature utilizes 12 single-gene Pg differential lines, with three subsets of four lines organized into a hexadecimal system that has 16 possible combinations of low (L) or high (H) reaction for each letter (Table 2). Each isolate was designated with a three letter race code based on its reaction on the differential hosts (Fetch and Jin 2007). The infection types (ITs) of the subsets were represented by the first letter, the second and third letters. For instance, high ITs on the four Pg genes in a set are assigned the letter ‘T’, while low ITs on the four hosts, are assigned a letter ‘B’. Hence, if an isolate produces high infection types (susceptible reactions) on the 12 Pg genes, the race will be assigned the three letter code ‘TTT’. An isolate which produces low ITs (resistant reactions) on the 12 Pg genes has a race code ‘BBB’.

Table 2 Letter code designations for races of P. graminis f. sp. avenae using 12 differential lines in three ordered subsets of lines each
Full size table

Inoculation of oat stem rust differential hosts

Sixteen Pg single-gene differentials were orderly planted in 12 cm diameter porcelain pots. Four holes were evenly dug into the growing medium in each pot and 5–6 seeds for each differential were sown in a hole. Marvellous was used as the susceptible control to ascertain the success of inoculation to each differential host set. Inoculation, incubation and disease development conditions were the same as previously described above. Fourteen days after inoculation, ITs were assessed following the 0–4 scale described by Stewart and Roberts (1970). ITs were grouped into two categories, namely, the ITs 0,;, 1, 1+, 2 and 2+ were considered as low (resistance) while the ITs 3-, 3, 3+ and 4, were considered as high (susceptible) (Stewart and Roberts 1970).

All the experiments were carried out in the College of Plant Protection, Shenyang Agricultural University.

Results

Race identification

Twenty-six isolates were obtained from 11 accessions. Three races, namely TKR, TJM and TKM, of Pga were characterized and the occurrence frequencies were 61.6, 30.8 and 7.6 %, respectively, as shown in Table 3. TKR was predominant and TJM was less -dominant. It was worth noting the special infection types (Fig. 2) produced on oat lines with Pg10 by some isolates of TKR, where the epidermis of the primary leaves was ruptured by uredinia and outward rolled. The pustule sizes were big enough to be scored into IT three or four (Fig. 2) and some pustules began to coalesce, which we scored as susceptible ITs though there was some resistant feature or chlorosis around uredinia (Fig. 2).

Table 3 Infection type on P. graminis f. sp. avenae genes and frequency for 3 races of oat stem rust
Full size table
Fig. 2
figure 2

Infection types on Pg10 produced by four different isolates of race TKR. Note: Ym8, Ym11, Ym13 and Ym26 were single pustule codes

Full size image

The virulence formulae of three races, namely, TKR, TJM, and TKM, to the 16 Pg genes showed that Pg6 and Pg15 were highly effective to all of the races identified in this study. Pg10, Pg13, and Pg-a were moderately resistant while Pg1, Pg2, Pg3, Pg4, Pg8, Pg9, Pg12, Pg16, Pg9-R and Marvellous were highly susceptible to all of the races. Races TJM and TKM showed opposite ITs on Pg10, ie., TKM was virulent to Pg10 while TJM, avirulent. Similarly, races TKM and TKR were opposite in their virulence to Pg13 and Pg-a, TKR being virulent to Pg13 and avirulent to Pg-a, but TKM was avirulent to Pg13 and virulent to Pg-a. Furthermore, race TKR was virulent to Pg10 and Pg13, and avirulent to Pg-a, while TJM was virulent to Pg-a and avirulent to Pg10 and Pg13.

Virulence frequencies of the 26 isolates to the16 differential hosts

Virulences to resistance genes Pg1, Pg2, Pg3, Pg4, Pg8, Pg9, Pg12, Pg16 and the gene(s) in Marvellous were detected at 100 % frequency; to Pg10 at 69.2, Pg13 at 61.6, Pg-a at 61.6 %, and the gene(s) in Rodney 0 at 84.6 %; and no virulence to Pg6 and Pg15 was detected (Table 4).

Table 4 Virulence frequency of 26 P. graminis f. sp. avenae isolates on 16 differentials
Full size table

Reaction of oat cultivars to the oat stem rust races

The resistance test results are shown in Table 5. Thirteen (37.1 %) of the 35 tested oat cultivars showed different resistance levels to races TKR, TKM and TJM at the seedling stage. Of thirteen resistant varieties, twelve were highly resistant (ITs;1,1) and one immune. The remaining 22 (62.9 %) were all moderately to highly susceptible (Table 5).

Table 5 Reaction of oat cultivars to three races of P. graminis f. sp. avenae
Full size table

Discussion

The races of Pga in China

In this study, 26 single-uredium isolates of Pga obtained from one of China’s major oat production regions (Hebei province), were tested following the way adopted in North America, i.e., using the single Pg gene differentials of oat and the three-letter Pga-code nomenclature system (Fetch and Jin 2007). Three races, TKR, TJM and TKM, were found. This is the first published report for the physiological races and virulence of Pga in China. All of the reported races in North America have no virulence on gene Pg10 and only races NA1 and NA70 have virulence on Pg6 (Fetch and Jin 2007). In China, among the 26 isolates tested, none had virulence to Pg6 and Pg15 while up to 69.2 % showed virulence on Pg10. Therefore, it is obvious that races and virulence of Pga in China and North America are considerably different, similar to the wheat stem rust pathogen, P. graminis f. sp. tritici, in China and North America (Cao and Chen 2010).

The virulence of oat stem rust races and effective Pg genes in China

According to this study, Pg6 and Pg15 were resistant to all Chinese isolates tested. The studies conducted in the Rust Laboratory, previously located in Winnipeg, Canada, showed that Pg6 is useful for differentiating isolates of Pga in Canada. Most North American races were avirulent to Pg6 while the virulence to Saia (Pg6) was frequent in Australia (Adhikari et al. 2000). Pg15 is ineffective to most North American races, but in our study, Pg15 was immune or nearly immune (‘;’ or ‘;1=’) to most isolates.. Pg10, Pg13, Pg-a and Rodney had different levels of resistance to the three identified races, while Pg1, Pg2, Pg3, Pg4, Pg8, Pg9, Pg12, Pg16, Pg9-R and Marvellous were ineffective. Most of the Pg genes were ineffective to the tested isolates, indicating that the Chinese Pga population has a relatively wide spectrum of virulence.

Resistance level of Chinese oat cultivars

The seedling resistance genes are usually effective at the adult plant stage and confer a strong resistance response (Singh et al. 2008). In this study, seedling resistance evaluation of 35 cultivars was carried out using the three races we obtained. Only 13 varieties, including Bayan 1, Bayou 9 and Baiyan derivatives have different degrees of resistance and 22 oat cultivars were highly susceptible, with nearly 100 % severity. The results show that the majority of the cultivars are susceptible to oat stem rust. Thus, it is urgent to develop oat varieties with resistance to stem rust (Zhao et al. 2007a, b). The information of the races, virulence patterns and resistant cultivars identified in the study will be useful for breeding for stem rust resistant oat varieties. As a pioneering study on the disease in China, this report can only cover a relatively small number of oat stem rust samples collected from one major oat growing region. Therefore it is necessary to keep on monitoring the disease and characterizing virulence and races from a large number of rust samples from all major oat-growing regions in China.