1,218
Views
15
CrossRef citations to date
0
Altmetric
Special Issue: Rust Diseases of Field Crops and Forest Trees

Physiologic specialization of Puccinia triticina, the causal agent of wheat leaf rust, in Canada in 2015–2019

, , , , &
Pages S333-S346 | Accepted 05 Feb 2021, Published online: 27 Sep 2021
2

Abstract

Wheat leaves infected with Puccinia triticina, the causal agent of wheat leaf rust, were collected annually throughout Canada from 2015 to 2019. There were 47, 75, 44, 38 and 54 different virulence phenotypes, respectively, found annually, representing 154 unique virulence phenotypes. From Alberta, there were three virulence phenotypes, MBDS, TDBJ, and TDBS, found in 2015 and eight in 2019, the most common being MBDS and TNBJ (both 20%). The most common virulence phenotypes found in Manitoba and Saskatchewan were MBDS (17.8%) and TNBG (16.3%) in 2015, MNPS (17.2%) and MBDS (15.9%) in 2016, MNPS (41.9%) and TBBG (17.2%) in 2017, MNPS (35.3%) and TBBG (34.8%) in 2018, and MNPS (54.7%) and MBDS (10.7%) in 2019. In Ontario, the most common virulence phenotypes found were MBTN (25%) and MBDS (16.7%) in 2015, MCQQ (13.6%) in 2016, MBTN (33.3%) and PBDG (19.0%) in 2017, TFPJ (13.6%) in 2018, and MCTN (14.5%) and MBTN (10.9%) in 2019. In Quebec the most common virulence phenotypes found were TBBG (66.7%) and MLDS (33.3%) in 2015, TCGJ (23.1%) and MBTN (15.4%) in 2017, MCQH (26.3%) and FCPT (21.1%) in 2018, and MBTN and MCRS (both 18.2%) in 2019. The frequencies of virulence varied on all resistance genes over these years. Within Canada, virulence on Lr21 peaked in 2018 at 39.9% and then declined in 2019, with a similar trend noticed for virulence on Lr2a and Lr2c. There was no virulence detected on Lr19, Lr29, Lr32, Lr52, and Lr22a, while virulence on Lr25 was rare.

Résumé

Des feuilles de blé infectées par Puccinia triticina, l’agent causal de la rouille brune, ont été collectées annuellement partout au Canada de 2015 à 2019. On a trouvé annuellement 47, 75, 44, 38 et 54 différents phénotypes de virulence, respectivement, ce qui a représenté 154 phénotypes de virulence uniques. En Alberta, en 2015, on a trouvé trois phénotypes de virulence, MBDS, TDBJ et TDBS, et huit en 2019, les plus courants étant MBDS and TNBJ (20% chacun). Les phénotypes de virulence les plus courants trouvés au Manitoba et en Saskatchewan étaient MBDS (17,8%) et TNBG (16,3%) en 2015, MNPS (17,2%) et MBDS (15,9%) en 2016, MNPS (41,9%) et TBBG (17,2%) en 2017, MNPS (35,3%) et TBBG (34,8%) en 2018 ainsi que MNPS (54,7%) et MBDS (10,7%) en 2019. En Ontario, les phénotypes de virulence les plus courants étaient MBTN (25%) et MBDS (16,7%) en 2015, MCQQ (13,6%) en 2016, MBTN (33,3%) et PBDG (19,0%) en 2017, TFPJ (13,6%) en 2018 ainsi que MCTN (14,5%) et MBTN (10,9%) en 2019. Au Québec, les phénotypes de virulence les plus courants étaient TBBG (66,7%) et MLDS (33,3%) en 2015, TCGJ (23,1%) et MBTN (15,4%) en 2017, MCQH (26,3%) et FCPT (21,1%) en 2018 ainsi que MBTN et MCRS (18,2% chacun) en 2019. Durant ces années, les fréquences de virulence ont varié en fonction de tous les gènes de résistance. Au Canada, la virulence à l’égard du gène Lr21 a culminé en 2018 à 39,9%, puis a baissé en 2019. Une tendance similaire a été observée à l’égard des gènes Lr2a et Lr2c. Aucune virulence n’a été détectée à l’égard des gènes Lr19, Lr29, Lr32, Lr52 et Lr22a, tandis que, en ce qui concerne le gène Lr25, elle était rare.

Introduction

Wheat leaf rust, caused by Puccinia triticina Eriks. (Anikster et al. Citation1997) (syn. P. recondita Rob. ex Desmaz. f. sp. tritici) is one of the most common and damaging diseases of wheat worldwide (Huerta-Espino et al. Citation2011) and causes epidemics annually on wheat in Canada. Leaf rust is normally found widespread throughout eastern Saskatchewan, Manitoba, Ontario, Quebec and the Maritime provinces. The severity of the disease varies from year to year and region to region and is affected by the amount of inoculum in the form of urediniospores entering each region of Canada from the neighbouring USA states, the temperature, rainfall, and the relative susceptibility of the wheat crop.

In field surveys conducted in Manitoba from June to September annually from 2015 to 2019 the wheat crop was evaluated for the incidence and severity of wheat leaf rust. Leaves infected with P. triticina were sampled, dried and stored, then subsequently isolates were made from these infected leaves to analyze for virulence. Annual virulence surveys for P. triticina have been done in Canada since 1931, using the same set of Thatcher near isogenic lines, each containing a different leaf rust-resistance gene, providing a continuous record of the annual virulence spectrum of the P. triticina population in Canada (McCallum et al. Citation2016a). This continuous record facilitates monitoring of virulence on the most common genes in Canadian wheat germplasm including Lr1, Lr2a, Lr10, Lr13, Lr14a, Lr16, Lr21, and Lr34 (McCallum et al. Citation2016a; Toth et al. Citation2018) and other genes that are found occasionally or could be used in future wheat cultivars. Annual changes in the virulence phenotypes found and the frequency of virulence on many resistance genes can be seen over time (Wang et al. Citation2010; McCallum et al. Citation2016a). The number of virulence phenotypes found each year depends on the number of isolates analyzed and the diversity of the population, but it ranged from 22 to 78 over the years 1995 to 2015 with an average of 40.7 (B. McCallum unpublished). The populations from western Canada were normally different from those found in Ontario and Quebec, which were also different from that found in Prince Edward Island, though there were virulence phenotypes that were in common between the regions.

The objective of this study was to get a representative sample of isolates for the Canadian P. triticina population annually from 2015 to 2019 and then analyze the virulence spectrum for these isolates, and frequency of virulence to key resistance genes within these populations. Results were compared between regions and with previous years from the same region to detect trends in virulence changes and the predominant virulence phenotypes.

Materials and methods

Virulence on the standard seedling differential lines

Infected wheat leaves were collected from individual fields and nurseries from June to September during 2015 to 2019 in various locations throughout Alberta, Saskatchewan, Manitoba, Ontario, and Quebec. The leaves were air dried at 20–27°C for approximately 12 to 24 h then stored separately for 2–4 months at 5°C. Puccinia triticina urediniospores from individual collections were then scraped off from infected leaves using a metal spatula and a small amount of water, and inoculated onto the susceptible wheat cultivar ‘Little Club’ by rubbing the leaves with urediniospores, water, and Tween 20 mixture as described previously (McCallum and Seto-Goh Citation2005). Each pot of ‘Little Club’ seedlings was pretreated with 50 mL maleic hydrazide solution (0.36 g L−1) approximately 5 d after seeding to prevent the emergence of secondary leaves and to produce larger uredinia with abundant sporulation. A plastic cone, approximately 25 cm in height with an open top, was placed over each pot to minimize cross-contamination. Inoculated plants were placed into a dew chamber (Model I-60D, Percival Scientific, Perry, IA) with nearly 100% relative humidity for approximately 17 h to allow the urediniospores to germinate and initiate the infection process, then placed into a greenhouse at 20 ± 4°C with supplemental high-pressure sodium lighting, resulting in a photoperiod of approximately 16 h. Approximately 7 d after inoculation, chlorotic spots appeared, indicating areas of infection. The leaves were then trimmed so that a single isolated uredinium remained on the upper edge of each trimmed leaf. Cross-contamination was minimized by removing all extra leaves. At approximately 14 d after inoculation, urediniospores were collected from a single isolated uredinium into a 00 gelatin capsule using a vacuum suction micro-collector, mixed with a light mineral oil (Soltrol 170 Isoparaffin, Chevron Phillips Chemical Co., TX, USA), and sprayed onto a 7-d-old set of wheat seedlings which included 16 single resistance gene ‘Thatcher’ near-isogenic lines and the susceptible line ‘Thatcher’ to test virulence, and a pot of ‘Thatcher’ plants for urediniospore increase. Two single uredinial isolates were typically evaluated from each rust collection, although sometimes one or three isolates per collection were analyzed. The inoculated increase pot of ‘Thatcher’ was kept isolated with a plastic cone on top of the pot, and urediniospores were vacuum collected for subsequent inoculations. Approximately 12 seeds of each ‘Thatcher’ near-isogenic line were planted in a clump, and the clumps were evenly spaced in a fibre flat (25 × 15 cm). After inoculation, the plants were allowed to dry for at least 1 h to allow the oil to evaporate and then incubated and maintained as described above. Infection types produced on the 12 standard leaf rust (Lr) differential lines [Set 1: Lr1 (RL6003a), Lr2a (RL6016), Lr2c (RL6047), Lr3 (RL6002); Set 2: Lr9 (RL6010), Lr16 (RL6005), Lr24 (RL6064), Lr26 (RL6078); Set 3: Lr3ka (RL6007), Lr11 (RL6053), Lr17 (RL6008), Lr30 (RL6049)] were used to determine the three letter code according to the virulence phenotype nomenclature (Long and Kolmer Citation1989). Four supplemental differential lines [Set 4: LrB (RL6051), Lr10 (RL6004), Lr14a (RL6013), Lr18 (RL6009)] were added to provide additional virulence information about the isolates, resulting in a four letter code. All isolates were also inoculated onto ‘Thatcher’, Thatcher-Lr21 (RL6043), and LrCen (RL6003b). The resistance gene temporarily named LrCen was previously identified in the Thatcher-Lr1 near-isogenic line RL6003 (McCallum and Seto-Goh Citation2006b). Infection types to all the differential near-isogenic lines were rated 12 d after inoculation. Isolates that produced infection types ‘;’ (hypersensitive flecks), ‘1’ (small uredinia with necrosis), and ‘2’ (small to medium sized uredinia with chlorosis) were considered avirulent to the differential line, and those that produced infection types ‘3’ (medium sized uredinia without chlorosis or necrosis) and ‘4’ (large uredinia without chlorosis or necrosis) were considered virulent to the line (Long Citation1989). Inoculations were repeated if the infection response was not clear.

Virulence on adult plant differential lines and additional seedling differential lines

At least one isolate from most of the unique virulence phenotypes identified was inoculated onto adult plants of ‘Thatcher’ and six ‘Thatcher’ near-isogenic lines (Lr12 [RL6011], Lr13 [RL4031], Lr21 [RL6043], Lr22a [RL6044], Lr35 [RL6082], and Lr37 [RL6081]), using urediniospores that were increased as described previously. Single plants of each ‘Thatcher’ near-isogenic line and ‘Thatcher’ were grown together in a 15-cm-diameter pot in a greenhouse at day/night temperatures of 25/18°C with supplemental high-pressure sodium lighting. Plants were trimmed so that only two or three culms per plant remained. Flag leaves of all the plants within a pot were inoculated with a single pustule isolate, as described previously for the seedling inoculation. Inoculated plants were dried for over one hour to prevent cross-contamination and then incubated overnight in a dew chamber and grown in the greenhouse as described previously for seedling inoculation. Infection types were evaluated 14 d after inoculation. This subset of isolates was also tested on 12 additional ‘Thatcher’ near-isogenic lines at the seedling stage (Lr2b [RL6019], Lr3bg [RL6042], Lr14b [RL6006], Lr15 [RL6052], Lr19 [RL6040], Lr20 [RL6092], Lr23 [RL6012], Lr25 [RL6084], Lr28 [RL6079], Lr29 [RL6080], Lr32 [RL6086], and Lr52) and retested on Lr21 (RL6043). These isolates were also retested on seedling plants of the set of 16 ‘Thatcher’ near-isogenic lines mentioned previously to confirm their infection types, as many of these resistance genes (particularly Lr18 and LrB) are sensitive to temperature and other conditions. Inoculation, incubation, and rating were as described previously for seedling evaluation.

Results and discussion

Virulence on the standard seedling differential lines

Canada. From 2015 to 2019, 223, 255, 258, 225 and 235 isolates were analyzed annually, and from these isolates there were 47, 75, 44, 38 and 54 virulence phenotypes identified, respectively (Supplemental ). Many virulence phenotypes were found in multiple years, and overall there were 154 different virulence phenotypes found over these years (Supplemental ).

Table 1. Frequency (%) of predominant Puccinia triticina virulence phenotypes in Canada from 2002 to 2019

In 2015, leaf rust occurred at a low levels of severity and developed late in the growing season in Manitoba and Saskatchewan (McCallum et al. Citation2016b) and was also found in 23 of 33 fields surveyed in Ontario at a mean severity of 2.6% (Xue and Chen Citation2016). The most common virulence phenotypes in 2015 were MBDS (18.1%), TNBG (15.0%), and MLPS (8.4%) which were all relatively common in 2014 and previous years ().

In 2016, the wheat crop was generally seeded early in Manitoba and Saskatchewan and the leaf rust epidemic built up throughout the growing season to moderate levels of severity at the end of the growing season on untreated plots (McCallum et al. Citation2017a). In Ontario leaf rust was found at a severity of 1.0% among 10 infected fields from 26 surveyed (Xue et al. Citation2017). The most frequently isolated virulence phenotypes were MNPS (15.7%), MBDS (14.5%), and MPPS (7.1%). Of these MNPS increased significantly from 2.6% in 2015 to 15.7% in 2016, but was not common in the previous years ().

In 2017, the growing season was cool in May and June in Manitoba and Saskatchewan, but hot and relatively dry in July and August (McCallum et al. Citation2018a), leaf rust occurred commonly but only reached higher levels late in the growing season. In Ontario leaf rust was found in 15 of 32 fields surveyed with an average severity of 1.8% (Xue and Chen Citation2018), although in Quebec 2017 was a year with high incidence and severity (S. Rosa, unpublished). The most commonly identified virulence phenotypes in 2017 were MNPS (37.6%), TBBG (15.5%), and MBDS (7.0%). TBBG was common in previous years () however, this was one of a few virulence phenotypes that gained virulence to Lr21 in 2011 and therefore posed a risk to Canadian wheat cultivars with Lr21.

In 2018, wheat leaf rust was found at all locations surveyed throughout southern Manitoba and Saskatchewan with an average severity of 3.5% in Manitoba and 13.1% in eastern Saskatchewan in untreated trial plots (McCallum et al. Citation2019). The crop year in Eastern Canada was dry in 2018 and leaf rust was only found in two of 29 fields surveyed at a mean severity of 3.0% (Xue and Chen Citation2019). The most common virulence phenotypes were MNPS (29.8%), TBBG (28.9%) and TNBG (4.9%). TBBG and TNBG, along with TBBJ (3.6%) were virulent to Lr21 which led to 2018 having the highest level of virulence to Lr21 (39.9%) from 2014 to 2019 ().

Table 2. Frequency (%) of virulence to 17 Thatcher near isogenic wheat lines with different leaf rust resistance genes

In 2019, in Manitoba and Saskatchewan leaf rust was found in 67% of the winter wheat locations surveyed, at a severity level of trace to 5%, whereas in spring wheat it was found at 91% of the locations surveyed generally at trace levels (McCallum et al. Citation2020aa). In Ontario leaf rust was found in 16 of 38 fields surveyed with a mean disease severity of 1.8% (Xue and Chen Citation2020). The most frequently identified virulence phenotypes were MNPS (39.1%), MBDS (8.1%), and TBBJ (6.0%). The relatively lower level of T _ _ _ virulence phenotypes, many of which are virulent on Lr21, led to a decline in the level of Lr21 virulence between 2018 (39.9%) and 2019 (6.4%) ().

Alberta. Only a few isolates from Alberta were analyzed in 2015 and 2019, due to the fact that infected leaf samples were not sent from Alberta in most years, likely due to low infection levels. In 2015 there were three virulence phenotypes, MBDS (two isolates), TDBJ, and TDBS (one isolate each). Collections from Alberta have been rare in recent years, though in the past the population from Alberta was similar to that in Manitoba and Saskatchewan, and both MBDS and TDBJ were also found in Manitoba and Saskatchewan in 2015 but TDBS was not.

In 2019 there were eight virulence phenotypes found among ten isolates; MBDS and TNBJ (two isolates each), MFNS, MLPS, TMBJ, TNCJ, TNMJ and TNRJ (all one isolate each). Of these only MLPS was found in the larger sample from Manitoba and Saskatchewan in 2019. The source of inoculum for the 2019 Alberta population appeared to be different from that for Manitoba and Saskatchewan because of the differences between these populations.

Manitoba and Saskatchewan. There were 41, 60, 28, 17 and 19 virulence phenotypes found among 208, 233, 224, 184 and 159 isolates, in 2015 to 2019, respectively. In 2015 the most common virulence phenotypes were MBDS (17.8%), TNBG, (16.3%) and MLPS (9.1%) (). Of these MBDS has been a common virulence phenotype throughout 2000 to 2019 (), whereas TNBG was common from 2011 to 2018 and MLPS from 2013 to 2019. In 2014 the virulence phenotypes TBDG (17.9%), TBBG (14.7%), and MLPS and TNBG (both at 8.4%) (McCallum et al. Citation2020b) were the most frequent, though MBDS was found among 6.3% of the isolates, so the frequencies of the most common virulence phenotypes did not change dramatically. In the north central USA, bordering Manitoba and Saskatchewan, the most common virulence phenotypes found in 2015 were TBBGS (31.1%), TNBJJ (11.4%) and MBDSD (9.6%) (Kolmer and Hughes Citation2017), which is similar to Manitoba and Saskatchewan, as is normally the case since the inoculum moves northward from this region into Canada. The frequency of virulence to Lr21 increased from 2.1% in 2014 to 10.1% in 2015 (), similar to the levels found in 2012 and 2013 (). Similarly the frequency of virulence to Lr24 increased from 24.2% in 2014 to 48.6% in 2015 (), returning to similar levels found in previous years ().

Table 3. Frequency and distribution of virulence phenotypes of Puccinia triticina identified in Manitoba and Saskatchewan from 2015 to 2019 by infection types to selected resistance genes

Fig. 1 Frequency of virulence (%) from 2002–2019a in the Manitoba and Saskatchewan population of P. triticina to near-isogenic lines containing Lr2a, Lr9, Lr16, Lr24, Lr17 or Lr21.

Fig. 1 Frequency of virulence (%) from 2002–2019a in the Manitoba and Saskatchewan population of P. triticina to near-isogenic lines containing Lr2a, Lr9, Lr16, Lr24, Lr17 or Lr21.

In 2016, the most common virulence phenotypes were MNPS (17.2%), MBDS (15.9%) and MPPS (7.7%). MNPS was one of the most common virulence phenotypes from 2016 to 2019 () and MPPS, which is similar, was also common during these years. In the USA states that border Manitoba and Saskatchewan the most common virulence phenotypes in 2016 were MBDSD (20.8%), TNBJJ (15.0%) and MJBJG (8.3%) (Kolmer and Hughes Citation2018). While MBDS was also common in Manitoba and Saskatchewan, TNBJ (2.1%) and MJBJ (0.4%) were much less common than in the USA. The frequency of virulence to Lr21 declined again in 2016 (4.3%) compared to 2015 (10.1%) (), while the frequency of virulence to Lr16 (16.7%) was higher than previous or subsequent years ().

In 2017, MNPS (41.9%), TBBG (17.2%) and MBDS (7.5%) were the most common. TBBG was frequently found between 2010 to 2018 in this region. The most common virulence phenotypes found in the neighbouring USA states were TNBJJ (15.7%), TBBGS (13.4%) MPPSD and TNBGJ (both at 9.4%) (Kolmer Citation2019a). TNBJ, TNBG and MPPS were also found in Manitoba and Saskatchewan in 2017, though at lower levels of 2.2%, 3.1%, and 5.7% (). The frequency of virulence to Lr21 increased in 2017 to 25.6% compared with 4.3% in 2016 (, ) and the frequency of virulence to Lr9 (57.7%) reached its highest level in recent years ().

In 2018, the most common virulence phenotypes were MNPS (35.3%), TBBG (34.8%), and TNBG (5.9%), these were similar to 2017, TNBG was a common virulence phenotype from 2012 to 2018. However, the overall level of virulence to Lr21 increased significantly from 25.6% in 2017 to 48.7% in 2018 (). This was the highest level reached in recent years (). This was due to a number of similar virulence phenotypes that were virulent to Lr21 primarily TBBG, TBBJ, TBBQ, TBBS, TNBG, and TPBG (). Similarly the frequency of virulence to Lr2a in 2018 (51.9%) was higher than in the three previous years (, ).

In the neighbouring USA, the most common virulence phenotypes were MNPSD (39.8%), MBDSD (18.1%), MPPSD (12.0%), and TBBGS (10.8%) (Kolmer Citation2018). Virulence phenotypes MBDS and MPPS were also found in Manitoba and Saskatchewan, although at lower levels of 1.6% and 3.2%, respectively.

In 2019, the common virulence phenotypes were MNPS (54.7%), MBDS (10.7%), and TBBJ (8.2%). TBBJ was frequently found from 2001 to 2006 and then again in 2018 and 2019, and both MNPS and MBDS were also common in recent years (). The virulence frequencies for Lr2a, Lr2c, and Lr21 all declined significantly from 2018 (, ), due to the relatively higher frequency of M_ _ _ virulence phenotypes, which are avirulent to Lr2a, Lr2c, and Lr21, and lower frequency of T _ _ _ phenotypes, which are virulent to Lr2a and Lr2c, and many of which are virulent to Lr21. The frequencies of virulence increased in 2019 for Lr9, Lr17, and Lr24 compared with 2018 (, ), with the frequency of virulence to Lr9 at 75.5%, the highest in recent years (). Virulence phenotypes MNPSD (37.1%), TBBGS (22.6%), and MPPS (16.1%) were the most frequently found in the neighbouring USA states (Kolmer Citation2019b).

The most popular wheat cultivars in Manitoba and Saskatchewan over the years 2015–2019 were ‘AAC Brandon’ and ‘Stettler’ (https://www.grainscanada.gc.ca/en/grain-research/statistics/varieties-by-acreage/). ‘AAC Brandon’ was reported to have molecular markers indicating the presence of Lr16, Lr21, and Lr34 (Toth et al. Citation2018) while ‘Stettler’ is thought to have Lr2a (Bokore et al. Citation2020) and other resistance genes. While Lr16 and Lr34 remained effective at reducing leaf rust during these years, Lr21 has become less effective and the frequency of virulence to Lr2a was generally high. Selection on ‘AAC Brandon’ and other cultivars carrying Lr21 could have led to higher levels of virulence on Lr21 in Manitoba and Saskatchewan.

Ontario. In 2015 there were nine virulence phenotypes found from 12 isolates from Ontario (). The most common being MBTN (three isolates) and MBDS (two isolates), both of these were also found in Manitoba and Saskatchewan in 2015. MBTN was not found in Ontario in 2014, though the sample size was very small (seven isolates), however MBDS was found in this 2014 sample from Ontario. MBTN was also the most commonly found virulence phenotype in Ontario in 2010, 2011, 2012 and 2013 (McCallum et al. Citation2020c, Citation2017b, Citation2018b, Citation2020c). Similarly in the USA states south of Ontario the most common virulence phenotypes in 2015 were MBTNB (50%) and MCTNB (25.9%) (Kolmer and Hughes Citation2017). Virulence to Lr21 was detected for the first time in Ontario in 2015 (9.5%), although this was represented by only a single isolate of virulence phenotype TDBJ.

Table 4. Frequency and distribution of virulence phenotypes of Puccinia triticina identified in Ontario from 2015 to 2019 by infection types to selected resistance genes

In 2016, there were 17 virulence phenotypes found among 22 isolates with the most frequent being MCQQ (13.6%), MBTN, PCJQ, and TCJJ (9.1% each), of these MBTN was found in previous years in Ontario (). Of these 17 virulence phenotypes only four were found in the larger sample from Manitoba and Saskatchewan, demonstrating the difference between the population found in Ontario from that in Manitoba and Saskatchewan. In the neighbouring USA, the most frequently found virulence phenotypes were MBTNB (72.9%), and MCTNB (8.5%) (Kolmer and Hughes Citation2018). The frequency of virulence to Lr21 increased from 2015 (9.5%) to 2016 (13.6%) although the samples sizes for both years were relatively small.

In 2017, there were 12 virulence phenotypes found among 21 isolates with the most common being MBTN (33.3%) and PBDG (19.0%) (). Of these only three were also found in Manitoba and Saskatchewan. In the Ohio Valley area of the USA the most common virulence phenotypes were MBTNB (44.7%) and MCTNB (18.4%) and all three of the isolates tested from the northeastern states were PBDGG (Kolmer Citation2019a).

There were 17 virulence phenotypes found among 22 isolates tested from 2018 (). The most common were TFPJ (13.6%) and MCQK, TBJS, and TCGJ (9.1% each), of these TFPJ and TBJS have been found previously in Ontario. Only one of these virulence phenotypes was found among the larger sample from Manitoba and Saskatchewan, demonstrating the differentiation of the Ontario population from that in western Canada. In 2018 all isolates analyzed from northeast USA came from New York State and they were TCJBB (75%) and TFTSB (25%) (Kolmer Citation2018).

Thirty virulence phenotypes were found from 55 isolates tested in 2019, with the most common being MCTN (14.5%) and MBTN (10.9%) (), both of these virulence phenotypes were found previously in Ontario. However, only ten of these virulence phenotypes were found in the larger sample from Manitoba and Saskatchewan. In the Ohio Valley area the most commonly found virulence phenotypes in 2019 were MBTNB (40%), and MCTNB (16%) (Kolmer Citation2019b). This was very similar to those found in Ontario.

Ontario differed from Manitoba and Saskatchewan in the virulence phenotypes that were found but also in the frequencies of virulence to some resistance genes. The frequency of virulence to Lr18 was higher in Ontario than in Manitoba and Saskatchewan whereas that for Lr21 was lower.

Quebec. There were just three isolates analyzed from Quebec in 2015, two of these were TBBG and one was MLDS, both of which have been found previously in Quebec (). There were no isolates in 2016, but in 2017 ten virulence phenotypes were found among 13 isolates (). The most common of these were TCGJ (three isolates) and MBTN (two isolates), neither of which were found in Quebec recently. MBTNB (44.7%) was the most common virulence phenotype in the Ohio Valley region in 2017.

Table 5. Frequency and distribution of virulence phenotypes of Puccinia triticina identified in Quebec from 2015 to 2019 by infection types to selected resistance genes

In 2018, nine virulence phenotypes were found among 19 isolates and the most frequently found were MCQH (26.3%) and FCPT (21.1%), which were not found recently in Quebec. Of these nine virulence phenotypes five were only found in Quebec in 2018.

In 2019, nine virulence phenotypes were found among 11 isolates and the most common were MBTN an MCRS (two isolates each). MBTN was found in Quebec in 2017 but MCRS was not found recently. Of these nine virulence phenotypes only three were unique to Quebec. In New York state in 2018 there were just two virulence phenotypes found TCJBB (75%) and TFTSB (25%) (Kolmer Citation2018).

Virulence on additional seedling and adult plant differential lines

For each year, a sub-set of isolates was selected to represent the range of virulence phenotypes found on the standard set of differential lines. These annual sub-sets were then tested on additional Thatcher near isogenic lines, some at the seedling stage (Lr2b, Lr3bg, Lr14b, Lr15, Lr19, Lr20, Lr23, Lr25, Lr28, Lr29, Lr32 and Lr52) and some at the adult plant stage (Lr12, Lr13, Lr22a, Lr35, Lr37). There were 56, 79, 52, 41 and 56 isolates tested on these additional wheat lines from 2015 to 2019, respectively. No virulence was detected to Lr19, Lr29, Lr32, Lr52, or Lr22a (). There was no virulence detected to Lr25 until 2018 and 2019 where it was found only at low levels of 2.4% and 7.1%, respectively. Alternatively the frequencies of virulence were very high to Lr14b, Lr15, Lr20, Lr12, and Lr13 (). Virulence to Lr37 was initially low in 2015 (12.5%), but subsequently was at high levels, near 70%, from 2016 to 2019. The frequencies of virulence to Lr2b, Lr3bg, Lr23, Lr28, and Lr35 were intermediate and did not show strong directional changes from 2015 to 2019 ().

Table 6. Frequency (%) of virulence for a sub-set of isolates selected to represent the virulence phenotypes found annually to 17 additional Thatcher near isogenic wheat lines with different resistance genes to leaf rust

Summary

Overall the P. triticina population found in Canada in 2015 was different from that found in 2014 and previous years. The most commonly found virulence phenotypes shifted from those found in 2014 () and there were also 11 new virulence phenotypes (FBDJ, MJMS, MNPS, MRPS, TBBP, TBTJ, TDTJ, TFKD, TFTJ, TPBQ, and SBDB), which had not been found during the years 1995 to 2014, which is a relatively high number. Virulence frequencies to many key genes increased in 2015 compared to 2014 including Lr9, Lr24, Lr26, Lr3ka, Lr17, Lr30, LrB, Lr10, Lr14a and Lr21 (). Common genes in Canadian wheat germplasm include Lr1, Lr2a, Lr10, Lr13, Lr14a, Lr16, Lr21, and Lr34 (McCallum et al. Citation2016a; Toth et al. Citation2018). Increases in virulence to Lr21 could result in some cultivars with Lr21 becoming more susceptible, and that has been seen in the cultivars ‘McKenzie’ and ‘Glenn’ (B.McCallum unpublished), though other Canadian cultivars such as ‘Vesper’, ‘Cardale’, ‘AAC Warman’ also have Lr21 (Toth et al. Citation2018) and have not yet become susceptible, likely due to the presence of other resistance genes.

The P. triticina population found in Canada was fairly diverse in 2016 with 75 virulence phenotypes found among the 255 isolates tested. The predominant virulence phenotypes found in 2016 were similar to 2015, though the frequencies of virulence changed on many key resistance genes, there were decreases in the frequencies of virulence to Lr2a, Lr2c, and Lr21, whereas there were increases in the frequencies of virulence to Lr16, Lr24, Lr26, Lr3ka, Lr11, Lr17, Lr30, Lr10 and Lr14a.

In 2017, 44 virulence phenotypes were identified among 258 isolates. The most frequent virulence phenotypes MNPS, avirulent to Lr21, and TBBG, virulent to Lr21, increased significantly from 2016, and the overall frequency of virulence to Lr21 increased from 6.7% in 2016 to 23.4% in 2017 (). Compared to 2016 there were also small increases in virulence to Lr2a, Lr2c, Lr9, Lr3ka, and Lr30 and decreases in virulence to Lr16, LrB and Lr14a ().

In 2018, the frequency of virulence to Lr21 reached its highest level (39.9%) (), as many of the most common virulence phenotypes were virulent to Lr21, particularly TBBG (28.9%) and TNBG (4.9%), and TBBJ (3.6%) (, Supplemental ). Compared with 2017 there were also increases in virulence to Lr2a, Lr2c, Lr26, and Lr18, whereas there were decreases in virulence to Lr9, Lr24, Lr3ka, Lr17, Lr30, LrB and Lr14a.

In 2019 the most frequently identified virulence phenotypes were MNPS (39.1%), MBDS (8.1%), and TBBJ (6.0%), the relatively lower level of T _ _ _ virulence phenotypes, many of which are virulent on Lr21, led to a decline in the level of Lr21 virulence between 2018 (39.9%) and 2019 (6.4%). There were also decreases to the frequencies of virulence to Lr2a, Lr2c, Lr10, and Lr18, whereas there were increases in virulence to Lr9, Lr24, Lr3ka, Lr11, Lr17, Lr30, LrB, and Lr14a.

Changes in the virulence phenotypes and frequencies of virulence found in Canada can result from changes in the P. triticina population arriving annually from the neighbouring USA states as well as selection in Canada. Generally the shifts in the Canadian P. triticina population follow those observed in the neighbouring USA states. This is reflected in the differences found between the populations found in Manitoba and Saskatchewan and those from Ontario and Quebec as their sources of initial inoculum from the USA are different. The wheat cultivars grown in western Canada are also different from those grown in Ontario and Quebec. The rise in the frequency of Lr21 virulence could reflect its frequent use in wheat cultivars on both sides of this border.

Supplemetary material

Supplemental data for this article can be accessed online here: https://doi.org/10.1080/07060661.2021.1888156.

Acknowledgements

We thank Francis Kirigwi and Harpinder Randhawa, who sent samples for analysis, and other cooperators who grew trap rust nurseries.

References