Anti-resistance strategies for fungicides against wheat pathogen Zymoseptoria tritici with focus on DMI fungicides

Crop Protection 99 (2017) 108e117

Contents lists available at ScienceDirect

Crop Protection journal homepage: www.elsevier.com/locate/cropro

Anti-resistance strategies for fungicides against wheat pathogen Zymoseptoria tritici with focus on DMI fungicides Thies Marten Heick*, Annemarie Fejer Justesen, Lise Nistrup Jørgensen Aarhus University, Department of Agroecology, Flakkebjerg, Forsøgsvej 1, 4200 Slagelse, Denmark

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 November 2016 Received in revised form 5 May 2017 Accepted 7 May 2017

Septoria tritici blotch (STB) caused by the fungal pathogen Zymoseptoria tritici is a global threat to sustainable wheat production. Applications of fungicide against STB are regarded as an essential means to minimise yield losses, however, fungicide resistance is developing and affecting fungicide efficacy greatly. Only a few fungicide classes are available for STB control. DMI fungicides are seen as the main group, but increasing problems with resistance is challenging their use. In 2015 and 2016, field trials were conducted testing fungicide spray strategies using one, two or three applications, alternations, mixtures of different DMIs and DMI mixed with other modes of action including a SDHI and a multi-site inhibitor. The strategies were tested for disease control, their impact on yield and their effect on resistance build-up, measured as CYP51 alterations responsible for decreased sensitivity and efficacy of DMIs. Strategies consisting of three treatments provided adequate control of STB and significant yield increases. The best results with regard to yield and control were attained by a diversified DMI strategy also including the SDHI boscalid and the multi-site inhibitor folpet. Spraying once or twice lowered selection yet compromised STB control and yield. CYP51 alteration I381V was the most predominated in all samples. Frequencies of alterations D134G, V136A and S524T increased significantly following applications with DMI fungicides. The more diversified a strategy, the less it selected for CYP51 alterations. EC50 values for epoxiconazole showed a tendency to be higher post-treatment in 2015. The results presented in this study encourage the adoption of mixing and alternating fungicides into spray strategies to minimise the risk of resistance build-up and to prolong the effective life of fungicides. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Bacillus subtilis CYP51 Disease control Folpet Mycophaerella graminicola SDHI Selection

1. Introduction Winter wheat (Triticum aestivum) is one of the major cereal crops in Europe. In 2014, the crop was cultivated on approximately 27 million hectares, achieving yields of seven to nine tonnes per hectare in highly intensified cropping systems, as known in northwestern Europe (Anonymous, 2016b). An essential factor ensuring high yields is well-timed disease management throughout the growing season. For many years Parastagonospora nodorum was the dominant disease in winter wheat in Europe (Bearchell et al., 2005). Since the early 1980s, the ascomycete Zymoseptoria tritici (Z. tritici) causing septoria tritici blotch (STB) has taken over and is now regarded as the most important disease in wheat (O'Driscoll et al., 2014) with yield losses in the range of 10e50% depending on the region and yearly disease pressure. Several agronomical practices

* Corresponding author. E-mail address: [email protected] (T.M. Heick). http://dx.doi.org/10.1016/j.cropro.2017.05.009 0261-2194/© 2017 Elsevier Ltd. All rights reserved.

e.g. late sowing and planting of resistant cultivars (Thomas et al., 1989; Gladders et al., 2001; Brown et al., 2015) have shown some potential keeping disease levels low and minimising epidemics. In most years, however, STB management is heavily reliant on frequent fungicide applications (O'Driscoll et al., 2014). In north-western Europe, two to four fungicide sprays are commonly applied to winter wheat for the control of fungal diseases such as STB (Jørgensen et al., 2008; Fones and Gurr, 2015). For disease management of STB in winter wheat, compounds belonging to four fungicide classes are available in Europe: (1) quinone outside inhibitors (QoIs), (2) sterol 14a-demethylation inhibitors (DMIs), (3) succinate dehydrogenase inhibitors (SDHIs) and (4) multi-site inhibitors. For many years now, DMI fungicides epoxiconazole and prothioconazole have been the most widely used fungicides for STB control throughout Europe. In recent years, declining field efficacies have been observed for both fungicides (Anonymous, 2014; Wieczorek et al., 2016) which raises major concern of a total loss of efficacy of this important fungicide class.

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

Fungicide resistance to DMIs has been associated with three molecular mechanisms: (1) mutations in the DMI target gene CYP51 causing amino acid alterations in the CYP51 enzyme (Cools and Fraaije, 2013), (2) overexpression of the target gene CYP51 (Chassot et al., 2008; Cools et al., 2012) and (3) an enhanced efflux of the cell (Omrane et al., 2015). Some of these CYP51 alterations have shown to affect the performance of specific DMIs e.g. I381V for tebuconazole (Fraaije et al., 2007); others show a more broad effect on the efficacy of several DMIs (Mullins et al., 2011). All three mechanisms can occur in combination in a single strain at the same time, rendering it difficult to control (Kirikyali et al., 2017). In the northern European Z. tritici population, alterations of the CYP51 contribute the most to reduced sensitivity to DMIs, whereas overexpression and enhanced efflux are rare (Heick et al., 2017). The rapid development and spread of resistant strains are a real threat to wheat production, however, yearly efficacy testing of fungicide compounds carried out in most European countries shows that DMI fungicides still provide significant control of STB in the field (Jørgensen et al., 2017). The evolution of fungicide resistance in Z. tritici poses a severe threat to profitability and sustainability of wheat production worldwide. Numerous studies have identified the major drivers for fungicide resistance and how to mitigate them in anti-fungicideresistance management strategies in theory (van den Bosch et al., 2014; Hobbelen et al. 2010, 2011a, 2013, 2014, 2011b, van den Bosch et al., 2011; Grimmer et al., 2014). Key principles of antiresistance strategies without compromising disease control include the optimal application timing, the adjustment of the number of applications, the appropriate dose and the application of fungicide mixtures and/or alterations of active ingredients with different modes of action (MoAs) (van den Bosch et al., 2014; Dooley et al., 2016b; van den Berg et al., 2013; Mavroeidi and Shaw, 2006). All principles should be taken into account when designing a practicable anti-resistance strategy. In addition, the sensitivity status of the pathogen population is another important factor that can influence the success of any strategy. Van den Bosch et al. (2011) defined three phases of fungicide resistance evolution: an initial ‘emergence phase’, in which a resistant strain arises for the first time by spontaneous mutation or invasion from another population. Followed by a ‘selection phase’, in which the resistant strain is present in the population and increases in frequency over time due to selection pressure imposed by fungicide applications. In the final ‘adjustment phase’, the resistant strain has established itself and accounts for a large proportion of the population. According to the authors, anti-resistance strategy can only be successful when employed in the ‘selection phase’; a prevention of the rise of a resistant strain is impossible and once established in a population, resistant strains can only be managed to some degree by agronomical practices (van den Bosch et al., 2011). There is an urge to extend the effective life of those active ingredients that are available at the moment. Further decline in efficacy, new restricted criteria for the registration of new active ingredients imposed by authorities (endocrine disruptors; hazard

109

vs. risk) and the development of highly resistant pathogens will constrain the current situation further (Hillocks, 2012; Jess et al., 2014). There is a general lack of applied studies showing the effect of anti-fungicide-resistance strategies in practice, and only a few studies have been published measuring the effect of spray strategies in the field (Dooley et al., 2016b, 2016c). In this work we use the Danish disease management situation, i.e. a limited armoury of fungicides and a moderately DMI-resistant Z. tritici population (Heick et al., 2017), as starting point to assess the following hypotheses: (1) a limited spectrum of fungicides provides good control against Z. tritici; (2) Diversified spray strategies, i.e. using fungicide mixtures and alteration of fungicides, counteract the selection of CYP51 alterations, and thus selection for DMI insensitivity. To test these hypotheses, field trials were carried out testing commercially available fungicides including DMIs, a SDHI, a multisite inhibitor and the biofungicide Bacillus subtilis (B. subtilis), in different combinations and at different sites. The effect of the different fungicide strategies on CYP51 alterations was assessed by pyrosequencing and qPCR. 2. Material and methods 2.1. Trial design and fungicide application A total of four field trials was conducted during 2014/15 and 2015/16 at two locations in Denmark (Table 1). All trials were laid out as complete randomised block design with four replicates containing nine treatments and an untreated control. Plot size was 12.5e22 m2. Table 2 gives an overview of the six products and their active ingredients used in this study. Table 3 shows the different treatments and their timings; growth stage (GS) 31e32, GS 37e39 and GS 59e65 (Zadoks et al., 1974). Application rates were half the label rate, as commonly recommended in Denmark. All fungicides were applied using a plot sprayer in 150 L ha
1 water at low pressure with flat fan nozzles. With respect to other management issues the crop was treated using standard cultural practices. 2.2. Disease and yield assessments Zymoseptoria tritici developed naturally at all sites and was the dominating foliar disease in 2015 and 2016. In both years, disease severity was medium to severe. Field trials at the Flakkebjerg site were irrigated to avoid drought and to promote disease development during a dry period from May to June in both years. Foliar diseases were assessed several times during the season as per cent diseased leaf area on flag leaf and 2nd leaf. Disease assessments at GS 75 and GS 77, the latter as per cent green leaf area (GLA), are regarded as the key assessments for statistical analyses. No other diseases were present and influenced the results in the four trials. All trials were harvested and yield (adjusted to 15% moisture content) and yield increase (both dt ha
1) were measured.

Table 1 Trial site details, cultivar information, dates and growth stage (GS), at which fungicide applications and disease assessments took place. Site

Year

Cultivar

Susceptibility

Coordinates (lat, lon)

Flakkebjerg Hadsten Flakkebjerg Horsens

55.326724, 11.390078 2014/15 Hereford very susceptible 56.323629, 10.101383 2014/15 Nakskov moderate 55.331065, 11.384772 2015/16 Hereford very susceptible 55.858976, 2015/16 Hereford very susceptible 9.756632

a

www.sortsinfo.dk.

Attack of STB First application Second application Third application Disease assessment

to STBa

Name

moderate high high high

Date

GS Date

GS

Date

GS

Date

GS

24-04-2015 22-04-2015 04-05-2016 01-05-2016

31 31 32 32

33 37 37 37

04-06-2015 16-06-2015 07-06-2016 06-06-2016

51 59 65 55

11-07-2015 16-07-2015 09-07-2016 07-07-2016

75 75 75 75

06-05-2015 22-05-2015 19-05-2016 18-05-2016

110

T.M. Heick et al. / Crop Protection 99 (2017) 108e117 Table 2 Fungicide products used in the trials. Product name

Company

Active ingredients

Armure 300 EC® Bell® Folpan 500 SC® Proline EC 250® Prosaro 250 EC® Serenade ASO®

Syngenta Nordics A/S BASF A/S Adama Northern Europe B.V. Bayer A/S Bayer A/S Bayer A/S

150 g L
1 difenoconazole þ 150 g L
1 propiconazole 67 g L
1 epoxiconazole þ 233 g L
1 boscalid 500 g L
1 folpet 250 g L
1 prothioconazole 125 g L
1 tebuconazole þ 125 g L
1 prothioconazole Bacillus subtilis strain QST 713 - 1012 cfu L
1

Table 3 Fungicide application patterns at timings GS 31e32, GS 37e39 and GS 59e65, doses rates, total amount of a.i. (g) applied. Treatment

1 2 3 4 5 6 7 8 9 10

GS 31-32

Untreated prothioconazole prothioconazole prothioconazole prothioconazole prothioconazole þ folpet folpet Bacillus subtilis þ prothioconazole e e

Dose

GS 37-39

L ha
1

a.i. (g)

e 0.4 0.4 0.4 0.4 0.4 þ 1.0

e 100 100 100 100 100/500

1.5 2.0 þ 0.4

750 100

e e

e e

e prothioconazole epoxiconazole þ boscalid epoxiconazole þ boscalid epoxiconazole þ boscalid epoxiconazole þ boscalid þ folpet epoxiconazole þ boscalid Bacillus subtilis þ epoxiconazole þ boscalid epoxiconazole þ boscalid epoxiconazole þ boscalid

2.3. Analyses of amino acid alterations associated with fungicide resistance At each site and for each plot, leaf samples consisting of 20 flag leaves were collected at GS 75 to determine CYP51 alterations, Sdh alterations and QoI mutation G143A in the cytochrome b complex of the post-treatment Z. tritici population. Leaves were dried at room temperature and stored until further use. Treatment samples were bulked for each trial site, i.e. leaves from all four replicates from one site were regarded as one sample. Leaves were cut into three cm pieces and ground to powder in the presence of ten steel balls (ø 5 mm) using a Geno/Grinder® 2010 (Spex®SamplePrep, Stanmore, U.K.) for 5  2 min at 1,500 rpm. Genomic DNA was extracted from a total of 30 mg of pulverised leaf/fungus material using Qiagen DNeasy Plant Mini extraction kit (Qiagen GmbH, Hildesheim, Germany) and eluted in 100 ml elution buffer, according to the producer's protocol. Leaf samples were analysed at BASF's laboratory facilities at Limburgerhof, Germany. Pyrosequencing as described by Stammler et al. (2008) was used to assess the frequency of CYP51 alterations D134G, V136A/C, Y137F, A379G, I381V and alterations at amino acid position 459e461 (deletion Y459, deletion Y460, Y459C/D/S and Y461H/S). In the Sdh subunit c, alterations C-N86K and C-G90R were investigated and C-H152R only in 2016 (Stammler, pers. comm.). Briefly, forward primers of each primer pair were biotinylated at the 50 end. Standard PCR reactions amplifying the corresponding fragments were performed using Maxima Hot Start PCR Master Mix (Thermo Scientific, Darmstadt, Germany), according to the following programme: initial heating at 95  C for 15 s followed by 40 cycles of 15 s denaturation at 94  C, 30 s annealing at 72  C, and final elongation step at 72  C for 5 min. Every reaction was performed in duplicate. The PCR fragments were immobilised with Streptavidin Sepharose beads (GE Healthcare, Chalfont St Giles, UK), cleaned in 70% ethanol, denatured in 0.2 M sodium hydroxide and finally washed in 10 mM tris-acetate. These steps were carried out using a Vacuum Prep Worktable (Qiagen GmbH, Hildesheim, Germany). The single-stranded DNA

Dose

GS 59-65

L ha
1

a.i. (g)

e 0.4 0.5 0.5 0.5 0.5 þ 1.0

e 100 33.5/116.5 33.5/116.5 33.5/116.5 33.5/116.5/500

0.5 2.0 þ 0.5

33.5/116.5 33.5/116.5

0.5 1.0

33.5/116.5 67/233

Dose L ha
1

a.i. (g)

e prothioconazole prothioconazole tebuconazole þ prothioconazole difenoconazole þ propiconazole tebuconazole þ prothioconazole

e 0.4 0.4 0.5 0.4 0.5

e 100 100 62.5/62.5 60/60 62.5/62.5

tebuconazole þ prothioconazole Bacillus subtilis þ tebuconazole þ prothioconazole tebuconazole þ prothioconazole e

0.5 2.0 þ 0.5

62.5/62.5 62.5/62.5

0.5 e

62.5/62.5 e

were mixed with an annealing buffer and the corresponding sequencing primer, and incubated at 80  C for 3 min to avoid any loop formation. Pyrosequencing was performed on a PSG 96MA machine using PyroMark®Gold Q96 reagents (all Qiagen GmbH, Hildesheim, Germany). The frequencies of alterations S524T in the CYP51 (Stammler and Semar 2011), G143A in the cyt b (Taher et al., 2014) and C-T79N in the SdhC (Stammler, pers. comm.) were detected by qPCR on a Rotor-Gene-Q (Qiagen, GmbH, Hildesheim, Germany) using Takyon™ No Rox Probe MasterMix dTTP (Eurogentec, Seraing, Belgium). 2.4. Fungal isolates and in vitro sensitivity test Single pycnidium isolates were produced from leaf samples of treatments 1, 2, 4, 6 and 10; it was aimed at producing ten isolates per site, year and treatment. Leaves were kept on moistened filter paper in petri dishes at high humidity for 24 h. Using a sterile needle, cirrhi from single pycnidium were transferred onto Potato Dextrose Agar supplemented with 0.01% streptomycin, and incubated at 20  C in 12 h white light/12 h darkness for five days. A total of 167 Z. tritici isolates were produced and approximately six to ten isolates were tested per treatment and site. In order to test the sensitivity of isolates to the DMI epoxiconazole, spore suspensions were produced by scraping off six-day-old Z. tritici spores and transferring them into 10 mL of sterile, demineralised water. Spore suspensions were subsequently vortexed and adjusted to 2.5  104 spores mL
1. Aliquots of 100 ml spore suspension and 100 ml fungicide solution were added to nunc™ 96-deep well microtitre plate (ThermoFisher, Roskilde, Denmark). Epoxiconazole was mixed separately with 2  Potato Dextrose Broth to obtain the following final microtitre plate fungicide concentrations (mg L
1): 10, 3.3, 1.0, 0.3, 0.1, 0.03, 0.01, 0. Technical duplicates of each isolates were performed on the same plate and the Dutch isolate IPO323 (DMI-sensitive) and Irish isolate OP15.1 (moderately DMI-resistant) were included as references. Microtitre plates were wrapped in aluminium foil and incubated in the dark at 20  C for six days. Plates

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

were visually checked for bacterial or fungal contamination prior to the analysis in an iMark™ Microplate Absorbance Reader (Bio-Rad, Copenhagen, Denmark) at wavelength 620 nm. Fungicide sensitivities were calculated as the concentration of a fungicidal compound, at which fungal growth in vitro is inhibited by 50% (EC50) by non-linear regression (curve fit) using GraphPad Prism (GraphPad software, La Jolla, CA, USA). 2.5. Statistical analysis Disease assessments, yield increase, effects of treatments and difference in CYP51 alteration frequencies were analysed across years by an ANOVA Grand Means using ARM 2016.4. Summary Across Trials software (Gylling Data Management, Inc., Brookings, SD, USA). EC50 values were analysed by R software (RCoreTeam, 2014), assuming a gamma distribution. Green leaf area (GLA) was plotted against yield increase and ‘selection index’. ‘Selection index’ is defined as sum of frequencies of alterations D134G, V136A/C and S524T, and gives an indication of the extent of selection imposed by different applications of fungicides. 3. Results 3.1. Disease and yield assessments STB epidemics were moderate to high in both years. On average STB attacks were 73% on flag leaf and 93% on 2nd leaf assessed at GS 75e77 in the untreated control. All fungicide treatments reduced attacks significantly compared to the untreated control (Table 4), however significant differences in levels of reduction were observed between the different treatments. Treatments 2e5, which comprised three applications of DMIs in various combinations, reduced STB attacks similarly. The addition of folpet (treatment 6) at GS 31e32 and GS 37e39 improved disease control on 2nd leaf. However, when folpet was applied as a solo product at GS 31e32 STB attacks were significantly higher. Applying fungicides only one or two times resulted in less effective disease control. Addition of the biofungicide B. subtilis to the spray programme

111

reduced control compared to the same treatment without B. subtilis (treatment 4 vs. 8). Average yield of the untreated control across all four trials was 90.2 dt ha
1. Applications of fungicides led to significantly increased yields. Plots treated with the most diversified fungicide strategies (treatment 5 and 6) achieved on average the highest yields in both years. Treatments, which included only one or two sprays as well as combinations with B. subtilis, yielded significantly less compared to other treated entries. Fungicide application resulted in yield increases varying from 8.3 dt ha
1 when treated one time at GS 37e39 (treatment 10) and 17.1 dt ha
1 (treatment 6). Major part of the yield increase was linked to higher thousand grain yields (data not shown). 3.2. Analyses of mutations associated with fungicide resistance Pyrosequencing of bulked Z. tritici samples showed significant differences between the Z. tritici populations in Flakkebjerg and Jutland (Hadsten and Horsens); the Flakkebjerg populations showed lower frequencies of CYP51 alterations D134G and V136A compared to localities in Jutland (Table 5). Alterations D134G, V136A and S524T were significantly affected by the application of fungicides compared to the untreated control. In both years, treatment 2 (3  prothioconazole) caused the highest changes in frequencies of the alterations. Diversification of the spray strategies led to less selection (treatment 3e5) of those alterations. V136C occurred at a low level (3e11%) and was mainly unaffected by treatments. Alteration I381V remained at a high level and was not significantly influenced by the treatments; the frequency of A379G was generally low and was significantly selected for by treatment 5. Whereas I381V was present between 81% and 100% in all treatments regardless of year or location, the amounts of A379G were lower in 2016 than in 2015. Frequencies of D134G, V136A/C and S524T, on the other hand were significantly increased in 2016 compared to 2015 (Table 5). Table 6 illustrates the effect of treatments 2e10 on CYP51 alterations. The classification summarises alteration data of all four field trials and is done by significant statistical difference of the

Table 4 Averaged disease severity (measured as average per cent STB on flag leaf and 2nd leaf) at GS 75e77 and average yield (dt ha
1) and relative yield (per cent) of four field trials carried out in 2015 and 2016. Treatments within the same category with the same letter do not significantly differ (p ¼ 0.05). Treatment

Active ingredient (ha
1)

STB attack (%)a

Efficacy

Yield

Relative yield

GS 31-32

GS 37-39

GS 59-65

2nd leaf

flag leaf

%b

dt ha
1

%

1 2 3

e 100 g prothioconazole 100 g prothioconazole

e 100 g prothioconazole 100 g prothioconazole

96 a 58 e 59 e

73 a 25 cd 21 de

e 66 71

90.2 f 103.6 bc 104.4 b

100 115 116

4

100 g prothioconazole

21 de

71

104.1 b

115

100 g prothioconazole

54 e

16 e

78

105.6 ab

117

6

100 g prothioconazole 500 g folpet

62.5 g tebuconazole 62.5 g prothioconazole 60 g difenoconazole 60 g propiconazole 62.5 g tebuconazole 62.5 g prothioconazole

54 e

5

42 f

16 e

78

107.4 a

119

7

750 g folpet

72 cd

32 c

56

100.2 de

111

8

100 g prothioconazole Bacillus subtilisc

69 d

27 cd

63

101.9 cd

113

9

e

80 bc

40 b

45

101.2 d

112

10

e

e 100 g prothioconazole 33.5 g epoxiconazole 116.5 g boscalid 33.5 g epoxiconazole 116.5 g boscalid 33.5 g epoxiconazole 116.5 g boscalid 33.5 g epoxiconazole 116.5 g boscalid 500 g folpet 33.5 g epoxiconazole 116.5 g boscalid 33.5 g epoxiconazole 116.5 g boscalid Bacillus subtilisc 33.5 g epoxiconazole 116.5 g boscalid 67 g epoxiconazole 233 g boscalid

86 b

46 b

37

98.5 e

109

a b c

Disease assessments were carried out at GS 75. Efficacy (%) on flag leaf at GS 75. B. subtilis 2 L ha
1.

62.5 g tebuconazole 62.5 g prothioconazole 62.5 g tebuconazole 62.5 g prothioconazole Bacillus subtilisc 62.5 g tebuconazole 62.5 g prothioconazole e

112

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

Table 5 Z.tritici CYP51 alteration frequencies (%) post-treatment in four field trials in 2015 and 2016.

Treatment Flakkebjerg 2015

Hadsten 2015

D134G

V136A

V136C

A379G

I381V

S524T

D134G

V136A

V136C

A379G

I381V

1

0%

16%

6%

49%

94%

2%

19%

30%

5%

37%

92%

S524T 1%

2

43%

54%

8%

23%

81%

7%

68%

78%

0%

13%

96%

20%

3

22%

41%

5%

26%

85%

10%

48%

56%

5%

21%

97%

10%

4

24%

28%

5%

43%

98%

1%

29%

48%

11%

32%

94%

15%

5

0%

23%

0%

55%

91%

1%

19%

36%

0%

52%

88%

2%

6

21%

24%

8%

48%

96%

6%

30%

39%

0%

24%

98%

5%

7

0%

14%

11%

42%

91%

4%

0%

17%

0%

52%

96%

5%

8

14%

21%

5%

42%

96%

1%

19%

24%

0%

39%

100%

3%

9

14%

19%

8%

48%

94%

2%

19%

24%

10%

34%

92%

2%

10

20%

27%

7%

38%

92%

1%

11%

24%

6%

50%

90%

2%

S524T

Horsens 2016

Treatment Flakkebjerg 2016 D134G

V136A

V136C

A379G

I381V

S524T

D134G

V136A

V136C

A379G

I381V

1

24%

39%

3%

29%

90%

4%

32%

53%

11%

31%

98%

9%

2

45%

58%

6%

15%

97%

15%

67%

70%

9%

11%

98%

20%

3

43%

52%

11%

22%

94%

11%

56%

71%

14%

15%

95%

21%

4

29%

38%

8%

31%

99%

9%

44%

54%

10%

21%

100%

22%

5

20%

24%

5%

50%

97%

4%

30%

44%

11%

35%

98%

15%

6

18%

31%

12%

31%

93%

11%

35%

48%

15%

23%

100%

10%

7

20%

27%

13%

42%

98%

12%

42%

48%

9%

30%

99%

10%

8

23%

36%

7%

37%

98%

7%

41%

54%

12%

23%

96%

15%

9

17%

28%

7%

35%

95%

7%

20%

37%

11%

34%

98%

10%

10

30%

35%

10%

34%

94%

1%

40%

49%

9%

31%

94%

8%

treatments across locations and years compared to the untreated control (4). In general, CYP51 alternations were most selected for following treatments 2, 3 and 4, compared to in the untreated control; the more diversified the spray strategy (treatments 5), the less it selected for CYP51 alterations. The addition of folpet (treatment 6) at GS 31e32 and GS 37e39 had a slight effect on CYP51 alteration compared to treatment 4, which contained the same DMIs. Treatment 7 selected only moderately for A379G, but had no effect on the other alterations tested except for V136A. Treatment 8, which included B. subtilis, and treatments 9 and 10, both of which contained only one or two DMI applications, selected overall less than treatments 2, 3 and 4. The presence of Y137F and alterations at the position 459e461 were investigated in addition to the CYP51 alterations described

above. Y137F was present in neither year. Alterations in the 459e461 region were detected in all samples. However, due to limitations of the pyrosequencing assay, different alterations at those positions cannot be distinguished from each other. Mutation V136G was not included in the testing. All samples were tested positive for the presence of QoI mutation G143A at frequencies between 88 and 100%. Sdh alterations C-T79N, C-N89K and C-G90R were not detected in any sample. C-H152R, only tested in 2016, was not found either. 3.3. Correlation of % GLA and CYP51 selection and yield Fig. 1 illustrates correlations between percent green leaf area (GLA) and yield increase (dt ha
1), and ‘selection index’, as sum of

Table 6 Evaluation scheme of CYP51 alteration selection caused by different spray strategies. The total effect on resistance to DMIs of each spray strategy was ranked according to the selection of CYP51 alterations. Combined data of four field trials over two years. Treatment

1 2 3 4 5 6 7 8 9 10

Effect on resistancea

CYP51 mutation D134G

V136A

V136C

A379G

I381V

S524T

4 [ [ b 4 4 4 4 a 4 ***

4 [ [ 4 Y a a 4 Y 4 ***

4 4 b 4 4 b 4 4 4 4 ns

4 Y a 4 [ Y b 4 4 4 ***

4 4 4 4 4 4 4 4 4 4 ns

4 [ [ [ 4 [ 4 4 4 4 ***

e 2.0 3.0 1.5 0 0 0 0
1.5 0

Frequency of alterations as compared to the untreated: unchanged (4), increased ([), moderately increased (b), moderate decreased (a), decrease (Y) in comparison to the untreated check 1 (0). ns ¼ not significant; *** ¼ significant (p ¼ 0.05). a As compared to treatment 1. ([) ¼ 1, (b) ¼ 0.5, (4) ¼ 0, (a) ¼
0.5, (Y) ¼
1 highest possible value is 5.

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

113

Fig. 1. Linear regressions of green leaf area (GLA) in per cent and yield increase (dt ha
1) (left) and GLA and ‘selection index’ (right) across four field trials carried out in 2015 and 2016.

the alterations D134G, V136A/C and S524T. Across all trials there was a strong correlation between GLA and yield increase (R2 ¼ 0.953). Plotting GLA against ‘selection index’ showed a marginal trend where increasing GLA also led to higher selection for resistance. Repeated applications of treatments with few active ingredients as in spray strategies 2e4, showed high GLA and a high ‘selection index’. Strategies with a low ‘selection index’ (<80%) were those treated less (treatments 1, 9 and 10) or with the most diversified spray scheme (treatments 5, 6, 7 and 8). 3.4. In vitro sensitivity test In 2015, EC50 values for epoxiconazole varied from 0.01 to 4.04 ppm (mean 0.50 ppm). In Flakkebjerg, mean EC50 values were lower than in Jutland, with 0.37 and 0.66 ppm respectively. Ten isolates showed EC50 values > 1 ppm, all of which were found in treated plots. In 2016, a general shift was seen in EC50 values for epoxiconazole, ranging from 0.05 to 10.00 ppm (mean 1.25 ppm); in this year, 27 isolates out of a total 74 had EC50 values > 1 ppm. As in the previous year, the average values were higher in Jutland (1.52 ppm) than in Flakkebjerg (0.93 ppm) in 2016. Overall, EC50 values from 2016 were significantly higher than in 2015 (p < 0.01). In both years, the reference isolates IPO323 had an EC50 value of 0.01 ppm. There was no significant difference between EC50 values of isolates from treated plots compared to isolates from untreated plots in either year (Fig. 2). 4. Discussion The present study examined the efficacy of spray strategies, based on a limited assortment of fungicides as it is available in Denmark (Heick et al., 2017), and their impact on fungicide resistance in STB. It is a challenge to carry out effective fungicide resistance management with only a few active ingredients available and with the goal of minimising the risk of resistance build-up in the local STB population. The starting point of this investigation was an untreated control and further a strategy, which contained three sprays with the same fungicide (prothioconazole), being subsequently replaced by other products

(Table 1). The application of the same fungicide more than twice is normally not encouraged and current recommendations follow those of the Fungicide Resistance Action Committee FRAC (Anonymous, 2016c), avoiding repetitive use of a single MoA. Thus, this scenario served more as a non-practical counterexample. The efficacy results showed that STB control was satisfactory and similar for all treatments containing any DMI fungicide application at three timings (66e78% control of the flag leaf at GS 75 to 77). Yield increases were corresponding to the efficacy of the treatments. The addition of SDHI boscalid in a mixture with DMI epoxiconazole increased efficacy marginally compared to three sprays of prothioconazole. Highest efficacy was measured for the most diverse treatment (treatment 5) and a standard DMI-dominated spray strategy with additional folpet at GS 31e32 and GS 37e39 (treatment 6), which indicated that multi-site inhibitors actually offer additional control under rapid disease progression. A reduction of number of sprays, substituting one DMI spray with a multi-site inhibitor or supplementing with B. subtilis at all three timings led to inferior control and consequently lower yield. Bacillus subtilis (strain QST 713), an anti-fungal bacterium branded as biofungicide Serenade ASO®, has shown promising results in vegetable crops such as tomatoes grown in greenhouse against grey mould (Botrytis cinerea) and Pseudomonas syringae (Hinarejos et al., 2016; Punja et al., 2016), but more variable control at field level against e.g. yellow rust (Puccinia striiformis) on wheat (Reiss and Jørgensen, 2017). In this study no beneficial effect of B. subtilis on STB was observed. Adverse effects of the fungicide tank mix with B. subtilis might be part of the explanation why the addition of B. subtilis did not work. It is unknown whether B. subtilis and/or its formulation have an antagonistic effect on prothioconazole or the other way around. More specific field trials testing B. subtilis as a solo product at different timings and doses would be needed to gain more knowledge on whether B. subtilis qualifies for the control of STB. Especially as part of an IPM approach, this product deserves further attention as an addition to classic fungicides. The application of fungicides changed the frequencies of CYP51 alterations indicating reduced sensitivity towards DMI fungicides of the post-treatment population. This is in line with Dooley et al.

114

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

Fig. 2. Box whisker plots representing sensitivity values to epoxiconazole (as EC50 log10 mg L
1) of Z. tritici field isolates (n ¼ 167) collected in 2015 and 2016. Z. tritici isolates from five treatments after application of fungicides were tested. The dot through the box represents the median, the circles outliers.

(2016a), who showed that the applications of full rates of DMIs metconazole and epoxiconazole selected for Z. tritici strains with increased EC50 values for these fungicides. In our study, the decreased sensitivity was not reflected by increased EC50 values, as only a tendency towards increased EC50 values of epoxiconazole was detected in treated plots in 2015 but not for 2016. Still a general increase of EC50 values from 2015 to 2016 occurred (in the presented study and at national level), along with an increase in CYP51 alteration frequencies, indicating a strong link between these two factors. Dooley et al. (2016a) tested doses at half the label rate and found a clear difference in terms of selection at half rate. As higher doses have been associated with fungicide resistance development (Mavroeidi and Shaw, 2006; van den Bosch et al., 2011), the reduced dose might partly explain, why no differences in EC50 values were observed. For the in vitro sensitivity test, solely epoxiconazole was used, even though prothioconazole was the main fungicide used in this investigation. Epoxiconazole was chosen because results have shown a greater stability in the laboratory. In addition, there has been an ongoing discussion on whether to utilise prothioconazole, a profungicide, or its fungicidally active metabolite prothioconazole-destio. Nevertheless, as epoxiconazole and prothioconazole show a strong cross-resistance pattern, the use of epoxiconazole can be justified in the present case. Wieczorek et al. (2015) demonstrated that the effect of fungicides on Z. tritici CYP51 alterations was detectable using posttreatment leaf samples and pyrosequencing and qPCR. In their study, alterations V136A and D134G were selected by treatments consisting of two sprays of DMIs. Those two alterations are often present in the same strain (Cools and Fraaije, 2013; Buitrago et al.,

2014), and reduce sensitivity to epoxiconazole and prothioconazole, but maintain sensitivity to difenoconazole and tebuconazole (Leroux and Walker, 2011). In the present study D134G and V136A were significantly selected by prothioconazole. The more diverse the spray strategy, the less the selection of those two alterations. One component used in this investigation was a mixture of SDHI boscalid and DMI epoxiconazole. It has been suggested that DMI/ SDHI mixtures might help to lower resistance to DMIs and thereby slow down resistance build-up (Fraaije et al., 2012). This was confirmed in field trials testing the effect of mixtures of SDHI and DMIs on Z. tritici resistance (Dooley et al., 2016c). The authors found that including isopyrazam increased the number of epoxiconazolesensitive Z. tritici strains compared to epoxiconazole used alone. The same trend was seen for CYP51 alteration selection by two mixtures in field trials (bixafen þ prothioconazole; boscalid þ epoxiconazole) (Wieczorek et al., 2015). Treatment 5, which finished with a mixture of propiconazole and difenoconazole at GS 59e65, did not select for D134G and kept V136A at a lower level, indicating that those fungicides impose a higher efficacy on these alterations and provide a different selection pressure than prothioconazole. Alteration S524T was highly selected by DMI fungicides, especially prothioconazole. Strategy 4, although selecting marginally less than 2 and 3, was still selecting more than 5, which did not include either epoxiconazole or prothioconazole as a last spray. Previous studies have demonstrated that crossresistance patterns exist for several DMIs; cyproconazole, epoxiconazole and prothioconazole sharing one group and difenoconazole and tebuconazole another (Buitrago et al., 2014; Leroux and Walker, 2011). Those strong relationships also seem to be apparent in the selection patterns. Finishing a spray strategy off with a

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

fungicide or fungicide mixture representing an alternative crossresistance pattern neutralised or selected less for alterations selected by e.g. prothioconazole and epoxiconazole. Even though tebuconazole similarly to difenoconazole selects different strains, which do not harbour D134G, V136A and S524T, this effect is reduced when the product is used in mixture with prothioconazole at GS 57e65. Although sprays at those growth stages are seen to have major impact on selection the data from treatments 7 and 9 verify that omitting the use of DMI at GS 31e32 or replacing it with e.g. the multi-site inhibitor folpet helps to keep down the selection for D134G, V136A and S524T significantly. This additional information gives an indication that spray strategies, which are based on only one active ingredient or active ingredients having crossresistance, are prone to select into one direction. These findings highlight the need to diversify spray programmes as much as possible. The diversification of spray programmes by mixing or alternating DMI fungicides might, however, eventually increase selection for even more complex Z. tritici variants. Examples of such haplotypes have been already found in Europe (Kirikyali et al., 2017). Thus, monitoring of pathogen populations for changes in haplotype composition and EC50 values for a range of fungicides is vital to provide an accurate overview on the fungicide sensitivity situation and the need to adapt spray programmes to a new situation. Latest work on the prevalence of CYP51 variants, indicate that alterations D134G, V136A and S524T are becoming more established in the European Z. tritici population (Buitrago et al., 2014; Kildea et al., 2014; Anonymous, 2016a). Buitrago et al. (2014) saw an increase in strains harbouring alterations D134G and V136A/C in combination with I381V and S524T, investigating changes of CYP51 alterations of European Z. tritici isolates in the populations from 2009 to 2011. Those strains were significantly associated with an impact on DMI sensitivity. The frequencies of these alterations have also increased since 2009 in Denmark and Sweden (Heick et al., 2017). The comparison of pyrosequencing results of the untreated samples for 2015 and 2016 suggests a further shift into the direction of a higher prevalence of D134G, V136A/C and S524T in the Danish Z. tritici population. A general shift in frequencies of these alterations might partly explain the significant increase in EC50 values seen in 2016. This gives reasons to believe that the Scandinavian Z. tritici population is currently developing into the direction seen in other western European countries. In recent years several alterations in the Sdh gene have been found in Z. tritici field isolates in the United Kingdom, France, Germany and the Republic of Ireland reducing sensitivity to SDHI fungicides (Dooley et al., 2016a; Kirikyali et al., 2017). In this study, no Sdh alterations were found in Z. tritici isolates in Denmark in 2015 or 2016, suggesting a slower development of strains harbouring these alterations, as previously seen for the development for CYP51 alterations the Danish Z. tritici population (Heick et al., 2017). The frequency of QoI mutation G143A, however, has been at the same high level as in previous years. A strong correlation between GLA and yield increase has been reported in the literature (Gooding et al., 2000). The results of this study indicated the same trend that yields increase if higher proportion of leaf area remains photosynthetically active for as long as possible. Well-coordinated fungicide applications are the only means protecting the crop effectively under high disease pressure and reducing the negative effect on yields. In terms of selection, however, this could have negative consequences. Previous studies have established that effective disease control contrasts the intentions of effective resistance management (van den Berg et al., 2016; van den Bosch et al., 2014; van den Bosch et al., 2011). In the present study, it was demonstrated that high disease control

115

increased selection for fungicide resistance significantly. Less selection for fungicide resistance was seen following lower disease control as spray strategies 9 and 10, which were only treated once or twice. The addition of B. subtilis also compromised disease control, which might have equally led to the reduction in selection pressure. The diversification of treatments resulted in higher disease control and hence higher GLA. The correlation between GLA and ‘selection index’ confirmed a marginal trend that better disease control promoted selection, confirming the assumption that a more efficient disease control increases resistance. Similar findings were described by Dooley et al.’s (2016b), who showed that decreased doses and consequently reduced control maintain sensitivity, measured as EC50. This leads to the controversial advice that farmers should not keep their crops too clean as this will increase the risk of resistance. 5. Conclusion Whenever a fungicide is applied, selection pressure is imposed and selects for more resistant strains of a pathogen. Supposing there is no fitness penalty of those strains, the majority of a population will become resistant over time and eventually the fungicide loses its effect (Hobbelen et al., 2014). Hence, practical approaches to prolong the effective life of fungicides are greatly needed. This study was set out to employ spray strategies in the field to delay the build-up of fungicide resistance. It was shown that adjusting the number of applications decreased selection, yet compromised efficacy and yields. Selection for fungicide resistance was notably reduced alternating and diversifying the spray strategy, including mixing of different MoAs as well as DMI fungicides. These principles are, therefore, appropriate for fungicide resistance management of STB in the field and farmers should be encouraged to adopt them. The current study has examined spray strategies against STB in a region with limited product availability. Even in such a constrained situation, resistance management has proven to be practicable. It should be even easier to adopt the proposed antiresistance strategies, having more active ingredients available. The results support the hypothesis that STB control still can be achieved with the few products that have been on the market for quite some time. However, the introduction of new chemistry is much welcomed, in order to maintain disease control and yields in the future, but also to get more flexibility when designing spray strategies. Further experimental investigations are needed to estimate the effects on selection of other MoAs. In order to be able to control STB in the future, other IPM solutions besides fungicide applications should be encouraged, including varietal resistance or the use cultivar mixtures. Acknowledgements The authors would like to give gratitude to the companies BASF SE, Bayer CropScience, Syngenta and Adama for financial support of the field trials. Special thanks go to Dr Gerd Stammler for offering the opportunity to carry out parts of the experiments in his laboratory at BASF SE, Limburgerhof, Germany. Further, the authors would like to thank the technicians at CROP-Flakkebjerg assisting in the field and the laboratory for their great commitment in realising this work. This work was funded by Innovation Fund Denmark EvoPPM; 0603-00516B. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.cropro.2017.05.009.

116

T.M. Heick et al. / Crop Protection 99 (2017) 108e117

References Anonymous, 2014. HGCA. http://www.hgca.com/media/547811/Fungicideperformance-in-wheat-2014-V2.pdf. Accessed 07.February.2017. Anonymous, 2016a. EUROWHEAT. www.eurowheat.au.dk. Accessed 22.November.2016 2016. Anonymous, 2016b. FAOSTAT. http://faostat.fao.org/site/567/DesktopDefault.aspx? PageID¼567#ancor. Accessed 11.January.2016. Anonymous, 2016c. FRAC. www.frac.info. Accessed 22.November.2016 2016. Bearchell, S.J., Fraaije, B.A., Shaw, M.W., Fitt, B.D.L., 2005. Wheat archive links longterm fungal pathogen population dynamics to air pollution [Article] Proc. Natl. Acad. Sci. U. S. A. 102 (15), 5438e5442. http://dx.doi.org/10.1073/ pnas.0501596102. Brown, J.K.M., Chartrain, L., Lasserre-Zuber, P., Saintenac, C., 2015. Genetics of resistance to Zymoseptoria tritici and applications to wheat breeding [Review] Fungal Genet. Biol. 79, 33e41. http://dx.doi.org/10.1016/j.fgb.2015.04.017. Buitrago, C., Frey, R., Wullschleger, J., Sierotzki, H., 2014. An update on the genetic changes in the CYP51 gene of Mycosphaerella graminicola and their relationship to DMI fungicide sensitivity. In: Dehne, H.W., Deising, B., Fraaije, U., Gisi, D., Hermann, D., Mehl, A., et al. (Eds.), Modern Fungicides and Antifungal Compounds VII. DPG Spectrum Phytomedizin, Braunschweig, Germany, pp. 103e110. Chassot, C., Hugelshofer, U., Sierotzki, H., Gisi, U., 2008. Sensitivity of CYP51 genotypes to DMI fungicides in Mycosphaerella graminicola. In: Deising, H.B., Dehne, H.W., Gisi, U., Kuck, K.H., Russell, P.E., Stammler, G., et al. (Eds.), Modern Fungicides and Antifungal Compounds V. DPG Spectrum Phytomedizin, Braunschweig, Germany, pp. 129e136. Cools, H.J., Bayon, C., Atkins, S., Lucas, J.A., Fraaije, B.A., 2012. Overexpression of the sterol 14 alpha-demethylase gene (MgCYP51) in Mycosphaerella graminicola isolates confers a novel azole fungicide sensitivity phenotype [Article] Pest Manag. Sci. 68 (7), 1034e1040. http://dx.doi.org/10.1002/ps.3263. Cools, H.J., Fraaije, B.A., 2013. Update on mechanisms of azole resistance in Mycosphaerella graminicola and implications for future control [Review] Pest Manag. Sci. 69 (2), 150e155. http://dx.doi.org/10.1002/ps.3348. Dooley, H., Shaw, M.W., Mehenni-Ciz, J., Spink, J., Kildea, S., 2016a. Detection of Zymoseptoria tritici SDHI-insensitive field isolates carrying the SdhC-H152R and SdhD-R47W substitutions. Pest Manag. Sci. 72 (12), 2203e2207. http:// dx.doi.org/10.1002/ps.4269. Dooley, H., Shaw, M.W., Spink, J., Kildea, S., 2016b. Effect of azole fungicide mixtures, alternations and dose on azole sensitivity in the wheat pathogen Zymoseptoria tritici. Plant Pathol. 65 (1), 124e136. http://dx.doi.org/10.1111/ppa.12395. Dooley, H., Shaw, M.W., Spink, J., Kildea, S., 2016c. The effect of succinate dehydrogenase inhibitor/azole mixtures on selection of Zymoseptoria tritici isolates with reduced sensitivity [Article] Pest Manag. Sci. 72 (6), 1150e1159. http:// dx.doi.org/10.1002/ps.4093. Fones, H., Gurr, S., 2015. The impact of Septoria tritici Blotch disease on wheat: an EU perspective. Fungal Genet. Biol. 79, 3e7. http://dx.doi.org/10.1016/ j.fgb.2015.04.004. Fraaije, B.A., Bayon, C., Atkins, S., Cools, H.J., Lucas, J.A., Fraaije, M.W., 2012. Risk assessment studies on succinate dehydrogenase inhibitors, the new weapons in the battle to control Septoria leaf blotch in wheat [Article] Mol. Plant Pathol. 13 (3), 263e275. http://dx.doi.org/10.1111/j.1364-3703.2011.00746.x. Fraaije, B.A., Cools, H.J., Kim, S.H., Motteram, J., Clark, W.S., Lucas, J.A., 2007. A novel substitution I381V in the sterol 14 alpha-demethylase (CYP51) of Mycosphaerella graminicola is differentially selected by azole fungicides [Article] Mol. Plant Pathol. 8 (3), 245e254. http://dx.doi.org/10.1111/j.13643703.2007.00388.X. Gladders, P., Paveley, N.D., Barrie, I.A., Hardwick, N.V., Hims, M.J., Langton, S., et al., 2001. Agronomic and meteorological factors affecting the severity of leaf blotch caused by Mycosphaerella graminicola in commercial wheat crops in England [Article] Ann. Appl. Biol. 138 (3), 301e311. http://dx.doi.org/10.1111/j.17447348.2001.tb00115.x. Gooding, M.J., Dimmock, J., France, J., Jones, S.A., 2000. Green leaf area decline of wheat flag leaves: the influence of fungicides and relationships with mean grain weight and grain yield. Ann. Appl. Biol. 136 (1), 77e84. http://dx.doi.org/ 10.1111/j.1744-7348.2000.tb00011.x. Grimmer, M.K., van den Bosch, F., Powers, S.J., Paveley, N.D., 2014. Evaluation of a matrix to calculate fungicide resistance risk [Article] Pest Manag. Sci. 70 (6), 1008e1016. http://dx.doi.org/10.1002/ps.3646. Heick, T.M., Justesen, A.F., Jørgensen, L.N., 2017. Resistance of wheat pathogen Zymoseptoria tritici to DMI and QoI fungicides in the Nordic-Baltic region - a status. Eur. J. Plant Pathol. http://dx.doi.org/10.1007/s10658-017-1216-7. Hillocks, R.J., 2012. Farming with fewer pesticides: EU pesticide review and resulting challenges for UK agriculture. Crop Prot. 31 (1), 85e93. http:// dx.doi.org/10.1016/j.cropro.2011.08.008. Hinarejos, E., Castellano, M., Rodrigo, I., Belles, J.M., Conejero, V., Lopez-Gresa, M.P., et al., 2016. Bacillus subtilis IAB/BS03 as a potential biological control agent [Article] Eur. J. Plant Pathol. 146 (3), 597e608. http://dx.doi.org/10.1007/ s10658-016-0945-3. Hobbelen, P.H.F., Paveley, N.D., Fraaije, B.A., Lucas, J.A., van den Bosch, F., 2011a. Derivation and testing of a model to predict selection for fungicide resistance [Article] Plant Pathol. 60 (2), 304e313. http://dx.doi.org/10.1111/j.13653059.2010.02380.x. Hobbelen, P.H.F., Paveley, N.D., Oliver, R.P., van den Bosch, F., 2013. The usefulness of

fungicide mixtures and alternation for delaying the selection for resistance in populations of mycosphaerella graminicola on winter wheat: a modeling analysis. Phytopathology 103 (7), 690e707. http://dx.doi.org/10.1094/phyto-0612-0142-r. Hobbelen, P.H.F., Paveley, N.D., Van den Bosch, F., 2010. The usefulness of mixtures of a single-site and a multi-site fungicide as resistance management strategy [Meeting Abstract] Phytopathology 100 (6). S51eS51. Hobbelen, P.H.F., Paveley, N.D., van den Bosch, F., 2011b. Delaying selection for fungicide insensitivity by mixing fungicides at a low and high risk of resistance development: a modeling analysis [Article] Phytopathology 101 (10), 1224e1233. http://dx.doi.org/10.1094/phyto-10-10-0290. Hobbelen, P.H.F., Paveley, N.D., van den Bosch, F., 2014. The emergence of resistance to fungicides [Article] Plos One 9 (3), e91910. http://dx.doi.org/10.1371/ journal.pone.0091910. Jess, S., Kildea, S., Moody, A., Rennick, G., Murchie, A.K., Cooke, L.R., 2014. European Union policy on pesticides: implications for agriculture in Ireland. Pest Manag. Sci. 70 (11), 1646e1654. http://dx.doi.org/10.1002/ps.3801. Jørgensen, L.N., Matzen, N., Semaskiene, R., Korbas, M., Danielewicz, J., Glazek, M., et al., 2017. Azoles have different strengths and perform diversely across Europe. In: Deising, H.B., Fraaije, B., Mehl, A., Oerke, E.C., Sierotzki, H., Stammler, G. (Eds.), Modern Fungicides and Antifungal Compounds, vol. VIII. Deutsche Phytomedizinische Gesellschaft, Braunschweig, pp. 129e134. Jørgensen, L.N., Nielsen, G.C., Orum, J.E., Jensen, J.E., Pinnschmidt, H.O., 2008. Integrating disease control in winter wheat - optimizing fungicide input [article] Outlooks Pest Manag. 19 (5), 206e213. http://dx.doi.org/10.1564/ 19oct04.
Kildea, S., Mehenni-Ciz, J., Spink, J., OSullivan, E., 2014. Changes in the frequency of Irish Mycophaerella graminicola CYP51 variants 2006-2011. In: Dehne, H.W., Deising, H.B., Fraaije, B., Gisi, U., Hermann, D., Mehl, A., et al. (Eds.), 17th International Reinhardsbrunn Symposium. DPG Spectrum Phytomedizin, Braunschweig, Germany, pp. 143e144. Kirikyali, N., Diez, P., Luo, J., Hawkins, N.J., Fraaije, B.A., 2017. Azole and SDHI sensitivity status of Zymoseptoria tritici field populations sampled in France, Germany and the UK during 2015. In: Deising, H.B., Fraaije, B.A., Mehl, A., Oerke, E.C., Sierotzki, H., Stammler, G. (Eds.), Modern Fungicides and Antifungal Compounds, vol. VIII. Deutsche Phytomedizinische Gesellschaft, Braunschweig, pp. 153e158. Leroux, P., Walker, A.S., 2011. Multiple mechanisms account for resistance to sterol 14 alpha-demethylation inhibitors in field isolates of Mycosphaerella graminicola [Article] Pest Manag. Sci. 67 (1), 44e59. http://dx.doi.org/10.1002/ps.2028. Mavroeidi, V.I., Shaw, M.W., 2006. Effects of fungicide dose and mixtures on selection for triazole resistance in Mycosphaerella graminicola under field conditions [Article] Plant Pathol. 55 (6), 715e725. http://dx.doi.org/10.1111/j.13653059.2006.01441.x. Mullins, J.G.L., Parker, J.E., Cools, H.J., Togawa, R.C., Lucas, J.A., Fraaije, B.A., et al., 2011. Molecular modelling of the emergence of azole resistance in Mycosphaerella graminicola [Article] Plos One 6 (6), 11. http://dx.doi.org/10.1371/ journal.pone.0020973. O'Driscoll, A., Kildea, S., Doohan, F., Spink, J., Mullins, E., 2014. The wheat-Septoria conflict: a new front opening up? [Review] Trends Plant Sci. 19 (9), 602e610. http://dx.doi.org/10.1016/j.tplants.2014.04.011. Omrane, S., Sghyer, H., Audeon, C., Lanen, C., Duplaix, C., Walker, A.S., et al., 2015. Fungicide efflux and the MgMFS1 transporter contribute to the multidrug resistance phenotype in Zymoseptoria tritici field isolates. Environ. Microbiol. 17 (8), 2805e2823. http://dx.doi.org/10.1111/1462-2920.12781. Punja, Z.K., Rodriguez, G., Tirajoh, A., 2016. Effects of Bacillus subtilis strain QST 713 and storage temperatures on post-harvest disease development on greenhouse tomatoes [Article] Crop Prot. 84, 98e104. http://dx.doi.org/10.1016/ j.cropro.2016.02.011. RCoreTeam, 2014. R: A Language and Environment for Statistical Computing. Vienna, Austria. Reiss, A., Jørgensen, L.N., 2017. Biological control of yellow rust of wheat (Puccinia striiformis) with Serenade®ASO (Bacillus subtilis strain QST713). Crop Prot. 93, 1e8. http://dx.doi.org/10.1016/j.cropro.2016.11.009. Stammler, G., Carstensen, M., Koch, A., Semar, M., Strobel, D., Schlehuber, S., 2008. Frequency of different CYP51-haplotypes of Mycosphaerella graminicola and their impact on epoxiconazole-sensitivity and -field efficacy [Article] Crop Prot. 27 (11), 1448e1456. http://dx.doi.org/10.1016/j.cropro.2008.07.007. Stammler, G., Semar, M., 2011. Sensitivity of Mycosphaerella graminicola (anamorph: septoria tritici) to DMI fungicides across Europe and impact on field performance. EPPO Bull. 48 (2), 149e155. http://dx.doi.org/10.1111/j.13652338.2011.02454.x. Taher, K., Graf, S., Fakhfakh, M.M., Salah, H.B.H., Yahyaoui, A., Rezgui, S., et al., 2014. Sensitivity of Zymoseptoria tritici isolates from Tunisia to pyraclostrobin, fluxapyroxad, epoxiconazole, metconazole, prochloraz and tebuconazole [Article] J. Phytopathol. 162 (7e8), 442e448. http://dx.doi.org/10.1111/ jph.12210. Thomas, M.R., Cook, R.J., King, J.E., 1989. Factors affecting development of septoriatritici in winter-wheat and its effect on yield [Article] Plant Pathol. 38 (2), 246e257. http://dx.doi.org/10.1111/j.1365-3059.1989.tb02140.x. van den Berg, F., Paveley, N.D., van den Bosch, F., 2016. Dose and number of applications that maximize fungicide effective life exemplified by Zymoseptoria tritici on wheat e a model analysis. Plant Pathol. 65 (8), 1380e1389. http:// dx.doi.org/10.1111/ppa.12558.

T.M. Heick et al. / Crop Protection 99 (2017) 108e117 van den Berg, F., van den Bosch, F., Paveley, N.D., 2013. Optimal fungicide application timings for disease control are also an effective anti-resistance strategy: a case study for Zymoseptoria tritici (Mycosphaerella graminicola) on wheat [Article] Phytopathology 103 (12), 1209e1219. http://dx.doi.org/10.1094/phyto03-13-0061-r. van den Bosch, F., Oliver, R., van den Berg, F., Paveley, N., 2014. Governing principles can guide fungicide-resistance management Tactics. In: VanAlfen, N.K. (Ed.), Annual Review of Phytopathology, vol. 52, pp. 175e195. van den Bosch, F., Paveley, N., Shaw, M., Hobbelen, P., Oliver, R., 2011. The dose rate debate: does the risk of fungicide resistance increase or decrease with dose? [Review] Plant Pathol. 60 (4), 597e606. http://dx.doi.org/10.1111/j.13653059.2011.02439.x.

117

_ R., Mehl, A., Sierotzki, H., Stammler, G., et al., Wieczorek, T.M., Berg, G., Semaskiene, 2015. Impact of DMI and SDHI fungicides on disease control and CYP51 mutations in populations of Zymoseptoria tritici from Northern Europe [journal article] Eur. J. Plant Pathol. 143 (4), 861e871. http://dx.doi.org/10.1007/s10658015-0737-1. Wieczorek, T.M., Jørgensen, L.N., Christiansen, H.-B., Olsen, B.B., 2016. Fungicide resistance-related investigations. In: Jørgensen, L.N., Nielsen, B.J., Jensen, P.K., Hartvig, P., Wieczorek, T.M., Kaiser, C. (Eds.), Applied Crop Protection 2015, vol. 74. DCA - Nationalt Center for Fødevarer og Jordbrug, pp. 81e88. Zadoks, J.C., Chang, C.C., Konzak, C.F., 1974. A decimal code for the growth stages of cereals. Weed Res. 14 (6), 415e421. http://dx.doi.org/10.1111/j.13653180.1974.tb01084.x.