Mixtures as a fungicide resistance management tactic.

Page 1 of 39

1 2 3 4 5

Mixtures as a fungicide resistance management tactic Frank van den Bosch1, Neil Paveley2, Femke van den Berg1, Peter Hobbelen1and Richard Oliver3. 1

Phytopathology "First Look" paper • http://dx.doi.org/10.1094/PHYTO-04-14-0121-RVW • posted 08/20/2014 This paper has been peer reviewed and accepted for publication but has not yet been copyedited or proofread. The final published version may differ.

6

Rothamsted Research, West Common, Harpenden, AL5 2JQ, United Kingdom. 2

7 8 9

3

ADAS High Mowthorpe, Dugglesby YO17 8BP, United Kingdom

Environment & Agriculture, Centre for Crop and Disease Management (CCDM), Curtin

University, Bentley, Washington 6102, Australia

10 11 12

ABSTRACT

13

We have reviewed the experimental and modelling evidence on the use of mixtures of fungicides of

14

differing modes of action as a resistance management tactic. The evidence supports the following

15

conclusions:

16

1. Adding a mixing partner to a fungicide which is at-risk of resistance (without lowering the dose

17

of the at-risk fungicide) reduces the rate of selection for fungicide resistance. This holds for the

18

use of mixing partner fungicides that have either multi-site or single-site modes of action. The

19

resulting predicted increase in the effective life of the at-risk fungicide is large enough to be of

20

practical relevance. The more effective the mixing partner (due to inherent activity and/or

21

dose), the larger the reduction in selection and the larger the increase in effective life of the at-

22

risk fungicide.

23

2. Adding a mixing partner whilst lowering the dose of the at-risk fungicide reduces the selection

24

for fungicide resistance, without compromising effective disease control. The very few studies

25

existing suggest that the reduction in selection is more sensitive to lowering the dose of the at-

26

risk fungicide than to increasing the dose of the mixing partner.

27 28

3. Although there are very few studies, the existing evidence suggests that mixing two at-risk fungicides is also a useful resistance management tactic.

29

Amongst the aspects that have received too little attention to draw generic conclusions about the

30

effectiveness of fungicide mixtures as resistance management strategies are: (i) The relative effect

31

of the dose of the two mixing partners on selection for fungicide resistance, (ii) The effect of mixing


Page 2 of 39

32

on the effective life of a fungicide (the time from introduction of the fungicide mode of action to the

33

time point where the fungicide can no longer maintain effective disease control), (iii) polygenically

34

determined resistance, (iv) Mixtures of two at-risk fungicides, (v) The emergence phase of

35

resistance evolution and the effects of mixtures during this phase, and (vi) Mono-cyclic diseases

36

and non-foliar diseases. The lack of studies on these aspects of mixture use of fungicides should

37

be a warning against over-interpreting the findings in this review.

38 39

Keywords:

40

Evolution, Fungicide resistance, Selection, Emergence, Dose, Mixtures.

41 42

Acknowledgements:

43

Rothamsted Research receives support from the Biotechnological and Biological Research Council

44

(BBSRC) of the UK. This work was partly funded by the Chemicals Regulations Directorate (CRD)

45

and Department for Environment Food and Rural Affairs (DEFRA) of the UK, and by the Australian

46

Grains Research and Development (GRDC) of Australia.

47 48


Page 3 of 39

49

INTRODUCTION

50

The production of many crop species is dependent on fungicides to control diseases. Evolved

51

resistance is a significant threat to the efficacy of current fungicides and the optimisation of tactics

52

to extend their useful life is a priority. The effectiveness of mixtures of two or more different modes

53

of action as a fungicide resistance management method is subject of an extensive discussion since

54

the late 1970s. Some authors favour mixtures, whilst others have largely dismissed mixtures as a

55

resistance management tactic (2, 53). The development of sound resistance management tactics

56

requires exploration of the validity of these contrasting views and in this review we ask whether it is

57

possible to draw generic conclusions or that different situations require different tactics?

58

Resistance management methods are urgently discussed when new resistances are

59

discovered and when a new mode of action is introduced. Unfortunately, in discussions on

60

resistance management, it is often not explicit what the evidence is, what is opinion and what is

61

speculation. In this review we present an overview of the existing evidence on mixtures as a

62

method to prevent or delay the development of fungicide resistance. We will also identify the

63

evidence that is missing currently.

64 65

There are four main reasons for considering the use of mixtures in a fungicide application program

66

(5).

67

1. Broadening the spectrum. Pathogens are differentially sensitive to different active

68

substances. If a spray programme aims to control two or more pathogens it can, from a

69

disease control viewpoint, be useful to spray mixtures of fungicides.

70

2.

71

pathogen.

72

3. As an insurance against resistance. Mixtures may be used as insurance against failure

73

of one of the mixture components. If the pathogen to be controlled develops resistance to

74

one of the mixture components, the other fungicide (of a different mode of action) can

75

ensure that effective control is maintained.

Improved disease control. Adding a mixing partner may improve the control of a


Page 4 of 39

76

4. Resistance management. Mixtures are frequently advocated as a resistance

77

management tactic, i.e. specifically to extend the effective life of a fungicide by slowing the

78

rate of evolution of the pathogen population.

79

In this review we exclusively discuss the use of mixtures for resistance management. Hence, what

80

we present here is only one aspect of what needs to be considered when the use of mixtures is

81

discussed.

82 83

In a recent review ( 64) we have shown that there is a simple governing principle which we have

84

found to be a good predictor of whether particular resistance management tactics are likely to be

85

effective. The determinant of effectiveness is whether a tactic reduces the selection coefficient and

86

exposure time

87

= −

(1)

88

Where s = rS - rR is the selection coefficient, rS and rR are the intrinsic rate of increase of the

89

sensitive and the resistant strain, respectively, ( 40) and T is the time the pathogen is exposed to

90

the fungicide. The intrinsic rate of increase is the net increase in pathogen density per cycle of

91

infection. We will loosely describe this parameter as the epidemic growth rate. In population

92

genetics, the fitness of a pathogen strain is defined as the intrinsic rate of increase, r, (13) of that

93

pathogen strain. The selection coefficient provides a measure of the increase in the resistant

94

strain, as a fraction of the total pathogen population, per cycle of infection, and is related (non-

95

linearly) to the period of time that a new mode of action will remain effective. Here, we use the

96

governing principle of selection coefficient and exposure time to explore resistance management

97

using mixtures, and compare the results with published experimental evidence.

98 99

INSIGHTS FROM THE GOVERNING PRINCIPLE.

100

The effectiveness of mixtures as a resistance management tactic is affected by the doses of the

101

mixture components. Calculation of the selection coefficient, s, in a mixture of an at-risk fungicide

102

and a mixing partner gives some initial insight into the effect of dosages on the rate of selection for

103

fungicide resistance. If the density of the resistant strain is low the sensitive strain is the key

104

determinant of disease control. The growth rate of the sensitive strain, rS, then is a measure of the


Page 5 of 39

105

increase in disease severity. We thus use this growth rate as a measure of the level of disease

106

control achieved by the fungicide applications. We assume, as the simplest case, that the resistant

107

strain is completely resistant to the at-risk fungicide, suffers no fitness cost (in the absence of any

108

fungicide) and that the mixing partner has the same effect on the fitness of both strains. Thus

109

=

(2)

110

=

(3)

111

where r is the fitness of the strains in the absence of the fungicides, DA is the dose of the at-risk

112

fungicide, DM is the dose of the mixing partner, θ(DA) is the fractional reduction of the fitness of the

113

sensitive strain when using dose DA of the to-be-protected fungicide and α(DM) is the fractional

114

reduction of the fitness of both the sensitive and the resistant strain when using dose DM of the

115

mixing partner. Note that we modelled the effect of the fungicides as multiplicative, to represent

116

each fungicide acting independently on the intrinsic rate of increased. As dose response curves we

117

use exponentially decreasing fitness with dose with a lower asymptote, as this curve fits currently

118

published dose response curves from field experiments. We thus have

119

= 1 − 1 −

(4)

120

= 1 − 1 −

(5)

121

Substituting these, gives the dependence of the selection coefficient on the dose of the at-risk

122

fungicide and the mixing partner

123 124 125

, = 1 − 1 − 1 −

(6)

and as rate of epidemic increase = 1 − 1 − 1 − 1 − . (7)

126

Figure 1 shows the effect of the dose of the at-risk fungicide and the dose of the mixing partner on

127

the selection coefficient and on the growth rate of the epidemic. The lines in the graph are contour

128

lines for values of the selection coefficient, s, and the epidemic growth rate, rS. Both are expressed

129

as fractions of the fitness of the pathogen population in the absence of the fungicides, r.

130

The figure shows that increasing the dose of the mixing partner (arrow 1) decreases the

131

selection for fungicide resistance and increases the control of the pathogen (i.e. decreases the rate

132

of epidemic increase). It is also possible to decrease the selection for fungicide resistance by


Page 6 of 39

133

decreasing the dose of the at-risk fungicide (arrow 3), though this would decrease the level of

134

pathogen control. The effect of dose on selection has been determined ( 63, 64). The figure also

135

shows that it is possible to increase the dose of the mixing partner, decrease the dose of the at-risk

136

fungicide (arrow 2) and maintain disease control whilst decreasing the selection for fungicide

137

resistance. In the next sections we focus on arrow 1 and arrow 2 in more detail and compare

138

predictions against published evidence for polycyclic diseases; (we consider monocyclic diseases

139

in Box 1).

140 141

142 143

FOOTNOTE: In fact, fungicides do not directly affect the intrinsic rate of increase, r, of the

144

pathogen strains, but rather affect the life-cycle parameters of the pathogen such as infection

145

efficiency, latent period, spore production rate and infectious period. These life-cycle parameters

146

form the fundamental building blocks of the intrinsic rate of increase. In e-Xtra 1 we show that

147

qualitatively the same results emerge when we model the effect of the fungicides on life-cycle

148

parameters and use these to calculate the consequences for the intrinsic rate of increase.

149 150

THE EXPERIMENTAL EVIDENCE.

151

Adding a mixing component to an at-risk fungicide (arrow 1).

152

Whether adding a mixing partner to an at-risk fungicide to which resistance is developing

153

without reducing the dose of the at-risk fungicide, is a valid resistance management tactic has

154

been the subject of intensive discussions since the 1970s. The discussion initially involved

155

modelling studies. Several researchers have suggested that ‘mixing only reduces the build-up of

156

pesticide resistance by reducing the required dose of the pesticides that are mixed’ (2). Others

157

suggested that arrow 1 is a valid resistance management method. Over the last few decades

158

sufficient experimental studies have been published to allow a test of these contradictory views.

159

Relevant studies have been reported by (3, 4, 8, 9, 11, 15, 16, 17, 20, 21, 22, 24, 30, 31, 35, 37,

160

38, 39, 41, 42, 43, 46, 47, 48, 49, 57, 58, 59). In our analysis we separate studies with mixing

161

partners that are multi-site acting from mixing partners that are single-site acting fungicides. In the


Page 7 of 39

162

former case, the risk of resistance in the mixing partner can be largely discounted, whereas single-

163

site fungicide mixing partners are also at-risk. The results are given in the Online Appendix 2, but

164

in summary (Note: the dose of the fungicide against which the development of resistance is

165

measured is not changed in these treatments):

166

• Mixing partner is a multi-site fungicide: We found 17 publications with a total of 27

167

pathogen-crop-fungicide mixtures tested for their effect on fungicide resistance selection as

168

compared to the use of the at-risk solo fungicides. Of these 27 combinations 24 showed

169

reduced selection for fungicide resistance in the at-risk fungicide when used in mixture.

170

• Mixing partner is a single-site fungicide: We found 17 publications with a total 24

171

pathogen-crop-fungicide mixtures tested for their effect on the selection of resistance. In 20

172

cases the mixture delayed the build-up of resistance compared to the use of the at-risk

173

fungicide as a solo product.

174

The extensive experimental evidence thus shows that adding a mixing component to an at-risk

175

fungicide, not changing the dose of the at-risk fungicide, does reduce the selection for fungicide

176

resistance. There is no difference between mixing with a multi-site mixing partner and mixing with a

177

single site mixing partner to which no resistance is developing. With a total of 50 experiments of

178

which 86% show a reduced selection for fungicide resistance in a mixture even though the dose of

179

the at-risk fungicide is not reduced, we conclude that the evidence suggests that this is a valid

180

resistance management method in many cases. Only one experiment found an increased selection

181

in the mixture. We refer to Box 2 for a discussion of this case. We also refer to Box 1 for a

182

discussion on mono-cyclic pathogens.

183 184

Adding a mixing component and lowering the dose of the at-risk fungicide (arrow 2).

185

Lowering the dose of the at-risk fungicide in a mixture may compromise effective control. However,

186

our study of the selection coefficient (Figure 1) predicts that it should be possible to combine

187

mixing and dose reduction in such a way that effective control is not compromised. We have found

188

13 papers that discuss this scenario, summarised in Online Appendix 3 (7, 17, 18, 29, 30, 35, 44,

189

46, 48, 49, 55, 59, 66). The comparison is between the selection for fungicide resistance imposed


Page 8 of 39

190

by the use of the at-risk fungicide as solo product at the initial dose and the mixture with the

191

reduced dose of the at-risk fungicide.

192

• Mixing partner is a multi-site fungicide: A total of 8 publications with 11 pathogen-crop-

193

fungicide cases have been studied. In 8 cases it was found that the mixture delays the

194

selection for fungicide resistance in the at-risk fungicide compared to using the at-risk

195

fungicide as pure product. In 6 of the 8 publications the disease severity was measured

196

during the experiments and in all these cases effective control was similar in the solo use

197

and the mixture treatment.

198

• Mixing partner is a single-site fungicide: We found 6 publications with a total of 9

199

pathogen-crop-fungicide mixtures tested. In 7 cases the selection for fungicide resistance of

200

the at-risk fungicide was reduced by the mixture. In all studies the authors confirm that

201

disease control was at a similar level as in the solo product treatment.

202

The experimental evidence substantiates the predictions from the selection coefficient (Figure 1).

203

Both adding a mixture partner and reducing the dose of the at-risk fungicide will reduce selection

204

(see Box 2 for a discussion on the exceptions). It is possible to decrease the dose of the at-risk

205

fungicide, maintain the level of control and decrease the selection for fungicide resistance (arrow 2

206

in figure 1). We stress here however that for each mixture that may enter the market, field

207

experiments are needed to prove that effective control will not be compromised.

208 209

Further evidence on dose of the mixing components.

210

We have shown that adding a mixing partner to an at-risk fungicide to which resistance is

211

developing will reduce the rate of selection for fungicide resistance. Does this imply that the larger

212

the dose of the mixing partner, the larger the reduction in selection for resistance? This question is

213

of importance in making decisions on dose of the mixing components when developing mixtures.

214

There are only 4 papers (See Online Appendix 4 (3, 16, 20, 37)) that estimate the frequency of

215

resistance for sprays with mixtures with different dosages of the mixing partner. These papers

216

study, in total, 9 pathogen/fungicide/mixing-partner combinations. Although the amount of evidence

217

is small all experiments suggest that increasing the dose of the mixing partner decreases the

218

selection for fungicide resistance. Very likely this further decrease with increasing dose will be


Page 9 of 39

219

limited by the dose-response curve approaching its limiting value, but clearly all experiments were

220

done with dosages below such values. In the next section we will return to this topic when

221

describing model studies.

222

What has a larger effect on selection for fungicide resistance, increasing the dose of the

223

mixing partner or decreasing the dose of the at-risk fungicide? Obviously when we decrease the

224

dose of the at-risk fungicide disease control efficacy decreases, and when increasing the dose of

225

the mixing component disease control efficacy (and cost) increases. We thus have to make sure

226

that changing the doses in the mixture does not compromise effective control. Figure 1 clearly

227

shows that the answer depends on the dosage of both fungicides. Further exploration of equation 6

228

shows that the shape of the dose response curve also affect whether reducing the dose of the at-

229

risk fungicide or increasing the dose of the mixing partner has the larger effect on reducing the

230

selection rate (results not shown). Hence, optimising doses of mixtures remains a matter of case-

231

by-case study.

232

There are 4 papers describing field experiments with treatments reducing the dose of the

233

at-risk fungicide as well as increasing the dose of the mixing partner. (See Online-Appendix 5 (30,

234

31, 37, 42)) These papers describe 6 pathogen/fungicide/mixing-partner combinations. Generally

235

these experiments involve the full dose of the at-risk, with a full dose of the mixing partner or a full

236

dose of the at-risk plus a half or a full dose of the mixing partner.

237

All 6 pathogen/fungicide/mixing-partner combinations show that reducing the dose of the at-

238

risk fungicide reduces the selection for resistance more than increasing the dose of the mixing

239

partner. All these experiments relate to field relevant treatment programmes. The evidence base is

240

quite small, so we cannot conclude that reducing the dose of the at-risk fungicide is the more

241

effective resistance management tactic. More work in this area is needed and it may very well turn

242

out to be a question that needs to be studied on a case-by-case basis.

243 244 245 246 247

FROM QUALITATIVE TO QUANTITATIVE ANALYSIS: Is the reduction in selection large enough to matter?

248

The discussion so far has described consistent qualitative trends relating to mixing per se, mixing

249

multi-site or single-site fungicides and the relative doses of the two fungicides. We now address


Page 10 of 39

250

whether the reduction in selection for fungicide resistance from the use of mixtures is large enough

251

to be of practical relevance.

252

The reduction in selection for resistance certainly depends on the crop-pathogen-

253

fungicide combination and so the practical relevance will require case-by-case studies. How can

254

we quantify the delay in the build-up of resistance? Measurements of selection, the selection

255

coefficient, or a related quantity, from field experiments or model studies gives some insight, but it

256

does not tell us how much more durable the effective disease control from an at-risk fungicide will

257

be due to using it in a mixture. To be relevant, mixtures must increase effective disease control by

258

at least 1 growing season. We therefore need to measure the success of a resistance

259

management method in terms of years of effective disease control after the fungicide is introduced

260

(25, 26, 27, 60, 62). The models calculating the number of years of effective disease control (25,

261

26, 27, 60) were parameterised for M. graminicola on winter wheat, and the fungicides

262

azoxystrobin as the at-risk fungicide and chlorothalonil as the mixing partner. Here have done

263

additional calculations with two different dose response curves for the mixing component (Figure

264

2), one for chlorothalonil and one representing a hypothetical mixing partner with a lower disease

265

control efficacy. The findings are summarised in Tables 1 and 2.

266 267

268 269

Adding a mixing component to an at-risk fungicide, arrow 1: As found in the experimental evidence

270

discussed earlier, the model calculations show that adding a mixing partner to an at-risk fungicide

271

without lowering the dose of the at-risk fungicide reduces selection for resistance (Table 1).

272

Moreover the results show that the larger the dose of the mixing partner the further the selection

273

coefficient is reduced, as was also found in the few papers presenting experimental evidence on

274

this. Comparing the reduction of selection for resistance between mixing with chlorothalonil and

275

mixing with a lower efficacy fungicide with clearly shows that a mixing partner with lower efficacy is

276

less effective in lowering the rate of selection for fungicide resistance.


Page 11 of 39

277

The calculations of the effective-life of the mixtures show that a reduced rate of

278

selection goes paired with an increased effective life; using a mixing partner can increase the

279

effective life of an at-risk fungicide by several years.

280

Two key conclusions thus are that to reduce the rate of selection for fungicide

281

resistance and thereby increase the effective life of a fungicide we need to (i) use the highest

282

possible dose of the mixing partner, and (ii) use the mixing partner that has the highest disease

283

control efficacy. However, see the next section where two at-risk fungicides are mixed in the

284

scenario where resistance is developing against both. Hence using a mixing partner that is subject

285

to significant resistance and correspondingly poor efficacy is not a useful resistance management

286

tactic. This in turns means that the development of resistance to each active must be monitored on

287

an ongoing basis.

288

Adding a mixing partner and lowering the dose of the at-risk fungicide, arrow 2:

289

The last column of Tables 1 and 2 show that introducing a mixing partner and reducing the dose of

290

the at-risk fungicide can further increase the effective life of the at-risk fungicide whilst maintaining

291

effective control. Therefore we can conclude that at least for the crop/pathogen/fungicide cases

292

considered it may be possible to reduce the dose of the at-risk fungicide and therewith maximise

293

the effectiveness of mixing as a resistance management tactic.

294 295

We note here, however, that there is only one modelling study on the practical consequences of

296

using mixtures as a resistance management tactic. There is clearly a need for more studies to

297

investigate the effect of mixtures on the effective life of fungicides.

298 299

MIXING TWO AT-RISK FUNGICIDES.

300

An important question not answered so far is whether mixing two at-risk fungicides, against each of

301

which resistance is developing or can develop, is a useful resistance management tactic. This is a

302

highly relevant question as there are only a handful of multi-site, low risk fungicides and these are

303

generally of lower efficacy than single-site acting fungicides. Hence, for example, the new SDHI

304

fungicides are being introduced in mixtures with triazoles (12, 54). There are two opposing forces

305

to consider here. We have shown that adding a mixing partner B to an at-risk fungicide A, will slow


Page 12 of 39

306

down the selection for resistance to A. However adding an at-risk mixing partner B creates

307

selection pressure for resistance against B. The question thus is whether the gain, in terms of

308

slowing down selection, from adding an at-risk fungicide to the spray programme outweighs the

309

loss of putting an additional fungicide at risk of resistance.

310 311

We have found four papers on experimental work studying the development of resistance when the

312

pathogen develops resistance to both mixing components and the selection is measured for both

313

(4, 39, 57, 65). Of these the experiment presented by (65) did not include the control where the

314

fungicides were used as solo active substance and can therefore not answer our questions.

315

• Brent et al. (4) measured the concurrent selection for resistance to tridemorph and

316

ethirimol by barley powdery mildew (Blumeria graminis f. sp. hordei). Comparing the

317

by solo tridemorph with ethirimol + tridemorph showed that the mixture had a

318

rate for tridemorph resistance.

to

selection

lower

selection

319

• Lorenz et al. (39) compared mixtures and solo use of fenpropimorph and triadimenol for

320

wheat powdery mildew (Blumeria graminis f. sp. tritici). They showed a lower rate of

321

resistance selection for resistance to either component by the mixture than the solo

322

applications.

323 324

• Thygesen et al. (57) showed that mixing reduces the selection for both pyroclostrobin and epoxiconazole resistance in M. graminicola of wheat.

325

Though the number of applicable studies is only 3 they all show that mixing two at risk fungicides

326

slows down the selection for resistance in both fungicides. However these experiments did not

327

address the net effect of putting a second fungicide at risk, as discussed above.

328 329

The work developed by Hobbelen et al (26) is the only modelling study to assess the effective life

330

of a mixture of two at risk fungicides as compared to solo use of these fungicides (on different

331

fields at the same time) and sequential use (where use of one fungicide is delayed until the end of

332

the other fungicide’s effective life). Shaw (52) who also studied the development of resistance to

333

two at-risk fungicides, did not compare selection due to solo product use with mixture use but

334

compared mixture versus alternation. There are different scenarios to be considered here:

Page 13 of 39

335

(i)

resistant to fungicide A and a type resistant to fungicide B.

336 337


338

the pathogen population consist of three strains: a wild type sensitive, a type

(ii)

(ii) the pathogen population consists of four strains, a wild type sensitive, a type resistant to A, a type resistant to B and a strain resistant to A and B.

339

Hobbelen et al. (27) only considered the second scenario. They modelled two sets of fields

340

interconnected by spore exchange between crop growing seasons (the model was parameterised

341

for Mycosphaerella graminicola in wheat, but such spore exchange between seasons is

342

characteristic for a much broader range of foliar pathogens). Each set of fields could receive a

343

different spray programme. Two fungicides of different modes of action were available to control

344

the pathogen and the pathogen evolved resistance to both. Hobbelen et al. (27) then compared the

345

effective life of the two fungicides when used as a mixture of the fungicides with the use of the solo

346

fungicides each on half the fields (concurrent use). We also calculated (not in the publication) the

347

effective live for sequential use where one of the products is withheld from use, the other product

348

used as solo product till its effective life is ended and then introducing the other fungicide

349

(sequential use). They allowed the dose to be chosen in each case separately such that the

350

effective life of the fungicides was maximised.

351

Key conclusions from this modelling study were that mixing two at-risk fungicides always

352

gave an equal or higher effective fungicide life compared to either concurrent or sequential use of

353

the solo products. Thus mixing two at-risk fungicides is a useful anti-resistance management

354

method. The difference in the maximum effective life between the mixture and the concurrent use

355

tactic decreased when the frequency of the double resistant strain at the start of a treatment tactic

356

decreased. The difference in the maximum effective life between the mixture and the concurrent

357

use tactic increased with increasing fitness costs of resistance and increasing sensitivity of the

358

resistant strains to the high-risk fungicides. The maximum effective life of the mixture tactic was

359

usually obtained by applying both high-risk fungicides at dose rates that provided just enough

360

control of a severe epidemic and that resulted in an approximately equal selection pressure on

361

both single resistant strains.

362


Page 14 of 39

363

In summary we can say that the available evidence suggests that mixing two at-risk fungicides is a

364

useful resistance management tactic in the case where a strain resistant to both fungicides is

365

present in the pathogen population, but that the number of studies on which this conclusion is

366

based is too small to warrant generalisation and the scale of the increased effective life is not

367

always clear. The subject of mixtures of two at-risk fungicides is in need of further work. In Box 3

368

we also discuss a hypothetical situation where two at-risk fungicides applied as mixture on a

369

sexually reproducing pathogen may lead to the use of high dose strategies.

370 371

RESISTANCE EMERGENCE.

372

So far we have discussed the situation where resistance is present in the pathogen population and

373

is selected for by the use of fungicides; the selection phase (62). When a new fungicide mode-of-

374

action is introduced, if no target-site resistant strains are present in the pathogen population

375

resistance has to emerge through mutants derived from the sensitive population. Once a resistant

376

mutant has built up a sub-population that is large enough not to die out due to demographic

377

stochasticity, the resistant strain is said to have emerged (emergence phase, 62). The time

378

between the introduction of the fungicide mode-of-action and the moment the sub-population is

379

large enough not to die out by chance is termed the emergence time (28).

380

The importance of the emergence phase is currently unknown. No data are available on

381

pre-existence of resistance in the field. It is however of key relevance to develop an understanding

382

of the relevance of the emergence phase and the effect of resistance management strategies on

383

the length of the emergence phase. This is important because it is possible that a different set of

384

resistance management strategies is effective in the emergence phase, compared to the

385

resistance management strategies that are effective in the selection phase.

386

Studying the emergence phase in field populations will be extremely difficult and may be

387

impossible in practice. There is however a range of lab studies looking into the development of

388

mutants under UV treatment. This type of studies can possibly be adapted to study the emergence

389

phase for a range of fungicide treatment methods. Also microcosm studies in glasshouses may be

390

feasible. Due to the difficulties of field experimentation modelling studies mimicking field conditions

391

will be a key tool.


Page 15 of 39

392

Several authors have studied simple stochastic models for the emergence of escape

393

mutants (32, 50, 51), and shown the effect of the pathogen’s life-cycle parameters on the

394

emergence of new pathogen strains. The models do not however lend themselves to quantification

395

of the effect of fungicide spray programmes on the emergence of fungicide resistance due to their

396

abstract nature, not including key drivers such as seasonality in host density, and the fact the

397

selection pressure (the fungicide) is not constant through time. We have shown that such drivers

398

have strong effects on the outcome of selection processes (25, 26, 27, 62).

399

There is currently only one modelling study looking into the emergence phase of fungicide

400

resistance (28). The model was derived from the models of Hobbelen et al. (25, 26, 27) as

401

discussed above by representing the dynamics of the resistant strain by a stochastic event-driven

402

model (6). The model was parameterized for Mycosphaerella graminicola on winter wheat (Triticum

403

aestivum). Analysis of mixture treatment strategies with the model showed that adding a not-at-risk

404

fungicide to an at-risk fungicide may, as in the selection phase, also be effective in prolonging the

405

emergence phase in the evolution of resistance to high-risk fungicides.

406

It is not possible to come to general conclusions about resistance management in the

407

emergence phase of fungicide resistance evolution on the basis of one modelling study. Further

408

work is needed in this area.

409 410

DISCUSSION.

411

The experimental and theoretical evidence on the use of mixtures as a fungicide resistance

412

management tactic leads to the following conclusions:

413

Conclusion 1: Adding a mixing partner to an at-risk fungicide (without lowering the dose of

414

the at-risk fungicide) reduces the rate of selection for fungicide resistance. This holds for the use of

415

multi-site as well as for the use of single-site fungicide as the mixing partner. A first model study

416

suggests that the reduction of selection for resistance results in an increase of the effective life of

417

an at-risk fungicide which is large enough to be of practical relevance. All existing evidence

418

suggests that the larger the dose of the mixing partner the larger the reduction in selection and the

419

larger the increase in effective life of the at-risk fungicide. A mixing partner fungicide with a lower

420

efficacy will give a smaller reduction in selection rate.


Page 16 of 39

421

Conclusion 2: Adding a mixing partner to an at-risk fungicide as well as lowering the dose

422

of the at-risk fungicide also reduces the selection for fungicide resistance. The very few studies

423

existing suggest that the reduction in selection is more sensitive to lowering the dose of the at-risk

424

fungicide than to increasing the dose of the mixing partner. Several experimental as well as

425

modelling studies show that it is possible to lower the dose of the at-risk fungicide without

426

compromising effective disease control. However, effective disease control is a critical parameter

427

that needs to be rigorously established when developing fungicide mixtures.

428

Conclusion 3: Although there is little evidence about the usefulness of mixing two at-risk

429

fungicides as a resistance management tactic, the evidence suggests that mixing two at-risk

430

fungicides, without intrinsic positive cross-resistance, is a useful resistance management tactic.

431

However: Most of the evidence available on mixtures to date relates to the selection for

432

resistance of a fungicide to which a considerable resistance develops in one single (mutation) step,

433

in mixture with a fungicide to which no resistance is developing, or at least the resistance

434

development of the mixing partner is not considered. The review made clear that for this situation

435

mixing is a useful resistance management tactic. This is, however, just a subset of the practically

436

relevant cases of fungicide resistance development. We have discussed the mixture of two

437

fungicides to both of which resistance is developing as a practically important and clearly different

438

example that has received virtually no attention. Moreover many pathogens develop resistance to

439

fungicides through a sequence of small steps (sequence of mutations) each adding to ( or

440

occasionally detracting from (19) ) the resistance. The evolution of azole resistance is a key

441

example for which there is inadequate published evidence on mixture effects.

442

There are a number of examples in Botrytis cinerea where resistance has been attributed to

443

the overexpression of efflux pumps (36; 45; 23). This is analogous to many causes of resistance to

444

herbicides (10) and to clinical antibiotics (1). The current status of resistance in Botrytis has

445

recently been reviewed. The impact of efflux pump mutations on the management of resistance to

446

two or more fungicides that are susceptible to resistance by this mechanism has not been

447

modelled and little or no data is available currently on which to test such a model. The contrasting

448

impacts of efflux pump-mediated cross resistance coupled with potential fitness penalties needs


Page 17 of 39

449

further study. Hence we caution that our conclusions should not be over-interpreted as they do not

450

cover all possible relevant cases.

451

Conclusions 1 to 3 are generally well supported by evidence, but nevertheless are

452

accompanied by some caveats. Results of experiments where no difference between treatments

453

are measured have a smaller chance of being published than results of experiments where a

454

difference is found. This may cause the effectiveness of mixing as resistance management method

455

to be overestimated. Further, most of the experimental evidence relates to poly-cyclic pathogens

456

and foliar-applied fungicides. In Box 2 we described a possible difference between mono-cyclic

457

and poly-cyclic pathogens, but no evidence exists as yet. A final aspect is that all experimental

458

evidence relates to cases where a single mutation (or other genetic change) confers a high level of

459

resistance. Slow-shifting resistance is grossly absent from the mixtures literature. As also

460

concluded in van den Bosch et al (64) ‘’Industry and regulatory decisions about fungicide

461

resistance management often cannot wait for the accumulation of new evidence, so decisions

462

should be taken by weighing the existing evidence and making judgments about the consequences

463

should decisions prove to be wrong. With this review we hope to make a contribution towards such

464

evidence-based resistance management.’’

465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487

REFERENCES: 1.. Arias CA, Murray BE. 2009. Antibiotic-resistant bugs in the 21st century—a clinical superchallenge. New England Journal of Medicine 360(5): 439-443. 2. Birch, C. P. D., and Shaw, M. W. 1997. When can reduced doses and pesticide mixtures delay the build-up of pesticide resistance? A mathematical model. J. Appl. Ecol. 34:1032–42. 3. Bolton, N. J. E., and Smith, J. M. 1988. Strategies to combat fungicide resistance in barley powdery mildew. British Crop Protection Conference, Pests & Diseases 367-372. 4. Brent, K.J., Carter, G.A., Hollomon, D. W., Hunter, T., Locke, T., and Proven, M. 1989. Factors affecting build-up of fungicide resistance in powdery mildew in spring barley. Netherlands Journal of Plant Pathology 95 (S1):31-41. 5. Brent, K.J., and Hollomon, D. W. 2007. Fungicide resistance in crop pathogens: How can it be managed? Fungicide Resistance Action Committee. 6. Cao, Y., Gillespie, D. T., Petzold, L. R. 2006. Efficient step size selection for the tauleaping simulation method. J. Chem. Phys. 124: 044109.


Page 18 of 39

488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538

7. Carnegie, S.F., Cameron, A.M., and Haddon, P. 2008. Effects of fungicide and rate of application on the development of isolates of Polyscytalum pustulans resitant to thiabendazole and on the control of skin spot. Potato Research 51:113-129. 8. Cohen, Y., and Samoucha, Y. 1989. Selection for matalaxyl reistance in potato crops infected with Pyhotophthora infestans: effects of fungicides and initial frequency of resistant sporangia. Plant Pathology 38:382-390. 9. Cohen, Y., and Samoucha, Y. 1990. Competition between oxadixyl-sensitive and -resistant field isolates of Phytophthora infestans on fungicide-treated potato crops. Crop Protection 9:15-20. 10. Conte SS, Lloyd AM. 2011. Exploring multiple drug and herbicide resistance in plants— Spotlight on transporter proteins. Plant Science 180(2): 196-203. 11. Cooke, L. R., Locke, T., Lockley, K. D., Phillips, A. N., Sadiq, M. D. S., Coll, R., Black, L., Taggart, P. J., and Mercer, P. C. 2004. The effect of fungicide programmes based on epoxiconazole on the control and DMI sensitivity of Rhynchosporium secalis in winter barley. Crop Protection 23: 393-406. 12. Cools, H. J., Hawink, N. L., Fraaije, B. A. 2013. Constraints on the evolution of azole resistance in plant pathogenic fungi. Plant Pathology 62:36-42. 13. Crow, J. F., and Kimura, M. 1970. An introduction to Population Genetic Theory. Harper and Row, New York. 14. De Waard, M. A., Nistelrooy, J. G. M. 1979. Mechanism of resistance to fenarimol in Aspergillus nidulans. Pesticide Biochemistry and Physiology 10:219-229. 15. Dovas, C., Skylakakis, G., and Georgopoulos, S. G. 1976. The adaptability of the benomylresistant population of Cercospora beticola in Northern Greece. Phytopathology 66:1452-1456. 16. English, A. R., and van Halsema, G. 1954. A note on the delay in the emergence of resistant xanthomonas and erwinia strains by the use of streptomycin plus terramycin combinations. Plant Disease Reporter 38:429-431. 17. Forster, B., and Staub, T. 1996. Basis for use strategies of anilinopyrimidineand phenylpyrrole fungicides against Botrytis cinerea. Crop Protection 15:529-537. 18. Forster, B., Chavaillaz, D., Steden, C., Radtke, W., Käsbohrer, M., and Kühl, A. 1994. Influence of split application of fenpropimorph mixtures on disease control and on the sensitivity of Erysiphe graminis f. sp. tritici. British Crop Protection Council Monographs No. 60, 331-335. 19. Fraaije BA, Cools HJ, Kim SH, Motteram J, Clark WS, Lucas JA. 2007. A novel substitution I381V in the sterol 14α-demethylase (CYP51) of Mycosphaerella graminicola is differentially selected by azole fungicides. Molecular Plant Pathology 8(3): 245-254. 20. Genet, J-L., Jaworska, G., and Deparis, F. 2006. Effect of dose rate and mixtures of fungicides on selection for QoI resistance in populations of Plasmopara viticola. Pest Management Science 62:188-194.


Page 19 of 39

539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588

21. Genet, J-L., and Jaworska, G. 2013. Characterization of European Plasmopara viticola isolates with reduced sensitivity to cymoxanil. European Journal of Plant Pathology 135: 383-393. 22. Grabski, C., and Gisi, U. 1985. Mixtures of fungicides with synergistic interactions for protection against phenylamide resistance in Phytophthora. British Crop Protection Conference Monographs No. 31, 315-318. 23. Hahn M. 2014. The rising threat of fungicide resistance in plant pathogenic fungi: Botrytis as a case study. Journal of Chemical Biology. 24. Hoare, F. A., Hunter, T., and Jordan, V. W. L. 1986. Influence of spray programmes on development of fungicide resistance in the eyespot pathogen of wheat, Pseudocercosporella herpotrichoides. Plant Pathology 35:506-511. 25. Hobbelen, P. H. F., Fraaije, B., Lucas, J. A., Paveley, N. D., and van den Bosch, F. 2011a. Derivation and testing of a model to predict selection for fungicide resistance. Plant Pathology 60:304-313. 26. Hobbelen, P. H. F., Paveley, N. D., and van den Bosch, F. 2011b. Delaying selection for fungicide insensitivity by mixing fungicides at a low and high risk of resistance development: a modelling analysis. Phytopathology 101:1224-1233. 27. Hobbelen, P. H. F., Paveley, N. D., and van den Bosch, F. 2013. The Value of Alternation or Mixtures of Fungicides for Delaying the Selection of Resistance Against Two Modes of Action in Populations of Mycosphaerella graminicola on Winter Wheat. Phytopathology, 101:690-707. 28. Hobbelen, P. H. F., Paveley, N. D., and van den Bosch, F. 2014. The emergence of resistance to fungicides. PlosOne. DOI: 10.1371/journal.pone.0091910 29. Horsten, J. A. H. M. 1979. Acquired resistance to systemic fungicides of Septoria nodorum and Cercosporella herpotrichoides in cereals, PhD thesis. Wageningen University, the Netherlands. 30. Hunter, T., Brent, K. J., Carter, G. A., and Hutcheon, J.A. 1987. Effects of fungicide spray regimes on incidence of dicarboximide resistance in grey mould (Botrytis cinerea) on strawberry plants. Annals of Applied Biology 110:515-525. 31. Hunter, T., Brent, K. J., and Carter, G. A. 1984. Effects of fungicide regimes on sensitivity and control of barley mildew. British Crop Protection Conference, Pests & Diseases 471-476. 32. Iwasa, Y., Michor, F., and Nowak, M. A. 2003. Evolutionary dynamics of invasion and escape. J. Theor. Biol. 226:205-214. 33. Kalamarakis, A. E., de Waard, M. A., Ziogas, B. N., and Georgopoulos, 1991. Resistance to fenarimol in Nectria haematococca var. cucurbitae. Pesticide Biochemistry and Physiology 40:212-220. 34. Koenraadt, H., Somerville, C., and Jones, A. L. 1992. Characterization of mutations in the betatubuline gene of benomyl-resistant field strains of Venturia inaequalis and other plant pathogenic fungi. Phytopathology 82:1348-1354.


Page 20 of 39

589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638

35. Köller, W., and Wilcox, W. F. 1999. Evaluation of tactics for managing resistance of Venturia inaequalis to sterol demethylation inhibitors. Plant Disease 83:857-863. 36. Kretschmer M, Leroch M, Mosbach A, Walker AS, Fillinger S, Mernke D, Schoonbeek HJ, Pradier JM, Leroux P, De Waard MA, et al. 2009. Fungicide-driven evolution and molecular basis of multidrug resistance in field populations of the grey mould fungus Botrytis cinerea. PLoS Pathogens 5(12). 37. Lalancette, N., Hickey, K.D., and Cole, Jr. H. 1987. Effects of mixtures of benomyl and mancozeb on build-up of benomyl-resistant Venturia inaequalis. Phytopathology 77:86-91. 38. LaMondia, J. A. 2001. Management of Euonymus anthracnose and fungicide resistance in Colletotrichum gloeosporioides by alternating or mixing fungicides Journal of Environmental Horticulture 19:51-55. 39. Lorenz, G., Saur, R., Schelberger, K., Forster, B., Kung, R., and Zobrist, P. 1992. Long term monitoring results of wheat powdery mildew sensitivity towards fenpropimorph and strategies to avoid the development of resistance. Brighton Crop Protection Conference: Pests and Diseases, Vols. 1, 2 and 3 Pages: 3) 171-176 40. Madden, L.V., Hughes, G. and van den Bosch, F. 2007. The Study of Plant Disease Epidemics APS Press. 41. Maraite, H., Meunier, S., Gilles, G., and Bal, E. 1985. Evolution de la resistance aux imides cycliques chez Botrytis cinerea sur fraisiers en Belgique. Bulletin OEPP 15:387-394. 42. Mavroeidi, V.I., and Shaw, M.W. 2006. Effects of fungicide dose and mixtures on selection for triazole resistance in Mycosphaerella graminicola under field conditions. Plant Pathology 55:715725. 43. McCartney, C., Mercer, P. C., Cooke, L. R., and Fraaije, B. A. 2007. Effects of a stobilurinbased spray programme on disease control, green leaf area, yield and development of fungicideresistance in Mycosphaerella graminicola in Northern Ireland. Crop Protection 26:1272-1280. 44. McGee, D. C., and Zuck, M. G. 1981. Competition between benomyl-resistant and sensitive strains of Venturia inaequalis on apple seedlings. Phytopathology 71:529-532.

45. Mernke D, Dahm S, Walker AS, Lalève A, Fillinger S, Leroch M, Hahn M. 2011. Two promoter rearrangements in a drug efflux transporter gene are responsible for the appearance and spread of multidrug resistance phenotype MDR2 in Botrytis cinerea isolates in French and German vineyards. Phytopathology 101(10): 1176-1183. 46. Northhover, J., and Matteoni, J. A. 1986. Resistance of Botrytis cinerea to benomyl and iprodione in vineyards and greenhouses after exposure to fungicides alone or mixed with captan. Plant Disease 70:398-402. 47. Oxley, S. J. P., Burnett, F., Hunter, T., Fraaije, B. A., Cooke, L. R., Mercer, C., and Gilchrist, A. 2008. Understanding fungicide mixtures to control Rhynchosporium in barley and minimise resistance shifts: HGCA Project Report No. 436, August 200894pp.


Page 21 of 39

639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687

48. Samoucha, Y., and Gisi, U. 1987. Use of two- and three-way mixtures to prevent build-up of resistance to phenylamide fungicides in Phytophthora and Plasmopara. Phytopathology 77:14051409. 49. Sanders, P. L., Houser, W. L., Parish, P. J., and Cole, H. 1985. Reduced-rate fungicide mixtures to delay fungicide resistance and to control selected turfgrass disease. Plant Disease 69:939-943. 50. Serra, M. C. 2006. On waiting time to escape. J. Appl. Probab. 43:296-302. 51. Serra, M. C., and Haccou, P. 2007. Dynamics of escape mutants. Theor. Popul. Biol. 72:167178. 52. Shaw, M. W. 1993. Theoretical-analysis of the effect of interacting activities on the rate of selection for combined resistance to fungicide mixtures. Crop Protection 12:120-126. 53. Shaw, M. W. 2006. Is there such a thing as a fungicide resistancestrategy? A modeller’s perspective. Aspects of Applied Biology 78:37-44.

54. Sierotzki, H., and Scalliet, G. 2013. A review of current knowledge of resistance aspects for the next-generation succinate dehydrgenase inhibitor fungicides. Phytopathology 103:880-887. 55. Taggart, P. J., Cooke, L. R., Mercer, P. C., and Shaw, M. W. 1998. Effects of fungicides used to control Rhynchosporium secalis where benzimidazole resistance is present. Crop Protection 17:727-734. 56. Tate, K. G., and Samuels, G. J. 1976. Benzimidazole tolerance in Venturia inaequalis in New Zealand. Plant Disease Reporter 60:706-710. 57. Thygesen, K., Jorgensen, L. N., Jensen, K. S., and Munk, L. 2009. Spatial and temporal impact of fungicide spray strategies on fungicide sensitivity of Mycosphaerella graminicola in winter wheat. European Journal of Plant Pathology 123:435-447. 58. Turechek, W. W., and Köller, W. 2004. Managing resistance of Venturia inaequalis to the strobilurin fungicides. Plant Health Progress doi: 10.1094/PHP-2004-0908-01RS. 59. Vali, R. J., and Moorman, G. W. 1992. Influence of selected fungicide regimes on frequency of dicarboximide-resistant and dicarboximide-sensitive strains of Botrytis cinerea. Plant Disease 76:919-924. 60. van den Berg, F., van den Bosch, F., and 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. Phytopathology, in press. 61. van den Bosch, F., McRoberts, N., van den Berg, F. and Madden, L. 2008. The basic reproduction number of plant pathogens: Matrix approaches to complex dynamics. Phytopathology, 98:239-249.


Page 22 of 39

688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708

62. van den Bosch, F., and Gilligan, C. A. 2008. Models of fungicide resistance. Annual review of Phytopathology 46:123-147. 63. van den Bosch, F., Paveley, N. D., Shaw, M. W., Hobbelen, P., and Oliver, R. 2011. The dose rate debate: does the risk of fungicide resistance increase or decrease with dose? Plant Pathology 60:597-606. 64. van den Bosch, F., Oliver, R., van den Berg, F., and Paveley, N. D. 2014. Governing principles can guide fungicide resistance management tactic. Annual Review of Phytopathology, in press.

65. Zhang, C-Q., Zhang, Y., and Zhu, G. N. 2008. The mixture of kresoxim-methyl and boscalis, an excellent alternative controlling grey mold caused by Botrytis cinerea. Annales of Applied Biology 153:205-213. 66. Zhu, S. S., Liu, X. L., Wang, Y., Wu, X. H., Liu, P. F., Li, J. Q., Yuan, S. K., and Si, N. G. 2007. Resistance of Pseudoperonospora cubensis to flumorph on cucumber in plastic houses. Plant Pathology 56: 967-975.


Page 23 of 39

709

BOX 1: Monocyclic versus polycyclic pathogens.

710

In this box we consider a possible difference between mono- and poly-cyclic pathogens in the

711

effect on selection for fungicide resistance of adding a mixing partner to an at-risk fungicide. The

712

selection coefficient, defined as the difference in the exponential growth rates of the resistant and

713

the sensitive strain, is derived from a model where pathogen generations overlap due to the

714

continuous reproduction and continuous death of lesions. Thus the simple model mimics a poly-

715

cyclic pathogen. In this case adding a mixing partner to an at-risk fungicide is expected to reduce

716

the rate of selection for fungicide resistance. More detailed models for poly-cyclic pathogens, such

717

as discussed in the online appendix 1 and in the model by Hobbelen et al (26) suggest the same

718

conclusion. Does this conclusion also hold for mono-cyclic pathogens?

719

If we denote by PS,n and PR,n the density of the sensitive and the resistant strain in

720

generation n, the equivalent of the exponential growth model, equations 1.1 and 1.2, for a discrete

721

time model are

722

,

!

= "# ,

(2.1)

723

,

!

= "#, ,

(2.2)

724

where R0S and R0R are the net-reproductive number of the sensitive and the resistant strain,

725

respectively. The net-reproductive number is the number of daughter lesions per mother lesion

726

when availability of susceptible host is not limiting (61 for an introduction to this topic). The

727

frequency of resistance in generation n+1, ρn, is calculated from

728

$

!

=

%& '&( %& '&(

%) ')(

(2.3)

729

The at-risk fungicide affects the net-reproduction number of the sensitive strain and the mixing

730

partner affects the net-reproductive number of both pathogen strains. We can then write

731

"#* = "#

(2.4)

732

"# = "#

(2.5)

733

where R0 is the net-reproductive number of both strains in the absence of fungicide use, θ(DA) is

734

the fractional reduction of the net-reproduction of the sensitive strain when using dose DA of the to-

735

be-protected fungicide and α(DM) is the fractional reduction of the net-reproduction of both the

Page 24 of 39

736

sensitive and the resistant strain when using dose DM of the mixing partner. Substituting (2.4) and

737

(2.5) into (2.3) we find


738

$

!

=

% + '&( % + , ')( % + '&(

=

% '&( % , ')( % '&(

(2.6)

739

From which we see that the effect of the mixing partner cancels out, implying the that mixing

740

partner has no effect on selection for resistance and we conclude that, in our simple model for a

741

mono-cyclic pathogen, adding a mixing partner to an at-risk fungicide does not reduce the

742

selection for resistance.

743

The reason for the difference between the poly-cyclic and the mono-cyclic pathogens is that

744

in the poly-cyclic case several pathogen generations live at the same moment and each generation

745

has been under a different selection history, with ‘’older’’ generations having experienced less

746

selection than ‘’younger’’ generations. This can be seen by calculating, using equations (2.1) and

747

(2.2), two sequential generations. Then assuming that a fraction µ of the first generation survives

748

and lives at the same time as the second generation it is easily seen that the effect of the mixing

749

partner no longer cancels out from the calculation of the frequency of resistance.

750

The reasoning above suggests that adding a mixing partner to an at-risk fungicide may not

751

reduce selection for mono-cyclic species. Interestingly the only experiment on mixing using

752

Pseudocercosporella herpotrichoides causing eyespot disease in wheat (24) suggests that adding

753

a mixing partner does reduce selection.

754

We conclude that there is an urgent need to develop models and conduct resistance

755

experiments on mono-cyclic pathogens, because the effect of fungicide mixtures on selection for

756

resistance may or may not be different from the effect of mixing in poly-cyclic species.

757 758 759 760


Page 25 of 39

761

BOX 2: A closer look at the exceptions.

762

In this box we discuss the two cases that contradict the predictions from the governing principle

763

coefficient. Köller and Wilcox (35) studied the effect of mixtures on the selection for fenarimol

764

resistance in Venturia inaequalis, (causal organism of apple scab). Resistance to fenarimol is

765

polygenically determined in Nectria haematococca var. cucurbitae with at least nine loci involved

766

(33) and there are phenotypes known that show a reduced uptake of fenarimol in Aspergillus

767

nidulans (14). There are strains with low and with high levels of resistance. Secondly, McGee and

768

Zuck (44) studied the effect of mixtures on the selection for benomyl resistance in the same

769

pathogen species. High levels of resistance are conferred by a single gene mutation (34) but there

770

are also low resistance strains known (34, 56). The genetics of the low resistance strains is not

771

known but might be related to efflux pump mechanisms.

772

We have shown previously (63) that low resistance levels/partial resistance may be associated

773

with dose response curves of the resistant and the sensitive pathogen strains that converge at

774

higher dose. If this is the case it is possible that the selection for fungicide resistance is lower at

775

high dose than at low fungicide dose. The two cases may be examples of this mechanism. There

776

are however no data available for any fungicide-pathogen combination that allow us to see if dose

777

response curves converge within the range of doses permitted. We have discussed (63) that

778

research in this area is urgently needed because it may, in some cases, have consequences for

779

resistance management. We reiterate here the need for experimental and modelling work on the

780

existence of converging dose response curves for cases where the pathogen has partially resistant

781

strains.

782


Page 26 of 39

783

BOX 3: Mixtures of two at-risk fungicides and the dose rate debate.

784

In this box we consider an exceptional case where a high dose of two at-risk fungicides may be

785

required to reduce selection for resistance in a haploid sexual pathogen species. In our previous

786

work we have shown that the vast majority of the evidence suggests that for plant pathogens high

787

dosage leads to high rates of selection (63), but that there are several mechanisms not well

788

investigated. This box adds another possible exception that is in need of further research.

789

Consider a haploid pathogen species and two at-risk fungicides, A and B, to which high

790

levels of resistance can develop through by a mutation in the genome. There is a gene with alleles

791

making the pathogen sensitive, AS, and resistant, AR, to fungicide A. There is another gene with

792

alleles making the pathogen sensitive, BS, and resistant, BR, to fungicide B. Now consider the use

793

of a, say 50:50 mixture of fungicides A and B. The dose of this mixture used is plotted on the x-axis

794

in figure 5.1. On the y-axis is the infection efficiency of the pathogen with dose-response lines for

795 796

797 798

each of the possible genotypes. When the pathogen is sexually reproducing and the frequency of

799

both the resistance alleles is small, most resistance alleles will be in pathogen individuals that are

800

sensitive to the other fungicide. This is because any individual with genotype ARBR that appears in

801

the population will almost certainly mate an ASBS individual because individuals with this genotype

802

make up the vast majority of the pathogen population, and produce genotypes with one resistance

803

allele combined with sensitivity to the other fungicide. Just as in the case of diploid sexual insects,

804

as we discussed in van den Bosch et al. (63), higher dosages of the mixture may remove more

805

resistance alleles from the pathogen population. Thus delaying the development of individuals with

806

a resistance allele for both fungicides.

807

We note here again that this is speculation and there is no existing evidence for this

808

situation. The case of sexually reproducing pathogen species to which mixtures of two at-risk

809

fungicides are used is in need of further work. Moreover fungi are capable of asexual

810

recombination and this may affect the dynamics of resistance genes as well.

811

Page 27 of 39

812


813 814

Table 1:

815 816 817 818 819

The selection coefficient for resistance to the at-risk fungicide (azoxistrobin) in a range of mixtures with chlorothalonil and a hypothetical fungicide with a lower efficacy (Figure 2). The calculations were done with the model of van den Berg et al (60) parameterised for Mycosphaerella graminicola on wheat. Exponential curves were fitted to the resistance frequency at harvest, the parameter then is the selection coefficient in units of per year.

Dose of mixing partner 0 20 40 60 80 100

Mixing partner is chlorothalonil Dose of at Dose of atDose of atrisk 25% risk 50% risk 100%

Mixing partner with a lower efficacy Dose of at Dose of at- Dose of atrisk 25% risk 50% risk 100%

0.70 0.50 0.45 0.42 0.41 0.41

0.68 0.55 0.53 0.52 0.51 0.51

0.84 0.61 0.55 0.52 0.50 0.49

0.96 0.72 0.65 0.62 0.61 0.59

0.84 0.71 0.68 0.66 0.64 0.64

0.96 0.82 0.79 0.77 0.75 0.75

820 821 822

Table 2:

823 824 825 826

The effective life of mixtures of the at-risk fungicide (azoxistrobin) in a range of mixtures with chlorothalonil and a hypothetical fungicide with a lower efficacy (Figure 2). The calculations were done with the model of van den Berg et al (60) parameterised for Mycosphaerella graminicola on wheat. The effective life has units of years.

Dose of mixing partner 0 20 40 60 80 100 827 828 829 830 831

Mixing partner is chlorothalonil Dose of at Dose of atDose of atrisk 25% risk 50% risk 100%

Mixing partner with a lower efficacy Dose of at Dose of at- Dose of atrisk 25% risk 50% risk 100%

6 9 11 12 13 13

6 7 8 8 8 8

5 8 9 10 11 11

5 7 8 9 9 9

5 7 7 7 8 8

5 6 6 6 7 7

Page 28 of 39

832

FIGURE LEGENDS.


833 834

Figure 1. The selection coefficient, S, expressed as fraction of the intrinsic rate of increase, r, as

835

dependent on the dose of the at-risk fungicide and the dose of the mixing partner. The rate of

836

epidemic increase, rS , expressed as fraction of the intrinsic rate of increase, r, as dependent on

837

the dose of the to-be-protected fungicide and the dose of the mixing partner. Parameter values

838

used are: kM=kA=2, αmax=0.95, θmax=0.75. The three arrows represent options for changing the

839

doses of the two fungicides (see text for further explanation).

840 841

Figure 2: The dose response curves for azoxistrobin sensitive strains and for chlorothalonil and a

842

hypothetical mixing partner with a lower efficacy than chlorothalonil.

843 844

Figure 5.1: The infection efficiency of the four pathogen strains as function of fungicide dose. The

845

AS,BS strain is sensitive to both fungicide A and B, the AS,BR strain is sensitive to fungicide A and

846

resistant to fungicide B, the AR,BS strain is resistant to fungicide A and sensitive to fungicide B, the

847

AR,BR strain is resistant to both fungicide A and B.

848

Dose of the at-risk fungicide


Page 29 of 39

849

850

1.0 s/r=0.6

851

852

853

854

855 856 FIGURE 1.

0.8

1

0.6

0.0 0.2

s/r=0.4

2 s/r=0.3

0.4

3 2

0.4 0.6

rs/r=0.05

3 s/r=0.2

1 s/r=0.1

rs/r=0.3 0.8

Dose of the mixing component rs/r=0.1

0.2 rs/r=0.2

0.0 1.0

Disease severity at grain filling


Page 30 of 39

857

864

865

866

867 Figure 2

858

25

859

860

861

862

863

Figure 5.1 20 Less effective mixing partner Pyraclostrobin Chlorothalonil

15

10

5

0 0.0 0.2 0.4 0.6 0.8

Fraction of the label dose 1.0

Page 31 of 39


e-Xtra 1: r and the life-cycle components, The selection coefficient calculated for fungicides affecting the life-cycle parameters. We show here the example of two fungicides (one at-risk and one not-at-risk) that affect the pathogen infection efficiency. Similar results are obtained when the at-risk fungicide also affects the latent period, as is usual for systemic fungicides, but the calculations become more complex. We show the case where the resistant strain is absolute resistant to the at-risk fungicide. We use an H-E-I-R model to describe the dynamics of the epidemic (35). =

, , , −

=

−

(A.2)

=

−

(A.3)

=

α

(A.4)

(A.1)

Where H is the density of healthy/uninfected host tissue, E is the density of exposed/latently infected host tissue, I is the density of infectious/sporulating host tissue and R is the density of removed/past-sporulating host tissue. The pathogen has a transmission rate, β, p = 1/γ is the mean length of the latent period and i = 1/α is the mean length of the infectious period. The pathogen transmission rate is a composite parameter (35) and is the product of the infection efficiency, the spore production rate and the probability that a spore lands on a leaf. We will thus use the transmission rate as the parameter affected by the fungicides. From this model we can derive that the rate of growth of the epidemic, r, is given by (35)

= + − 1 − where

=

(A.5)

(A.6)

.Each strain of the pathogen has its own growth rate, rS and rR, and depends on the fungicide applied. The transmission rate, β: A dose DA of fungicide A results in a transmission rate

= ! "1 − #$ %1 − & '()* +* ,-

(A.7)

for the sensitive strain and = ! (A.8) for the insensitive strain, where β0 is the maximum transmission rate when no fungicide is used, βmSA and βmRA are the maximum fractional reduction of the transmission rate (i.e. at infinite dose) of the sensitive and the insensitive strain respectively under applications of fungicide A. For the mixtures we get + . = ! "1 − #$ %1 − & '()* +* ,- "1 − #/ %1 − & '()0 +0 ,-

(A.9)

And when the insensitive strain is absolutely resistant to fungicide A we have + . = ! "1 − #/ %1 − & '(10 +0 ,-

(A.10)

Substituting (A.9) in (A.5) we find the rate of increase of the sensitive strain, rS. Substituting (A.10) in (A.5) we find the rate of increase of the resistant strain, rR. Asd in the main text we plot both

Page 32 of 39

contour lines for the selection coefficient, s=rR-rS, and contour lines for the rate of increase of the sensitive strain, rS.

The figure shows that the results of these calculations are qualitatively similar to the calculations in the main text, where it is assumed that the fungicide affects rS and rR.

1.0 s=0.075

s=0.055 s=0.035

Dose of the at-risk fungicice


We use parameters that mimic the case of Zymoseptoria graminicola on wheat with the at-risk fungicide pyraclostrobin and the not-at-risk fungicide chlorotalonil (Hobbelen et al 2011). p=15, i=25, R0=12, ksA=3, ksB =krB =2, we take the βmsA, βmsB and βmrB to be 1, for ease of calculation.

0.8 s=0.015

0.6

rS=0.001 s=0.005 rS=0.02

0.4 rS=0.05 0.2

rS=0.1

0.0 0.0

0.2

0.4

0.6

0.8

1.0

Dose of ther mixing component

Figure A1: Contour lines of the selection coefficient, s, and of the growth rate of the sensitive strain, rS, as function of the dose of the at-risk fungicide in the mixture and the dose of the mixing component.

Page 33 of 39

Online appendix 2, the table for adding a mixing component.


Adding a second fungicide to a to-be-protected fungicide. Experimental evidence of the effect of fungicide mixing on the selection for fungicide insensitivity. The dose of the to-be-protected fungicide is the same in the two compared treatments. Pathogen and host names as listed in the publication. Above the double line are publications where a statistical analysis is presented. Below the double line are publications where such a statistical analysis is not presented. PA = phenylamides; QoI = quinone outside inhibitor; MBC = methyl benzimidazole carbamate.

Mixing component is multi-site acting fungicide Pathogen 1

Host 2

Botrytis cinerea

Strawberry

Botrytis cinerea

Grapevine

Phytophthora infestans

Mixing component

Potato

Fungicide to be protected procymidone (Dicarboximides) iprodione (Dicarboximides) metalaxyl (PA)

dichlofluanid (sulfamides)

Effect of mixtures on resistance selection No effect

captan (Phthalimides)

Decreased

mancozeb (Dithiocarbamates)

Decreased

Phytophthora infestans

Potato

oxadixyl (PA)


Decreased

Phytophthora infestans Phytophthora infestans

Potato

oxadixyl (PA)

Decreased

Tomato & potato

oxadixyl (PA)

Plasmopara viticola

Grapevine

oxadixyl (PA)

Mycosphaerella graminicola Plasmopara viticola Botrytis cinerea

Winter wheat Grapevine

azoxystrobin (QoI)

mancozeb (Dithiocarbamates) mancozeb (Dithiocarbamates) and cymoxanil (Cyanoacetamideoxime) mancozeb (Dithiocarbamates) and cymoxanil (Cyanoacetamideoxime) chlorothalonil (Chloronitriles) folpet (phthalimides)

Strawberry

Botrytis cinerea

Geranium

Botrytis cinerea

Geranium

Venturia inaequalis Venturia inaequalis Erwinia amylovora

cymoxanil (cyanoacetamideoxime) Not specified (Several dicarboximides) vinclozolin (Dicarboximides)

Decreased

Reference Hunter et al. (1987) Northhover & Matteoni (1986) Cohen & Samoucha (1989) Cohen & Samoucha (1990) Grabski & Gisi (1985) Samoucha & Gisi (1987)

Decreased

Samoucha & Gisi (1987)

No effect

McCartney et al. (2007) Genet & Jaworska (2013) Maraite et al. (1985) Vali & Moorman (1992) Vali & Moorman (1992) Köller & Wilcox (1999) Köller & Wilcox (1999) English & van Halsema (1954)

Decreased

Not specified (Several multi-sites) copper hydroxide (inorganic)

Decreased

vinclozolin (Dicarboximides)

chlorothalonil (chloronitriles)

Decreased

Apple

fenarimol (DMI)

Increased

Apple

fenarimol (DMI)

mancozeb (Dithiocarbamates) dodine (guanidines)

Decreased

Pepper

terramycin (antibiotic)

Decreased


Decreased

English & van Halsema (1954)


Decreased

Lalancette et al. (1987)


Decreased

LaMondia (2001)

Xanthomonas vesicatoria

Pepper

Venturia inaequalis

Apple

streptomycin (glucopyranosyl antibiotic) streptomycin (glucopyranosyl antibiotic) benomyl (MBC)

Colletotrichum gloeosporioides

Euonymus fortunei

thiophanate-methyl (MBC)

Decreased


Page 34 of 39


Euonymus fortunei



Euonymus fortunei


Pythium aphanidermatum

Ryegrass

metalaxyl (PA)

Plasmopara viticola

Grapevine

Plasmopara viticola

Ethylene-bis dithiocarbamate (Dithiocarbamates) copper hydroxide (inorganic)

Decreased

LaMondia (2001)

Decreased

LaMondia (2001)


Decreased

Sanders et al. (1985)

famoxadone (QoI)

folpet (phthalimides)

Decreased

Genet et al. (2006)

Grapevine

famoxadone (QoI)


Decreased

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Decreased

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Decreased

Genet et al. (2006)

Venturia inaequalis

Apple

trifloxystrobin (QoI)


Decreased

Turechek & Köller (2004)

Rhychosporium secalis

Spring & winter barley

protioconazole (triazolinthione)


Decreased

Oxley et al. (2008)

Mixing component

Effect of mixtures on resistance selection Decreased

Reference

Mixing component is single-site acting fungicide Pathogen

1

Erysiphe graminis f. sp. hordei Erysiphe graminis f. sp. hordei Erysiphe graminis f. sp. hordei Erysiphe graminis f. sp. hordei Mycospaerella graminicola Mycosphaerella graminicola Phytophthora infestans

Host

2

Barley Spring barley Spring barley Spring barley Wheat

Fungicide to be protected ethirimol (hydroxy(2-amino-)pyrimidines) triadimenol (DMI) triadimenol (DMI)

triadimefon (DMI) ethirimol (hydroxy(2-amino-)Pyrimidines) tridemorph (Amines)

Decreased Decreased

ethirimol (hydroxy-(2amino-)Pyrimidines) fluquinconazole (DMI)

triadimenol (DMI)

Decreased

azoxystrobin (QoI)

Exp 1: no effect Exp. 2: decreased No effect

Winter wheat Potato

epoxiconazole (DMI)

pyraclostrobin (QoI)

metalaxyl (PA)

cymoxanil (Cyanoacetamideoxime)

Decreased

Phytophthora infestans Mycosphaerella graminicola Botrytis cinerea

Potato

oxadixyl (PA)

Decreased

Winter wheat Grapevine

azoxystrobin (QoI)

cymoxanil (Cyanoacetamideoxime) epoxiconazole (DMI)

cyprodinil (AP)

fludioxonil (PP)

Decreased

Erysiphe graminis

Spring barley Winter barley Winter barley Winter barley Wheat

propiconazole (DMI)

tridemorph (Amines)

Decreased

epoxiconazole (DMI)

fenpropimorph (Amines)

Decreased

epoxiconazole (DMI)

cyprodinil (AP)

Decreased

epoxiconazole (DMI)

azoxystrobin (QoI)

Decreased

triadimenol (DMI)


Decreased

Rhynchosporium secalis Rhynchosporium secalis Rhynchosporium secalis Erysiphe graminis f. sp. tritici

No effect

Hunter et al. (1984) Brent et al. (1989) Brent et al. (1989) Brent et al. (1989) Mavroeidi & Shaw (2006) Thygesen et al. (2009) Cohen & Samoucha (1989) Grabski & Gisi (1985) McCartney et al. (2007) Forster & Staub (1996) Bolton & Smith (1988) Cooke et al. (2004) Cooke et al. (2004) Cooke et al. (2004) Lorenz et al. (1992)


Page 35 of 39

Cercospora beticola

Beet

benomyl (MBC)

triphenyltin hydroxide

No effect

Dovas et al. (1976)

Pseudocercospor ella herpotrichoides Pythium aphanidermatum

Wheat

carbendazim (MBC)

prochloraz (DMI)

Decreased

Hoare et al. (1986)

Ryegrass

metalaxyl (PA)

propamocarb (carbamates)

Decreased


Plasmopara viticola

Grapevine

famoxadone (QoI)

fosetyl-al (Phosphonates)

Decreased

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Decreased

Genet et al. (2006)

Plasmopara viticola

Grapevine

Not specified (QoIs)

Not specified (non-QoIs)

Decreased

Toffolatti et al. (2011)

Rhynchosporium secalis

Spring & winter barley Spring & winter barley Spring & winter barley

prothioconazole (triazolinthione)

cyprodinil (AP)

Decreased

Oxley et al. (2008)


pyraclostrobin (QoI)

Decreased

Oxley et al. (2008)



Decreased

Oxley et al. (2008)

Rhynchosporium secalis Rhynchosporium secalis

1

Pathogen synonyms: Blumeria graminis = Erysiphe graminis; Mycosphaerella graminicola = Septoria tritici = Zymoseptoria tritici. Venturia inaequalis anamorphs have been described under the names Fusicladium dendriticum and Spilocaea pomi. Tapesia yallundae is the causal agent for a range of cereal and forage grass diseases. The anamorph o f T. yallundae is the W-type strain of Pseudocercosporella herpotrichoides; the R-type strain of Pseudocercosporella herpotrichoides is now known as Tapesia acuformis. Colletotrichum gloeosporioides is the asexual stage of Glomerella cingulata. 2

Latin host names: wheat = Triticum aestivum; barley = Hordeum vulgare; apple = Malus domestica; grapevine = Vitis; potato = Solanum tuberosum; ryegrass = Lolium species; strawberry = Fragaria; tomato = Solanum lycopersicum; wintercreeper = Euonymus fortunei, beet = Beta vulgaris; pepper = Piper; Geranium = Pelargonium x hortorum.

Page 36 of 39

Online appendix 3, the table for adding a mixing component as well as reducing the dose of


the at-risk Practicable fungicide mixture strategies. Experimental evidence of the effect of fungicide mixing on the selection for fungicide insensitivity. Comparison between the selection for insensitivity when using the to-be-protected fungicide as a single product versus reducing the dose of the to-beprotected and adding a second fungicide to the to-be-protected fungicide. The total dose of the tobe-protected fungicide is not the same in the two compared treatments. Pathogen and host names as listed in the publication. Above the double line are publications where a statistical analysis is presented. Below the double line are publications where such a statistical analysis is not presented.

Mixing component is multi-site fungicide Pathogen 1

Host 2

Mixing component ⱡ

Apple

Fungicide to be protected procymidone (Dicarboximides) iprodione (Dicarboximides) benomyl (MBC)

Botrytis cinerea

Strawberry

Botrytis cinerea

Grapevine

Venturia inaequalis Pseudoperonospo ra cubensis Botrytis cinerea

Cucumber

flumorph (CAA)

Geranium

Apple

vinclozolin (Dicarboximides) vinclozolin (Dicarboximides) fenarimol (DMI)

Botrytis cinerea

Geranium

Venturia inaequalis Venturia inaequalis Phytophthora infestans

Apple

fenarimol (DMI)

Tomato & potato

oxadixyl (PA)

Plasmopara viticola

Grapevine

oxadixyl (PA)


Ryegrass

metalaxyl (PA)

Effect of mixtures on resistance selection Decreased

dichlofluanid (Sulfamides) captan (phthalimides)

No effect

captan (phthalimides)

Increased

mancozeb (Dithiocarbamates) Copper hydroxide (inorganic) chlorothalonil (Chloronitriles) mancozeb (Dithiocarbamates) dodine (guanidines)

Decreased

mancozeb (Dithiocarbamates) and cymoxanil (cyanoacetamideoxime) mancozeb (Dithiocarbamates) and cymoxanil (cyanoacetamideoxime) mancozeb (Dithiocarbamates)

Decreased

Decreased Decreased No effect Decreased

Reference Hunter et al. (1987) Northhover & Matteoni (1986) McGee & Zuck (1981) Zhu et al. (2007) Vali & Moorman (1992) Vali & Moorman (1992) Köller & Wilcox (1999) Köller & Wilcox (1999) Samoucha & Gisi (1987)

Decreased

Samoucha & Gisi (1987)

Decreased


Effect of mixtures on resistance selection No effect

Reference

Mixing component is single-site fungicide Pathogen

1

Blumeria graminis Septoria nodorum Cercosporella

Host

2

Wheat Wheat Wheat

Fungicide to be protected fenpropimorph (Amines) carbendazim (MBC) carbendazim

Mixing component

ⱡ

propiconazole (DMI) edifenphos (Phosphoro-thiolates) edifenphos

Decreased

Forster et al. (1994) Horsten (1979)

Decreased

Horsten (1979)

Page 37 of 39


herpotrichoides Botrytis cinerea

Grapevine

(MBC) cyprodinil (AP)

(Phosphoro-thiolates) fludioxonil (PP)

thiabendazole (MBC) thiabendazole (MBC)

imazalil (DMI) imazalil (DMI)

Exp. 1: decreased Exp. 2: no effect Decreased

Decreased

Forster & Staub (1996) Carnegie et al. (2008) Carnegie et al. (1994)

Polyscytalum pustulans Polyscytalum pustulans

Potato

Helminthosporiu m solani

Potato

thiabendazole (MBC)

imazalil (DMI)

Decreased

Carnegie et al. (1994)

Rhychosporium secalis

Winter barley

carbendazim (MBC)

propiconazole (DMI)

Taggert et al. (1998)


Ryegrass

metalaxyl (PA)

propamocarb (Carbamates)

Authors suggest decreased, but retrospective analysis suggests no sign. Effect Decreased

Potato


1

Pathogen synonyms: Blumeria graminis = Erysiphe graminis; Polyscytalum pustulans = Oospora pustulans. Venturia inaequalis anamorphs have been described under the names Fusicladium dendriticum and Spilocaea pomi. Tapesia yallundae is the causal agent for a range of cereal and forage grass diseases. The anamorph o f T. yallundae is the W-type strain of Pseudocercosporella herpotrichoides; the R-type strain of Pseudocercosporella herpotrichoides is now known as Tapesia acuformis. 2

Latin host names: wheat = Triticum; apple = Malus domestica; grapevine = Vitis; potato = Solanum tuberosum; ryegrass = Lolium; strawberry = Fragaria; tomato = Solanum lycopersicum; Geranium = Pelargonium x hortorum; cucumber = Cucurbita.

Page 38 of 39

Online appendix 4, is increasing dose of mixing partner decreasing the selection further.


Does an increasing dose further reduce selection?

Pathogen

Host

Fungicide to be protected

Mixing component

Does increasing dose reduce selection

Reference

Plasmopara viticola

Grapevine

famoxadone (QoI)


Yes

Genet et al. (2006)

Plasmopara viticola

Grapevine

famoxadone (QoI)


Yes

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Yes

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Yes

Genet et al. (2006)

Erwinia amylovora

Pepper


yes

English & van Halsema (1954)

Venturia inaequalis

Apple

streptomycin (glucopyranosyl antibiotic) benomyl (MBC)


yes

Lalancette et al. (1987)

Plasmopara viticola

Grapevine

famoxadone (QoI)


Yes

Genet et al. (2006)

Plasmopara viticola

Grapevine

azoxystrobin (QoI)


Yes

Genet et al. (2006)

Erysiphe graminis

Spring barley

propiconazole (DMI)

tridemorph (Amines)

Yes

Bolton & Smith (1988)

Single site

Although Bolton & Smith the two dosages of the mixing partner were used in different years. So not an ideal comparison.

Page 39 of 39

Online appendix 5, ‘’is it more effective to reduce the dose of the at-risk or increase the dose of the mixing partner?’’ Three papers which address this question have been omitted from the table below, specifically:


Koller et all 1999 (which is considered as an exception in Box 3), Genet 2006 due to difficulty recalculating data into dose rates and Sanders et al 1985 as this is a laboratory experiment and dose is not comparable to field.

Pathogen

Host

Erysiphe graminis f. sp. hordei Erysiphe graminis f. sp. hordei Mycospaerella graminicola Venturia inaequalis

Barley

Botrytis cinerea

Barley

Fungicide to be protected ethirimol (hydroxy(2-amino-)Pyrimidines) triadimefon (DMI)

Wheat

fluquinconazole (DMI)

Apple

benomyl (MBC)

Strawberry

procymidone (Dicarboximides)

Mixing component triadimefon (DMI) ethirimol (hydroxy(2-amino-)Pyrimidines) azoxystrobin (QoI) mancozeb (Dithiocarbamates) dichlofluanid (Sulfamides)

Lowering at-risk more effective? yes yes yes yes

Exp 1 unconvincing due to high % res. Exp 2 ‘yes’.

Reference Hunter et al. (1984) Hunter et al. (1984) Mavroeidi & Shaw (2006) Lalancette et al. (1987) Hunter et al. (1987)

Governing principles can guide fungicide-resistance management tactics.

The evolution of fungicide resistance.

Does farm fungicide use induce azole resistance in Aspergillus fumigatus?

Identification of Genes Related to Fungicide Resistance in Fusarium fujikuroi.

Sanctions as a tactic used in partner conflicts: theoretical, operational, and preliminary findings.

A test of male infanticide as a reproductive tactic in a cichlid fish.

Life as a bachelor: quantifying the success of an alternative reproductive tactic in male blue monkeys.

Resistance and resilience responses of a range of soil eukaryote and bacterial taxa to fungicide application.

A fungicide-responsive kinase as a tool for synthetic cell fate regulation.

Epidemiological and evolutionary management of plant resistance: optimizing the deployment of cultivar mixtures in time and space in agricultural landscapes.

Environmental azole fungicide, prochloraz, can induce cross-resistance to medical triazoles in Candida glabrata.

Suspension Array for Multiplex Detection of Eight Fungicide-Resistance Related Alleles in Botrytis cinerea.

Spatial distribution of single-nucleotide polymorphisms related to fungicide resistance and implications for sampling.

Fungicide resistance risk assessment based on traits associated with the rate of pathogen evolution.

Mechanisms of Resistance to an Azole Fungicide in the Grapevine Powdery Mildew Fungus, Erysiphe necator.

Does fungicide application in vineyards induce resistance to medical azoles in Aspergillus species?

Enhanced fungicide resistance in Isaria fumosorosea following ionizing radiation-induced mutagenesis.

Allelopathic interactions in crop-weed mixtures : Applications for weed management.

Non-target impact of fungicide tetraconazole on microbial communities in soils with different agricultural management.

Comparative analyses of fungicide sensitivity and SSR marker variations indicate a low risk of developing azoxystrobin resistance in Phytophthora infestans.

Chemical management in fungicide sensitivity of Mycosphaerella fijiensis collected from banana fields in México.

Botrytis fragariae, a New Species Causing Gray Mold on Strawberries, Shows High Frequencies of Specific and Efflux-Based Fungicide Resistance.

A pleiotropic drug resistance transporter is involved in reduced sensitivity to multiple fungicide classes in Sclerotinia homoeocarpa (F.T. Bennett).

Microbial conversion of fungicide vinclozolin.